{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.7980221756068326, "eval_steps": 500, "global_step": 30000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.005993407252022775, "grad_norm": 2.2955188751220703, "learning_rate": 2.996703626011388e-07, "loss": 0.2171, "step": 100 }, { "epoch": 0.01198681450404555, "grad_norm": 0.9237516522407532, "learning_rate": 5.993407252022776e-07, "loss": 0.139, "step": 200 }, { "epoch": 0.017980221756068324, "grad_norm": 0.4022263288497925, "learning_rate": 8.990110878034162e-07, "loss": 0.1111, "step": 300 }, { "epoch": 0.0239736290080911, "grad_norm": 0.28793156147003174, "learning_rate": 1.1986814504045551e-06, "loss": 0.1079, "step": 400 }, { "epoch": 0.029967036260113874, "grad_norm": 0.479831337928772, "learning_rate": 1.4983518130056939e-06, "loss": 0.1078, "step": 500 }, { "epoch": 0.03596044351213665, "grad_norm": 0.2760140895843506, "learning_rate": 1.7980221756068325e-06, "loss": 0.1052, "step": 600 }, { "epoch": 0.041953850764159424, "grad_norm": 0.44269922375679016, "learning_rate": 2.0976925382079712e-06, "loss": 0.1052, "step": 700 }, { "epoch": 0.0479472580161822, "grad_norm": 0.25097307562828064, "learning_rate": 2.3973629008091102e-06, "loss": 0.1054, "step": 800 }, { "epoch": 0.05394066526820498, "grad_norm": 0.3348759412765503, "learning_rate": 2.6970332634102492e-06, "loss": 0.1042, "step": 900 }, { "epoch": 0.05993407252022775, "grad_norm": 0.11476008594036102, "learning_rate": 2.9967036260113878e-06, "loss": 0.104, "step": 1000 }, { "epoch": 0.06592747977225052, "grad_norm": 0.1612076610326767, "learning_rate": 3.2963739886125264e-06, "loss": 0.1041, "step": 1100 }, { "epoch": 0.0719208870242733, "grad_norm": 0.3466070592403412, "learning_rate": 3.596044351213665e-06, "loss": 0.1034, "step": 1200 }, { "epoch": 0.07791429427629608, "grad_norm": 0.41668832302093506, "learning_rate": 3.8957147138148035e-06, "loss": 0.1023, "step": 1300 }, { "epoch": 0.08390770152831885, "grad_norm": 0.3885754346847534, "learning_rate": 4.1953850764159425e-06, "loss": 0.1038, "step": 1400 }, { "epoch": 0.08990110878034162, "grad_norm": 0.3091404438018799, "learning_rate": 4.4950554390170815e-06, "loss": 0.1022, "step": 1500 }, { "epoch": 0.0958945160323644, "grad_norm": 0.24296438694000244, "learning_rate": 4.7947258016182205e-06, "loss": 0.1014, "step": 1600 }, { "epoch": 0.10188792328438717, "grad_norm": 0.2710008919239044, "learning_rate": 5.0943961642193594e-06, "loss": 0.1014, "step": 1700 }, { "epoch": 0.10788133053640996, "grad_norm": 0.10421634465456009, "learning_rate": 5.3940665268204984e-06, "loss": 0.1006, "step": 1800 }, { "epoch": 0.11387473778843273, "grad_norm": 0.35064780712127686, "learning_rate": 5.6937368894216374e-06, "loss": 0.1006, "step": 1900 }, { "epoch": 0.1198681450404555, "grad_norm": 0.3558998703956604, "learning_rate": 5.9934072520227756e-06, "loss": 0.101, "step": 2000 }, { "epoch": 0.12586155229247828, "grad_norm": 0.13521741330623627, "learning_rate": 6.2930776146239146e-06, "loss": 0.101, "step": 2100 }, { "epoch": 0.13185495954450105, "grad_norm": 0.23776298761367798, "learning_rate": 6.592747977225053e-06, "loss": 0.1013, "step": 2200 }, { "epoch": 0.13784836679652382, "grad_norm": 0.12444166839122772, "learning_rate": 6.892418339826191e-06, "loss": 0.1008, "step": 2300 }, { "epoch": 0.1438417740485466, "grad_norm": 0.13655610382556915, "learning_rate": 7.19208870242733e-06, "loss": 0.1019, "step": 2400 }, { "epoch": 0.1498351813005694, "grad_norm": 0.27996689081192017, "learning_rate": 7.491759065028469e-06, "loss": 0.1014, "step": 2500 }, { "epoch": 0.15582858855259216, "grad_norm": 0.30384400486946106, "learning_rate": 7.791429427629607e-06, "loss": 0.1005, "step": 2600 }, { "epoch": 0.16182199580461493, "grad_norm": 0.06711748242378235, "learning_rate": 8.091099790230747e-06, "loss": 0.1001, "step": 2700 }, { "epoch": 0.1678154030566377, "grad_norm": 0.20208857953548431, "learning_rate": 8.390770152831885e-06, "loss": 0.1011, "step": 2800 }, { "epoch": 0.17380881030866047, "grad_norm": 0.5147479772567749, "learning_rate": 8.690440515433025e-06, "loss": 0.1012, "step": 2900 }, { "epoch": 0.17980221756068324, "grad_norm": 0.2111215740442276, "learning_rate": 8.990110878034163e-06, "loss": 0.1002, "step": 3000 }, { "epoch": 0.18579562481270603, "grad_norm": 0.3195187449455261, "learning_rate": 9.289781240635301e-06, "loss": 0.1005, "step": 3100 }, { "epoch": 0.1917890320647288, "grad_norm": 0.1801885962486267, "learning_rate": 9.589451603236441e-06, "loss": 0.0999, "step": 3200 }, { "epoch": 0.19778243931675157, "grad_norm": 0.07530693709850311, "learning_rate": 9.889121965837579e-06, "loss": 0.1019, "step": 3300 }, { "epoch": 0.20377584656877434, "grad_norm": 0.13042743504047394, "learning_rate": 9.97902307461792e-06, "loss": 0.1003, "step": 3400 }, { "epoch": 0.2097692538207971, "grad_norm": 0.26616647839546204, "learning_rate": 9.94572636766224e-06, "loss": 0.1007, "step": 3500 }, { "epoch": 0.2157626610728199, "grad_norm": 0.35827577114105225, "learning_rate": 9.912429660706557e-06, "loss": 0.1007, "step": 3600 }, { "epoch": 0.22175606832484268, "grad_norm": 0.30124035477638245, "learning_rate": 9.879132953750875e-06, "loss": 0.1, "step": 3700 }, { "epoch": 0.22774947557686545, "grad_norm": 0.34060993790626526, "learning_rate": 9.845836246795192e-06, "loss": 0.1013, "step": 3800 }, { "epoch": 0.23374288282888822, "grad_norm": 0.14318062365055084, "learning_rate": 9.812539539839511e-06, "loss": 0.1008, "step": 3900 }, { "epoch": 0.239736290080911, "grad_norm": 0.310312420129776, "learning_rate": 9.779242832883829e-06, "loss": 0.0999, "step": 4000 }, { "epoch": 0.24572969733293376, "grad_norm": 0.1943434178829193, "learning_rate": 9.745946125928146e-06, "loss": 0.0999, "step": 4100 }, { "epoch": 0.25172310458495656, "grad_norm": 0.22400355339050293, "learning_rate": 9.712649418972464e-06, "loss": 0.1016, "step": 4200 }, { "epoch": 0.2577165118369793, "grad_norm": 0.13477976620197296, "learning_rate": 9.679352712016781e-06, "loss": 0.101, "step": 4300 }, { "epoch": 0.2637099190890021, "grad_norm": 0.15206047892570496, "learning_rate": 9.6460560050611e-06, "loss": 0.1009, "step": 4400 }, { "epoch": 0.2697033263410249, "grad_norm": 0.060487061738967896, "learning_rate": 9.61275929810542e-06, "loss": 0.0996, "step": 4500 }, { "epoch": 0.27569673359304764, "grad_norm": 0.16848477721214294, "learning_rate": 9.579462591149737e-06, "loss": 0.0998, "step": 4600 }, { "epoch": 0.28169014084507044, "grad_norm": 0.10402075946331024, "learning_rate": 9.546165884194054e-06, "loss": 0.1012, "step": 4700 }, { "epoch": 0.2876835480970932, "grad_norm": 0.12113852798938751, "learning_rate": 9.512869177238371e-06, "loss": 0.1008, "step": 4800 }, { "epoch": 0.293676955349116, "grad_norm": 0.23850949108600616, "learning_rate": 9.479572470282689e-06, "loss": 0.1005, "step": 4900 }, { "epoch": 0.2996703626011388, "grad_norm": 0.3209283649921417, "learning_rate": 9.446275763327008e-06, "loss": 0.1005, "step": 5000 }, { "epoch": 0.3056637698531615, "grad_norm": 0.6144764423370361, "learning_rate": 9.412979056371325e-06, "loss": 0.1, "step": 5100 }, { "epoch": 0.3116571771051843, "grad_norm": 0.04221544787287712, "learning_rate": 9.379682349415643e-06, "loss": 0.0997, "step": 5200 }, { "epoch": 0.31765058435720706, "grad_norm": 0.07346853613853455, "learning_rate": 9.346385642459962e-06, "loss": 0.1011, "step": 5300 }, { "epoch": 0.32364399160922985, "grad_norm": 0.3777712285518646, "learning_rate": 9.31308893550428e-06, "loss": 0.1008, "step": 5400 }, { "epoch": 0.3296373988612526, "grad_norm": 0.07576871663331985, "learning_rate": 9.279792228548597e-06, "loss": 0.1001, "step": 5500 }, { "epoch": 0.3356308061132754, "grad_norm": 0.194047212600708, "learning_rate": 9.246495521592916e-06, "loss": 0.1003, "step": 5600 }, { "epoch": 0.3416242133652982, "grad_norm": 0.09262324869632721, "learning_rate": 9.213198814637233e-06, "loss": 0.0995, "step": 5700 }, { "epoch": 0.34761762061732093, "grad_norm": 0.5301212668418884, "learning_rate": 9.17990210768155e-06, "loss": 0.0999, "step": 5800 }, { "epoch": 0.35361102786934373, "grad_norm": 0.04346400871872902, "learning_rate": 9.146605400725868e-06, "loss": 0.0995, "step": 5900 }, { "epoch": 0.3596044351213665, "grad_norm": 0.16581957042217255, "learning_rate": 9.113308693770187e-06, "loss": 0.0995, "step": 6000 }, { "epoch": 0.36559784237338927, "grad_norm": 0.37740394473075867, "learning_rate": 9.080011986814505e-06, "loss": 0.1005, "step": 6100 }, { "epoch": 0.37159124962541207, "grad_norm": 0.08017630875110626, "learning_rate": 9.046715279858824e-06, "loss": 0.1, "step": 6200 }, { "epoch": 0.3775846568774348, "grad_norm": 0.1393900364637375, "learning_rate": 9.013418572903141e-06, "loss": 0.1007, "step": 6300 }, { "epoch": 0.3835780641294576, "grad_norm": 0.11385737359523773, "learning_rate": 8.980121865947459e-06, "loss": 0.1001, "step": 6400 }, { "epoch": 0.38957147138148035, "grad_norm": 0.1457832008600235, "learning_rate": 8.946825158991776e-06, "loss": 0.1, "step": 6500 }, { "epoch": 0.39556487863350315, "grad_norm": 0.03458058834075928, "learning_rate": 8.913528452036094e-06, "loss": 0.0999, "step": 6600 }, { "epoch": 0.40155828588552595, "grad_norm": 0.4892089366912842, "learning_rate": 8.880231745080413e-06, "loss": 0.1, "step": 6700 }, { "epoch": 0.4075516931375487, "grad_norm": 0.11881079524755478, "learning_rate": 8.84693503812473e-06, "loss": 0.1009, "step": 6800 }, { "epoch": 0.4135451003895715, "grad_norm": 0.29624250531196594, "learning_rate": 8.813638331169047e-06, "loss": 0.1007, "step": 6900 }, { "epoch": 0.4195385076415942, "grad_norm": 0.18601752817630768, "learning_rate": 8.780341624213367e-06, "loss": 0.1002, "step": 7000 }, { "epoch": 0.425531914893617, "grad_norm": 0.017990827560424805, "learning_rate": 8.747044917257684e-06, "loss": 0.0998, "step": 7100 }, { "epoch": 0.4315253221456398, "grad_norm": 0.08424385637044907, "learning_rate": 8.713748210302001e-06, "loss": 0.1005, "step": 7200 }, { "epoch": 0.43751872939766256, "grad_norm": 0.40114349126815796, "learning_rate": 8.68045150334632e-06, "loss": 0.0994, "step": 7300 }, { "epoch": 0.44351213664968536, "grad_norm": 0.10041668266057968, "learning_rate": 8.647154796390638e-06, "loss": 0.0999, "step": 7400 }, { "epoch": 0.4495055439017081, "grad_norm": 0.4256216585636139, "learning_rate": 8.613858089434955e-06, "loss": 0.101, "step": 7500 }, { "epoch": 0.4554989511537309, "grad_norm": 0.2503170371055603, "learning_rate": 8.580561382479273e-06, "loss": 0.0996, "step": 7600 }, { "epoch": 0.46149235840575364, "grad_norm": 0.09117023646831512, "learning_rate": 8.547264675523592e-06, "loss": 0.0987, "step": 7700 }, { "epoch": 0.46748576565777644, "grad_norm": 0.12042222917079926, "learning_rate": 8.51396796856791e-06, "loss": 0.1002, "step": 7800 }, { "epoch": 0.47347917290979924, "grad_norm": 0.13141508400440216, "learning_rate": 8.480671261612228e-06, "loss": 0.101, "step": 7900 }, { "epoch": 0.479472580161822, "grad_norm": 0.20164711773395538, "learning_rate": 8.447374554656546e-06, "loss": 0.101, "step": 8000 }, { "epoch": 0.4854659874138448, "grad_norm": 0.09953387826681137, "learning_rate": 8.414077847700863e-06, "loss": 0.0998, "step": 8100 }, { "epoch": 0.4914593946658675, "grad_norm": 0.03517329692840576, "learning_rate": 8.38078114074518e-06, "loss": 0.1008, "step": 8200 }, { "epoch": 0.4974528019178903, "grad_norm": 0.01899944245815277, "learning_rate": 8.347484433789498e-06, "loss": 0.0995, "step": 8300 }, { "epoch": 0.5034462091699131, "grad_norm": 0.20903293788433075, "learning_rate": 8.314187726833817e-06, "loss": 0.0996, "step": 8400 }, { "epoch": 0.5094396164219359, "grad_norm": 0.28464946150779724, "learning_rate": 8.280891019878135e-06, "loss": 0.1007, "step": 8500 }, { "epoch": 0.5154330236739586, "grad_norm": 0.2716582417488098, "learning_rate": 8.247594312922452e-06, "loss": 0.1001, "step": 8600 }, { "epoch": 0.5214264309259814, "grad_norm": 0.07964222133159637, "learning_rate": 8.21429760596677e-06, "loss": 0.1, "step": 8700 }, { "epoch": 0.5274198381780042, "grad_norm": 0.0397951677441597, "learning_rate": 8.181000899011089e-06, "loss": 0.1003, "step": 8800 }, { "epoch": 0.533413245430027, "grad_norm": 0.31348341703414917, "learning_rate": 8.147704192055406e-06, "loss": 0.1001, "step": 8900 }, { "epoch": 0.5394066526820498, "grad_norm": 0.0627601370215416, "learning_rate": 8.114407485099725e-06, "loss": 0.1003, "step": 9000 }, { "epoch": 0.5454000599340725, "grad_norm": 0.2744080424308777, "learning_rate": 8.081110778144043e-06, "loss": 0.1002, "step": 9100 }, { "epoch": 0.5513934671860953, "grad_norm": 0.14465458691120148, "learning_rate": 8.04781407118836e-06, "loss": 0.0998, "step": 9200 }, { "epoch": 0.5573868744381181, "grad_norm": 0.21187633275985718, "learning_rate": 8.014517364232677e-06, "loss": 0.0999, "step": 9300 }, { "epoch": 0.5633802816901409, "grad_norm": 0.06333980709314346, "learning_rate": 7.981220657276996e-06, "loss": 0.0998, "step": 9400 }, { "epoch": 0.5693736889421637, "grad_norm": 0.3627989888191223, "learning_rate": 7.947923950321314e-06, "loss": 0.0996, "step": 9500 }, { "epoch": 0.5753670961941864, "grad_norm": 0.11346852034330368, "learning_rate": 7.914627243365631e-06, "loss": 0.1008, "step": 9600 }, { "epoch": 0.5813605034462092, "grad_norm": 0.1832839399576187, "learning_rate": 7.88133053640995e-06, "loss": 0.1, "step": 9700 }, { "epoch": 0.587353910698232, "grad_norm": 0.034973230212926865, "learning_rate": 7.848033829454268e-06, "loss": 0.1, "step": 9800 }, { "epoch": 0.5933473179502547, "grad_norm": 0.11853715777397156, "learning_rate": 7.814737122498585e-06, "loss": 0.0991, "step": 9900 }, { "epoch": 0.5993407252022775, "grad_norm": 0.13182196021080017, "learning_rate": 7.781440415542903e-06, "loss": 0.0993, "step": 10000 }, { "epoch": 0.6053341324543002, "grad_norm": 0.09213405847549438, "learning_rate": 7.748143708587222e-06, "loss": 0.0997, "step": 10100 }, { "epoch": 0.611327539706323, "grad_norm": 0.06769004464149475, "learning_rate": 7.71484700163154e-06, "loss": 0.1008, "step": 10200 }, { "epoch": 0.6173209469583458, "grad_norm": 0.020326493307948112, "learning_rate": 7.681550294675857e-06, "loss": 0.0993, "step": 10300 }, { "epoch": 0.6233143542103686, "grad_norm": 0.04122787341475487, "learning_rate": 7.648253587720174e-06, "loss": 0.0997, "step": 10400 }, { "epoch": 0.6293077614623914, "grad_norm": 0.1741415411233902, "learning_rate": 7.614956880764492e-06, "loss": 0.0998, "step": 10500 }, { "epoch": 0.6353011687144141, "grad_norm": 0.028361910954117775, "learning_rate": 7.5816601738088115e-06, "loss": 0.0992, "step": 10600 }, { "epoch": 0.6412945759664369, "grad_norm": 0.10101690888404846, "learning_rate": 7.548363466853129e-06, "loss": 0.0987, "step": 10700 }, { "epoch": 0.6472879832184597, "grad_norm": 0.1292373090982437, "learning_rate": 7.515066759897447e-06, "loss": 0.101, "step": 10800 }, { "epoch": 0.6532813904704825, "grad_norm": 0.25996342301368713, "learning_rate": 7.4817700529417646e-06, "loss": 0.0998, "step": 10900 }, { "epoch": 0.6592747977225052, "grad_norm": 0.08856533467769623, "learning_rate": 7.448473345986083e-06, "loss": 0.0993, "step": 11000 }, { "epoch": 0.665268204974528, "grad_norm": 0.11309339106082916, "learning_rate": 7.4151766390304e-06, "loss": 0.099, "step": 11100 }, { "epoch": 0.6712616122265508, "grad_norm": 0.28294652700424194, "learning_rate": 7.381879932074718e-06, "loss": 0.1009, "step": 11200 }, { "epoch": 0.6772550194785736, "grad_norm": 0.18077927827835083, "learning_rate": 7.348583225119036e-06, "loss": 0.1004, "step": 11300 }, { "epoch": 0.6832484267305964, "grad_norm": 0.03750494495034218, "learning_rate": 7.315286518163355e-06, "loss": 0.0996, "step": 11400 }, { "epoch": 0.6892418339826191, "grad_norm": 0.11316617578268051, "learning_rate": 7.2819898112076725e-06, "loss": 0.1003, "step": 11500 }, { "epoch": 0.6952352412346419, "grad_norm": 0.06800729036331177, "learning_rate": 7.248693104251991e-06, "loss": 0.0998, "step": 11600 }, { "epoch": 0.7012286484866647, "grad_norm": 0.024295495823025703, "learning_rate": 7.215396397296308e-06, "loss": 0.1006, "step": 11700 }, { "epoch": 0.7072220557386875, "grad_norm": 0.03560725226998329, "learning_rate": 7.1820996903406256e-06, "loss": 0.0996, "step": 11800 }, { "epoch": 0.7132154629907103, "grad_norm": 0.08191557973623276, "learning_rate": 7.148802983384944e-06, "loss": 0.0998, "step": 11900 }, { "epoch": 0.719208870242733, "grad_norm": 0.1713048815727234, "learning_rate": 7.115506276429261e-06, "loss": 0.1, "step": 12000 }, { "epoch": 0.7252022774947557, "grad_norm": 0.04900109022855759, "learning_rate": 7.0822095694735795e-06, "loss": 0.0996, "step": 12100 }, { "epoch": 0.7311956847467785, "grad_norm": 0.04907857999205589, "learning_rate": 7.048912862517897e-06, "loss": 0.0999, "step": 12200 }, { "epoch": 0.7371890919988013, "grad_norm": 0.07542908936738968, "learning_rate": 7.015616155562216e-06, "loss": 0.0988, "step": 12300 }, { "epoch": 0.7431824992508241, "grad_norm": 0.0938209518790245, "learning_rate": 6.9823194486065335e-06, "loss": 0.0992, "step": 12400 }, { "epoch": 0.7491759065028468, "grad_norm": 0.02947254106402397, "learning_rate": 6.949022741650852e-06, "loss": 0.0983, "step": 12500 }, { "epoch": 0.7551693137548696, "grad_norm": 0.4940189719200134, "learning_rate": 6.915726034695169e-06, "loss": 0.0992, "step": 12600 }, { "epoch": 0.7611627210068924, "grad_norm": 2.7082247734069824, "learning_rate": 6.8824293277394874e-06, "loss": 0.0988, "step": 12700 }, { "epoch": 0.7671561282589152, "grad_norm": 0.427272766828537, "learning_rate": 6.849132620783805e-06, "loss": 0.0999, "step": 12800 }, { "epoch": 0.773149535510938, "grad_norm": 0.11108270287513733, "learning_rate": 6.815835913828123e-06, "loss": 0.1006, "step": 12900 }, { "epoch": 0.7791429427629607, "grad_norm": 0.05636779963970184, "learning_rate": 6.7825392068724405e-06, "loss": 0.0998, "step": 13000 }, { "epoch": 0.7851363500149835, "grad_norm": 0.2918495237827301, "learning_rate": 6.749242499916758e-06, "loss": 0.0999, "step": 13100 }, { "epoch": 0.7911297572670063, "grad_norm": 0.08755291998386383, "learning_rate": 6.715945792961077e-06, "loss": 0.1006, "step": 13200 }, { "epoch": 0.7971231645190291, "grad_norm": 0.11561553925275803, "learning_rate": 6.682649086005395e-06, "loss": 0.0994, "step": 13300 }, { "epoch": 0.8031165717710519, "grad_norm": 0.03734762221574783, "learning_rate": 6.649352379049713e-06, "loss": 0.0989, "step": 13400 }, { "epoch": 0.8091099790230746, "grad_norm": 0.07186517119407654, "learning_rate": 6.61605567209403e-06, "loss": 0.099, "step": 13500 }, { "epoch": 0.8151033862750974, "grad_norm": 0.29982343316078186, "learning_rate": 6.5827589651383484e-06, "loss": 0.1, "step": 13600 }, { "epoch": 0.8210967935271202, "grad_norm": 0.09515996277332306, "learning_rate": 6.549462258182666e-06, "loss": 0.0992, "step": 13700 }, { "epoch": 0.827090200779143, "grad_norm": 0.07374672591686249, "learning_rate": 6.516165551226984e-06, "loss": 0.0992, "step": 13800 }, { "epoch": 0.8330836080311658, "grad_norm": 0.20344111323356628, "learning_rate": 6.4828688442713015e-06, "loss": 0.1004, "step": 13900 }, { "epoch": 0.8390770152831885, "grad_norm": 0.43719688057899475, "learning_rate": 6.44957213731562e-06, "loss": 0.0994, "step": 14000 }, { "epoch": 0.8450704225352113, "grad_norm": 0.06093699112534523, "learning_rate": 6.416275430359938e-06, "loss": 0.0997, "step": 14100 }, { "epoch": 0.851063829787234, "grad_norm": 0.25451895594596863, "learning_rate": 6.382978723404256e-06, "loss": 0.0991, "step": 14200 }, { "epoch": 0.8570572370392568, "grad_norm": 0.09084396809339523, "learning_rate": 6.349682016448574e-06, "loss": 0.1, "step": 14300 }, { "epoch": 0.8630506442912796, "grad_norm": 0.1228547915816307, "learning_rate": 6.316385309492892e-06, "loss": 0.0997, "step": 14400 }, { "epoch": 0.8690440515433023, "grad_norm": 0.27622973918914795, "learning_rate": 6.2830886025372095e-06, "loss": 0.0989, "step": 14500 }, { "epoch": 0.8750374587953251, "grad_norm": 0.11149102449417114, "learning_rate": 6.249791895581528e-06, "loss": 0.1002, "step": 14600 }, { "epoch": 0.8810308660473479, "grad_norm": 0.10650242865085602, "learning_rate": 6.216495188625845e-06, "loss": 0.0994, "step": 14700 }, { "epoch": 0.8870242732993707, "grad_norm": 0.46797627210617065, "learning_rate": 6.1831984816701626e-06, "loss": 0.1005, "step": 14800 }, { "epoch": 0.8930176805513935, "grad_norm": 0.46854981780052185, "learning_rate": 6.149901774714481e-06, "loss": 0.0988, "step": 14900 }, { "epoch": 0.8990110878034162, "grad_norm": 0.12268362194299698, "learning_rate": 6.1166050677588e-06, "loss": 0.0988, "step": 15000 }, { "epoch": 0.905004495055439, "grad_norm": 0.13641001284122467, "learning_rate": 6.083308360803117e-06, "loss": 0.099, "step": 15100 }, { "epoch": 0.9109979023074618, "grad_norm": 0.014134293422102928, "learning_rate": 6.050011653847435e-06, "loss": 0.0983, "step": 15200 }, { "epoch": 0.9169913095594846, "grad_norm": 0.1600305587053299, "learning_rate": 6.016714946891753e-06, "loss": 0.0984, "step": 15300 }, { "epoch": 0.9229847168115073, "grad_norm": 0.04153376817703247, "learning_rate": 5.9834182399360705e-06, "loss": 0.0998, "step": 15400 }, { "epoch": 0.9289781240635301, "grad_norm": 0.364000529050827, "learning_rate": 5.950121532980389e-06, "loss": 0.099, "step": 15500 }, { "epoch": 0.9349715313155529, "grad_norm": 0.10466318577528, "learning_rate": 5.916824826024706e-06, "loss": 0.0997, "step": 15600 }, { "epoch": 0.9409649385675757, "grad_norm": 0.2283148467540741, "learning_rate": 5.883528119069024e-06, "loss": 0.0996, "step": 15700 }, { "epoch": 0.9469583458195985, "grad_norm": 0.07752685993909836, "learning_rate": 5.850231412113343e-06, "loss": 0.1004, "step": 15800 }, { "epoch": 0.9529517530716212, "grad_norm": 0.0329587459564209, "learning_rate": 5.816934705157661e-06, "loss": 0.0991, "step": 15900 }, { "epoch": 0.958945160323644, "grad_norm": 0.1508065164089203, "learning_rate": 5.783637998201978e-06, "loss": 0.1006, "step": 16000 }, { "epoch": 0.9649385675756668, "grad_norm": 0.029579168185591698, "learning_rate": 5.750341291246297e-06, "loss": 0.0991, "step": 16100 }, { "epoch": 0.9709319748276896, "grad_norm": 0.029158690944314003, "learning_rate": 5.717044584290614e-06, "loss": 0.0989, "step": 16200 }, { "epoch": 0.9769253820797124, "grad_norm": 0.08389827609062195, "learning_rate": 5.683747877334932e-06, "loss": 0.0997, "step": 16300 }, { "epoch": 0.982918789331735, "grad_norm": 0.13760380446910858, "learning_rate": 5.65045117037925e-06, "loss": 0.099, "step": 16400 }, { "epoch": 0.9889121965837578, "grad_norm": 0.06386566907167435, "learning_rate": 5.617154463423567e-06, "loss": 0.0995, "step": 16500 }, { "epoch": 0.9949056038357806, "grad_norm": 0.03102271445095539, "learning_rate": 5.583857756467885e-06, "loss": 0.0999, "step": 16600 }, { "epoch": 1.0, "eval_accuracy": 0.6769802089043807, "eval_f1": 0.6622248214359344, "eval_loss": 0.0996900200843811, "eval_precision": 0.6552669320986634, "eval_recall": 0.6769802089043807, "eval_runtime": 539.1847, "eval_samples_per_second": 27.506, "eval_steps_per_second": 3.439, "step": 16685 }, { "epoch": 1.0008990110878033, "grad_norm": 0.4820581376552582, "learning_rate": 5.5505610495122045e-06, "loss": 0.0985, "step": 16700 }, { "epoch": 1.0068924183398262, "grad_norm": 0.35515159368515015, "learning_rate": 5.517264342556522e-06, "loss": 0.0984, "step": 16800 }, { "epoch": 1.012885825591849, "grad_norm": 0.043349362909793854, "learning_rate": 5.483967635600839e-06, "loss": 0.0986, "step": 16900 }, { "epoch": 1.0188792328438718, "grad_norm": 0.03270228952169418, "learning_rate": 5.450670928645158e-06, "loss": 0.099, "step": 17000 }, { "epoch": 1.0248726400958945, "grad_norm": 0.027137137949466705, "learning_rate": 5.417374221689475e-06, "loss": 0.099, "step": 17100 }, { "epoch": 1.0308660473479172, "grad_norm": 0.016054777428507805, "learning_rate": 5.384077514733793e-06, "loss": 0.0985, "step": 17200 }, { "epoch": 1.0368594545999401, "grad_norm": 0.14413999021053314, "learning_rate": 5.350780807778111e-06, "loss": 0.0988, "step": 17300 }, { "epoch": 1.0428528618519628, "grad_norm": 0.12687167525291443, "learning_rate": 5.317484100822429e-06, "loss": 0.0983, "step": 17400 }, { "epoch": 1.0488462691039857, "grad_norm": 0.04264242947101593, "learning_rate": 5.2841873938667464e-06, "loss": 0.0971, "step": 17500 }, { "epoch": 1.0548396763560084, "grad_norm": 0.31686630845069885, "learning_rate": 5.2508906869110655e-06, "loss": 0.0998, "step": 17600 }, { "epoch": 1.060833083608031, "grad_norm": 0.01217142678797245, "learning_rate": 5.217593979955383e-06, "loss": 0.099, "step": 17700 }, { "epoch": 1.066826490860054, "grad_norm": 0.020899010822176933, "learning_rate": 5.184297272999701e-06, "loss": 0.0998, "step": 17800 }, { "epoch": 1.0728198981120767, "grad_norm": 0.06026081368327141, "learning_rate": 5.151000566044019e-06, "loss": 0.0981, "step": 17900 }, { "epoch": 1.0788133053640996, "grad_norm": 0.040572892874479294, "learning_rate": 5.117703859088337e-06, "loss": 0.0982, "step": 18000 }, { "epoch": 1.0848067126161223, "grad_norm": 0.018748696893453598, "learning_rate": 5.084407152132654e-06, "loss": 0.0989, "step": 18100 }, { "epoch": 1.090800119868145, "grad_norm": 0.16863317787647247, "learning_rate": 5.051110445176972e-06, "loss": 0.0984, "step": 18200 }, { "epoch": 1.0967935271201679, "grad_norm": 0.238959401845932, "learning_rate": 5.01781373822129e-06, "loss": 0.0977, "step": 18300 }, { "epoch": 1.1027869343721906, "grad_norm": 0.05871168524026871, "learning_rate": 4.984517031265608e-06, "loss": 0.0981, "step": 18400 }, { "epoch": 1.1087803416242135, "grad_norm": 0.07862433046102524, "learning_rate": 4.951220324309926e-06, "loss": 0.0991, "step": 18500 }, { "epoch": 1.1147737488762361, "grad_norm": 0.033283356577157974, "learning_rate": 4.917923617354244e-06, "loss": 0.0981, "step": 18600 }, { "epoch": 1.1207671561282588, "grad_norm": 0.07558907568454742, "learning_rate": 4.884626910398562e-06, "loss": 0.0986, "step": 18700 }, { "epoch": 1.1267605633802817, "grad_norm": 0.045116860419511795, "learning_rate": 4.85133020344288e-06, "loss": 0.0989, "step": 18800 }, { "epoch": 1.1327539706323044, "grad_norm": 0.2039572298526764, "learning_rate": 4.818033496487198e-06, "loss": 0.099, "step": 18900 }, { "epoch": 1.1387473778843273, "grad_norm": 0.026863813400268555, "learning_rate": 4.784736789531515e-06, "loss": 0.0995, "step": 19000 }, { "epoch": 1.14474078513635, "grad_norm": 0.08295423537492752, "learning_rate": 4.751440082575834e-06, "loss": 0.0993, "step": 19100 }, { "epoch": 1.1507341923883727, "grad_norm": 0.012610992416739464, "learning_rate": 4.718143375620152e-06, "loss": 0.1, "step": 19200 }, { "epoch": 1.1567275996403956, "grad_norm": 0.15242983400821686, "learning_rate": 4.684846668664469e-06, "loss": 0.0993, "step": 19300 }, { "epoch": 1.1627210068924183, "grad_norm": 0.03405880555510521, "learning_rate": 4.6515499617087876e-06, "loss": 0.098, "step": 19400 }, { "epoch": 1.168714414144441, "grad_norm": 0.09746590256690979, "learning_rate": 4.618253254753105e-06, "loss": 0.0994, "step": 19500 }, { "epoch": 1.174707821396464, "grad_norm": 0.01837330497801304, "learning_rate": 4.584956547797423e-06, "loss": 0.0993, "step": 19600 }, { "epoch": 1.1807012286484866, "grad_norm": 0.06496419757604599, "learning_rate": 4.5516598408417415e-06, "loss": 0.0992, "step": 19700 }, { "epoch": 1.1866946359005095, "grad_norm": 0.014315762557089329, "learning_rate": 4.518363133886059e-06, "loss": 0.0988, "step": 19800 }, { "epoch": 1.1926880431525322, "grad_norm": 0.05952730402350426, "learning_rate": 4.485066426930376e-06, "loss": 0.0985, "step": 19900 }, { "epoch": 1.198681450404555, "grad_norm": 0.18996870517730713, "learning_rate": 4.4517697199746955e-06, "loss": 0.0985, "step": 20000 }, { "epoch": 1.2046748576565778, "grad_norm": 0.05301952362060547, "learning_rate": 4.418473013019013e-06, "loss": 0.1, "step": 20100 }, { "epoch": 1.2106682649086005, "grad_norm": 0.15663863718509674, "learning_rate": 4.38517630606333e-06, "loss": 0.0992, "step": 20200 }, { "epoch": 1.2166616721606234, "grad_norm": 0.2653251588344574, "learning_rate": 4.3518795991076486e-06, "loss": 0.0982, "step": 20300 }, { "epoch": 1.222655079412646, "grad_norm": 0.11596585810184479, "learning_rate": 4.318582892151967e-06, "loss": 0.0984, "step": 20400 }, { "epoch": 1.2286484866646687, "grad_norm": 0.12228166311979294, "learning_rate": 4.285286185196284e-06, "loss": 0.0987, "step": 20500 }, { "epoch": 1.2346418939166917, "grad_norm": 0.19990171492099762, "learning_rate": 4.2519894782406025e-06, "loss": 0.0982, "step": 20600 }, { "epoch": 1.2406353011687143, "grad_norm": 0.13501687347888947, "learning_rate": 4.21869277128492e-06, "loss": 0.099, "step": 20700 }, { "epoch": 1.2466287084207373, "grad_norm": 0.14756691455841064, "learning_rate": 4.185396064329238e-06, "loss": 0.0989, "step": 20800 }, { "epoch": 1.25262211567276, "grad_norm": 0.04468319192528725, "learning_rate": 4.1520993573735565e-06, "loss": 0.099, "step": 20900 }, { "epoch": 1.2586155229247828, "grad_norm": 0.026138102635741234, "learning_rate": 4.118802650417874e-06, "loss": 0.0986, "step": 21000 }, { "epoch": 1.2646089301768055, "grad_norm": 0.22098702192306519, "learning_rate": 4.085505943462192e-06, "loss": 0.0981, "step": 21100 }, { "epoch": 1.2706023374288282, "grad_norm": 0.027339546009898186, "learning_rate": 4.05220923650651e-06, "loss": 0.0992, "step": 21200 }, { "epoch": 1.2765957446808511, "grad_norm": 0.01603206805884838, "learning_rate": 4.018912529550828e-06, "loss": 0.0988, "step": 21300 }, { "epoch": 1.2825891519328738, "grad_norm": 0.008753624744713306, "learning_rate": 3.985615822595146e-06, "loss": 0.0984, "step": 21400 }, { "epoch": 1.2885825591848965, "grad_norm": 0.019739218056201935, "learning_rate": 3.9523191156394635e-06, "loss": 0.098, "step": 21500 }, { "epoch": 1.2945759664369194, "grad_norm": 0.8729631304740906, "learning_rate": 3.919022408683781e-06, "loss": 0.0983, "step": 21600 }, { "epoch": 1.300569373688942, "grad_norm": 0.06987379491329193, "learning_rate": 3.885725701728099e-06, "loss": 0.0987, "step": 21700 }, { "epoch": 1.306562780940965, "grad_norm": 0.11344248056411743, "learning_rate": 3.8524289947724175e-06, "loss": 0.0991, "step": 21800 }, { "epoch": 1.3125561881929877, "grad_norm": 0.08928582072257996, "learning_rate": 3.819132287816735e-06, "loss": 0.0974, "step": 21900 }, { "epoch": 1.3185495954450106, "grad_norm": 0.08214009553194046, "learning_rate": 3.785835580861053e-06, "loss": 0.0988, "step": 22000 }, { "epoch": 1.3245430026970333, "grad_norm": 0.026567021384835243, "learning_rate": 3.752538873905371e-06, "loss": 0.0983, "step": 22100 }, { "epoch": 1.330536409949056, "grad_norm": 0.1904483139514923, "learning_rate": 3.7192421669496893e-06, "loss": 0.0984, "step": 22200 }, { "epoch": 1.3365298172010789, "grad_norm": 0.15923745930194855, "learning_rate": 3.685945459994007e-06, "loss": 0.0982, "step": 22300 }, { "epoch": 1.3425232244531016, "grad_norm": 0.1682073026895523, "learning_rate": 3.652648753038325e-06, "loss": 0.0975, "step": 22400 }, { "epoch": 1.3485166317051243, "grad_norm": 0.1236729547381401, "learning_rate": 3.6193520460826424e-06, "loss": 0.0986, "step": 22500 }, { "epoch": 1.3545100389571472, "grad_norm": 0.2880057096481323, "learning_rate": 3.586055339126961e-06, "loss": 0.0991, "step": 22600 }, { "epoch": 1.3605034462091699, "grad_norm": 0.07628654688596725, "learning_rate": 3.5527586321712785e-06, "loss": 0.0993, "step": 22700 }, { "epoch": 1.3664968534611928, "grad_norm": 0.0549270398914814, "learning_rate": 3.5194619252155963e-06, "loss": 0.0983, "step": 22800 }, { "epoch": 1.3724902607132154, "grad_norm": 0.010687282308936119, "learning_rate": 3.486165218259914e-06, "loss": 0.0992, "step": 22900 }, { "epoch": 1.3784836679652384, "grad_norm": 0.32308948040008545, "learning_rate": 3.452868511304232e-06, "loss": 0.0985, "step": 23000 }, { "epoch": 1.384477075217261, "grad_norm": 0.10801893472671509, "learning_rate": 3.4195718043485503e-06, "loss": 0.0981, "step": 23100 }, { "epoch": 1.3904704824692837, "grad_norm": 0.035334184765815735, "learning_rate": 3.386275097392868e-06, "loss": 0.0981, "step": 23200 }, { "epoch": 1.3964638897213066, "grad_norm": 0.3928769826889038, "learning_rate": 3.352978390437186e-06, "loss": 0.0987, "step": 23300 }, { "epoch": 1.4024572969733293, "grad_norm": 0.03784007951617241, "learning_rate": 3.319681683481504e-06, "loss": 0.0983, "step": 23400 }, { "epoch": 1.408450704225352, "grad_norm": 0.16313448548316956, "learning_rate": 3.286384976525822e-06, "loss": 0.0992, "step": 23500 }, { "epoch": 1.414444111477375, "grad_norm": 0.7129063606262207, "learning_rate": 3.25308826957014e-06, "loss": 0.0985, "step": 23600 }, { "epoch": 1.4204375187293976, "grad_norm": 0.08185921609401703, "learning_rate": 3.2197915626144578e-06, "loss": 0.0985, "step": 23700 }, { "epoch": 1.4264309259814205, "grad_norm": 0.07759444415569305, "learning_rate": 3.1864948556587756e-06, "loss": 0.0987, "step": 23800 }, { "epoch": 1.4324243332334432, "grad_norm": 0.15154729783535004, "learning_rate": 3.1531981487030935e-06, "loss": 0.0996, "step": 23900 }, { "epoch": 1.4384177404854661, "grad_norm": 0.10120932757854462, "learning_rate": 3.1199014417474117e-06, "loss": 0.0979, "step": 24000 }, { "epoch": 1.4444111477374888, "grad_norm": 0.042903125286102295, "learning_rate": 3.0866047347917296e-06, "loss": 0.0988, "step": 24100 }, { "epoch": 1.4504045549895115, "grad_norm": 0.05083679407835007, "learning_rate": 3.053308027836047e-06, "loss": 0.0979, "step": 24200 }, { "epoch": 1.4563979622415344, "grad_norm": 0.055504605174064636, "learning_rate": 3.020011320880365e-06, "loss": 0.0974, "step": 24300 }, { "epoch": 1.462391369493557, "grad_norm": 0.10161745548248291, "learning_rate": 2.9867146139246835e-06, "loss": 0.0985, "step": 24400 }, { "epoch": 1.4683847767455798, "grad_norm": 0.17584453523159027, "learning_rate": 2.953417906969001e-06, "loss": 0.0992, "step": 24500 }, { "epoch": 1.4743781839976027, "grad_norm": 0.028774550184607506, "learning_rate": 2.9201212000133188e-06, "loss": 0.0995, "step": 24600 }, { "epoch": 1.4803715912496254, "grad_norm": 0.15757901966571808, "learning_rate": 2.8868244930576366e-06, "loss": 0.0984, "step": 24700 }, { "epoch": 1.486364998501648, "grad_norm": 0.08022894710302353, "learning_rate": 2.853527786101955e-06, "loss": 0.0988, "step": 24800 }, { "epoch": 1.492358405753671, "grad_norm": 0.023040054365992546, "learning_rate": 2.8202310791462727e-06, "loss": 0.0982, "step": 24900 }, { "epoch": 1.4983518130056939, "grad_norm": 0.032170914113521576, "learning_rate": 2.7869343721905906e-06, "loss": 0.0983, "step": 25000 }, { "epoch": 1.5043452202577166, "grad_norm": 0.04139495640993118, "learning_rate": 2.7536376652349084e-06, "loss": 0.0986, "step": 25100 }, { "epoch": 1.5103386275097392, "grad_norm": 0.11350516229867935, "learning_rate": 2.7203409582792263e-06, "loss": 0.0986, "step": 25200 }, { "epoch": 1.5163320347617621, "grad_norm": 0.20275545120239258, "learning_rate": 2.6870442513235445e-06, "loss": 0.0994, "step": 25300 }, { "epoch": 1.5223254420137848, "grad_norm": 0.013372101821005344, "learning_rate": 2.6537475443678624e-06, "loss": 0.0982, "step": 25400 }, { "epoch": 1.5283188492658075, "grad_norm": 0.21790750324726105, "learning_rate": 2.6204508374121802e-06, "loss": 0.098, "step": 25500 }, { "epoch": 1.5343122565178304, "grad_norm": 0.1723981350660324, "learning_rate": 2.587154130456498e-06, "loss": 0.0981, "step": 25600 }, { "epoch": 1.5403056637698531, "grad_norm": 0.14081990718841553, "learning_rate": 2.5538574235008163e-06, "loss": 0.0986, "step": 25700 }, { "epoch": 1.5462990710218758, "grad_norm": 0.1418023258447647, "learning_rate": 2.520560716545134e-06, "loss": 0.0977, "step": 25800 }, { "epoch": 1.5522924782738987, "grad_norm": 0.010309278964996338, "learning_rate": 2.487264009589452e-06, "loss": 0.0982, "step": 25900 }, { "epoch": 1.5582858855259216, "grad_norm": 0.06128934770822525, "learning_rate": 2.4539673026337694e-06, "loss": 0.0973, "step": 26000 }, { "epoch": 1.5642792927779443, "grad_norm": 0.1110377237200737, "learning_rate": 2.4206705956780877e-06, "loss": 0.0992, "step": 26100 }, { "epoch": 1.570272700029967, "grad_norm": 0.2141459882259369, "learning_rate": 2.3873738887224055e-06, "loss": 0.0989, "step": 26200 }, { "epoch": 1.57626610728199, "grad_norm": 0.12720057368278503, "learning_rate": 2.3540771817667234e-06, "loss": 0.0986, "step": 26300 }, { "epoch": 1.5822595145340126, "grad_norm": 0.0766364112496376, "learning_rate": 2.3207804748110412e-06, "loss": 0.0993, "step": 26400 }, { "epoch": 1.5882529217860353, "grad_norm": 0.3568774163722992, "learning_rate": 2.2874837678553595e-06, "loss": 0.0981, "step": 26500 }, { "epoch": 1.5942463290380582, "grad_norm": 0.2806101143360138, "learning_rate": 2.2541870608996773e-06, "loss": 0.0987, "step": 26600 }, { "epoch": 1.6002397362900809, "grad_norm": 0.09340783953666687, "learning_rate": 2.220890353943995e-06, "loss": 0.0982, "step": 26700 }, { "epoch": 1.6062331435421036, "grad_norm": 0.027837200090289116, "learning_rate": 2.187593646988313e-06, "loss": 0.0968, "step": 26800 }, { "epoch": 1.6122265507941265, "grad_norm": 0.12748104333877563, "learning_rate": 2.154296940032631e-06, "loss": 0.0983, "step": 26900 }, { "epoch": 1.6182199580461494, "grad_norm": 0.23494853079319, "learning_rate": 2.1210002330769487e-06, "loss": 0.0984, "step": 27000 }, { "epoch": 1.624213365298172, "grad_norm": 0.18360479176044464, "learning_rate": 2.0877035261212665e-06, "loss": 0.0989, "step": 27100 }, { "epoch": 1.6302067725501947, "grad_norm": 0.08933253586292267, "learning_rate": 2.054406819165585e-06, "loss": 0.0988, "step": 27200 }, { "epoch": 1.6362001798022177, "grad_norm": 0.01740170828998089, "learning_rate": 2.0211101122099027e-06, "loss": 0.0997, "step": 27300 }, { "epoch": 1.6421935870542403, "grad_norm": 0.0467255637049675, "learning_rate": 1.9878134052542205e-06, "loss": 0.0975, "step": 27400 }, { "epoch": 1.648186994306263, "grad_norm": 0.008605794049799442, "learning_rate": 1.9545166982985383e-06, "loss": 0.0989, "step": 27500 }, { "epoch": 1.654180401558286, "grad_norm": 0.14202085137367249, "learning_rate": 1.9212199913428566e-06, "loss": 0.0992, "step": 27600 }, { "epoch": 1.6601738088103086, "grad_norm": 0.09536932408809662, "learning_rate": 1.8879232843871742e-06, "loss": 0.0981, "step": 27700 }, { "epoch": 1.6661672160623313, "grad_norm": 0.09608137607574463, "learning_rate": 1.8546265774314923e-06, "loss": 0.0986, "step": 27800 }, { "epoch": 1.6721606233143542, "grad_norm": 0.022448230534791946, "learning_rate": 1.8213298704758101e-06, "loss": 0.0978, "step": 27900 }, { "epoch": 1.6781540305663771, "grad_norm": 0.018090438097715378, "learning_rate": 1.7880331635201282e-06, "loss": 0.0986, "step": 28000 }, { "epoch": 1.6841474378183996, "grad_norm": 0.016484877094626427, "learning_rate": 1.7547364565644458e-06, "loss": 0.098, "step": 28100 }, { "epoch": 1.6901408450704225, "grad_norm": 0.024596238508820534, "learning_rate": 1.7214397496087637e-06, "loss": 0.0986, "step": 28200 }, { "epoch": 1.6961342523224454, "grad_norm": 0.06954281032085419, "learning_rate": 1.6881430426530817e-06, "loss": 0.097, "step": 28300 }, { "epoch": 1.702127659574468, "grad_norm": 0.31834906339645386, "learning_rate": 1.6548463356973996e-06, "loss": 0.0976, "step": 28400 }, { "epoch": 1.7081210668264908, "grad_norm": 0.1658525913953781, "learning_rate": 1.6215496287417176e-06, "loss": 0.0974, "step": 28500 }, { "epoch": 1.7141144740785137, "grad_norm": 0.030118102207779884, "learning_rate": 1.5882529217860355e-06, "loss": 0.0997, "step": 28600 }, { "epoch": 1.7201078813305364, "grad_norm": 0.16278527677059174, "learning_rate": 1.5549562148303535e-06, "loss": 0.098, "step": 28700 }, { "epoch": 1.726101288582559, "grad_norm": 0.024548674002289772, "learning_rate": 1.5216595078746714e-06, "loss": 0.0986, "step": 28800 }, { "epoch": 1.732094695834582, "grad_norm": 0.40551477670669556, "learning_rate": 1.4883628009189894e-06, "loss": 0.0972, "step": 28900 }, { "epoch": 1.7380881030866049, "grad_norm": 0.05455062910914421, "learning_rate": 1.455066093963307e-06, "loss": 0.0976, "step": 29000 }, { "epoch": 1.7440815103386273, "grad_norm": 0.455079585313797, "learning_rate": 1.421769387007625e-06, "loss": 0.0995, "step": 29100 }, { "epoch": 1.7500749175906503, "grad_norm": 0.6706792116165161, "learning_rate": 1.388472680051943e-06, "loss": 0.0979, "step": 29200 }, { "epoch": 1.7560683248426732, "grad_norm": 0.05078733339905739, "learning_rate": 1.3551759730962608e-06, "loss": 0.097, "step": 29300 }, { "epoch": 1.7620617320946959, "grad_norm": 0.25584468245506287, "learning_rate": 1.3218792661405788e-06, "loss": 0.0981, "step": 29400 }, { "epoch": 1.7680551393467185, "grad_norm": 0.12247732281684875, "learning_rate": 1.2885825591848967e-06, "loss": 0.0986, "step": 29500 }, { "epoch": 1.7740485465987414, "grad_norm": 0.016069240868091583, "learning_rate": 1.2552858522292147e-06, "loss": 0.0987, "step": 29600 }, { "epoch": 1.7800419538507641, "grad_norm": 0.1197570264339447, "learning_rate": 1.2219891452735326e-06, "loss": 0.0981, "step": 29700 }, { "epoch": 1.7860353611027868, "grad_norm": 0.035120606422424316, "learning_rate": 1.1886924383178504e-06, "loss": 0.099, "step": 29800 }, { "epoch": 1.7920287683548097, "grad_norm": 0.1425538957118988, "learning_rate": 1.1553957313621683e-06, "loss": 0.0989, "step": 29900 }, { "epoch": 1.7980221756068326, "grad_norm": 0.027721386402845383, "learning_rate": 1.1220990244064863e-06, "loss": 0.0989, "step": 30000 } ], "logging_steps": 100, "max_steps": 33370, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 3.663876369096977e+17, "train_batch_size": 8, "trial_name": null, "trial_params": null }