diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,81893 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.8000547233052876, + "eval_steps": 500, + "global_step": 11696, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 6.840413160954922e-05, + "grad_norm": 19.005646318396554, + "learning_rate": 2.2779043280182232e-08, + "loss": 1.2551, + "step": 1 + }, + { + "epoch": 0.00013680826321909844, + "grad_norm": 19.825164309346295, + "learning_rate": 4.5558086560364464e-08, + "loss": 0.8688, + "step": 2 + }, + { + "epoch": 0.00020521239482864764, + "grad_norm": 28.98854364816896, + "learning_rate": 6.83371298405467e-08, + "loss": 1.2541, + "step": 3 + }, + { + "epoch": 0.00027361652643819687, + "grad_norm": 12.726041910800626, + "learning_rate": 9.111617312072893e-08, + "loss": 1.3166, + "step": 4 + }, + { + "epoch": 0.0003420206580477461, + "grad_norm": 32.47354504715627, + "learning_rate": 1.1389521640091118e-07, + "loss": 1.0256, + "step": 5 + }, + { + "epoch": 0.0004104247896572953, + "grad_norm": 26.14034306202561, + "learning_rate": 1.366742596810934e-07, + "loss": 1.5129, + "step": 6 + }, + { + "epoch": 0.00047882892126684454, + "grad_norm": 22.59197644966778, + "learning_rate": 1.5945330296127564e-07, + "loss": 1.2663, + "step": 7 + }, + { + "epoch": 0.0005472330528763937, + "grad_norm": 25.5670147105785, + "learning_rate": 1.8223234624145786e-07, + "loss": 1.4021, + "step": 8 + }, + { + "epoch": 0.000615637184485943, + "grad_norm": 19.27808738750586, + "learning_rate": 2.0501138952164012e-07, + "loss": 0.9687, + "step": 9 + }, + { + "epoch": 0.0006840413160954922, + "grad_norm": 11.116110861484534, + "learning_rate": 2.2779043280182236e-07, + "loss": 1.166, + "step": 10 + }, + { + "epoch": 0.0007524454477050414, + "grad_norm": 20.307456555365246, + "learning_rate": 2.5056947608200455e-07, + "loss": 1.1607, + "step": 11 + }, + { + "epoch": 0.0008208495793145906, + "grad_norm": 12.058120090485795, + "learning_rate": 2.733485193621868e-07, + "loss": 1.1394, + "step": 12 + }, + { + "epoch": 0.0008892537109241398, + "grad_norm": 29.53469411959909, + "learning_rate": 2.961275626423691e-07, + "loss": 0.9815, + "step": 13 + }, + { + "epoch": 0.0009576578425336891, + "grad_norm": 22.863198826841863, + "learning_rate": 3.189066059225513e-07, + "loss": 1.0778, + "step": 14 + }, + { + "epoch": 0.0010260619741432383, + "grad_norm": 17.711071469911264, + "learning_rate": 3.416856492027335e-07, + "loss": 1.2028, + "step": 15 + }, + { + "epoch": 0.0010944661057527875, + "grad_norm": 23.553529524293335, + "learning_rate": 3.644646924829157e-07, + "loss": 0.9792, + "step": 16 + }, + { + "epoch": 0.0011628702373623367, + "grad_norm": 14.014274948913696, + "learning_rate": 3.8724373576309803e-07, + "loss": 1.1274, + "step": 17 + }, + { + "epoch": 0.001231274368971886, + "grad_norm": 15.060242796380724, + "learning_rate": 4.1002277904328024e-07, + "loss": 1.1117, + "step": 18 + }, + { + "epoch": 0.001299678500581435, + "grad_norm": 18.10445634136811, + "learning_rate": 4.3280182232346246e-07, + "loss": 0.8063, + "step": 19 + }, + { + "epoch": 0.0013680826321909843, + "grad_norm": 19.88694931792778, + "learning_rate": 4.555808656036447e-07, + "loss": 1.2076, + "step": 20 + }, + { + "epoch": 0.0014364867638005335, + "grad_norm": 31.709343418336065, + "learning_rate": 4.783599088838269e-07, + "loss": 1.1243, + "step": 21 + }, + { + "epoch": 0.0015048908954100827, + "grad_norm": 14.322136912072493, + "learning_rate": 5.011389521640091e-07, + "loss": 1.0473, + "step": 22 + }, + { + "epoch": 0.001573295027019632, + "grad_norm": 15.407758596349403, + "learning_rate": 5.239179954441914e-07, + "loss": 1.1462, + "step": 23 + }, + { + "epoch": 0.0016416991586291811, + "grad_norm": 17.653636853777154, + "learning_rate": 5.466970387243736e-07, + "loss": 0.7997, + "step": 24 + }, + { + "epoch": 0.0017101032902387303, + "grad_norm": 12.819934604614861, + "learning_rate": 5.694760820045558e-07, + "loss": 0.7004, + "step": 25 + }, + { + "epoch": 0.0017785074218482795, + "grad_norm": 16.404739513799814, + "learning_rate": 5.922551252847382e-07, + "loss": 1.1528, + "step": 26 + }, + { + "epoch": 0.001846911553457829, + "grad_norm": 20.489541339944363, + "learning_rate": 6.150341685649204e-07, + "loss": 0.9264, + "step": 27 + }, + { + "epoch": 0.0019153156850673782, + "grad_norm": 15.10654212249387, + "learning_rate": 6.378132118451026e-07, + "loss": 1.1305, + "step": 28 + }, + { + "epoch": 0.001983719816676927, + "grad_norm": 11.814154776465195, + "learning_rate": 6.605922551252848e-07, + "loss": 0.9956, + "step": 29 + }, + { + "epoch": 0.0020521239482864766, + "grad_norm": 13.07191952655166, + "learning_rate": 6.83371298405467e-07, + "loss": 0.6871, + "step": 30 + }, + { + "epoch": 0.0021205280798960256, + "grad_norm": 10.183949325669186, + "learning_rate": 7.061503416856492e-07, + "loss": 1.1098, + "step": 31 + }, + { + "epoch": 0.002188932211505575, + "grad_norm": 10.159981004141525, + "learning_rate": 7.289293849658314e-07, + "loss": 0.7531, + "step": 32 + }, + { + "epoch": 0.002257336343115124, + "grad_norm": 11.160497033320771, + "learning_rate": 7.517084282460136e-07, + "loss": 0.7946, + "step": 33 + }, + { + "epoch": 0.0023257404747246734, + "grad_norm": 11.595618249120902, + "learning_rate": 7.744874715261961e-07, + "loss": 0.9285, + "step": 34 + }, + { + "epoch": 0.0023941446063342224, + "grad_norm": 14.112149224597388, + "learning_rate": 7.972665148063783e-07, + "loss": 0.6524, + "step": 35 + }, + { + "epoch": 0.002462548737943772, + "grad_norm": 9.682423448956861, + "learning_rate": 8.200455580865605e-07, + "loss": 0.7987, + "step": 36 + }, + { + "epoch": 0.002530952869553321, + "grad_norm": 8.09964246078515, + "learning_rate": 8.428246013667427e-07, + "loss": 1.0615, + "step": 37 + }, + { + "epoch": 0.00259935700116287, + "grad_norm": 10.817647425345854, + "learning_rate": 8.656036446469249e-07, + "loss": 0.9523, + "step": 38 + }, + { + "epoch": 0.0026677611327724196, + "grad_norm": 8.998123556132862, + "learning_rate": 8.883826879271071e-07, + "loss": 0.6709, + "step": 39 + }, + { + "epoch": 0.0027361652643819686, + "grad_norm": 6.850653523922662, + "learning_rate": 9.111617312072894e-07, + "loss": 1.0962, + "step": 40 + }, + { + "epoch": 0.002804569395991518, + "grad_norm": 12.190153566665263, + "learning_rate": 9.339407744874717e-07, + "loss": 0.9311, + "step": 41 + }, + { + "epoch": 0.002872973527601067, + "grad_norm": 6.213900932761143, + "learning_rate": 9.567198177676538e-07, + "loss": 0.5645, + "step": 42 + }, + { + "epoch": 0.0029413776592106164, + "grad_norm": 7.108373282149988, + "learning_rate": 9.79498861047836e-07, + "loss": 0.9261, + "step": 43 + }, + { + "epoch": 0.0030097817908201654, + "grad_norm": 10.518919022143347, + "learning_rate": 1.0022779043280182e-06, + "loss": 0.5782, + "step": 44 + }, + { + "epoch": 0.003078185922429715, + "grad_norm": 7.764676932291131, + "learning_rate": 1.0250569476082005e-06, + "loss": 0.6455, + "step": 45 + }, + { + "epoch": 0.003146590054039264, + "grad_norm": 6.521873460229345, + "learning_rate": 1.0478359908883828e-06, + "loss": 0.3617, + "step": 46 + }, + { + "epoch": 0.0032149941856488133, + "grad_norm": 18.063747653307693, + "learning_rate": 1.0706150341685651e-06, + "loss": 0.6493, + "step": 47 + }, + { + "epoch": 0.0032833983172583622, + "grad_norm": 9.220471237618497, + "learning_rate": 1.0933940774487472e-06, + "loss": 0.9514, + "step": 48 + }, + { + "epoch": 0.0033518024488679117, + "grad_norm": 7.445975018992441, + "learning_rate": 1.1161731207289296e-06, + "loss": 0.5799, + "step": 49 + }, + { + "epoch": 0.0034202065804774607, + "grad_norm": 7.466263840089492, + "learning_rate": 1.1389521640091117e-06, + "loss": 1.4957, + "step": 50 + }, + { + "epoch": 0.00348861071208701, + "grad_norm": 7.85539824512469, + "learning_rate": 1.161731207289294e-06, + "loss": 0.6961, + "step": 51 + }, + { + "epoch": 0.003557014843696559, + "grad_norm": 7.915840333343226, + "learning_rate": 1.1845102505694763e-06, + "loss": 0.9127, + "step": 52 + }, + { + "epoch": 0.0036254189753061085, + "grad_norm": 7.225193236251746, + "learning_rate": 1.2072892938496584e-06, + "loss": 0.6309, + "step": 53 + }, + { + "epoch": 0.003693823106915658, + "grad_norm": 8.294950416467335, + "learning_rate": 1.2300683371298407e-06, + "loss": 0.5625, + "step": 54 + }, + { + "epoch": 0.003762227238525207, + "grad_norm": 6.585241188174048, + "learning_rate": 1.252847380410023e-06, + "loss": 0.5428, + "step": 55 + }, + { + "epoch": 0.0038306313701347563, + "grad_norm": 6.969663654186199, + "learning_rate": 1.2756264236902052e-06, + "loss": 0.5, + "step": 56 + }, + { + "epoch": 0.0038990355017443053, + "grad_norm": 5.529118056995139, + "learning_rate": 1.2984054669703875e-06, + "loss": 0.3135, + "step": 57 + }, + { + "epoch": 0.003967439633353854, + "grad_norm": 7.12123855217604, + "learning_rate": 1.3211845102505696e-06, + "loss": 0.6864, + "step": 58 + }, + { + "epoch": 0.004035843764963404, + "grad_norm": 7.080102741067122, + "learning_rate": 1.343963553530752e-06, + "loss": 0.3519, + "step": 59 + }, + { + "epoch": 0.004104247896572953, + "grad_norm": 9.304129044871978, + "learning_rate": 1.366742596810934e-06, + "loss": 0.9068, + "step": 60 + }, + { + "epoch": 0.004172652028182502, + "grad_norm": 8.629972158057784, + "learning_rate": 1.3895216400911163e-06, + "loss": 0.6226, + "step": 61 + }, + { + "epoch": 0.004241056159792051, + "grad_norm": 6.120841151942936, + "learning_rate": 1.4123006833712984e-06, + "loss": 0.9269, + "step": 62 + }, + { + "epoch": 0.004309460291401601, + "grad_norm": 5.75185766480899, + "learning_rate": 1.4350797266514807e-06, + "loss": 0.3918, + "step": 63 + }, + { + "epoch": 0.00437786442301115, + "grad_norm": 5.178097442019518, + "learning_rate": 1.4578587699316629e-06, + "loss": 0.1852, + "step": 64 + }, + { + "epoch": 0.004446268554620699, + "grad_norm": 7.078527881196362, + "learning_rate": 1.4806378132118452e-06, + "loss": 0.3746, + "step": 65 + }, + { + "epoch": 0.004514672686230248, + "grad_norm": 6.076177486072526, + "learning_rate": 1.5034168564920273e-06, + "loss": 0.4731, + "step": 66 + }, + { + "epoch": 0.004583076817839798, + "grad_norm": 5.309470109272941, + "learning_rate": 1.5261958997722096e-06, + "loss": 0.3059, + "step": 67 + }, + { + "epoch": 0.004651480949449347, + "grad_norm": 11.840534750100392, + "learning_rate": 1.5489749430523921e-06, + "loss": 0.6966, + "step": 68 + }, + { + "epoch": 0.004719885081058896, + "grad_norm": 10.222125041426017, + "learning_rate": 1.5717539863325742e-06, + "loss": 0.55, + "step": 69 + }, + { + "epoch": 0.004788289212668445, + "grad_norm": 7.131042260797559, + "learning_rate": 1.5945330296127566e-06, + "loss": 0.8744, + "step": 70 + }, + { + "epoch": 0.004856693344277995, + "grad_norm": 6.28439698217826, + "learning_rate": 1.6173120728929387e-06, + "loss": 0.3941, + "step": 71 + }, + { + "epoch": 0.004925097475887544, + "grad_norm": 7.881210723948495, + "learning_rate": 1.640091116173121e-06, + "loss": 0.4078, + "step": 72 + }, + { + "epoch": 0.004993501607497093, + "grad_norm": 8.96401475052852, + "learning_rate": 1.662870159453303e-06, + "loss": 0.4546, + "step": 73 + }, + { + "epoch": 0.005061905739106642, + "grad_norm": 5.662264076690713, + "learning_rate": 1.6856492027334854e-06, + "loss": 0.508, + "step": 74 + }, + { + "epoch": 0.005130309870716191, + "grad_norm": 5.650398726445972, + "learning_rate": 1.7084282460136675e-06, + "loss": 0.565, + "step": 75 + }, + { + "epoch": 0.00519871400232574, + "grad_norm": 5.064481983036394, + "learning_rate": 1.7312072892938498e-06, + "loss": 0.4604, + "step": 76 + }, + { + "epoch": 0.005267118133935289, + "grad_norm": 5.656431500805364, + "learning_rate": 1.753986332574032e-06, + "loss": 0.6448, + "step": 77 + }, + { + "epoch": 0.005335522265544839, + "grad_norm": 6.727741346392739, + "learning_rate": 1.7767653758542143e-06, + "loss": 0.5996, + "step": 78 + }, + { + "epoch": 0.005403926397154388, + "grad_norm": 7.0610981115319404, + "learning_rate": 1.7995444191343964e-06, + "loss": 0.5485, + "step": 79 + }, + { + "epoch": 0.005472330528763937, + "grad_norm": 10.515029300861103, + "learning_rate": 1.8223234624145789e-06, + "loss": 0.3882, + "step": 80 + }, + { + "epoch": 0.005540734660373486, + "grad_norm": 6.913243614884951, + "learning_rate": 1.845102505694761e-06, + "loss": 0.7728, + "step": 81 + }, + { + "epoch": 0.005609138791983036, + "grad_norm": 7.44139695855309, + "learning_rate": 1.8678815489749433e-06, + "loss": 0.7166, + "step": 82 + }, + { + "epoch": 0.005677542923592585, + "grad_norm": 6.013853689222373, + "learning_rate": 1.8906605922551254e-06, + "loss": 0.5366, + "step": 83 + }, + { + "epoch": 0.005745947055202134, + "grad_norm": 7.164864581519447, + "learning_rate": 1.9134396355353075e-06, + "loss": 0.3289, + "step": 84 + }, + { + "epoch": 0.005814351186811683, + "grad_norm": 4.504062778899276, + "learning_rate": 1.93621867881549e-06, + "loss": 0.6583, + "step": 85 + }, + { + "epoch": 0.005882755318421233, + "grad_norm": 6.336368362696905, + "learning_rate": 1.958997722095672e-06, + "loss": 0.4917, + "step": 86 + }, + { + "epoch": 0.005951159450030782, + "grad_norm": 6.941150117440937, + "learning_rate": 1.9817767653758545e-06, + "loss": 0.3338, + "step": 87 + }, + { + "epoch": 0.006019563581640331, + "grad_norm": 5.375141053234146, + "learning_rate": 2.0045558086560364e-06, + "loss": 0.543, + "step": 88 + }, + { + "epoch": 0.006087967713249881, + "grad_norm": 4.387718813621196, + "learning_rate": 2.0273348519362187e-06, + "loss": 0.4088, + "step": 89 + }, + { + "epoch": 0.00615637184485943, + "grad_norm": 7.167311169971348, + "learning_rate": 2.050113895216401e-06, + "loss": 0.4818, + "step": 90 + }, + { + "epoch": 0.006224775976468979, + "grad_norm": 6.185913691772371, + "learning_rate": 2.0728929384965833e-06, + "loss": 0.4268, + "step": 91 + }, + { + "epoch": 0.006293180108078528, + "grad_norm": 5.868832008511959, + "learning_rate": 2.0956719817767656e-06, + "loss": 0.9069, + "step": 92 + }, + { + "epoch": 0.0063615842396880775, + "grad_norm": 6.800985421978269, + "learning_rate": 2.118451025056948e-06, + "loss": 0.5596, + "step": 93 + }, + { + "epoch": 0.0064299883712976265, + "grad_norm": 5.190847350978537, + "learning_rate": 2.1412300683371303e-06, + "loss": 0.754, + "step": 94 + }, + { + "epoch": 0.0064983925029071755, + "grad_norm": 5.612694886865947, + "learning_rate": 2.164009111617312e-06, + "loss": 0.2145, + "step": 95 + }, + { + "epoch": 0.0065667966345167245, + "grad_norm": 6.941990850815697, + "learning_rate": 2.1867881548974945e-06, + "loss": 0.3697, + "step": 96 + }, + { + "epoch": 0.006635200766126274, + "grad_norm": 7.2841488548206215, + "learning_rate": 2.209567198177677e-06, + "loss": 0.4138, + "step": 97 + }, + { + "epoch": 0.006703604897735823, + "grad_norm": 4.951824944216801, + "learning_rate": 2.232346241457859e-06, + "loss": 0.2513, + "step": 98 + }, + { + "epoch": 0.006772009029345372, + "grad_norm": 8.000708965835068, + "learning_rate": 2.255125284738041e-06, + "loss": 0.6468, + "step": 99 + }, + { + "epoch": 0.006840413160954921, + "grad_norm": 5.7999871220939, + "learning_rate": 2.2779043280182233e-06, + "loss": 0.3713, + "step": 100 + }, + { + "epoch": 0.006908817292564471, + "grad_norm": 6.147647385197564, + "learning_rate": 2.3006833712984057e-06, + "loss": 0.2845, + "step": 101 + }, + { + "epoch": 0.00697722142417402, + "grad_norm": 5.608769631195229, + "learning_rate": 2.323462414578588e-06, + "loss": 0.317, + "step": 102 + }, + { + "epoch": 0.007045625555783569, + "grad_norm": 6.103642928359892, + "learning_rate": 2.34624145785877e-06, + "loss": 0.4528, + "step": 103 + }, + { + "epoch": 0.007114029687393118, + "grad_norm": 5.298617726137569, + "learning_rate": 2.3690205011389526e-06, + "loss": 0.4712, + "step": 104 + }, + { + "epoch": 0.007182433819002668, + "grad_norm": 4.204705887622809, + "learning_rate": 2.3917995444191345e-06, + "loss": 0.2891, + "step": 105 + }, + { + "epoch": 0.007250837950612217, + "grad_norm": 4.968420341640101, + "learning_rate": 2.414578587699317e-06, + "loss": 0.3758, + "step": 106 + }, + { + "epoch": 0.007319242082221766, + "grad_norm": 5.563778976676936, + "learning_rate": 2.437357630979499e-06, + "loss": 0.5295, + "step": 107 + }, + { + "epoch": 0.007387646213831316, + "grad_norm": 7.90042876999444, + "learning_rate": 2.4601366742596815e-06, + "loss": 0.4192, + "step": 108 + }, + { + "epoch": 0.007456050345440865, + "grad_norm": 4.968652688279104, + "learning_rate": 2.4829157175398634e-06, + "loss": 0.2256, + "step": 109 + }, + { + "epoch": 0.007524454477050414, + "grad_norm": 4.555290300056485, + "learning_rate": 2.505694760820046e-06, + "loss": 0.2949, + "step": 110 + }, + { + "epoch": 0.007592858608659963, + "grad_norm": 8.712030795473783, + "learning_rate": 2.5284738041002284e-06, + "loss": 0.212, + "step": 111 + }, + { + "epoch": 0.007661262740269513, + "grad_norm": 6.010760612056703, + "learning_rate": 2.5512528473804103e-06, + "loss": 0.5523, + "step": 112 + }, + { + "epoch": 0.007729666871879062, + "grad_norm": 5.3342068324429, + "learning_rate": 2.5740318906605926e-06, + "loss": 0.6505, + "step": 113 + }, + { + "epoch": 0.007798071003488611, + "grad_norm": 4.600779818176575, + "learning_rate": 2.596810933940775e-06, + "loss": 0.3948, + "step": 114 + }, + { + "epoch": 0.00786647513509816, + "grad_norm": 5.298359688154018, + "learning_rate": 2.6195899772209573e-06, + "loss": 0.5686, + "step": 115 + }, + { + "epoch": 0.007934879266707709, + "grad_norm": 6.036839538684912, + "learning_rate": 2.642369020501139e-06, + "loss": 0.5582, + "step": 116 + }, + { + "epoch": 0.008003283398317258, + "grad_norm": 4.140851917886642, + "learning_rate": 2.6651480637813215e-06, + "loss": 0.6109, + "step": 117 + }, + { + "epoch": 0.008071687529926808, + "grad_norm": 5.633881593110654, + "learning_rate": 2.687927107061504e-06, + "loss": 0.6218, + "step": 118 + }, + { + "epoch": 0.008140091661536357, + "grad_norm": 4.289564746015437, + "learning_rate": 2.710706150341686e-06, + "loss": 0.3407, + "step": 119 + }, + { + "epoch": 0.008208495793145906, + "grad_norm": 5.727674930453696, + "learning_rate": 2.733485193621868e-06, + "loss": 0.5627, + "step": 120 + }, + { + "epoch": 0.008276899924755455, + "grad_norm": 6.303750328927804, + "learning_rate": 2.7562642369020503e-06, + "loss": 0.5755, + "step": 121 + }, + { + "epoch": 0.008345304056365004, + "grad_norm": 4.919276783907435, + "learning_rate": 2.7790432801822326e-06, + "loss": 0.7892, + "step": 122 + }, + { + "epoch": 0.008413708187974553, + "grad_norm": 5.745666819984284, + "learning_rate": 2.801822323462415e-06, + "loss": 0.4816, + "step": 123 + }, + { + "epoch": 0.008482112319584102, + "grad_norm": 3.3021973049739892, + "learning_rate": 2.824601366742597e-06, + "loss": 0.2087, + "step": 124 + }, + { + "epoch": 0.008550516451193653, + "grad_norm": 3.583496027217604, + "learning_rate": 2.847380410022779e-06, + "loss": 0.3031, + "step": 125 + }, + { + "epoch": 0.008618920582803202, + "grad_norm": 5.084549815637344, + "learning_rate": 2.8701594533029615e-06, + "loss": 0.4115, + "step": 126 + }, + { + "epoch": 0.008687324714412751, + "grad_norm": 5.169411939973952, + "learning_rate": 2.892938496583144e-06, + "loss": 0.73, + "step": 127 + }, + { + "epoch": 0.0087557288460223, + "grad_norm": 5.500460475372184, + "learning_rate": 2.9157175398633257e-06, + "loss": 0.7081, + "step": 128 + }, + { + "epoch": 0.008824132977631849, + "grad_norm": 4.785514338582568, + "learning_rate": 2.938496583143508e-06, + "loss": 0.293, + "step": 129 + }, + { + "epoch": 0.008892537109241398, + "grad_norm": 4.006016780712426, + "learning_rate": 2.9612756264236903e-06, + "loss": 0.3647, + "step": 130 + }, + { + "epoch": 0.008960941240850947, + "grad_norm": 4.057031560288478, + "learning_rate": 2.9840546697038727e-06, + "loss": 0.4324, + "step": 131 + }, + { + "epoch": 0.009029345372460496, + "grad_norm": 4.238679577088611, + "learning_rate": 3.0068337129840546e-06, + "loss": 0.5028, + "step": 132 + }, + { + "epoch": 0.009097749504070047, + "grad_norm": 8.267964080806344, + "learning_rate": 3.029612756264237e-06, + "loss": 0.742, + "step": 133 + }, + { + "epoch": 0.009166153635679596, + "grad_norm": 3.9999206356281434, + "learning_rate": 3.052391799544419e-06, + "loss": 0.5112, + "step": 134 + }, + { + "epoch": 0.009234557767289145, + "grad_norm": 5.47105228463376, + "learning_rate": 3.075170842824602e-06, + "loss": 0.5871, + "step": 135 + }, + { + "epoch": 0.009302961898898694, + "grad_norm": 4.607785859391537, + "learning_rate": 3.0979498861047843e-06, + "loss": 0.2264, + "step": 136 + }, + { + "epoch": 0.009371366030508243, + "grad_norm": 4.22974469160654, + "learning_rate": 3.120728929384966e-06, + "loss": 0.4083, + "step": 137 + }, + { + "epoch": 0.009439770162117792, + "grad_norm": 3.9994916890941803, + "learning_rate": 3.1435079726651485e-06, + "loss": 0.3025, + "step": 138 + }, + { + "epoch": 0.00950817429372734, + "grad_norm": 4.589387496023141, + "learning_rate": 3.1662870159453308e-06, + "loss": 0.4453, + "step": 139 + }, + { + "epoch": 0.00957657842533689, + "grad_norm": 3.9600038349730293, + "learning_rate": 3.189066059225513e-06, + "loss": 0.3212, + "step": 140 + }, + { + "epoch": 0.00964498255694644, + "grad_norm": 4.444664316235586, + "learning_rate": 3.211845102505695e-06, + "loss": 0.4255, + "step": 141 + }, + { + "epoch": 0.00971338668855599, + "grad_norm": 7.234159709454101, + "learning_rate": 3.2346241457858773e-06, + "loss": 0.6359, + "step": 142 + }, + { + "epoch": 0.009781790820165538, + "grad_norm": 3.9100144748346644, + "learning_rate": 3.2574031890660596e-06, + "loss": 0.3349, + "step": 143 + }, + { + "epoch": 0.009850194951775087, + "grad_norm": 5.202414409697801, + "learning_rate": 3.280182232346242e-06, + "loss": 0.2867, + "step": 144 + }, + { + "epoch": 0.009918599083384636, + "grad_norm": 3.6307094516345657, + "learning_rate": 3.302961275626424e-06, + "loss": 0.3751, + "step": 145 + }, + { + "epoch": 0.009987003214994185, + "grad_norm": 3.9965867589550914, + "learning_rate": 3.325740318906606e-06, + "loss": 0.4858, + "step": 146 + }, + { + "epoch": 0.010055407346603734, + "grad_norm": 2.951678784787292, + "learning_rate": 3.3485193621867885e-06, + "loss": 0.2077, + "step": 147 + }, + { + "epoch": 0.010123811478213285, + "grad_norm": 3.4251001218097823, + "learning_rate": 3.371298405466971e-06, + "loss": 0.2961, + "step": 148 + }, + { + "epoch": 0.010192215609822834, + "grad_norm": 3.287778735219955, + "learning_rate": 3.3940774487471527e-06, + "loss": 0.3616, + "step": 149 + }, + { + "epoch": 0.010260619741432383, + "grad_norm": 10.991617129669507, + "learning_rate": 3.416856492027335e-06, + "loss": 0.3236, + "step": 150 + }, + { + "epoch": 0.010329023873041932, + "grad_norm": 4.033355155831931, + "learning_rate": 3.4396355353075173e-06, + "loss": 0.4022, + "step": 151 + }, + { + "epoch": 0.01039742800465148, + "grad_norm": 4.3844776220404205, + "learning_rate": 3.4624145785876997e-06, + "loss": 0.611, + "step": 152 + }, + { + "epoch": 0.01046583213626103, + "grad_norm": 14.058672778349182, + "learning_rate": 3.4851936218678815e-06, + "loss": 0.5051, + "step": 153 + }, + { + "epoch": 0.010534236267870579, + "grad_norm": 7.80365163594154, + "learning_rate": 3.507972665148064e-06, + "loss": 0.3451, + "step": 154 + }, + { + "epoch": 0.010602640399480128, + "grad_norm": 4.125150331012009, + "learning_rate": 3.530751708428246e-06, + "loss": 0.4217, + "step": 155 + }, + { + "epoch": 0.010671044531089679, + "grad_norm": 5.039843855020962, + "learning_rate": 3.5535307517084285e-06, + "loss": 0.2837, + "step": 156 + }, + { + "epoch": 0.010739448662699227, + "grad_norm": 3.918100501267, + "learning_rate": 3.5763097949886104e-06, + "loss": 0.3118, + "step": 157 + }, + { + "epoch": 0.010807852794308776, + "grad_norm": 4.056138324816646, + "learning_rate": 3.5990888382687927e-06, + "loss": 0.4352, + "step": 158 + }, + { + "epoch": 0.010876256925918325, + "grad_norm": 4.966299904438864, + "learning_rate": 3.6218678815489755e-06, + "loss": 0.7583, + "step": 159 + }, + { + "epoch": 0.010944661057527874, + "grad_norm": 4.442087595940305, + "learning_rate": 3.6446469248291578e-06, + "loss": 0.4503, + "step": 160 + }, + { + "epoch": 0.011013065189137423, + "grad_norm": 7.13931382576582, + "learning_rate": 3.66742596810934e-06, + "loss": 0.3963, + "step": 161 + }, + { + "epoch": 0.011081469320746972, + "grad_norm": 4.112154234754447, + "learning_rate": 3.690205011389522e-06, + "loss": 0.3504, + "step": 162 + }, + { + "epoch": 0.011149873452356523, + "grad_norm": 3.555506165520702, + "learning_rate": 3.7129840546697043e-06, + "loss": 0.4485, + "step": 163 + }, + { + "epoch": 0.011218277583966072, + "grad_norm": 5.516215284710705, + "learning_rate": 3.7357630979498866e-06, + "loss": 0.4529, + "step": 164 + }, + { + "epoch": 0.011286681715575621, + "grad_norm": 4.261153144261622, + "learning_rate": 3.758542141230069e-06, + "loss": 0.4671, + "step": 165 + }, + { + "epoch": 0.01135508584718517, + "grad_norm": 4.115496916287671, + "learning_rate": 3.781321184510251e-06, + "loss": 0.4641, + "step": 166 + }, + { + "epoch": 0.011423489978794719, + "grad_norm": 5.17290816114968, + "learning_rate": 3.804100227790433e-06, + "loss": 0.6399, + "step": 167 + }, + { + "epoch": 0.011491894110404268, + "grad_norm": 4.391104403893715, + "learning_rate": 3.826879271070615e-06, + "loss": 0.4279, + "step": 168 + }, + { + "epoch": 0.011560298242013817, + "grad_norm": 8.335203901793443, + "learning_rate": 3.849658314350798e-06, + "loss": 0.4812, + "step": 169 + }, + { + "epoch": 0.011628702373623366, + "grad_norm": 4.434292856629052, + "learning_rate": 3.87243735763098e-06, + "loss": 0.702, + "step": 170 + }, + { + "epoch": 0.011697106505232917, + "grad_norm": 3.7322460953597947, + "learning_rate": 3.8952164009111624e-06, + "loss": 0.6439, + "step": 171 + }, + { + "epoch": 0.011765510636842466, + "grad_norm": 4.772309864666896, + "learning_rate": 3.917995444191344e-06, + "loss": 0.5708, + "step": 172 + }, + { + "epoch": 0.011833914768452015, + "grad_norm": 5.2641211040220535, + "learning_rate": 3.940774487471526e-06, + "loss": 0.678, + "step": 173 + }, + { + "epoch": 0.011902318900061564, + "grad_norm": 3.2721992281019676, + "learning_rate": 3.963553530751709e-06, + "loss": 0.2714, + "step": 174 + }, + { + "epoch": 0.011970723031671113, + "grad_norm": 6.668229440465964, + "learning_rate": 3.986332574031891e-06, + "loss": 1.0851, + "step": 175 + }, + { + "epoch": 0.012039127163280662, + "grad_norm": 4.790181525519879, + "learning_rate": 4.009111617312073e-06, + "loss": 0.2452, + "step": 176 + }, + { + "epoch": 0.01210753129489021, + "grad_norm": 4.55762579283424, + "learning_rate": 4.0318906605922555e-06, + "loss": 0.5185, + "step": 177 + }, + { + "epoch": 0.012175935426499761, + "grad_norm": 11.767477793918758, + "learning_rate": 4.054669703872437e-06, + "loss": 0.2694, + "step": 178 + }, + { + "epoch": 0.01224433955810931, + "grad_norm": 4.458236624955253, + "learning_rate": 4.07744874715262e-06, + "loss": 0.5203, + "step": 179 + }, + { + "epoch": 0.01231274368971886, + "grad_norm": 3.0751982973871077, + "learning_rate": 4.100227790432802e-06, + "loss": 0.3245, + "step": 180 + }, + { + "epoch": 0.012381147821328408, + "grad_norm": 4.7046434226199505, + "learning_rate": 4.123006833712984e-06, + "loss": 0.6176, + "step": 181 + }, + { + "epoch": 0.012449551952937957, + "grad_norm": 4.80791599074275, + "learning_rate": 4.145785876993167e-06, + "loss": 0.2326, + "step": 182 + }, + { + "epoch": 0.012517956084547506, + "grad_norm": 4.092231061018268, + "learning_rate": 4.168564920273349e-06, + "loss": 0.5002, + "step": 183 + }, + { + "epoch": 0.012586360216157055, + "grad_norm": 7.6722773118204675, + "learning_rate": 4.191343963553531e-06, + "loss": 0.8132, + "step": 184 + }, + { + "epoch": 0.012654764347766604, + "grad_norm": 3.296817715766381, + "learning_rate": 4.214123006833713e-06, + "loss": 0.254, + "step": 185 + }, + { + "epoch": 0.012723168479376155, + "grad_norm": 4.902526900681102, + "learning_rate": 4.236902050113896e-06, + "loss": 0.3638, + "step": 186 + }, + { + "epoch": 0.012791572610985704, + "grad_norm": 5.190085731801548, + "learning_rate": 4.259681093394078e-06, + "loss": 0.409, + "step": 187 + }, + { + "epoch": 0.012859976742595253, + "grad_norm": 4.835850531515545, + "learning_rate": 4.2824601366742606e-06, + "loss": 0.5272, + "step": 188 + }, + { + "epoch": 0.012928380874204802, + "grad_norm": 4.459400239484077, + "learning_rate": 4.3052391799544425e-06, + "loss": 0.2758, + "step": 189 + }, + { + "epoch": 0.012996785005814351, + "grad_norm": 6.026480770570654, + "learning_rate": 4.328018223234624e-06, + "loss": 0.5509, + "step": 190 + }, + { + "epoch": 0.0130651891374239, + "grad_norm": 3.8588851432502635, + "learning_rate": 4.350797266514807e-06, + "loss": 0.5614, + "step": 191 + }, + { + "epoch": 0.013133593269033449, + "grad_norm": 2.97172581953081, + "learning_rate": 4.373576309794989e-06, + "loss": 0.4809, + "step": 192 + }, + { + "epoch": 0.013201997400642998, + "grad_norm": 2.376027361871209, + "learning_rate": 4.396355353075171e-06, + "loss": 0.1223, + "step": 193 + }, + { + "epoch": 0.013270401532252549, + "grad_norm": 4.446738244083037, + "learning_rate": 4.419134396355354e-06, + "loss": 0.2533, + "step": 194 + }, + { + "epoch": 0.013338805663862098, + "grad_norm": 2.8521301540621264, + "learning_rate": 4.4419134396355355e-06, + "loss": 0.1751, + "step": 195 + }, + { + "epoch": 0.013407209795471647, + "grad_norm": 4.941495510094261, + "learning_rate": 4.464692482915718e-06, + "loss": 0.5595, + "step": 196 + }, + { + "epoch": 0.013475613927081196, + "grad_norm": 6.919960480533056, + "learning_rate": 4.4874715261959e-06, + "loss": 0.3075, + "step": 197 + }, + { + "epoch": 0.013544018058690745, + "grad_norm": 4.670698674847405, + "learning_rate": 4.510250569476082e-06, + "loss": 0.5307, + "step": 198 + }, + { + "epoch": 0.013612422190300294, + "grad_norm": 5.495798023024644, + "learning_rate": 4.533029612756265e-06, + "loss": 0.3068, + "step": 199 + }, + { + "epoch": 0.013680826321909843, + "grad_norm": 4.21326225243941, + "learning_rate": 4.555808656036447e-06, + "loss": 0.3253, + "step": 200 + }, + { + "epoch": 0.013749230453519393, + "grad_norm": 4.978391733345305, + "learning_rate": 4.578587699316629e-06, + "loss": 0.731, + "step": 201 + }, + { + "epoch": 0.013817634585128942, + "grad_norm": 4.980550608242736, + "learning_rate": 4.601366742596811e-06, + "loss": 0.5036, + "step": 202 + }, + { + "epoch": 0.013886038716738491, + "grad_norm": 4.650973870216513, + "learning_rate": 4.624145785876993e-06, + "loss": 0.8233, + "step": 203 + }, + { + "epoch": 0.01395444284834804, + "grad_norm": 3.5786284046787817, + "learning_rate": 4.646924829157176e-06, + "loss": 0.6272, + "step": 204 + }, + { + "epoch": 0.01402284697995759, + "grad_norm": 9.652753363531936, + "learning_rate": 4.669703872437358e-06, + "loss": 0.7241, + "step": 205 + }, + { + "epoch": 0.014091251111567138, + "grad_norm": 5.218372879564819, + "learning_rate": 4.69248291571754e-06, + "loss": 0.3955, + "step": 206 + }, + { + "epoch": 0.014159655243176687, + "grad_norm": 3.6903642263509435, + "learning_rate": 4.7152619589977225e-06, + "loss": 0.5157, + "step": 207 + }, + { + "epoch": 0.014228059374786236, + "grad_norm": 4.300377212657347, + "learning_rate": 4.738041002277905e-06, + "loss": 0.4585, + "step": 208 + }, + { + "epoch": 0.014296463506395787, + "grad_norm": 3.561884693079683, + "learning_rate": 4.760820045558087e-06, + "loss": 0.4034, + "step": 209 + }, + { + "epoch": 0.014364867638005336, + "grad_norm": 5.072530005861202, + "learning_rate": 4.783599088838269e-06, + "loss": 0.5482, + "step": 210 + }, + { + "epoch": 0.014433271769614885, + "grad_norm": 4.554559408123922, + "learning_rate": 4.806378132118452e-06, + "loss": 0.3215, + "step": 211 + }, + { + "epoch": 0.014501675901224434, + "grad_norm": 3.9709731956313132, + "learning_rate": 4.829157175398634e-06, + "loss": 0.2835, + "step": 212 + }, + { + "epoch": 0.014570080032833983, + "grad_norm": 3.7952870631553224, + "learning_rate": 4.851936218678816e-06, + "loss": 0.5455, + "step": 213 + }, + { + "epoch": 0.014638484164443532, + "grad_norm": 5.715299638984723, + "learning_rate": 4.874715261958998e-06, + "loss": 0.6453, + "step": 214 + }, + { + "epoch": 0.014706888296053081, + "grad_norm": 3.3834166570951445, + "learning_rate": 4.89749430523918e-06, + "loss": 0.2783, + "step": 215 + }, + { + "epoch": 0.014775292427662632, + "grad_norm": 3.7328468931137797, + "learning_rate": 4.920273348519363e-06, + "loss": 0.4554, + "step": 216 + }, + { + "epoch": 0.01484369655927218, + "grad_norm": 4.9791531728883545, + "learning_rate": 4.943052391799545e-06, + "loss": 0.8466, + "step": 217 + }, + { + "epoch": 0.01491210069088173, + "grad_norm": 2.980251409323108, + "learning_rate": 4.965831435079727e-06, + "loss": 0.3549, + "step": 218 + }, + { + "epoch": 0.014980504822491279, + "grad_norm": 4.395206409724005, + "learning_rate": 4.9886104783599095e-06, + "loss": 0.7749, + "step": 219 + }, + { + "epoch": 0.015048908954100828, + "grad_norm": 3.6082187303282236, + "learning_rate": 5.011389521640092e-06, + "loss": 0.4069, + "step": 220 + }, + { + "epoch": 0.015117313085710377, + "grad_norm": 3.522924973793695, + "learning_rate": 5.034168564920274e-06, + "loss": 0.4589, + "step": 221 + }, + { + "epoch": 0.015185717217319926, + "grad_norm": 6.392357132142796, + "learning_rate": 5.056947608200457e-06, + "loss": 0.6428, + "step": 222 + }, + { + "epoch": 0.015254121348929475, + "grad_norm": 3.1645410811562975, + "learning_rate": 5.079726651480639e-06, + "loss": 0.2201, + "step": 223 + }, + { + "epoch": 0.015322525480539025, + "grad_norm": 4.72759393204318, + "learning_rate": 5.102505694760821e-06, + "loss": 0.4195, + "step": 224 + }, + { + "epoch": 0.015390929612148574, + "grad_norm": 3.801655997033193, + "learning_rate": 5.125284738041003e-06, + "loss": 0.5512, + "step": 225 + }, + { + "epoch": 0.015459333743758123, + "grad_norm": 3.621387062415292, + "learning_rate": 5.148063781321185e-06, + "loss": 0.4901, + "step": 226 + }, + { + "epoch": 0.015527737875367672, + "grad_norm": 3.65722512829578, + "learning_rate": 5.170842824601367e-06, + "loss": 0.3665, + "step": 227 + }, + { + "epoch": 0.015596142006977221, + "grad_norm": 3.837425642504662, + "learning_rate": 5.19362186788155e-06, + "loss": 0.5216, + "step": 228 + }, + { + "epoch": 0.015664546138586772, + "grad_norm": 4.20028311081782, + "learning_rate": 5.216400911161732e-06, + "loss": 0.5288, + "step": 229 + }, + { + "epoch": 0.01573295027019632, + "grad_norm": 4.306887699574032, + "learning_rate": 5.2391799544419145e-06, + "loss": 0.5792, + "step": 230 + }, + { + "epoch": 0.01580135440180587, + "grad_norm": 4.66087557028672, + "learning_rate": 5.2619589977220964e-06, + "loss": 0.7369, + "step": 231 + }, + { + "epoch": 0.015869758533415417, + "grad_norm": 4.164492100067285, + "learning_rate": 5.284738041002278e-06, + "loss": 0.5648, + "step": 232 + }, + { + "epoch": 0.015938162665024968, + "grad_norm": 4.133466975233253, + "learning_rate": 5.307517084282461e-06, + "loss": 0.4188, + "step": 233 + }, + { + "epoch": 0.016006566796634515, + "grad_norm": 3.478063897815463, + "learning_rate": 5.330296127562643e-06, + "loss": 0.3895, + "step": 234 + }, + { + "epoch": 0.016074970928244066, + "grad_norm": 4.09787986007061, + "learning_rate": 5.353075170842825e-06, + "loss": 0.3413, + "step": 235 + }, + { + "epoch": 0.016143375059853617, + "grad_norm": 4.2430086763681185, + "learning_rate": 5.375854214123008e-06, + "loss": 0.438, + "step": 236 + }, + { + "epoch": 0.016211779191463164, + "grad_norm": 3.5746900904316177, + "learning_rate": 5.3986332574031895e-06, + "loss": 0.6505, + "step": 237 + }, + { + "epoch": 0.016280183323072715, + "grad_norm": 4.926562205761369, + "learning_rate": 5.421412300683372e-06, + "loss": 0.3022, + "step": 238 + }, + { + "epoch": 0.016348587454682262, + "grad_norm": 4.282850425800018, + "learning_rate": 5.444191343963554e-06, + "loss": 0.3774, + "step": 239 + }, + { + "epoch": 0.016416991586291813, + "grad_norm": 3.2225261315173337, + "learning_rate": 5.466970387243736e-06, + "loss": 0.3282, + "step": 240 + }, + { + "epoch": 0.01648539571790136, + "grad_norm": 4.37771862575557, + "learning_rate": 5.489749430523919e-06, + "loss": 0.5993, + "step": 241 + }, + { + "epoch": 0.01655379984951091, + "grad_norm": 3.6034791360470493, + "learning_rate": 5.512528473804101e-06, + "loss": 0.3754, + "step": 242 + }, + { + "epoch": 0.01662220398112046, + "grad_norm": 3.835251741042257, + "learning_rate": 5.5353075170842826e-06, + "loss": 0.6983, + "step": 243 + }, + { + "epoch": 0.01669060811273001, + "grad_norm": 4.331035585587104, + "learning_rate": 5.558086560364465e-06, + "loss": 0.98, + "step": 244 + }, + { + "epoch": 0.01675901224433956, + "grad_norm": 4.148519109293207, + "learning_rate": 5.580865603644647e-06, + "loss": 0.5085, + "step": 245 + }, + { + "epoch": 0.016827416375949106, + "grad_norm": 4.239177387619685, + "learning_rate": 5.60364464692483e-06, + "loss": 0.5053, + "step": 246 + }, + { + "epoch": 0.016895820507558657, + "grad_norm": 3.5281193029528324, + "learning_rate": 5.626423690205012e-06, + "loss": 0.4398, + "step": 247 + }, + { + "epoch": 0.016964224639168204, + "grad_norm": 3.960072846248373, + "learning_rate": 5.649202733485194e-06, + "loss": 0.4558, + "step": 248 + }, + { + "epoch": 0.017032628770777755, + "grad_norm": 3.0619879703272965, + "learning_rate": 5.6719817767653765e-06, + "loss": 0.2537, + "step": 249 + }, + { + "epoch": 0.017101032902387306, + "grad_norm": 4.5594895503379, + "learning_rate": 5.694760820045558e-06, + "loss": 0.8217, + "step": 250 + }, + { + "epoch": 0.017169437033996853, + "grad_norm": 4.251591426550482, + "learning_rate": 5.71753986332574e-06, + "loss": 0.5406, + "step": 251 + }, + { + "epoch": 0.017237841165606404, + "grad_norm": 4.23090121523519, + "learning_rate": 5.740318906605923e-06, + "loss": 0.5963, + "step": 252 + }, + { + "epoch": 0.01730624529721595, + "grad_norm": 3.7564978570138647, + "learning_rate": 5.763097949886105e-06, + "loss": 0.2759, + "step": 253 + }, + { + "epoch": 0.017374649428825502, + "grad_norm": 4.284384018925384, + "learning_rate": 5.785876993166288e-06, + "loss": 0.5485, + "step": 254 + }, + { + "epoch": 0.01744305356043505, + "grad_norm": 4.556332731084533, + "learning_rate": 5.8086560364464695e-06, + "loss": 0.7456, + "step": 255 + }, + { + "epoch": 0.0175114576920446, + "grad_norm": 10.270258352861504, + "learning_rate": 5.831435079726651e-06, + "loss": 0.7201, + "step": 256 + }, + { + "epoch": 0.017579861823654147, + "grad_norm": 3.2269714502715416, + "learning_rate": 5.854214123006834e-06, + "loss": 0.3734, + "step": 257 + }, + { + "epoch": 0.017648265955263698, + "grad_norm": 3.8008107579213957, + "learning_rate": 5.876993166287016e-06, + "loss": 0.2433, + "step": 258 + }, + { + "epoch": 0.01771667008687325, + "grad_norm": 3.978495774292149, + "learning_rate": 5.899772209567198e-06, + "loss": 0.5346, + "step": 259 + }, + { + "epoch": 0.017785074218482796, + "grad_norm": 4.062700758155044, + "learning_rate": 5.922551252847381e-06, + "loss": 0.39, + "step": 260 + }, + { + "epoch": 0.017853478350092346, + "grad_norm": 3.680067071147968, + "learning_rate": 5.945330296127563e-06, + "loss": 0.491, + "step": 261 + }, + { + "epoch": 0.017921882481701894, + "grad_norm": 3.8246688422043995, + "learning_rate": 5.968109339407745e-06, + "loss": 0.5016, + "step": 262 + }, + { + "epoch": 0.017990286613311444, + "grad_norm": 4.135148482510889, + "learning_rate": 5.990888382687927e-06, + "loss": 0.5702, + "step": 263 + }, + { + "epoch": 0.01805869074492099, + "grad_norm": 3.593063910962475, + "learning_rate": 6.013667425968109e-06, + "loss": 0.2668, + "step": 264 + }, + { + "epoch": 0.018127094876530542, + "grad_norm": 4.015810980738893, + "learning_rate": 6.036446469248292e-06, + "loss": 0.3187, + "step": 265 + }, + { + "epoch": 0.018195499008140093, + "grad_norm": 5.011331544756438, + "learning_rate": 6.059225512528474e-06, + "loss": 0.8526, + "step": 266 + }, + { + "epoch": 0.01826390313974964, + "grad_norm": 3.0209954623857835, + "learning_rate": 6.082004555808656e-06, + "loss": 0.2629, + "step": 267 + }, + { + "epoch": 0.01833230727135919, + "grad_norm": 3.4900134870055433, + "learning_rate": 6.104783599088838e-06, + "loss": 0.5662, + "step": 268 + }, + { + "epoch": 0.01840071140296874, + "grad_norm": 3.814086544862785, + "learning_rate": 6.127562642369021e-06, + "loss": 0.4613, + "step": 269 + }, + { + "epoch": 0.01846911553457829, + "grad_norm": 3.8774699214995976, + "learning_rate": 6.150341685649204e-06, + "loss": 0.6258, + "step": 270 + }, + { + "epoch": 0.018537519666187836, + "grad_norm": 3.3098242226974413, + "learning_rate": 6.173120728929386e-06, + "loss": 0.368, + "step": 271 + }, + { + "epoch": 0.018605923797797387, + "grad_norm": 2.93969036171499, + "learning_rate": 6.1958997722095685e-06, + "loss": 0.4095, + "step": 272 + }, + { + "epoch": 0.018674327929406938, + "grad_norm": 4.297064274173813, + "learning_rate": 6.21867881548975e-06, + "loss": 0.4979, + "step": 273 + }, + { + "epoch": 0.018742732061016485, + "grad_norm": 2.726229781259932, + "learning_rate": 6.241457858769932e-06, + "loss": 0.2681, + "step": 274 + }, + { + "epoch": 0.018811136192626036, + "grad_norm": 4.059936257341167, + "learning_rate": 6.264236902050115e-06, + "loss": 0.4854, + "step": 275 + }, + { + "epoch": 0.018879540324235583, + "grad_norm": 2.8492062278913792, + "learning_rate": 6.287015945330297e-06, + "loss": 0.4001, + "step": 276 + }, + { + "epoch": 0.018947944455845134, + "grad_norm": 3.7236105785411384, + "learning_rate": 6.309794988610479e-06, + "loss": 0.5337, + "step": 277 + }, + { + "epoch": 0.01901634858745468, + "grad_norm": 2.6859520290973746, + "learning_rate": 6.3325740318906616e-06, + "loss": 0.2388, + "step": 278 + }, + { + "epoch": 0.019084752719064232, + "grad_norm": 3.527352182033933, + "learning_rate": 6.3553530751708435e-06, + "loss": 0.3062, + "step": 279 + }, + { + "epoch": 0.01915315685067378, + "grad_norm": 4.1922463662091225, + "learning_rate": 6.378132118451026e-06, + "loss": 0.4103, + "step": 280 + }, + { + "epoch": 0.01922156098228333, + "grad_norm": 4.923881152184107, + "learning_rate": 6.400911161731208e-06, + "loss": 0.3328, + "step": 281 + }, + { + "epoch": 0.01928996511389288, + "grad_norm": 4.0054823885425614, + "learning_rate": 6.42369020501139e-06, + "loss": 0.7239, + "step": 282 + }, + { + "epoch": 0.019358369245502428, + "grad_norm": 3.72005271533306, + "learning_rate": 6.446469248291573e-06, + "loss": 0.283, + "step": 283 + }, + { + "epoch": 0.01942677337711198, + "grad_norm": 4.229505927786465, + "learning_rate": 6.469248291571755e-06, + "loss": 0.5943, + "step": 284 + }, + { + "epoch": 0.019495177508721526, + "grad_norm": 4.2738946464430425, + "learning_rate": 6.4920273348519365e-06, + "loss": 0.3508, + "step": 285 + }, + { + "epoch": 0.019563581640331076, + "grad_norm": 3.770529493275998, + "learning_rate": 6.514806378132119e-06, + "loss": 0.3137, + "step": 286 + }, + { + "epoch": 0.019631985771940624, + "grad_norm": 3.926920529067754, + "learning_rate": 6.537585421412301e-06, + "loss": 0.4062, + "step": 287 + }, + { + "epoch": 0.019700389903550174, + "grad_norm": 3.9972456822065126, + "learning_rate": 6.560364464692484e-06, + "loss": 0.616, + "step": 288 + }, + { + "epoch": 0.019768794035159725, + "grad_norm": 3.3365769856570715, + "learning_rate": 6.583143507972666e-06, + "loss": 0.5358, + "step": 289 + }, + { + "epoch": 0.019837198166769272, + "grad_norm": 3.2491968336235266, + "learning_rate": 6.605922551252848e-06, + "loss": 0.4092, + "step": 290 + }, + { + "epoch": 0.019905602298378823, + "grad_norm": 3.8847805845683165, + "learning_rate": 6.6287015945330304e-06, + "loss": 0.5143, + "step": 291 + }, + { + "epoch": 0.01997400642998837, + "grad_norm": 3.3325976553342707, + "learning_rate": 6.651480637813212e-06, + "loss": 0.51, + "step": 292 + }, + { + "epoch": 0.02004241056159792, + "grad_norm": 4.416943863801096, + "learning_rate": 6.674259681093394e-06, + "loss": 0.5608, + "step": 293 + }, + { + "epoch": 0.02011081469320747, + "grad_norm": 3.6480731169511116, + "learning_rate": 6.697038724373577e-06, + "loss": 0.7121, + "step": 294 + }, + { + "epoch": 0.02017921882481702, + "grad_norm": 3.8245682444643467, + "learning_rate": 6.719817767653759e-06, + "loss": 0.3721, + "step": 295 + }, + { + "epoch": 0.02024762295642657, + "grad_norm": 3.6028651407634533, + "learning_rate": 6.742596810933942e-06, + "loss": 0.3006, + "step": 296 + }, + { + "epoch": 0.020316027088036117, + "grad_norm": 3.8916437737869125, + "learning_rate": 6.7653758542141235e-06, + "loss": 0.534, + "step": 297 + }, + { + "epoch": 0.020384431219645668, + "grad_norm": 2.9868643449544185, + "learning_rate": 6.788154897494305e-06, + "loss": 0.1772, + "step": 298 + }, + { + "epoch": 0.020452835351255215, + "grad_norm": 3.9659200776492756, + "learning_rate": 6.810933940774488e-06, + "loss": 0.6089, + "step": 299 + }, + { + "epoch": 0.020521239482864766, + "grad_norm": 3.691094004832039, + "learning_rate": 6.83371298405467e-06, + "loss": 0.5175, + "step": 300 + }, + { + "epoch": 0.020589643614474313, + "grad_norm": 4.339155662781035, + "learning_rate": 6.856492027334852e-06, + "loss": 0.5534, + "step": 301 + }, + { + "epoch": 0.020658047746083864, + "grad_norm": 2.848978109947804, + "learning_rate": 6.879271070615035e-06, + "loss": 0.3531, + "step": 302 + }, + { + "epoch": 0.020726451877693414, + "grad_norm": 3.009061083558082, + "learning_rate": 6.9020501138952166e-06, + "loss": 0.4712, + "step": 303 + }, + { + "epoch": 0.02079485600930296, + "grad_norm": 4.499013660205827, + "learning_rate": 6.924829157175399e-06, + "loss": 0.6252, + "step": 304 + }, + { + "epoch": 0.020863260140912512, + "grad_norm": 3.6016508720462195, + "learning_rate": 6.947608200455581e-06, + "loss": 0.1842, + "step": 305 + }, + { + "epoch": 0.02093166427252206, + "grad_norm": 2.7585034355908298, + "learning_rate": 6.970387243735763e-06, + "loss": 0.3378, + "step": 306 + }, + { + "epoch": 0.02100006840413161, + "grad_norm": 3.1021671042180614, + "learning_rate": 6.993166287015946e-06, + "loss": 0.4481, + "step": 307 + }, + { + "epoch": 0.021068472535741158, + "grad_norm": 3.1072465354416106, + "learning_rate": 7.015945330296128e-06, + "loss": 0.3292, + "step": 308 + }, + { + "epoch": 0.02113687666735071, + "grad_norm": 3.831242580313289, + "learning_rate": 7.03872437357631e-06, + "loss": 0.5045, + "step": 309 + }, + { + "epoch": 0.021205280798960256, + "grad_norm": 5.871208743087901, + "learning_rate": 7.061503416856492e-06, + "loss": 0.5975, + "step": 310 + }, + { + "epoch": 0.021273684930569806, + "grad_norm": 3.8521955020239957, + "learning_rate": 7.084282460136674e-06, + "loss": 0.5092, + "step": 311 + }, + { + "epoch": 0.021342089062179357, + "grad_norm": 4.571616092842075, + "learning_rate": 7.107061503416857e-06, + "loss": 0.3806, + "step": 312 + }, + { + "epoch": 0.021410493193788904, + "grad_norm": 4.393180904187398, + "learning_rate": 7.129840546697039e-06, + "loss": 0.7288, + "step": 313 + }, + { + "epoch": 0.021478897325398455, + "grad_norm": 3.8126424153103007, + "learning_rate": 7.152619589977221e-06, + "loss": 0.7039, + "step": 314 + }, + { + "epoch": 0.021547301457008002, + "grad_norm": 3.4203134570723823, + "learning_rate": 7.1753986332574035e-06, + "loss": 0.6817, + "step": 315 + }, + { + "epoch": 0.021615705588617553, + "grad_norm": 3.023977660195016, + "learning_rate": 7.1981776765375854e-06, + "loss": 0.2983, + "step": 316 + }, + { + "epoch": 0.0216841097202271, + "grad_norm": 2.8267522662784503, + "learning_rate": 7.220956719817767e-06, + "loss": 0.2302, + "step": 317 + }, + { + "epoch": 0.02175251385183665, + "grad_norm": 3.3154795939266677, + "learning_rate": 7.243735763097951e-06, + "loss": 0.5082, + "step": 318 + }, + { + "epoch": 0.0218209179834462, + "grad_norm": 3.8429064639186055, + "learning_rate": 7.266514806378133e-06, + "loss": 0.3542, + "step": 319 + }, + { + "epoch": 0.02188932211505575, + "grad_norm": 4.278921245766677, + "learning_rate": 7.2892938496583155e-06, + "loss": 0.771, + "step": 320 + }, + { + "epoch": 0.0219577262466653, + "grad_norm": 3.106450425940979, + "learning_rate": 7.3120728929384974e-06, + "loss": 0.3979, + "step": 321 + }, + { + "epoch": 0.022026130378274847, + "grad_norm": 3.33226752924825, + "learning_rate": 7.33485193621868e-06, + "loss": 0.228, + "step": 322 + }, + { + "epoch": 0.022094534509884398, + "grad_norm": 5.085273980661899, + "learning_rate": 7.357630979498862e-06, + "loss": 0.3076, + "step": 323 + }, + { + "epoch": 0.022162938641493945, + "grad_norm": 3.4393584602841405, + "learning_rate": 7.380410022779044e-06, + "loss": 0.5177, + "step": 324 + }, + { + "epoch": 0.022231342773103496, + "grad_norm": 2.928862773524759, + "learning_rate": 7.403189066059227e-06, + "loss": 0.2724, + "step": 325 + }, + { + "epoch": 0.022299746904713046, + "grad_norm": 5.24084960419797, + "learning_rate": 7.425968109339409e-06, + "loss": 0.2973, + "step": 326 + }, + { + "epoch": 0.022368151036322594, + "grad_norm": 3.815850356974092, + "learning_rate": 7.4487471526195905e-06, + "loss": 0.3185, + "step": 327 + }, + { + "epoch": 0.022436555167932144, + "grad_norm": 3.7971008924311764, + "learning_rate": 7.471526195899773e-06, + "loss": 0.2887, + "step": 328 + }, + { + "epoch": 0.02250495929954169, + "grad_norm": 3.2605466461272963, + "learning_rate": 7.494305239179955e-06, + "loss": 0.3849, + "step": 329 + }, + { + "epoch": 0.022573363431151242, + "grad_norm": 3.5691999166848714, + "learning_rate": 7.517084282460138e-06, + "loss": 0.418, + "step": 330 + }, + { + "epoch": 0.02264176756276079, + "grad_norm": 11.506955893774379, + "learning_rate": 7.53986332574032e-06, + "loss": 0.5523, + "step": 331 + }, + { + "epoch": 0.02271017169437034, + "grad_norm": 3.9977086030783493, + "learning_rate": 7.562642369020502e-06, + "loss": 0.5796, + "step": 332 + }, + { + "epoch": 0.022778575825979887, + "grad_norm": 4.5493431995368905, + "learning_rate": 7.585421412300684e-06, + "loss": 0.6003, + "step": 333 + }, + { + "epoch": 0.022846979957589438, + "grad_norm": 4.003264376665545, + "learning_rate": 7.608200455580866e-06, + "loss": 0.8686, + "step": 334 + }, + { + "epoch": 0.02291538408919899, + "grad_norm": 3.1989387557361133, + "learning_rate": 7.630979498861048e-06, + "loss": 0.2725, + "step": 335 + }, + { + "epoch": 0.022983788220808536, + "grad_norm": 2.858402253497997, + "learning_rate": 7.65375854214123e-06, + "loss": 0.4511, + "step": 336 + }, + { + "epoch": 0.023052192352418087, + "grad_norm": 6.383144412054614, + "learning_rate": 7.676537585421414e-06, + "loss": 0.367, + "step": 337 + }, + { + "epoch": 0.023120596484027634, + "grad_norm": 4.3710183553866635, + "learning_rate": 7.699316628701596e-06, + "loss": 0.6829, + "step": 338 + }, + { + "epoch": 0.023189000615637185, + "grad_norm": 3.345858524096247, + "learning_rate": 7.722095671981777e-06, + "loss": 0.3344, + "step": 339 + }, + { + "epoch": 0.023257404747246732, + "grad_norm": 4.01325195046353, + "learning_rate": 7.74487471526196e-06, + "loss": 0.3982, + "step": 340 + }, + { + "epoch": 0.023325808878856283, + "grad_norm": 3.588061740306305, + "learning_rate": 7.767653758542141e-06, + "loss": 0.5188, + "step": 341 + }, + { + "epoch": 0.023394213010465834, + "grad_norm": 3.7438591028915096, + "learning_rate": 7.790432801822325e-06, + "loss": 0.5339, + "step": 342 + }, + { + "epoch": 0.02346261714207538, + "grad_norm": 5.065830559769252, + "learning_rate": 7.813211845102507e-06, + "loss": 0.7875, + "step": 343 + }, + { + "epoch": 0.02353102127368493, + "grad_norm": 2.7845650589850495, + "learning_rate": 7.835990888382689e-06, + "loss": 0.401, + "step": 344 + }, + { + "epoch": 0.02359942540529448, + "grad_norm": 2.6792730060898475, + "learning_rate": 7.85876993166287e-06, + "loss": 0.2864, + "step": 345 + }, + { + "epoch": 0.02366782953690403, + "grad_norm": 3.750514181966684, + "learning_rate": 7.881548974943052e-06, + "loss": 0.5005, + "step": 346 + }, + { + "epoch": 0.023736233668513577, + "grad_norm": 3.594812302524199, + "learning_rate": 7.904328018223234e-06, + "loss": 0.4001, + "step": 347 + }, + { + "epoch": 0.023804637800123127, + "grad_norm": 3.574369958899689, + "learning_rate": 7.927107061503418e-06, + "loss": 0.3995, + "step": 348 + }, + { + "epoch": 0.023873041931732678, + "grad_norm": 3.149186208599906, + "learning_rate": 7.9498861047836e-06, + "loss": 0.4498, + "step": 349 + }, + { + "epoch": 0.023941446063342225, + "grad_norm": 3.083526838304888, + "learning_rate": 7.972665148063782e-06, + "loss": 0.4633, + "step": 350 + }, + { + "epoch": 0.024009850194951776, + "grad_norm": 4.332822804760295, + "learning_rate": 7.995444191343964e-06, + "loss": 0.8546, + "step": 351 + }, + { + "epoch": 0.024078254326561323, + "grad_norm": 3.619934612534559, + "learning_rate": 8.018223234624145e-06, + "loss": 0.584, + "step": 352 + }, + { + "epoch": 0.024146658458170874, + "grad_norm": 3.7238167137836293, + "learning_rate": 8.041002277904329e-06, + "loss": 0.3757, + "step": 353 + }, + { + "epoch": 0.02421506258978042, + "grad_norm": 4.112077904696136, + "learning_rate": 8.063781321184511e-06, + "loss": 0.761, + "step": 354 + }, + { + "epoch": 0.024283466721389972, + "grad_norm": 4.4836903358448055, + "learning_rate": 8.086560364464693e-06, + "loss": 0.4796, + "step": 355 + }, + { + "epoch": 0.024351870852999523, + "grad_norm": 3.7059288355350253, + "learning_rate": 8.109339407744875e-06, + "loss": 0.5471, + "step": 356 + }, + { + "epoch": 0.02442027498460907, + "grad_norm": 3.0309375719911515, + "learning_rate": 8.132118451025057e-06, + "loss": 0.5134, + "step": 357 + }, + { + "epoch": 0.02448867911621862, + "grad_norm": 4.902699783588555, + "learning_rate": 8.15489749430524e-06, + "loss": 0.5135, + "step": 358 + }, + { + "epoch": 0.024557083247828168, + "grad_norm": 3.146452881686815, + "learning_rate": 8.177676537585422e-06, + "loss": 0.2271, + "step": 359 + }, + { + "epoch": 0.02462548737943772, + "grad_norm": 3.4669189517596486, + "learning_rate": 8.200455580865604e-06, + "loss": 0.5314, + "step": 360 + }, + { + "epoch": 0.024693891511047266, + "grad_norm": 3.4618266063667864, + "learning_rate": 8.223234624145786e-06, + "loss": 0.5446, + "step": 361 + }, + { + "epoch": 0.024762295642656817, + "grad_norm": 2.9157086956281377, + "learning_rate": 8.246013667425968e-06, + "loss": 0.5212, + "step": 362 + }, + { + "epoch": 0.024830699774266364, + "grad_norm": 4.102495941029923, + "learning_rate": 8.26879271070615e-06, + "loss": 0.4213, + "step": 363 + }, + { + "epoch": 0.024899103905875915, + "grad_norm": 4.3980161556530035, + "learning_rate": 8.291571753986333e-06, + "loss": 0.4604, + "step": 364 + }, + { + "epoch": 0.024967508037485465, + "grad_norm": 3.826347607972443, + "learning_rate": 8.314350797266515e-06, + "loss": 0.3827, + "step": 365 + }, + { + "epoch": 0.025035912169095013, + "grad_norm": 3.6339663242161184, + "learning_rate": 8.337129840546699e-06, + "loss": 0.4315, + "step": 366 + }, + { + "epoch": 0.025104316300704563, + "grad_norm": 3.8880355827867694, + "learning_rate": 8.35990888382688e-06, + "loss": 0.6917, + "step": 367 + }, + { + "epoch": 0.02517272043231411, + "grad_norm": 3.2629966014419987, + "learning_rate": 8.382687927107063e-06, + "loss": 0.2207, + "step": 368 + }, + { + "epoch": 0.02524112456392366, + "grad_norm": 3.0120840284042694, + "learning_rate": 8.405466970387244e-06, + "loss": 0.3117, + "step": 369 + }, + { + "epoch": 0.02530952869553321, + "grad_norm": 3.270519148361868, + "learning_rate": 8.428246013667426e-06, + "loss": 0.3724, + "step": 370 + }, + { + "epoch": 0.02537793282714276, + "grad_norm": 3.077599153769058, + "learning_rate": 8.45102505694761e-06, + "loss": 0.2692, + "step": 371 + }, + { + "epoch": 0.02544633695875231, + "grad_norm": 3.760271691286583, + "learning_rate": 8.473804100227792e-06, + "loss": 0.38, + "step": 372 + }, + { + "epoch": 0.025514741090361857, + "grad_norm": 3.3310427703241645, + "learning_rate": 8.496583143507974e-06, + "loss": 0.6353, + "step": 373 + }, + { + "epoch": 0.025583145221971408, + "grad_norm": 2.7123103115012217, + "learning_rate": 8.519362186788156e-06, + "loss": 0.173, + "step": 374 + }, + { + "epoch": 0.025651549353580955, + "grad_norm": 2.8690538282879334, + "learning_rate": 8.542141230068338e-06, + "loss": 0.2527, + "step": 375 + }, + { + "epoch": 0.025719953485190506, + "grad_norm": 2.8193627710654345, + "learning_rate": 8.564920273348521e-06, + "loss": 0.2424, + "step": 376 + }, + { + "epoch": 0.025788357616800053, + "grad_norm": 4.96406103607819, + "learning_rate": 8.587699316628703e-06, + "loss": 0.2991, + "step": 377 + }, + { + "epoch": 0.025856761748409604, + "grad_norm": 6.931880565197731, + "learning_rate": 8.610478359908885e-06, + "loss": 0.5233, + "step": 378 + }, + { + "epoch": 0.025925165880019155, + "grad_norm": 3.423776420555386, + "learning_rate": 8.633257403189067e-06, + "loss": 0.4978, + "step": 379 + }, + { + "epoch": 0.025993570011628702, + "grad_norm": 3.5539393381728153, + "learning_rate": 8.656036446469249e-06, + "loss": 0.5667, + "step": 380 + }, + { + "epoch": 0.026061974143238253, + "grad_norm": 8.00034344576688, + "learning_rate": 8.67881548974943e-06, + "loss": 0.4214, + "step": 381 + }, + { + "epoch": 0.0261303782748478, + "grad_norm": 3.277894275062366, + "learning_rate": 8.701594533029614e-06, + "loss": 0.3367, + "step": 382 + }, + { + "epoch": 0.02619878240645735, + "grad_norm": 2.321946990898113, + "learning_rate": 8.724373576309796e-06, + "loss": 0.2716, + "step": 383 + }, + { + "epoch": 0.026267186538066898, + "grad_norm": 3.0666897102108694, + "learning_rate": 8.747152619589978e-06, + "loss": 0.2733, + "step": 384 + }, + { + "epoch": 0.02633559066967645, + "grad_norm": 2.581751056536653, + "learning_rate": 8.76993166287016e-06, + "loss": 0.2582, + "step": 385 + }, + { + "epoch": 0.026403994801285996, + "grad_norm": 3.69392026034428, + "learning_rate": 8.792710706150342e-06, + "loss": 0.4831, + "step": 386 + }, + { + "epoch": 0.026472398932895547, + "grad_norm": 4.124971158475906, + "learning_rate": 8.815489749430525e-06, + "loss": 0.4475, + "step": 387 + }, + { + "epoch": 0.026540803064505097, + "grad_norm": 4.1570477401271555, + "learning_rate": 8.838268792710707e-06, + "loss": 0.4264, + "step": 388 + }, + { + "epoch": 0.026609207196114645, + "grad_norm": 3.761411446351577, + "learning_rate": 8.861047835990889e-06, + "loss": 0.4156, + "step": 389 + }, + { + "epoch": 0.026677611327724195, + "grad_norm": 3.6986484752669266, + "learning_rate": 8.883826879271071e-06, + "loss": 0.5276, + "step": 390 + }, + { + "epoch": 0.026746015459333743, + "grad_norm": 2.812726165367046, + "learning_rate": 8.906605922551253e-06, + "loss": 0.2916, + "step": 391 + }, + { + "epoch": 0.026814419590943293, + "grad_norm": 3.625073004678058, + "learning_rate": 8.929384965831437e-06, + "loss": 0.3811, + "step": 392 + }, + { + "epoch": 0.02688282372255284, + "grad_norm": 3.1087214391674864, + "learning_rate": 8.952164009111618e-06, + "loss": 0.5179, + "step": 393 + }, + { + "epoch": 0.02695122785416239, + "grad_norm": 4.16027113348425, + "learning_rate": 8.9749430523918e-06, + "loss": 0.5896, + "step": 394 + }, + { + "epoch": 0.027019631985771942, + "grad_norm": 3.483591924009424, + "learning_rate": 8.997722095671982e-06, + "loss": 0.5815, + "step": 395 + }, + { + "epoch": 0.02708803611738149, + "grad_norm": 3.0386850863089965, + "learning_rate": 9.020501138952164e-06, + "loss": 0.6145, + "step": 396 + }, + { + "epoch": 0.02715644024899104, + "grad_norm": 3.520198441243741, + "learning_rate": 9.043280182232346e-06, + "loss": 0.3, + "step": 397 + }, + { + "epoch": 0.027224844380600587, + "grad_norm": 3.1042388687868154, + "learning_rate": 9.06605922551253e-06, + "loss": 0.4012, + "step": 398 + }, + { + "epoch": 0.027293248512210138, + "grad_norm": 3.12746082689984, + "learning_rate": 9.088838268792711e-06, + "loss": 0.5832, + "step": 399 + }, + { + "epoch": 0.027361652643819685, + "grad_norm": 3.1097115425588857, + "learning_rate": 9.111617312072893e-06, + "loss": 0.2733, + "step": 400 + }, + { + "epoch": 0.027430056775429236, + "grad_norm": 2.888210422265504, + "learning_rate": 9.134396355353075e-06, + "loss": 0.3149, + "step": 401 + }, + { + "epoch": 0.027498460907038787, + "grad_norm": 2.7500043782286037, + "learning_rate": 9.157175398633257e-06, + "loss": 0.3753, + "step": 402 + }, + { + "epoch": 0.027566865038648334, + "grad_norm": 3.768223574233722, + "learning_rate": 9.17995444191344e-06, + "loss": 0.9369, + "step": 403 + }, + { + "epoch": 0.027635269170257885, + "grad_norm": 3.425670678541133, + "learning_rate": 9.202733485193623e-06, + "loss": 0.635, + "step": 404 + }, + { + "epoch": 0.027703673301867432, + "grad_norm": 3.4196473452073683, + "learning_rate": 9.225512528473805e-06, + "loss": 0.4468, + "step": 405 + }, + { + "epoch": 0.027772077433476983, + "grad_norm": 3.2052857624049675, + "learning_rate": 9.248291571753986e-06, + "loss": 0.7293, + "step": 406 + }, + { + "epoch": 0.02784048156508653, + "grad_norm": 3.8373972180049907, + "learning_rate": 9.271070615034168e-06, + "loss": 0.6246, + "step": 407 + }, + { + "epoch": 0.02790888569669608, + "grad_norm": 3.6417563649238, + "learning_rate": 9.293849658314352e-06, + "loss": 0.4313, + "step": 408 + }, + { + "epoch": 0.02797728982830563, + "grad_norm": 3.0761725822326818, + "learning_rate": 9.316628701594534e-06, + "loss": 0.6512, + "step": 409 + }, + { + "epoch": 0.02804569395991518, + "grad_norm": 3.7079495497499773, + "learning_rate": 9.339407744874716e-06, + "loss": 0.5228, + "step": 410 + }, + { + "epoch": 0.02811409809152473, + "grad_norm": 3.7035983302217588, + "learning_rate": 9.362186788154898e-06, + "loss": 0.5712, + "step": 411 + }, + { + "epoch": 0.028182502223134277, + "grad_norm": 3.4143242288299445, + "learning_rate": 9.38496583143508e-06, + "loss": 0.5704, + "step": 412 + }, + { + "epoch": 0.028250906354743827, + "grad_norm": 2.739781484339611, + "learning_rate": 9.407744874715261e-06, + "loss": 0.3394, + "step": 413 + }, + { + "epoch": 0.028319310486353375, + "grad_norm": 3.6047337899875713, + "learning_rate": 9.430523917995445e-06, + "loss": 0.6448, + "step": 414 + }, + { + "epoch": 0.028387714617962925, + "grad_norm": 3.0949549784798385, + "learning_rate": 9.453302961275629e-06, + "loss": 0.5478, + "step": 415 + }, + { + "epoch": 0.028456118749572473, + "grad_norm": 3.7213636798172005, + "learning_rate": 9.47608200455581e-06, + "loss": 0.2971, + "step": 416 + }, + { + "epoch": 0.028524522881182023, + "grad_norm": 3.2720712801437117, + "learning_rate": 9.498861047835992e-06, + "loss": 0.6113, + "step": 417 + }, + { + "epoch": 0.028592927012791574, + "grad_norm": 3.6928802695911265, + "learning_rate": 9.521640091116174e-06, + "loss": 0.4674, + "step": 418 + }, + { + "epoch": 0.02866133114440112, + "grad_norm": 2.1746544856741985, + "learning_rate": 9.544419134396356e-06, + "loss": 0.2738, + "step": 419 + }, + { + "epoch": 0.028729735276010672, + "grad_norm": 2.7947841906272792, + "learning_rate": 9.567198177676538e-06, + "loss": 0.6011, + "step": 420 + }, + { + "epoch": 0.02879813940762022, + "grad_norm": 2.879511791504836, + "learning_rate": 9.589977220956722e-06, + "loss": 0.3807, + "step": 421 + }, + { + "epoch": 0.02886654353922977, + "grad_norm": 3.352271798578965, + "learning_rate": 9.612756264236904e-06, + "loss": 0.4886, + "step": 422 + }, + { + "epoch": 0.028934947670839317, + "grad_norm": 3.0034583939740513, + "learning_rate": 9.635535307517085e-06, + "loss": 0.5573, + "step": 423 + }, + { + "epoch": 0.029003351802448868, + "grad_norm": 3.1738056170325692, + "learning_rate": 9.658314350797267e-06, + "loss": 0.6096, + "step": 424 + }, + { + "epoch": 0.02907175593405842, + "grad_norm": 2.4493024711353923, + "learning_rate": 9.68109339407745e-06, + "loss": 0.3639, + "step": 425 + }, + { + "epoch": 0.029140160065667966, + "grad_norm": 2.8357672390651323, + "learning_rate": 9.703872437357633e-06, + "loss": 0.3013, + "step": 426 + }, + { + "epoch": 0.029208564197277517, + "grad_norm": 2.86195582252171, + "learning_rate": 9.726651480637815e-06, + "loss": 0.4527, + "step": 427 + }, + { + "epoch": 0.029276968328887064, + "grad_norm": 2.889336809305718, + "learning_rate": 9.749430523917997e-06, + "loss": 0.3614, + "step": 428 + }, + { + "epoch": 0.029345372460496615, + "grad_norm": 1.765439040594671, + "learning_rate": 9.772209567198178e-06, + "loss": 0.1831, + "step": 429 + }, + { + "epoch": 0.029413776592106162, + "grad_norm": 2.9041918265476543, + "learning_rate": 9.79498861047836e-06, + "loss": 0.4614, + "step": 430 + }, + { + "epoch": 0.029482180723715713, + "grad_norm": 2.717476148996379, + "learning_rate": 9.817767653758544e-06, + "loss": 0.3071, + "step": 431 + }, + { + "epoch": 0.029550584855325263, + "grad_norm": 1.496299117014328, + "learning_rate": 9.840546697038726e-06, + "loss": 0.1715, + "step": 432 + }, + { + "epoch": 0.02961898898693481, + "grad_norm": 4.705775429759128, + "learning_rate": 9.863325740318908e-06, + "loss": 0.7501, + "step": 433 + }, + { + "epoch": 0.02968739311854436, + "grad_norm": 2.3308916869943745, + "learning_rate": 9.88610478359909e-06, + "loss": 0.3037, + "step": 434 + }, + { + "epoch": 0.02975579725015391, + "grad_norm": 2.2247337762011963, + "learning_rate": 9.908883826879272e-06, + "loss": 0.3008, + "step": 435 + }, + { + "epoch": 0.02982420138176346, + "grad_norm": 3.89882902575919, + "learning_rate": 9.931662870159453e-06, + "loss": 0.454, + "step": 436 + }, + { + "epoch": 0.029892605513373006, + "grad_norm": 3.0862241201447413, + "learning_rate": 9.954441913439637e-06, + "loss": 0.4607, + "step": 437 + }, + { + "epoch": 0.029961009644982557, + "grad_norm": 4.786859998429217, + "learning_rate": 9.977220956719819e-06, + "loss": 0.6094, + "step": 438 + }, + { + "epoch": 0.030029413776592104, + "grad_norm": 4.676960145592508, + "learning_rate": 1e-05, + "loss": 0.3554, + "step": 439 + }, + { + "epoch": 0.030097817908201655, + "grad_norm": 1.7808507170637213, + "learning_rate": 9.999999877287929e-06, + "loss": 0.1795, + "step": 440 + }, + { + "epoch": 0.030166222039811206, + "grad_norm": 3.032278437024821, + "learning_rate": 9.999999509151716e-06, + "loss": 0.5839, + "step": 441 + }, + { + "epoch": 0.030234626171420753, + "grad_norm": 3.022281287651799, + "learning_rate": 9.999998895591385e-06, + "loss": 0.4707, + "step": 442 + }, + { + "epoch": 0.030303030303030304, + "grad_norm": 3.5026035842989565, + "learning_rate": 9.999998036606964e-06, + "loss": 0.5391, + "step": 443 + }, + { + "epoch": 0.03037143443463985, + "grad_norm": 3.2818082062159686, + "learning_rate": 9.999996932198493e-06, + "loss": 0.4377, + "step": 444 + }, + { + "epoch": 0.030439838566249402, + "grad_norm": 2.9644564449760695, + "learning_rate": 9.999995582366028e-06, + "loss": 0.3288, + "step": 445 + }, + { + "epoch": 0.03050824269785895, + "grad_norm": 2.8180429937969365, + "learning_rate": 9.999993987109635e-06, + "loss": 0.2473, + "step": 446 + }, + { + "epoch": 0.0305766468294685, + "grad_norm": 3.3010059979912114, + "learning_rate": 9.999992146429393e-06, + "loss": 0.5265, + "step": 447 + }, + { + "epoch": 0.03064505096107805, + "grad_norm": 2.9658384501556077, + "learning_rate": 9.99999006032539e-06, + "loss": 0.4664, + "step": 448 + }, + { + "epoch": 0.030713455092687598, + "grad_norm": 1.9908315269972456, + "learning_rate": 9.999987728797733e-06, + "loss": 0.2147, + "step": 449 + }, + { + "epoch": 0.03078185922429715, + "grad_norm": 2.580728603379656, + "learning_rate": 9.999985151846532e-06, + "loss": 0.2969, + "step": 450 + }, + { + "epoch": 0.030850263355906696, + "grad_norm": 2.599969656932276, + "learning_rate": 9.999982329471915e-06, + "loss": 0.3298, + "step": 451 + }, + { + "epoch": 0.030918667487516246, + "grad_norm": 3.5494568288337636, + "learning_rate": 9.99997926167402e-06, + "loss": 0.8638, + "step": 452 + }, + { + "epoch": 0.030987071619125794, + "grad_norm": 2.93573446175713, + "learning_rate": 9.999975948452999e-06, + "loss": 0.5184, + "step": 453 + }, + { + "epoch": 0.031055475750735344, + "grad_norm": 2.575045885445724, + "learning_rate": 9.999972389809014e-06, + "loss": 0.3596, + "step": 454 + }, + { + "epoch": 0.031123879882344895, + "grad_norm": 3.435352547654266, + "learning_rate": 9.99996858574224e-06, + "loss": 0.5975, + "step": 455 + }, + { + "epoch": 0.031192284013954442, + "grad_norm": 2.1019723520921962, + "learning_rate": 9.999964536252861e-06, + "loss": 0.3655, + "step": 456 + }, + { + "epoch": 0.03126068814556399, + "grad_norm": 5.970973334719753, + "learning_rate": 9.99996024134108e-06, + "loss": 0.5215, + "step": 457 + }, + { + "epoch": 0.031329092277173544, + "grad_norm": 3.311366715093286, + "learning_rate": 9.999955701007104e-06, + "loss": 0.7659, + "step": 458 + }, + { + "epoch": 0.03139749640878309, + "grad_norm": 2.657825731875826, + "learning_rate": 9.99995091525116e-06, + "loss": 0.4734, + "step": 459 + }, + { + "epoch": 0.03146590054039264, + "grad_norm": 3.594246150058569, + "learning_rate": 9.999945884073478e-06, + "loss": 0.3776, + "step": 460 + }, + { + "epoch": 0.03153430467200219, + "grad_norm": 2.318633309998629, + "learning_rate": 9.99994060747431e-06, + "loss": 0.1815, + "step": 461 + }, + { + "epoch": 0.03160270880361174, + "grad_norm": 2.9719758527071263, + "learning_rate": 9.999935085453911e-06, + "loss": 0.3784, + "step": 462 + }, + { + "epoch": 0.03167111293522129, + "grad_norm": 2.43643607860207, + "learning_rate": 9.999929318012555e-06, + "loss": 0.2908, + "step": 463 + }, + { + "epoch": 0.031739517066830834, + "grad_norm": 3.575861268901555, + "learning_rate": 9.999923305150523e-06, + "loss": 0.476, + "step": 464 + }, + { + "epoch": 0.031807921198440385, + "grad_norm": 3.5346319747717305, + "learning_rate": 9.99991704686811e-06, + "loss": 0.5814, + "step": 465 + }, + { + "epoch": 0.031876325330049936, + "grad_norm": 2.6997709578922766, + "learning_rate": 9.999910543165623e-06, + "loss": 0.3302, + "step": 466 + }, + { + "epoch": 0.031944729461659487, + "grad_norm": 3.1771880690190586, + "learning_rate": 9.999903794043385e-06, + "loss": 0.259, + "step": 467 + }, + { + "epoch": 0.03201313359326903, + "grad_norm": 3.1329381161284546, + "learning_rate": 9.999896799501722e-06, + "loss": 0.3114, + "step": 468 + }, + { + "epoch": 0.03208153772487858, + "grad_norm": 2.709007092568346, + "learning_rate": 9.999889559540984e-06, + "loss": 0.3012, + "step": 469 + }, + { + "epoch": 0.03214994185648813, + "grad_norm": 3.071754990306721, + "learning_rate": 9.999882074161517e-06, + "loss": 0.5553, + "step": 470 + }, + { + "epoch": 0.03221834598809768, + "grad_norm": 3.60034241769772, + "learning_rate": 9.999874343363695e-06, + "loss": 0.5269, + "step": 471 + }, + { + "epoch": 0.03228675011970723, + "grad_norm": 2.8272688976708356, + "learning_rate": 9.999866367147897e-06, + "loss": 0.2089, + "step": 472 + }, + { + "epoch": 0.03235515425131678, + "grad_norm": 3.876798181776706, + "learning_rate": 9.999858145514513e-06, + "loss": 0.7591, + "step": 473 + }, + { + "epoch": 0.03242355838292633, + "grad_norm": 3.661939302691454, + "learning_rate": 9.999849678463949e-06, + "loss": 0.542, + "step": 474 + }, + { + "epoch": 0.03249196251453588, + "grad_norm": 3.443931821949108, + "learning_rate": 9.999840965996617e-06, + "loss": 0.5828, + "step": 475 + }, + { + "epoch": 0.03256036664614543, + "grad_norm": 2.663331599947143, + "learning_rate": 9.999832008112947e-06, + "loss": 0.284, + "step": 476 + }, + { + "epoch": 0.03262877077775498, + "grad_norm": 3.094973640052321, + "learning_rate": 9.999822804813378e-06, + "loss": 0.379, + "step": 477 + }, + { + "epoch": 0.032697174909364524, + "grad_norm": 2.9061791888193444, + "learning_rate": 9.999813356098363e-06, + "loss": 0.3004, + "step": 478 + }, + { + "epoch": 0.032765579040974074, + "grad_norm": 2.7008084640919456, + "learning_rate": 9.999803661968361e-06, + "loss": 0.3416, + "step": 479 + }, + { + "epoch": 0.032833983172583625, + "grad_norm": 3.1501273454931615, + "learning_rate": 9.999793722423855e-06, + "loss": 0.4789, + "step": 480 + }, + { + "epoch": 0.032902387304193176, + "grad_norm": 2.4121181781314722, + "learning_rate": 9.999783537465328e-06, + "loss": 0.3347, + "step": 481 + }, + { + "epoch": 0.03297079143580272, + "grad_norm": 4.221073697480436, + "learning_rate": 9.999773107093282e-06, + "loss": 0.4664, + "step": 482 + }, + { + "epoch": 0.03303919556741227, + "grad_norm": 3.732556830201586, + "learning_rate": 9.999762431308228e-06, + "loss": 0.6438, + "step": 483 + }, + { + "epoch": 0.03310759969902182, + "grad_norm": 3.4097579164097755, + "learning_rate": 9.99975151011069e-06, + "loss": 0.4665, + "step": 484 + }, + { + "epoch": 0.03317600383063137, + "grad_norm": 3.472983242984133, + "learning_rate": 9.999740343501203e-06, + "loss": 0.7539, + "step": 485 + }, + { + "epoch": 0.03324440796224092, + "grad_norm": 3.7602202063866947, + "learning_rate": 9.999728931480318e-06, + "loss": 0.3613, + "step": 486 + }, + { + "epoch": 0.033312812093850466, + "grad_norm": 2.636213531128844, + "learning_rate": 9.999717274048594e-06, + "loss": 0.3593, + "step": 487 + }, + { + "epoch": 0.03338121622546002, + "grad_norm": 3.362759823315215, + "learning_rate": 9.999705371206601e-06, + "loss": 0.5417, + "step": 488 + }, + { + "epoch": 0.03344962035706957, + "grad_norm": 2.9222648788877486, + "learning_rate": 9.999693222954928e-06, + "loss": 0.5347, + "step": 489 + }, + { + "epoch": 0.03351802448867912, + "grad_norm": 2.1015761084275515, + "learning_rate": 9.999680829294165e-06, + "loss": 0.2959, + "step": 490 + }, + { + "epoch": 0.03358642862028866, + "grad_norm": 2.8531063551361457, + "learning_rate": 9.999668190224926e-06, + "loss": 0.3278, + "step": 491 + }, + { + "epoch": 0.03365483275189821, + "grad_norm": 4.055530897710398, + "learning_rate": 9.999655305747829e-06, + "loss": 0.6176, + "step": 492 + }, + { + "epoch": 0.033723236883507764, + "grad_norm": 2.712243304324089, + "learning_rate": 9.999642175863504e-06, + "loss": 0.298, + "step": 493 + }, + { + "epoch": 0.033791641015117314, + "grad_norm": 3.3034662104755457, + "learning_rate": 9.9996288005726e-06, + "loss": 0.7193, + "step": 494 + }, + { + "epoch": 0.033860045146726865, + "grad_norm": 3.3013404688487125, + "learning_rate": 9.999615179875772e-06, + "loss": 0.9162, + "step": 495 + }, + { + "epoch": 0.03392844927833641, + "grad_norm": 2.816846381939879, + "learning_rate": 9.999601313773687e-06, + "loss": 0.5054, + "step": 496 + }, + { + "epoch": 0.03399685340994596, + "grad_norm": 2.509384328176374, + "learning_rate": 9.999587202267027e-06, + "loss": 0.3333, + "step": 497 + }, + { + "epoch": 0.03406525754155551, + "grad_norm": 3.1580517602245797, + "learning_rate": 9.999572845356484e-06, + "loss": 0.7081, + "step": 498 + }, + { + "epoch": 0.03413366167316506, + "grad_norm": 3.174894042387177, + "learning_rate": 9.999558243042763e-06, + "loss": 0.5182, + "step": 499 + }, + { + "epoch": 0.03420206580477461, + "grad_norm": 3.037093654972128, + "learning_rate": 9.999543395326583e-06, + "loss": 0.6484, + "step": 500 + }, + { + "epoch": 0.034270469936384156, + "grad_norm": 2.110055230980543, + "learning_rate": 9.999528302208668e-06, + "loss": 0.3962, + "step": 501 + }, + { + "epoch": 0.034338874067993706, + "grad_norm": 2.5062612388244694, + "learning_rate": 9.999512963689763e-06, + "loss": 0.5279, + "step": 502 + }, + { + "epoch": 0.03440727819960326, + "grad_norm": 2.7988892233782203, + "learning_rate": 9.999497379770617e-06, + "loss": 0.3023, + "step": 503 + }, + { + "epoch": 0.03447568233121281, + "grad_norm": 2.2315382856704087, + "learning_rate": 9.999481550451999e-06, + "loss": 0.3795, + "step": 504 + }, + { + "epoch": 0.03454408646282235, + "grad_norm": 2.7850692649836706, + "learning_rate": 9.999465475734685e-06, + "loss": 0.5132, + "step": 505 + }, + { + "epoch": 0.0346124905944319, + "grad_norm": 2.5851427164931096, + "learning_rate": 9.999449155619463e-06, + "loss": 0.3753, + "step": 506 + }, + { + "epoch": 0.03468089472604145, + "grad_norm": 3.4393924532320455, + "learning_rate": 9.999432590107133e-06, + "loss": 0.8243, + "step": 507 + }, + { + "epoch": 0.034749298857651004, + "grad_norm": 2.992353310236355, + "learning_rate": 9.99941577919851e-06, + "loss": 0.5794, + "step": 508 + }, + { + "epoch": 0.034817702989260554, + "grad_norm": 2.302296079900006, + "learning_rate": 9.99939872289442e-06, + "loss": 0.3954, + "step": 509 + }, + { + "epoch": 0.0348861071208701, + "grad_norm": 2.0260105857982094, + "learning_rate": 9.999381421195697e-06, + "loss": 0.2983, + "step": 510 + }, + { + "epoch": 0.03495451125247965, + "grad_norm": 2.03826300341275, + "learning_rate": 9.999363874103194e-06, + "loss": 0.277, + "step": 511 + }, + { + "epoch": 0.0350229153840892, + "grad_norm": 2.3748629179092577, + "learning_rate": 9.999346081617769e-06, + "loss": 0.3168, + "step": 512 + }, + { + "epoch": 0.03509131951569875, + "grad_norm": 3.2327303183182665, + "learning_rate": 9.999328043740296e-06, + "loss": 0.4241, + "step": 513 + }, + { + "epoch": 0.035159723647308294, + "grad_norm": 2.646355657508899, + "learning_rate": 9.999309760471663e-06, + "loss": 0.2505, + "step": 514 + }, + { + "epoch": 0.035228127778917845, + "grad_norm": 3.4596515703476225, + "learning_rate": 9.999291231812765e-06, + "loss": 0.2641, + "step": 515 + }, + { + "epoch": 0.035296531910527396, + "grad_norm": 3.3552837831544315, + "learning_rate": 9.99927245776451e-06, + "loss": 0.5704, + "step": 516 + }, + { + "epoch": 0.035364936042136946, + "grad_norm": 3.2245231261060674, + "learning_rate": 9.999253438327823e-06, + "loss": 0.8433, + "step": 517 + }, + { + "epoch": 0.0354333401737465, + "grad_norm": 2.2686388293265214, + "learning_rate": 9.999234173503636e-06, + "loss": 0.2878, + "step": 518 + }, + { + "epoch": 0.03550174430535604, + "grad_norm": 2.986830006172932, + "learning_rate": 9.999214663292896e-06, + "loss": 0.5959, + "step": 519 + }, + { + "epoch": 0.03557014843696559, + "grad_norm": 2.763118448549523, + "learning_rate": 9.999194907696559e-06, + "loss": 0.5187, + "step": 520 + }, + { + "epoch": 0.03563855256857514, + "grad_norm": 3.5112666432036144, + "learning_rate": 9.999174906715593e-06, + "loss": 0.3753, + "step": 521 + }, + { + "epoch": 0.03570695670018469, + "grad_norm": 3.1531185724743, + "learning_rate": 9.999154660350982e-06, + "loss": 0.4181, + "step": 522 + }, + { + "epoch": 0.035775360831794244, + "grad_norm": 3.283017853927983, + "learning_rate": 9.999134168603721e-06, + "loss": 0.8353, + "step": 523 + }, + { + "epoch": 0.03584376496340379, + "grad_norm": 2.757256568009885, + "learning_rate": 9.999113431474815e-06, + "loss": 0.488, + "step": 524 + }, + { + "epoch": 0.03591216909501334, + "grad_norm": 2.414352035442871, + "learning_rate": 9.99909244896528e-06, + "loss": 0.4905, + "step": 525 + }, + { + "epoch": 0.03598057322662289, + "grad_norm": 4.13071850114862, + "learning_rate": 9.99907122107615e-06, + "loss": 0.821, + "step": 526 + }, + { + "epoch": 0.03604897735823244, + "grad_norm": 2.7526150645771432, + "learning_rate": 9.99904974780846e-06, + "loss": 0.4349, + "step": 527 + }, + { + "epoch": 0.03611738148984198, + "grad_norm": 3.259450450400641, + "learning_rate": 9.99902802916327e-06, + "loss": 0.8564, + "step": 528 + }, + { + "epoch": 0.036185785621451534, + "grad_norm": 2.619671636199153, + "learning_rate": 9.999006065141645e-06, + "loss": 0.2213, + "step": 529 + }, + { + "epoch": 0.036254189753061085, + "grad_norm": 3.0937577401651986, + "learning_rate": 9.998983855744661e-06, + "loss": 0.6947, + "step": 530 + }, + { + "epoch": 0.036322593884670636, + "grad_norm": 2.580064754754653, + "learning_rate": 9.99896140097341e-06, + "loss": 0.4946, + "step": 531 + }, + { + "epoch": 0.036390998016280186, + "grad_norm": 2.647360287240753, + "learning_rate": 9.998938700828995e-06, + "loss": 0.5355, + "step": 532 + }, + { + "epoch": 0.03645940214788973, + "grad_norm": 1.9956000738442015, + "learning_rate": 9.998915755312528e-06, + "loss": 0.2757, + "step": 533 + }, + { + "epoch": 0.03652780627949928, + "grad_norm": 1.9908920189843855, + "learning_rate": 9.998892564425136e-06, + "loss": 0.3738, + "step": 534 + }, + { + "epoch": 0.03659621041110883, + "grad_norm": 2.2733361133109105, + "learning_rate": 9.998869128167957e-06, + "loss": 0.4844, + "step": 535 + }, + { + "epoch": 0.03666461454271838, + "grad_norm": 2.307397484057122, + "learning_rate": 9.998845446542142e-06, + "loss": 0.48, + "step": 536 + }, + { + "epoch": 0.036733018674327926, + "grad_norm": 2.151295546881275, + "learning_rate": 9.998821519548855e-06, + "loss": 0.283, + "step": 537 + }, + { + "epoch": 0.03680142280593748, + "grad_norm": 2.272447429464378, + "learning_rate": 9.998797347189267e-06, + "loss": 0.3994, + "step": 538 + }, + { + "epoch": 0.03686982693754703, + "grad_norm": 2.0659569538489015, + "learning_rate": 9.998772929464567e-06, + "loss": 0.315, + "step": 539 + }, + { + "epoch": 0.03693823106915658, + "grad_norm": 2.7324035185823776, + "learning_rate": 9.998748266375952e-06, + "loss": 0.6205, + "step": 540 + }, + { + "epoch": 0.03700663520076613, + "grad_norm": 2.4751243597393633, + "learning_rate": 9.998723357924635e-06, + "loss": 0.2799, + "step": 541 + }, + { + "epoch": 0.03707503933237567, + "grad_norm": 3.4702842257815103, + "learning_rate": 9.998698204111836e-06, + "loss": 0.7575, + "step": 542 + }, + { + "epoch": 0.03714344346398522, + "grad_norm": 2.57121249503979, + "learning_rate": 9.998672804938791e-06, + "loss": 0.4561, + "step": 543 + }, + { + "epoch": 0.037211847595594774, + "grad_norm": 3.0349831942952403, + "learning_rate": 9.998647160406745e-06, + "loss": 0.6722, + "step": 544 + }, + { + "epoch": 0.037280251727204325, + "grad_norm": 3.374075683542556, + "learning_rate": 9.99862127051696e-06, + "loss": 0.4214, + "step": 545 + }, + { + "epoch": 0.037348655858813876, + "grad_norm": 2.2828094626007656, + "learning_rate": 9.998595135270705e-06, + "loss": 0.3502, + "step": 546 + }, + { + "epoch": 0.03741705999042342, + "grad_norm": 2.5184155855078583, + "learning_rate": 9.998568754669261e-06, + "loss": 0.4058, + "step": 547 + }, + { + "epoch": 0.03748546412203297, + "grad_norm": 2.8893244214574754, + "learning_rate": 9.998542128713926e-06, + "loss": 0.5691, + "step": 548 + }, + { + "epoch": 0.03755386825364252, + "grad_norm": 2.8385683603810627, + "learning_rate": 9.998515257406005e-06, + "loss": 0.3873, + "step": 549 + }, + { + "epoch": 0.03762227238525207, + "grad_norm": 2.5095880170980367, + "learning_rate": 9.998488140746818e-06, + "loss": 0.4893, + "step": 550 + }, + { + "epoch": 0.037690676516861615, + "grad_norm": 2.6982966502792407, + "learning_rate": 9.998460778737695e-06, + "loss": 0.534, + "step": 551 + }, + { + "epoch": 0.037759080648471166, + "grad_norm": 2.7100450217459175, + "learning_rate": 9.998433171379979e-06, + "loss": 0.4553, + "step": 552 + }, + { + "epoch": 0.03782748478008072, + "grad_norm": 3.2383800809018397, + "learning_rate": 9.99840531867503e-06, + "loss": 0.6666, + "step": 553 + }, + { + "epoch": 0.03789588891169027, + "grad_norm": 3.3036823953674874, + "learning_rate": 9.998377220624206e-06, + "loss": 0.6522, + "step": 554 + }, + { + "epoch": 0.03796429304329982, + "grad_norm": 2.976066052345043, + "learning_rate": 9.998348877228894e-06, + "loss": 0.4082, + "step": 555 + }, + { + "epoch": 0.03803269717490936, + "grad_norm": 2.6571736581085856, + "learning_rate": 9.998320288490479e-06, + "loss": 0.4197, + "step": 556 + }, + { + "epoch": 0.03810110130651891, + "grad_norm": 2.961781036881187, + "learning_rate": 9.998291454410372e-06, + "loss": 0.4245, + "step": 557 + }, + { + "epoch": 0.038169505438128463, + "grad_norm": 2.620200873484101, + "learning_rate": 9.998262374989981e-06, + "loss": 0.2994, + "step": 558 + }, + { + "epoch": 0.038237909569738014, + "grad_norm": 2.3525478288448385, + "learning_rate": 9.998233050230737e-06, + "loss": 0.4274, + "step": 559 + }, + { + "epoch": 0.03830631370134756, + "grad_norm": 4.274524222774414, + "learning_rate": 9.998203480134079e-06, + "loss": 0.4827, + "step": 560 + }, + { + "epoch": 0.03837471783295711, + "grad_norm": 2.0130062057503215, + "learning_rate": 9.998173664701456e-06, + "loss": 0.3116, + "step": 561 + }, + { + "epoch": 0.03844312196456666, + "grad_norm": 3.2904788884168337, + "learning_rate": 9.998143603934337e-06, + "loss": 0.3079, + "step": 562 + }, + { + "epoch": 0.03851152609617621, + "grad_norm": 2.863684257675164, + "learning_rate": 9.998113297834191e-06, + "loss": 0.5328, + "step": 563 + }, + { + "epoch": 0.03857993022778576, + "grad_norm": 2.414545220811264, + "learning_rate": 9.998082746402512e-06, + "loss": 0.3098, + "step": 564 + }, + { + "epoch": 0.038648334359395305, + "grad_norm": 2.6388396711667506, + "learning_rate": 9.998051949640795e-06, + "loss": 0.2236, + "step": 565 + }, + { + "epoch": 0.038716738491004855, + "grad_norm": 2.7786709368119067, + "learning_rate": 9.998020907550552e-06, + "loss": 0.6232, + "step": 566 + }, + { + "epoch": 0.038785142622614406, + "grad_norm": 3.0323022215613746, + "learning_rate": 9.997989620133308e-06, + "loss": 0.4689, + "step": 567 + }, + { + "epoch": 0.03885354675422396, + "grad_norm": 2.9002164948978426, + "learning_rate": 9.997958087390598e-06, + "loss": 0.3252, + "step": 568 + }, + { + "epoch": 0.03892195088583351, + "grad_norm": 4.140328738572101, + "learning_rate": 9.997926309323972e-06, + "loss": 1.0776, + "step": 569 + }, + { + "epoch": 0.03899035501744305, + "grad_norm": 2.133246171618605, + "learning_rate": 9.997894285934985e-06, + "loss": 0.3407, + "step": 570 + }, + { + "epoch": 0.0390587591490526, + "grad_norm": 3.3400596504252804, + "learning_rate": 9.997862017225214e-06, + "loss": 0.3322, + "step": 571 + }, + { + "epoch": 0.03912716328066215, + "grad_norm": 2.6280921026794934, + "learning_rate": 9.99782950319624e-06, + "loss": 0.4054, + "step": 572 + }, + { + "epoch": 0.039195567412271703, + "grad_norm": 2.8167003314039825, + "learning_rate": 9.99779674384966e-06, + "loss": 0.5913, + "step": 573 + }, + { + "epoch": 0.03926397154388125, + "grad_norm": 3.599242035890027, + "learning_rate": 9.997763739187083e-06, + "loss": 0.6009, + "step": 574 + }, + { + "epoch": 0.0393323756754908, + "grad_norm": 3.024154755435828, + "learning_rate": 9.997730489210126e-06, + "loss": 0.5536, + "step": 575 + }, + { + "epoch": 0.03940077980710035, + "grad_norm": 2.9119806395538177, + "learning_rate": 9.997696993920422e-06, + "loss": 0.3893, + "step": 576 + }, + { + "epoch": 0.0394691839387099, + "grad_norm": 3.2738778448613557, + "learning_rate": 9.997663253319619e-06, + "loss": 0.6714, + "step": 577 + }, + { + "epoch": 0.03953758807031945, + "grad_norm": 2.942956553880227, + "learning_rate": 9.997629267409367e-06, + "loss": 0.4654, + "step": 578 + }, + { + "epoch": 0.039605992201928994, + "grad_norm": 2.7063062723665654, + "learning_rate": 9.997595036191338e-06, + "loss": 0.2547, + "step": 579 + }, + { + "epoch": 0.039674396333538545, + "grad_norm": 2.7554789388028587, + "learning_rate": 9.997560559667212e-06, + "loss": 0.5287, + "step": 580 + }, + { + "epoch": 0.039742800465148095, + "grad_norm": 3.6030479065587087, + "learning_rate": 9.997525837838681e-06, + "loss": 0.8492, + "step": 581 + }, + { + "epoch": 0.039811204596757646, + "grad_norm": 2.543487624806257, + "learning_rate": 9.997490870707448e-06, + "loss": 0.5553, + "step": 582 + }, + { + "epoch": 0.0398796087283672, + "grad_norm": 2.673079503590658, + "learning_rate": 9.997455658275232e-06, + "loss": 0.5063, + "step": 583 + }, + { + "epoch": 0.03994801285997674, + "grad_norm": 3.3773921153911437, + "learning_rate": 9.99742020054376e-06, + "loss": 0.4997, + "step": 584 + }, + { + "epoch": 0.04001641699158629, + "grad_norm": 2.3222469268123067, + "learning_rate": 9.997384497514771e-06, + "loss": 0.2615, + "step": 585 + }, + { + "epoch": 0.04008482112319584, + "grad_norm": 2.5452690531779347, + "learning_rate": 9.99734854919002e-06, + "loss": 0.4745, + "step": 586 + }, + { + "epoch": 0.04015322525480539, + "grad_norm": 2.4936130536272403, + "learning_rate": 9.997312355571268e-06, + "loss": 0.2558, + "step": 587 + }, + { + "epoch": 0.04022162938641494, + "grad_norm": 2.556537121920867, + "learning_rate": 9.997275916660295e-06, + "loss": 0.5697, + "step": 588 + }, + { + "epoch": 0.04029003351802449, + "grad_norm": 2.7911751489985512, + "learning_rate": 9.997239232458888e-06, + "loss": 0.472, + "step": 589 + }, + { + "epoch": 0.04035843764963404, + "grad_norm": 3.979641132960794, + "learning_rate": 9.997202302968849e-06, + "loss": 0.6806, + "step": 590 + }, + { + "epoch": 0.04042684178124359, + "grad_norm": 2.986236031947915, + "learning_rate": 9.99716512819199e-06, + "loss": 0.5448, + "step": 591 + }, + { + "epoch": 0.04049524591285314, + "grad_norm": 2.776223355061929, + "learning_rate": 9.997127708130134e-06, + "loss": 0.6911, + "step": 592 + }, + { + "epoch": 0.04056365004446268, + "grad_norm": 2.40648412494369, + "learning_rate": 9.99709004278512e-06, + "loss": 0.4274, + "step": 593 + }, + { + "epoch": 0.040632054176072234, + "grad_norm": 3.048839033890537, + "learning_rate": 9.997052132158797e-06, + "loss": 0.3227, + "step": 594 + }, + { + "epoch": 0.040700458307681785, + "grad_norm": 3.56946444299656, + "learning_rate": 9.997013976253022e-06, + "loss": 0.538, + "step": 595 + }, + { + "epoch": 0.040768862439291335, + "grad_norm": 2.94813325623227, + "learning_rate": 9.996975575069675e-06, + "loss": 0.5113, + "step": 596 + }, + { + "epoch": 0.04083726657090088, + "grad_norm": 2.264498094944943, + "learning_rate": 9.996936928610632e-06, + "loss": 0.4166, + "step": 597 + }, + { + "epoch": 0.04090567070251043, + "grad_norm": 2.73507415552398, + "learning_rate": 9.996898036877797e-06, + "loss": 0.606, + "step": 598 + }, + { + "epoch": 0.04097407483411998, + "grad_norm": 3.2205909075599113, + "learning_rate": 9.996858899873076e-06, + "loss": 0.626, + "step": 599 + }, + { + "epoch": 0.04104247896572953, + "grad_norm": 2.9534275969631008, + "learning_rate": 9.996819517598392e-06, + "loss": 0.5126, + "step": 600 + }, + { + "epoch": 0.04111088309733908, + "grad_norm": 3.2629405856888702, + "learning_rate": 9.996779890055675e-06, + "loss": 0.5215, + "step": 601 + }, + { + "epoch": 0.041179287228948626, + "grad_norm": 2.383158911867164, + "learning_rate": 9.996740017246873e-06, + "loss": 0.4916, + "step": 602 + }, + { + "epoch": 0.04124769136055818, + "grad_norm": 2.5124486808849764, + "learning_rate": 9.99669989917394e-06, + "loss": 0.5541, + "step": 603 + }, + { + "epoch": 0.04131609549216773, + "grad_norm": 2.194798976256575, + "learning_rate": 9.99665953583885e-06, + "loss": 0.3672, + "step": 604 + }, + { + "epoch": 0.04138449962377728, + "grad_norm": 2.325852254437846, + "learning_rate": 9.99661892724358e-06, + "loss": 0.238, + "step": 605 + }, + { + "epoch": 0.04145290375538683, + "grad_norm": 2.2719384229168664, + "learning_rate": 9.996578073390124e-06, + "loss": 0.4306, + "step": 606 + }, + { + "epoch": 0.04152130788699637, + "grad_norm": 2.301042720011005, + "learning_rate": 9.996536974280488e-06, + "loss": 0.3472, + "step": 607 + }, + { + "epoch": 0.04158971201860592, + "grad_norm": 3.612586586650408, + "learning_rate": 9.99649562991669e-06, + "loss": 0.6133, + "step": 608 + }, + { + "epoch": 0.041658116150215474, + "grad_norm": 2.7651720376498177, + "learning_rate": 9.996454040300758e-06, + "loss": 0.4095, + "step": 609 + }, + { + "epoch": 0.041726520281825025, + "grad_norm": 2.631246366974051, + "learning_rate": 9.996412205434734e-06, + "loss": 0.5595, + "step": 610 + }, + { + "epoch": 0.04179492441343457, + "grad_norm": 3.418192288309614, + "learning_rate": 9.996370125320672e-06, + "loss": 0.378, + "step": 611 + }, + { + "epoch": 0.04186332854504412, + "grad_norm": 2.7696900692101134, + "learning_rate": 9.996327799960636e-06, + "loss": 0.5094, + "step": 612 + }, + { + "epoch": 0.04193173267665367, + "grad_norm": 3.0955040419430007, + "learning_rate": 9.996285229356705e-06, + "loss": 0.8477, + "step": 613 + }, + { + "epoch": 0.04200013680826322, + "grad_norm": 2.818183914612734, + "learning_rate": 9.99624241351097e-06, + "loss": 0.5499, + "step": 614 + }, + { + "epoch": 0.04206854093987277, + "grad_norm": 2.3383336233078587, + "learning_rate": 9.996199352425528e-06, + "loss": 0.4062, + "step": 615 + }, + { + "epoch": 0.042136945071482315, + "grad_norm": 3.3528389978512823, + "learning_rate": 9.996156046102498e-06, + "loss": 0.8964, + "step": 616 + }, + { + "epoch": 0.042205349203091866, + "grad_norm": 2.8863824279360495, + "learning_rate": 9.996112494544002e-06, + "loss": 0.564, + "step": 617 + }, + { + "epoch": 0.04227375333470142, + "grad_norm": 2.8796617300833396, + "learning_rate": 9.996068697752179e-06, + "loss": 0.7033, + "step": 618 + }, + { + "epoch": 0.04234215746631097, + "grad_norm": 2.395218543631772, + "learning_rate": 9.996024655729177e-06, + "loss": 0.4949, + "step": 619 + }, + { + "epoch": 0.04241056159792051, + "grad_norm": 2.8317573204417394, + "learning_rate": 9.995980368477161e-06, + "loss": 0.5001, + "step": 620 + }, + { + "epoch": 0.04247896572953006, + "grad_norm": 2.218526224483877, + "learning_rate": 9.995935835998303e-06, + "loss": 0.3228, + "step": 621 + }, + { + "epoch": 0.04254736986113961, + "grad_norm": 2.159186968366268, + "learning_rate": 9.995891058294788e-06, + "loss": 0.4118, + "step": 622 + }, + { + "epoch": 0.04261577399274916, + "grad_norm": 2.3670233080987124, + "learning_rate": 9.995846035368815e-06, + "loss": 0.2063, + "step": 623 + }, + { + "epoch": 0.042684178124358714, + "grad_norm": 2.8945008506671512, + "learning_rate": 9.995800767222596e-06, + "loss": 0.4867, + "step": 624 + }, + { + "epoch": 0.04275258225596826, + "grad_norm": 3.0508162414654123, + "learning_rate": 9.995755253858348e-06, + "loss": 0.8304, + "step": 625 + }, + { + "epoch": 0.04282098638757781, + "grad_norm": 2.8033134526809538, + "learning_rate": 9.995709495278309e-06, + "loss": 0.2748, + "step": 626 + }, + { + "epoch": 0.04288939051918736, + "grad_norm": 3.097725423687516, + "learning_rate": 9.995663491484725e-06, + "loss": 0.5737, + "step": 627 + }, + { + "epoch": 0.04295779465079691, + "grad_norm": 3.2344539006141693, + "learning_rate": 9.995617242479852e-06, + "loss": 0.4775, + "step": 628 + }, + { + "epoch": 0.04302619878240646, + "grad_norm": 3.035171420379785, + "learning_rate": 9.995570748265962e-06, + "loss": 0.8025, + "step": 629 + }, + { + "epoch": 0.043094602914016004, + "grad_norm": 2.4669068675161054, + "learning_rate": 9.995524008845335e-06, + "loss": 0.5538, + "step": 630 + }, + { + "epoch": 0.043163007045625555, + "grad_norm": 2.4630473462911797, + "learning_rate": 9.995477024220268e-06, + "loss": 0.3589, + "step": 631 + }, + { + "epoch": 0.043231411177235106, + "grad_norm": 2.3204551036439365, + "learning_rate": 9.995429794393063e-06, + "loss": 0.3302, + "step": 632 + }, + { + "epoch": 0.04329981530884466, + "grad_norm": 3.2084931214080354, + "learning_rate": 9.995382319366044e-06, + "loss": 0.3595, + "step": 633 + }, + { + "epoch": 0.0433682194404542, + "grad_norm": 2.3007884778360976, + "learning_rate": 9.995334599141537e-06, + "loss": 0.4638, + "step": 634 + }, + { + "epoch": 0.04343662357206375, + "grad_norm": 2.180980032525098, + "learning_rate": 9.995286633721886e-06, + "loss": 0.329, + "step": 635 + }, + { + "epoch": 0.0435050277036733, + "grad_norm": 3.5162182455734374, + "learning_rate": 9.995238423109445e-06, + "loss": 0.5359, + "step": 636 + }, + { + "epoch": 0.04357343183528285, + "grad_norm": 2.928198509759213, + "learning_rate": 9.99518996730658e-06, + "loss": 0.6009, + "step": 637 + }, + { + "epoch": 0.0436418359668924, + "grad_norm": 3.0296130627148563, + "learning_rate": 9.99514126631567e-06, + "loss": 0.7303, + "step": 638 + }, + { + "epoch": 0.04371024009850195, + "grad_norm": 2.555734207071128, + "learning_rate": 9.995092320139106e-06, + "loss": 0.3533, + "step": 639 + }, + { + "epoch": 0.0437786442301115, + "grad_norm": 2.1888398290704445, + "learning_rate": 9.995043128779289e-06, + "loss": 0.3008, + "step": 640 + }, + { + "epoch": 0.04384704836172105, + "grad_norm": 3.9124688503743825, + "learning_rate": 9.994993692238634e-06, + "loss": 0.5021, + "step": 641 + }, + { + "epoch": 0.0439154524933306, + "grad_norm": 3.5969110270004214, + "learning_rate": 9.99494401051957e-06, + "loss": 0.8237, + "step": 642 + }, + { + "epoch": 0.04398385662494014, + "grad_norm": 2.506234144698145, + "learning_rate": 9.994894083624534e-06, + "loss": 0.4956, + "step": 643 + }, + { + "epoch": 0.044052260756549694, + "grad_norm": 2.773741503965097, + "learning_rate": 9.994843911555974e-06, + "loss": 0.3327, + "step": 644 + }, + { + "epoch": 0.044120664888159244, + "grad_norm": 2.394587343439484, + "learning_rate": 9.994793494316355e-06, + "loss": 0.3665, + "step": 645 + }, + { + "epoch": 0.044189069019768795, + "grad_norm": 2.5265861743674196, + "learning_rate": 9.994742831908153e-06, + "loss": 0.354, + "step": 646 + }, + { + "epoch": 0.044257473151378346, + "grad_norm": 2.9795803880968146, + "learning_rate": 9.994691924333853e-06, + "loss": 0.3726, + "step": 647 + }, + { + "epoch": 0.04432587728298789, + "grad_norm": 77.439673047676, + "learning_rate": 9.994640771595954e-06, + "loss": 0.7492, + "step": 648 + }, + { + "epoch": 0.04439428141459744, + "grad_norm": 2.6141084864208857, + "learning_rate": 9.994589373696966e-06, + "loss": 0.4597, + "step": 649 + }, + { + "epoch": 0.04446268554620699, + "grad_norm": 2.16582042340796, + "learning_rate": 9.994537730639416e-06, + "loss": 0.3088, + "step": 650 + }, + { + "epoch": 0.04453108967781654, + "grad_norm": 2.4609602215641377, + "learning_rate": 9.994485842425833e-06, + "loss": 0.3357, + "step": 651 + }, + { + "epoch": 0.04459949380942609, + "grad_norm": 2.1098056141750883, + "learning_rate": 9.994433709058767e-06, + "loss": 0.4539, + "step": 652 + }, + { + "epoch": 0.044667897941035636, + "grad_norm": 2.380594640945342, + "learning_rate": 9.994381330540777e-06, + "loss": 0.1808, + "step": 653 + }, + { + "epoch": 0.04473630207264519, + "grad_norm": 2.621319230979525, + "learning_rate": 9.994328706874434e-06, + "loss": 0.4131, + "step": 654 + }, + { + "epoch": 0.04480470620425474, + "grad_norm": 2.2953718482876018, + "learning_rate": 9.994275838062321e-06, + "loss": 0.4972, + "step": 655 + }, + { + "epoch": 0.04487311033586429, + "grad_norm": 2.2364972046481357, + "learning_rate": 9.994222724107032e-06, + "loss": 0.4803, + "step": 656 + }, + { + "epoch": 0.04494151446747383, + "grad_norm": 2.671900152344576, + "learning_rate": 9.994169365011176e-06, + "loss": 0.3094, + "step": 657 + }, + { + "epoch": 0.04500991859908338, + "grad_norm": 2.761624416914664, + "learning_rate": 9.99411576077737e-06, + "loss": 0.5746, + "step": 658 + }, + { + "epoch": 0.045078322730692934, + "grad_norm": 1.9179726579241747, + "learning_rate": 9.994061911408245e-06, + "loss": 0.3432, + "step": 659 + }, + { + "epoch": 0.045146726862302484, + "grad_norm": 2.343648647660058, + "learning_rate": 9.994007816906449e-06, + "loss": 0.3504, + "step": 660 + }, + { + "epoch": 0.045215130993912035, + "grad_norm": 2.6259365227690763, + "learning_rate": 9.99395347727463e-06, + "loss": 0.5878, + "step": 661 + }, + { + "epoch": 0.04528353512552158, + "grad_norm": 1.987978205791407, + "learning_rate": 9.99389889251546e-06, + "loss": 0.248, + "step": 662 + }, + { + "epoch": 0.04535193925713113, + "grad_norm": 2.3602937652620137, + "learning_rate": 9.993844062631616e-06, + "loss": 0.3261, + "step": 663 + }, + { + "epoch": 0.04542034338874068, + "grad_norm": 2.8301695025048508, + "learning_rate": 9.993788987625793e-06, + "loss": 0.359, + "step": 664 + }, + { + "epoch": 0.04548874752035023, + "grad_norm": 2.242038196816417, + "learning_rate": 9.99373366750069e-06, + "loss": 0.146, + "step": 665 + }, + { + "epoch": 0.045557151651959775, + "grad_norm": 3.1478773428082616, + "learning_rate": 9.993678102259024e-06, + "loss": 0.642, + "step": 666 + }, + { + "epoch": 0.045625555783569326, + "grad_norm": 2.236559193763139, + "learning_rate": 9.993622291903522e-06, + "loss": 0.2971, + "step": 667 + }, + { + "epoch": 0.045693959915178876, + "grad_norm": 3.1090123717514953, + "learning_rate": 9.993566236436925e-06, + "loss": 0.6911, + "step": 668 + }, + { + "epoch": 0.04576236404678843, + "grad_norm": 2.460662426788662, + "learning_rate": 9.993509935861983e-06, + "loss": 0.3352, + "step": 669 + }, + { + "epoch": 0.04583076817839798, + "grad_norm": 3.5008031996362585, + "learning_rate": 9.99345339018146e-06, + "loss": 0.9674, + "step": 670 + }, + { + "epoch": 0.04589917231000752, + "grad_norm": 2.7057948846457514, + "learning_rate": 9.993396599398133e-06, + "loss": 0.3806, + "step": 671 + }, + { + "epoch": 0.04596757644161707, + "grad_norm": 2.178913628533892, + "learning_rate": 9.993339563514785e-06, + "loss": 0.4046, + "step": 672 + }, + { + "epoch": 0.04603598057322662, + "grad_norm": 2.020270465021151, + "learning_rate": 9.993282282534222e-06, + "loss": 0.327, + "step": 673 + }, + { + "epoch": 0.046104384704836174, + "grad_norm": 3.6861103882453925, + "learning_rate": 9.99322475645925e-06, + "loss": 0.7445, + "step": 674 + }, + { + "epoch": 0.046172788836445725, + "grad_norm": 3.0652691330293536, + "learning_rate": 9.993166985292695e-06, + "loss": 0.6326, + "step": 675 + }, + { + "epoch": 0.04624119296805527, + "grad_norm": 2.3381368814924643, + "learning_rate": 9.993108969037392e-06, + "loss": 0.3128, + "step": 676 + }, + { + "epoch": 0.04630959709966482, + "grad_norm": 2.937616579299129, + "learning_rate": 9.993050707696191e-06, + "loss": 0.7199, + "step": 677 + }, + { + "epoch": 0.04637800123127437, + "grad_norm": 2.7850757977822287, + "learning_rate": 9.992992201271949e-06, + "loss": 0.406, + "step": 678 + }, + { + "epoch": 0.04644640536288392, + "grad_norm": 2.4071830946767965, + "learning_rate": 9.992933449767538e-06, + "loss": 0.6542, + "step": 679 + }, + { + "epoch": 0.046514809494493464, + "grad_norm": 2.0712216666199685, + "learning_rate": 9.992874453185845e-06, + "loss": 0.3555, + "step": 680 + }, + { + "epoch": 0.046583213626103015, + "grad_norm": 2.885336972362876, + "learning_rate": 9.992815211529763e-06, + "loss": 0.5226, + "step": 681 + }, + { + "epoch": 0.046651617757712566, + "grad_norm": 2.561879748781842, + "learning_rate": 9.9927557248022e-06, + "loss": 0.3612, + "step": 682 + }, + { + "epoch": 0.046720021889322116, + "grad_norm": 2.6925583464995855, + "learning_rate": 9.992695993006077e-06, + "loss": 0.4052, + "step": 683 + }, + { + "epoch": 0.04678842602093167, + "grad_norm": 2.20561965091021, + "learning_rate": 9.992636016144324e-06, + "loss": 0.4369, + "step": 684 + }, + { + "epoch": 0.04685683015254121, + "grad_norm": 2.4699892543161295, + "learning_rate": 9.992575794219885e-06, + "loss": 0.4462, + "step": 685 + }, + { + "epoch": 0.04692523428415076, + "grad_norm": 2.39179667264597, + "learning_rate": 9.992515327235718e-06, + "loss": 0.4878, + "step": 686 + }, + { + "epoch": 0.04699363841576031, + "grad_norm": 2.0006735904183026, + "learning_rate": 9.99245461519479e-06, + "loss": 0.2515, + "step": 687 + }, + { + "epoch": 0.04706204254736986, + "grad_norm": 2.703452658226581, + "learning_rate": 9.992393658100084e-06, + "loss": 0.5579, + "step": 688 + }, + { + "epoch": 0.047130446678979414, + "grad_norm": 2.627639034305805, + "learning_rate": 9.992332455954585e-06, + "loss": 0.5142, + "step": 689 + }, + { + "epoch": 0.04719885081058896, + "grad_norm": 2.2462994555939466, + "learning_rate": 9.992271008761304e-06, + "loss": 0.5269, + "step": 690 + }, + { + "epoch": 0.04726725494219851, + "grad_norm": 2.089011477587009, + "learning_rate": 9.992209316523253e-06, + "loss": 0.3665, + "step": 691 + }, + { + "epoch": 0.04733565907380806, + "grad_norm": 2.247532368900006, + "learning_rate": 9.992147379243463e-06, + "loss": 0.4374, + "step": 692 + }, + { + "epoch": 0.04740406320541761, + "grad_norm": 3.111987463755167, + "learning_rate": 9.992085196924972e-06, + "loss": 0.4248, + "step": 693 + }, + { + "epoch": 0.047472467337027154, + "grad_norm": 1.872581241077805, + "learning_rate": 9.992022769570834e-06, + "loss": 0.3093, + "step": 694 + }, + { + "epoch": 0.047540871468636704, + "grad_norm": 2.4891549676935876, + "learning_rate": 9.991960097184111e-06, + "loss": 0.4321, + "step": 695 + }, + { + "epoch": 0.047609275600246255, + "grad_norm": 2.2152716477295007, + "learning_rate": 9.991897179767881e-06, + "loss": 0.4684, + "step": 696 + }, + { + "epoch": 0.047677679731855806, + "grad_norm": 2.4562729118579245, + "learning_rate": 9.991834017325232e-06, + "loss": 0.4613, + "step": 697 + }, + { + "epoch": 0.047746083863465356, + "grad_norm": 2.8343595207826313, + "learning_rate": 9.991770609859264e-06, + "loss": 0.4355, + "step": 698 + }, + { + "epoch": 0.0478144879950749, + "grad_norm": 2.2997091220835033, + "learning_rate": 9.991706957373088e-06, + "loss": 0.5773, + "step": 699 + }, + { + "epoch": 0.04788289212668445, + "grad_norm": 2.4341599186992346, + "learning_rate": 9.99164305986983e-06, + "loss": 0.4633, + "step": 700 + }, + { + "epoch": 0.047951296258294, + "grad_norm": 2.451043916683134, + "learning_rate": 9.991578917352629e-06, + "loss": 0.6135, + "step": 701 + }, + { + "epoch": 0.04801970038990355, + "grad_norm": 1.8987344325151325, + "learning_rate": 9.991514529824629e-06, + "loss": 0.2861, + "step": 702 + }, + { + "epoch": 0.048088104521513096, + "grad_norm": 3.1873669596564325, + "learning_rate": 9.991449897288992e-06, + "loss": 0.8021, + "step": 703 + }, + { + "epoch": 0.04815650865312265, + "grad_norm": 1.41403324073216, + "learning_rate": 9.991385019748891e-06, + "loss": 0.1861, + "step": 704 + }, + { + "epoch": 0.0482249127847322, + "grad_norm": 2.6940323152876635, + "learning_rate": 9.99131989720751e-06, + "loss": 0.4305, + "step": 705 + }, + { + "epoch": 0.04829331691634175, + "grad_norm": 2.6421511207119934, + "learning_rate": 9.991254529668045e-06, + "loss": 0.383, + "step": 706 + }, + { + "epoch": 0.0483617210479513, + "grad_norm": 2.578921622687247, + "learning_rate": 9.991188917133705e-06, + "loss": 0.5983, + "step": 707 + }, + { + "epoch": 0.04843012517956084, + "grad_norm": 2.7748531972871557, + "learning_rate": 9.991123059607712e-06, + "loss": 0.5918, + "step": 708 + }, + { + "epoch": 0.048498529311170394, + "grad_norm": 1.8303020181562535, + "learning_rate": 9.991056957093296e-06, + "loss": 0.2221, + "step": 709 + }, + { + "epoch": 0.048566933442779944, + "grad_norm": 1.5931613854286044, + "learning_rate": 9.990990609593702e-06, + "loss": 0.3211, + "step": 710 + }, + { + "epoch": 0.048635337574389495, + "grad_norm": 1.3722136778952645, + "learning_rate": 9.990924017112192e-06, + "loss": 0.2018, + "step": 711 + }, + { + "epoch": 0.048703741705999046, + "grad_norm": 2.7610090495814066, + "learning_rate": 9.990857179652026e-06, + "loss": 0.5378, + "step": 712 + }, + { + "epoch": 0.04877214583760859, + "grad_norm": 2.3910887960185585, + "learning_rate": 9.99079009721649e-06, + "loss": 0.305, + "step": 713 + }, + { + "epoch": 0.04884054996921814, + "grad_norm": 2.540549303642118, + "learning_rate": 9.990722769808876e-06, + "loss": 0.6376, + "step": 714 + }, + { + "epoch": 0.04890895410082769, + "grad_norm": 2.5770040907481317, + "learning_rate": 9.99065519743249e-06, + "loss": 0.4524, + "step": 715 + }, + { + "epoch": 0.04897735823243724, + "grad_norm": 2.192526016146244, + "learning_rate": 9.990587380090646e-06, + "loss": 0.3743, + "step": 716 + }, + { + "epoch": 0.049045762364046785, + "grad_norm": 2.8858177910984826, + "learning_rate": 9.990519317786673e-06, + "loss": 0.387, + "step": 717 + }, + { + "epoch": 0.049114166495656336, + "grad_norm": 2.714700373973337, + "learning_rate": 9.990451010523913e-06, + "loss": 0.5199, + "step": 718 + }, + { + "epoch": 0.04918257062726589, + "grad_norm": 2.3115113729042016, + "learning_rate": 9.99038245830572e-06, + "loss": 0.2825, + "step": 719 + }, + { + "epoch": 0.04925097475887544, + "grad_norm": 2.4959692409068954, + "learning_rate": 9.990313661135459e-06, + "loss": 0.3625, + "step": 720 + }, + { + "epoch": 0.04931937889048499, + "grad_norm": 2.5614930884439095, + "learning_rate": 9.990244619016504e-06, + "loss": 0.5881, + "step": 721 + }, + { + "epoch": 0.04938778302209453, + "grad_norm": 2.396809258237056, + "learning_rate": 9.990175331952246e-06, + "loss": 0.3711, + "step": 722 + }, + { + "epoch": 0.04945618715370408, + "grad_norm": 2.8918421888994317, + "learning_rate": 9.990105799946084e-06, + "loss": 0.629, + "step": 723 + }, + { + "epoch": 0.049524591285313634, + "grad_norm": 2.6349744098984242, + "learning_rate": 9.990036023001433e-06, + "loss": 0.3662, + "step": 724 + }, + { + "epoch": 0.049592995416923184, + "grad_norm": 1.8653704249111125, + "learning_rate": 9.989966001121719e-06, + "loss": 0.2978, + "step": 725 + }, + { + "epoch": 0.04966139954853273, + "grad_norm": 2.3294231296773713, + "learning_rate": 9.989895734310376e-06, + "loss": 0.3409, + "step": 726 + }, + { + "epoch": 0.04972980368014228, + "grad_norm": 2.695817286450022, + "learning_rate": 9.989825222570852e-06, + "loss": 0.6138, + "step": 727 + }, + { + "epoch": 0.04979820781175183, + "grad_norm": 1.957626386482116, + "learning_rate": 9.989754465906613e-06, + "loss": 0.2175, + "step": 728 + }, + { + "epoch": 0.04986661194336138, + "grad_norm": 2.5346130764917785, + "learning_rate": 9.98968346432113e-06, + "loss": 0.4453, + "step": 729 + }, + { + "epoch": 0.04993501607497093, + "grad_norm": 1.7770963224120413, + "learning_rate": 9.989612217817886e-06, + "loss": 0.3539, + "step": 730 + }, + { + "epoch": 0.050003420206580475, + "grad_norm": 3.0927144542739304, + "learning_rate": 9.989540726400383e-06, + "loss": 0.4257, + "step": 731 + }, + { + "epoch": 0.050071824338190025, + "grad_norm": 3.0179322266448567, + "learning_rate": 9.989468990072123e-06, + "loss": 0.518, + "step": 732 + }, + { + "epoch": 0.050140228469799576, + "grad_norm": 2.4238084028426825, + "learning_rate": 9.989397008836632e-06, + "loss": 0.3067, + "step": 733 + }, + { + "epoch": 0.05020863260140913, + "grad_norm": 2.465734366250469, + "learning_rate": 9.989324782697444e-06, + "loss": 0.429, + "step": 734 + }, + { + "epoch": 0.05027703673301868, + "grad_norm": 2.827155141871908, + "learning_rate": 9.9892523116581e-06, + "loss": 0.5432, + "step": 735 + }, + { + "epoch": 0.05034544086462822, + "grad_norm": 2.7819262389685977, + "learning_rate": 9.989179595722163e-06, + "loss": 0.5961, + "step": 736 + }, + { + "epoch": 0.05041384499623777, + "grad_norm": 2.3817949342345943, + "learning_rate": 9.989106634893195e-06, + "loss": 0.5276, + "step": 737 + }, + { + "epoch": 0.05048224912784732, + "grad_norm": 1.939319970975725, + "learning_rate": 9.989033429174783e-06, + "loss": 0.3514, + "step": 738 + }, + { + "epoch": 0.050550653259456874, + "grad_norm": 2.9038731279349483, + "learning_rate": 9.98895997857052e-06, + "loss": 0.7627, + "step": 739 + }, + { + "epoch": 0.05061905739106642, + "grad_norm": 1.8759461479362605, + "learning_rate": 9.988886283084008e-06, + "loss": 0.3578, + "step": 740 + }, + { + "epoch": 0.05068746152267597, + "grad_norm": 2.901276711896071, + "learning_rate": 9.988812342718868e-06, + "loss": 0.903, + "step": 741 + }, + { + "epoch": 0.05075586565428552, + "grad_norm": 2.4568655507020747, + "learning_rate": 9.988738157478726e-06, + "loss": 0.39, + "step": 742 + }, + { + "epoch": 0.05082426978589507, + "grad_norm": 2.520447509706398, + "learning_rate": 9.988663727367224e-06, + "loss": 0.4957, + "step": 743 + }, + { + "epoch": 0.05089267391750462, + "grad_norm": 1.4480306595644827, + "learning_rate": 9.988589052388018e-06, + "loss": 0.1147, + "step": 744 + }, + { + "epoch": 0.050961078049114164, + "grad_norm": 1.8227452815189278, + "learning_rate": 9.98851413254477e-06, + "loss": 0.2682, + "step": 745 + }, + { + "epoch": 0.051029482180723715, + "grad_norm": 3.243068368586012, + "learning_rate": 9.988438967841162e-06, + "loss": 0.7346, + "step": 746 + }, + { + "epoch": 0.051097886312333266, + "grad_norm": 2.48420353063812, + "learning_rate": 9.988363558280877e-06, + "loss": 0.481, + "step": 747 + }, + { + "epoch": 0.051166290443942816, + "grad_norm": 2.737312020201155, + "learning_rate": 9.988287903867621e-06, + "loss": 0.6253, + "step": 748 + }, + { + "epoch": 0.05123469457555236, + "grad_norm": 2.2559065248260164, + "learning_rate": 9.988212004605106e-06, + "loss": 0.3267, + "step": 749 + }, + { + "epoch": 0.05130309870716191, + "grad_norm": 2.4478752673821544, + "learning_rate": 9.98813586049706e-06, + "loss": 0.4306, + "step": 750 + }, + { + "epoch": 0.05137150283877146, + "grad_norm": 2.72511732500692, + "learning_rate": 9.988059471547217e-06, + "loss": 0.5038, + "step": 751 + }, + { + "epoch": 0.05143990697038101, + "grad_norm": 2.0394130566220463, + "learning_rate": 9.98798283775933e-06, + "loss": 0.423, + "step": 752 + }, + { + "epoch": 0.05150831110199056, + "grad_norm": 2.409584323780169, + "learning_rate": 9.987905959137156e-06, + "loss": 0.4477, + "step": 753 + }, + { + "epoch": 0.05157671523360011, + "grad_norm": 2.6349878691080972, + "learning_rate": 9.987828835684472e-06, + "loss": 0.7607, + "step": 754 + }, + { + "epoch": 0.05164511936520966, + "grad_norm": 2.1071925140099443, + "learning_rate": 9.987751467405063e-06, + "loss": 0.3467, + "step": 755 + }, + { + "epoch": 0.05171352349681921, + "grad_norm": 2.85542160600604, + "learning_rate": 9.987673854302728e-06, + "loss": 0.6174, + "step": 756 + }, + { + "epoch": 0.05178192762842876, + "grad_norm": 2.9881257733557667, + "learning_rate": 9.987595996381272e-06, + "loss": 0.3889, + "step": 757 + }, + { + "epoch": 0.05185033176003831, + "grad_norm": 3.291177056312309, + "learning_rate": 9.987517893644522e-06, + "loss": 0.5497, + "step": 758 + }, + { + "epoch": 0.05191873589164785, + "grad_norm": 2.06907615453158, + "learning_rate": 9.987439546096309e-06, + "loss": 0.3001, + "step": 759 + }, + { + "epoch": 0.051987140023257404, + "grad_norm": 1.9762565768070317, + "learning_rate": 9.987360953740477e-06, + "loss": 0.2643, + "step": 760 + }, + { + "epoch": 0.052055544154866955, + "grad_norm": 2.6735072587775988, + "learning_rate": 9.987282116580888e-06, + "loss": 0.488, + "step": 761 + }, + { + "epoch": 0.052123948286476506, + "grad_norm": 2.293378764054159, + "learning_rate": 9.987203034621408e-06, + "loss": 0.4537, + "step": 762 + }, + { + "epoch": 0.05219235241808605, + "grad_norm": 2.047060555863476, + "learning_rate": 9.98712370786592e-06, + "loss": 0.2842, + "step": 763 + }, + { + "epoch": 0.0522607565496956, + "grad_norm": 1.9600886183102078, + "learning_rate": 9.987044136318318e-06, + "loss": 0.2664, + "step": 764 + }, + { + "epoch": 0.05232916068130515, + "grad_norm": 2.182754259691024, + "learning_rate": 9.98696431998251e-06, + "loss": 0.2499, + "step": 765 + }, + { + "epoch": 0.0523975648129147, + "grad_norm": 2.634885006201394, + "learning_rate": 9.986884258862409e-06, + "loss": 0.5682, + "step": 766 + }, + { + "epoch": 0.05246596894452425, + "grad_norm": 2.9250584360137935, + "learning_rate": 9.986803952961946e-06, + "loss": 0.6357, + "step": 767 + }, + { + "epoch": 0.052534373076133796, + "grad_norm": 2.0964572150608936, + "learning_rate": 9.986723402285066e-06, + "loss": 0.227, + "step": 768 + }, + { + "epoch": 0.05260277720774335, + "grad_norm": 2.0156558869826804, + "learning_rate": 9.98664260683572e-06, + "loss": 0.2949, + "step": 769 + }, + { + "epoch": 0.0526711813393529, + "grad_norm": 1.4929194752153676, + "learning_rate": 9.986561566617876e-06, + "loss": 0.2149, + "step": 770 + }, + { + "epoch": 0.05273958547096245, + "grad_norm": 2.2379553525982168, + "learning_rate": 9.98648028163551e-06, + "loss": 0.2732, + "step": 771 + }, + { + "epoch": 0.05280798960257199, + "grad_norm": 2.8390571102916256, + "learning_rate": 9.98639875189261e-06, + "loss": 0.5031, + "step": 772 + }, + { + "epoch": 0.05287639373418154, + "grad_norm": 162.95901353083352, + "learning_rate": 9.986316977393183e-06, + "loss": 0.6771, + "step": 773 + }, + { + "epoch": 0.05294479786579109, + "grad_norm": 2.585853281407233, + "learning_rate": 9.986234958141236e-06, + "loss": 0.4519, + "step": 774 + }, + { + "epoch": 0.053013201997400644, + "grad_norm": 2.272055130262403, + "learning_rate": 9.986152694140803e-06, + "loss": 0.4857, + "step": 775 + }, + { + "epoch": 0.053081606129010195, + "grad_norm": 3.1188697195677664, + "learning_rate": 9.986070185395917e-06, + "loss": 0.3322, + "step": 776 + }, + { + "epoch": 0.05315001026061974, + "grad_norm": 14.273416921081909, + "learning_rate": 9.98598743191063e-06, + "loss": 0.6058, + "step": 777 + }, + { + "epoch": 0.05321841439222929, + "grad_norm": 90.32843831537028, + "learning_rate": 9.985904433689e-06, + "loss": 0.6931, + "step": 778 + }, + { + "epoch": 0.05328681852383884, + "grad_norm": 2.9038720605873194, + "learning_rate": 9.985821190735104e-06, + "loss": 0.7252, + "step": 779 + }, + { + "epoch": 0.05335522265544839, + "grad_norm": 2.8529855808986078, + "learning_rate": 9.98573770305303e-06, + "loss": 0.3676, + "step": 780 + }, + { + "epoch": 0.05342362678705794, + "grad_norm": 9.161885945351129, + "learning_rate": 9.985653970646872e-06, + "loss": 0.3736, + "step": 781 + }, + { + "epoch": 0.053492030918667485, + "grad_norm": 2.194454833316623, + "learning_rate": 9.985569993520741e-06, + "loss": 0.411, + "step": 782 + }, + { + "epoch": 0.053560435050277036, + "grad_norm": 2.457699052698779, + "learning_rate": 9.985485771678762e-06, + "loss": 0.6232, + "step": 783 + }, + { + "epoch": 0.05362883918188659, + "grad_norm": 3.189070389554228, + "learning_rate": 9.985401305125065e-06, + "loss": 0.6159, + "step": 784 + }, + { + "epoch": 0.05369724331349614, + "grad_norm": 2.2224152472891996, + "learning_rate": 9.985316593863796e-06, + "loss": 0.4199, + "step": 785 + }, + { + "epoch": 0.05376564744510568, + "grad_norm": 2.4284306185336084, + "learning_rate": 9.985231637899117e-06, + "loss": 0.4189, + "step": 786 + }, + { + "epoch": 0.05383405157671523, + "grad_norm": 2.197259487567383, + "learning_rate": 9.985146437235197e-06, + "loss": 0.539, + "step": 787 + }, + { + "epoch": 0.05390245570832478, + "grad_norm": 2.463506090795793, + "learning_rate": 9.985060991876214e-06, + "loss": 0.408, + "step": 788 + }, + { + "epoch": 0.05397085983993433, + "grad_norm": 2.0385982525716497, + "learning_rate": 9.984975301826367e-06, + "loss": 0.4012, + "step": 789 + }, + { + "epoch": 0.054039263971543884, + "grad_norm": 3.136121776205271, + "learning_rate": 9.98488936708986e-06, + "loss": 0.6939, + "step": 790 + }, + { + "epoch": 0.05410766810315343, + "grad_norm": 2.0265186593307067, + "learning_rate": 9.984803187670908e-06, + "loss": 0.3778, + "step": 791 + }, + { + "epoch": 0.05417607223476298, + "grad_norm": 2.7865705550700772, + "learning_rate": 9.984716763573747e-06, + "loss": 0.6609, + "step": 792 + }, + { + "epoch": 0.05424447636637253, + "grad_norm": 2.0149837405327324, + "learning_rate": 9.984630094802617e-06, + "loss": 0.3446, + "step": 793 + }, + { + "epoch": 0.05431288049798208, + "grad_norm": 2.6307614654469527, + "learning_rate": 9.98454318136177e-06, + "loss": 0.6273, + "step": 794 + }, + { + "epoch": 0.05438128462959163, + "grad_norm": 3.3954050719434226, + "learning_rate": 9.984456023255473e-06, + "loss": 0.5424, + "step": 795 + }, + { + "epoch": 0.054449688761201175, + "grad_norm": 2.7899467605207766, + "learning_rate": 9.984368620488005e-06, + "loss": 0.6975, + "step": 796 + }, + { + "epoch": 0.054518092892810725, + "grad_norm": 2.7340147217048627, + "learning_rate": 9.984280973063656e-06, + "loss": 0.4965, + "step": 797 + }, + { + "epoch": 0.054586497024420276, + "grad_norm": 2.159237819874233, + "learning_rate": 9.984193080986729e-06, + "loss": 0.3692, + "step": 798 + }, + { + "epoch": 0.05465490115602983, + "grad_norm": 1.8709534690051823, + "learning_rate": 9.984104944261536e-06, + "loss": 0.3756, + "step": 799 + }, + { + "epoch": 0.05472330528763937, + "grad_norm": 2.725538586185041, + "learning_rate": 9.984016562892404e-06, + "loss": 0.3718, + "step": 800 + }, + { + "epoch": 0.05479170941924892, + "grad_norm": 1.9853872248143476, + "learning_rate": 9.983927936883673e-06, + "loss": 0.4354, + "step": 801 + }, + { + "epoch": 0.05486011355085847, + "grad_norm": 2.7453207498946446, + "learning_rate": 9.983839066239692e-06, + "loss": 0.6639, + "step": 802 + }, + { + "epoch": 0.05492851768246802, + "grad_norm": 2.8104338924769867, + "learning_rate": 9.983749950964821e-06, + "loss": 0.5508, + "step": 803 + }, + { + "epoch": 0.05499692181407757, + "grad_norm": 2.5538394582001507, + "learning_rate": 9.983660591063437e-06, + "loss": 0.391, + "step": 804 + }, + { + "epoch": 0.05506532594568712, + "grad_norm": 2.6075046711895395, + "learning_rate": 9.983570986539924e-06, + "loss": 0.478, + "step": 805 + }, + { + "epoch": 0.05513373007729667, + "grad_norm": 3.346116324951345, + "learning_rate": 9.983481137398683e-06, + "loss": 0.2269, + "step": 806 + }, + { + "epoch": 0.05520213420890622, + "grad_norm": 2.674969827178773, + "learning_rate": 9.983391043644123e-06, + "loss": 0.6751, + "step": 807 + }, + { + "epoch": 0.05527053834051577, + "grad_norm": 6.366397663578314, + "learning_rate": 9.983300705280665e-06, + "loss": 0.3584, + "step": 808 + }, + { + "epoch": 0.05533894247212531, + "grad_norm": 2.339915412989942, + "learning_rate": 9.983210122312745e-06, + "loss": 0.373, + "step": 809 + }, + { + "epoch": 0.055407346603734864, + "grad_norm": 2.4422770907808715, + "learning_rate": 9.983119294744809e-06, + "loss": 0.4731, + "step": 810 + }, + { + "epoch": 0.055475750735344415, + "grad_norm": 2.496590214221062, + "learning_rate": 9.983028222581315e-06, + "loss": 0.4656, + "step": 811 + }, + { + "epoch": 0.055544154866953965, + "grad_norm": 2.852359514028327, + "learning_rate": 9.982936905826733e-06, + "loss": 0.7777, + "step": 812 + }, + { + "epoch": 0.055612558998563516, + "grad_norm": 4.243843556245071, + "learning_rate": 9.982845344485544e-06, + "loss": 0.4172, + "step": 813 + }, + { + "epoch": 0.05568096313017306, + "grad_norm": 2.426962121234296, + "learning_rate": 9.982753538562244e-06, + "loss": 0.2729, + "step": 814 + }, + { + "epoch": 0.05574936726178261, + "grad_norm": 2.0760990248465463, + "learning_rate": 9.982661488061338e-06, + "loss": 0.3998, + "step": 815 + }, + { + "epoch": 0.05581777139339216, + "grad_norm": 2.651739206085146, + "learning_rate": 9.982569192987346e-06, + "loss": 0.6, + "step": 816 + }, + { + "epoch": 0.05588617552500171, + "grad_norm": 3.5017900657668455, + "learning_rate": 9.982476653344799e-06, + "loss": 0.6264, + "step": 817 + }, + { + "epoch": 0.05595457965661126, + "grad_norm": 2.4172610670865207, + "learning_rate": 9.982383869138235e-06, + "loss": 0.1731, + "step": 818 + }, + { + "epoch": 0.056022983788220806, + "grad_norm": 2.828606411757475, + "learning_rate": 9.982290840372212e-06, + "loss": 0.4789, + "step": 819 + }, + { + "epoch": 0.05609138791983036, + "grad_norm": 2.9248378643516912, + "learning_rate": 9.982197567051296e-06, + "loss": 0.6796, + "step": 820 + }, + { + "epoch": 0.05615979205143991, + "grad_norm": 2.523782771437659, + "learning_rate": 9.982104049180064e-06, + "loss": 0.5955, + "step": 821 + }, + { + "epoch": 0.05622819618304946, + "grad_norm": 2.394269421748327, + "learning_rate": 9.982010286763109e-06, + "loss": 0.3263, + "step": 822 + }, + { + "epoch": 0.056296600314659, + "grad_norm": 2.143070439100002, + "learning_rate": 9.981916279805029e-06, + "loss": 0.418, + "step": 823 + }, + { + "epoch": 0.05636500444626855, + "grad_norm": 3.2371897727397854, + "learning_rate": 9.98182202831044e-06, + "loss": 0.74, + "step": 824 + }, + { + "epoch": 0.056433408577878104, + "grad_norm": 2.113788452204745, + "learning_rate": 9.98172753228397e-06, + "loss": 0.4705, + "step": 825 + }, + { + "epoch": 0.056501812709487655, + "grad_norm": 1.8164161435749964, + "learning_rate": 9.981632791730256e-06, + "loss": 0.3659, + "step": 826 + }, + { + "epoch": 0.056570216841097205, + "grad_norm": 2.8742623730496275, + "learning_rate": 9.981537806653949e-06, + "loss": 0.6062, + "step": 827 + }, + { + "epoch": 0.05663862097270675, + "grad_norm": 2.4141550385846156, + "learning_rate": 9.98144257705971e-06, + "loss": 0.48, + "step": 828 + }, + { + "epoch": 0.0567070251043163, + "grad_norm": 2.1336083516780966, + "learning_rate": 9.981347102952215e-06, + "loss": 0.3841, + "step": 829 + }, + { + "epoch": 0.05677542923592585, + "grad_norm": 2.0026231937637218, + "learning_rate": 9.98125138433615e-06, + "loss": 0.3058, + "step": 830 + }, + { + "epoch": 0.0568438333675354, + "grad_norm": 2.2972223966332703, + "learning_rate": 9.98115542121621e-06, + "loss": 0.412, + "step": 831 + }, + { + "epoch": 0.056912237499144945, + "grad_norm": 2.010161135712434, + "learning_rate": 9.981059213597111e-06, + "loss": 0.357, + "step": 832 + }, + { + "epoch": 0.056980641630754496, + "grad_norm": 2.347010036398286, + "learning_rate": 9.980962761483573e-06, + "loss": 0.4568, + "step": 833 + }, + { + "epoch": 0.057049045762364047, + "grad_norm": 2.935307760138367, + "learning_rate": 9.980866064880328e-06, + "loss": 0.8245, + "step": 834 + }, + { + "epoch": 0.0571174498939736, + "grad_norm": 2.701310910655135, + "learning_rate": 9.980769123792125e-06, + "loss": 0.6482, + "step": 835 + }, + { + "epoch": 0.05718585402558315, + "grad_norm": 2.8914515576527267, + "learning_rate": 9.98067193822372e-06, + "loss": 0.4066, + "step": 836 + }, + { + "epoch": 0.05725425815719269, + "grad_norm": 1.599369035674673, + "learning_rate": 9.980574508179886e-06, + "loss": 0.1768, + "step": 837 + }, + { + "epoch": 0.05732266228880224, + "grad_norm": 2.1809426388978577, + "learning_rate": 9.980476833665405e-06, + "loss": 0.4907, + "step": 838 + }, + { + "epoch": 0.05739106642041179, + "grad_norm": 1.9199460630552194, + "learning_rate": 9.980378914685069e-06, + "loss": 0.3885, + "step": 839 + }, + { + "epoch": 0.057459470552021344, + "grad_norm": 2.7803658944699303, + "learning_rate": 9.980280751243687e-06, + "loss": 0.594, + "step": 840 + }, + { + "epoch": 0.057527874683630895, + "grad_norm": 2.3866013749350423, + "learning_rate": 9.980182343346075e-06, + "loss": 0.4699, + "step": 841 + }, + { + "epoch": 0.05759627881524044, + "grad_norm": 2.037734874691204, + "learning_rate": 9.980083690997066e-06, + "loss": 0.1492, + "step": 842 + }, + { + "epoch": 0.05766468294684999, + "grad_norm": 2.7033672417160584, + "learning_rate": 9.9799847942015e-06, + "loss": 0.4978, + "step": 843 + }, + { + "epoch": 0.05773308707845954, + "grad_norm": 2.5758861068010015, + "learning_rate": 9.979885652964233e-06, + "loss": 0.5276, + "step": 844 + }, + { + "epoch": 0.05780149121006909, + "grad_norm": 2.5809462701699513, + "learning_rate": 9.979786267290129e-06, + "loss": 0.3193, + "step": 845 + }, + { + "epoch": 0.057869895341678634, + "grad_norm": 2.7403183514322333, + "learning_rate": 9.979686637184069e-06, + "loss": 0.616, + "step": 846 + }, + { + "epoch": 0.057938299473288185, + "grad_norm": 2.920597164185486, + "learning_rate": 9.97958676265094e-06, + "loss": 0.6216, + "step": 847 + }, + { + "epoch": 0.058006703604897736, + "grad_norm": 2.600529222088412, + "learning_rate": 9.979486643695651e-06, + "loss": 0.6244, + "step": 848 + }, + { + "epoch": 0.058075107736507287, + "grad_norm": 2.8128668439887825, + "learning_rate": 9.979386280323109e-06, + "loss": 0.8038, + "step": 849 + }, + { + "epoch": 0.05814351186811684, + "grad_norm": 2.9082251061812268, + "learning_rate": 9.979285672538243e-06, + "loss": 0.4786, + "step": 850 + }, + { + "epoch": 0.05821191599972638, + "grad_norm": 2.518236717809791, + "learning_rate": 9.979184820345992e-06, + "loss": 0.514, + "step": 851 + }, + { + "epoch": 0.05828032013133593, + "grad_norm": 1.9239412545571917, + "learning_rate": 9.979083723751306e-06, + "loss": 0.2318, + "step": 852 + }, + { + "epoch": 0.05834872426294548, + "grad_norm": 2.8216691175447544, + "learning_rate": 9.978982382759147e-06, + "loss": 0.7058, + "step": 853 + }, + { + "epoch": 0.05841712839455503, + "grad_norm": 2.615041815439836, + "learning_rate": 9.978880797374488e-06, + "loss": 0.5574, + "step": 854 + }, + { + "epoch": 0.05848553252616458, + "grad_norm": 2.854483250418998, + "learning_rate": 9.978778967602319e-06, + "loss": 0.5944, + "step": 855 + }, + { + "epoch": 0.05855393665777413, + "grad_norm": 2.4382138399084705, + "learning_rate": 9.978676893447635e-06, + "loss": 0.2368, + "step": 856 + }, + { + "epoch": 0.05862234078938368, + "grad_norm": 3.0137050921030437, + "learning_rate": 9.978574574915448e-06, + "loss": 0.6996, + "step": 857 + }, + { + "epoch": 0.05869074492099323, + "grad_norm": 2.677444019337918, + "learning_rate": 9.978472012010778e-06, + "loss": 0.7086, + "step": 858 + }, + { + "epoch": 0.05875914905260278, + "grad_norm": 2.834336283292121, + "learning_rate": 9.97836920473866e-06, + "loss": 0.5982, + "step": 859 + }, + { + "epoch": 0.058827553184212324, + "grad_norm": 2.839807880075379, + "learning_rate": 9.978266153104144e-06, + "loss": 0.3241, + "step": 860 + }, + { + "epoch": 0.058895957315821874, + "grad_norm": 2.2422394563722845, + "learning_rate": 9.978162857112283e-06, + "loss": 0.5662, + "step": 861 + }, + { + "epoch": 0.058964361447431425, + "grad_norm": 2.6370935555143067, + "learning_rate": 9.978059316768152e-06, + "loss": 0.2694, + "step": 862 + }, + { + "epoch": 0.059032765579040976, + "grad_norm": 2.8843940903681786, + "learning_rate": 9.97795553207683e-06, + "loss": 0.5115, + "step": 863 + }, + { + "epoch": 0.05910116971065053, + "grad_norm": 2.2912083355309463, + "learning_rate": 9.977851503043412e-06, + "loss": 0.4323, + "step": 864 + }, + { + "epoch": 0.05916957384226007, + "grad_norm": 2.5422310967860002, + "learning_rate": 9.977747229673004e-06, + "loss": 0.3959, + "step": 865 + }, + { + "epoch": 0.05923797797386962, + "grad_norm": 2.606952502119045, + "learning_rate": 9.977642711970726e-06, + "loss": 0.5744, + "step": 866 + }, + { + "epoch": 0.05930638210547917, + "grad_norm": 2.4721633636835567, + "learning_rate": 9.977537949941704e-06, + "loss": 0.4233, + "step": 867 + }, + { + "epoch": 0.05937478623708872, + "grad_norm": 2.113839729678478, + "learning_rate": 9.977432943591086e-06, + "loss": 0.2799, + "step": 868 + }, + { + "epoch": 0.059443190368698266, + "grad_norm": 2.532085834069017, + "learning_rate": 9.977327692924021e-06, + "loss": 0.6615, + "step": 869 + }, + { + "epoch": 0.05951159450030782, + "grad_norm": 2.2524329052394014, + "learning_rate": 9.977222197945679e-06, + "loss": 0.5819, + "step": 870 + }, + { + "epoch": 0.05957999863191737, + "grad_norm": 3.083735634144449, + "learning_rate": 9.977116458661238e-06, + "loss": 0.5505, + "step": 871 + }, + { + "epoch": 0.05964840276352692, + "grad_norm": 2.506197091207657, + "learning_rate": 9.977010475075886e-06, + "loss": 0.4009, + "step": 872 + }, + { + "epoch": 0.05971680689513647, + "grad_norm": 3.246917363056583, + "learning_rate": 9.976904247194826e-06, + "loss": 0.3479, + "step": 873 + }, + { + "epoch": 0.05978521102674601, + "grad_norm": 2.617148418988763, + "learning_rate": 9.976797775023272e-06, + "loss": 0.5503, + "step": 874 + }, + { + "epoch": 0.059853615158355564, + "grad_norm": 2.720839963211659, + "learning_rate": 9.97669105856645e-06, + "loss": 0.4437, + "step": 875 + }, + { + "epoch": 0.059922019289965114, + "grad_norm": 2.615224540526721, + "learning_rate": 9.9765840978296e-06, + "loss": 0.4218, + "step": 876 + }, + { + "epoch": 0.059990423421574665, + "grad_norm": 3.160299224825634, + "learning_rate": 9.97647689281797e-06, + "loss": 0.653, + "step": 877 + }, + { + "epoch": 0.06005882755318421, + "grad_norm": 2.87779873956826, + "learning_rate": 9.976369443536822e-06, + "loss": 0.6902, + "step": 878 + }, + { + "epoch": 0.06012723168479376, + "grad_norm": 4.425432748700984, + "learning_rate": 9.976261749991433e-06, + "loss": 0.569, + "step": 879 + }, + { + "epoch": 0.06019563581640331, + "grad_norm": 3.392013085483725, + "learning_rate": 9.976153812187087e-06, + "loss": 0.8461, + "step": 880 + }, + { + "epoch": 0.06026403994801286, + "grad_norm": 2.6792650028729543, + "learning_rate": 9.97604563012908e-06, + "loss": 0.5172, + "step": 881 + }, + { + "epoch": 0.06033244407962241, + "grad_norm": 3.0190501255405278, + "learning_rate": 9.975937203822726e-06, + "loss": 0.5326, + "step": 882 + }, + { + "epoch": 0.060400848211231956, + "grad_norm": 2.659570457921847, + "learning_rate": 9.975828533273346e-06, + "loss": 0.2099, + "step": 883 + }, + { + "epoch": 0.060469252342841506, + "grad_norm": 2.422890579419945, + "learning_rate": 9.975719618486272e-06, + "loss": 0.4013, + "step": 884 + }, + { + "epoch": 0.06053765647445106, + "grad_norm": 2.87984430566128, + "learning_rate": 9.975610459466853e-06, + "loss": 0.8378, + "step": 885 + }, + { + "epoch": 0.06060606060606061, + "grad_norm": 2.6413678303823858, + "learning_rate": 9.975501056220445e-06, + "loss": 0.5809, + "step": 886 + }, + { + "epoch": 0.06067446473767016, + "grad_norm": 2.4731835018601, + "learning_rate": 9.97539140875242e-06, + "loss": 0.4113, + "step": 887 + }, + { + "epoch": 0.0607428688692797, + "grad_norm": 2.7005001785876495, + "learning_rate": 9.975281517068158e-06, + "loss": 0.2535, + "step": 888 + }, + { + "epoch": 0.06081127300088925, + "grad_norm": 2.8545666847073883, + "learning_rate": 9.975171381173054e-06, + "loss": 0.3207, + "step": 889 + }, + { + "epoch": 0.060879677132498804, + "grad_norm": 2.092682010933842, + "learning_rate": 9.975061001072512e-06, + "loss": 0.4148, + "step": 890 + }, + { + "epoch": 0.060948081264108354, + "grad_norm": 3.0444640769135987, + "learning_rate": 9.974950376771954e-06, + "loss": 0.3724, + "step": 891 + }, + { + "epoch": 0.0610164853957179, + "grad_norm": 2.260906014160596, + "learning_rate": 9.974839508276806e-06, + "loss": 0.373, + "step": 892 + }, + { + "epoch": 0.06108488952732745, + "grad_norm": 2.2006818298644246, + "learning_rate": 9.974728395592512e-06, + "loss": 0.3318, + "step": 893 + }, + { + "epoch": 0.061153293658937, + "grad_norm": 2.227229777952215, + "learning_rate": 9.974617038724526e-06, + "loss": 0.4155, + "step": 894 + }, + { + "epoch": 0.06122169779054655, + "grad_norm": 3.1734148337429517, + "learning_rate": 9.974505437678315e-06, + "loss": 0.3618, + "step": 895 + }, + { + "epoch": 0.0612901019221561, + "grad_norm": 2.6458504793908144, + "learning_rate": 9.974393592459353e-06, + "loss": 0.5052, + "step": 896 + }, + { + "epoch": 0.061358506053765645, + "grad_norm": 1.5936560649864613, + "learning_rate": 9.974281503073134e-06, + "loss": 0.2588, + "step": 897 + }, + { + "epoch": 0.061426910185375196, + "grad_norm": 2.195629244655586, + "learning_rate": 9.974169169525157e-06, + "loss": 0.2119, + "step": 898 + }, + { + "epoch": 0.061495314316984746, + "grad_norm": 2.4977393330109634, + "learning_rate": 9.974056591820937e-06, + "loss": 0.5096, + "step": 899 + }, + { + "epoch": 0.0615637184485943, + "grad_norm": 3.039358021702255, + "learning_rate": 9.973943769966e-06, + "loss": 0.536, + "step": 900 + }, + { + "epoch": 0.06163212258020385, + "grad_norm": 2.479564640147992, + "learning_rate": 9.973830703965887e-06, + "loss": 0.5054, + "step": 901 + }, + { + "epoch": 0.06170052671181339, + "grad_norm": 7.672584889309925, + "learning_rate": 9.973717393826142e-06, + "loss": 0.7662, + "step": 902 + }, + { + "epoch": 0.06176893084342294, + "grad_norm": 3.5469020724838836, + "learning_rate": 9.97360383955233e-06, + "loss": 0.3153, + "step": 903 + }, + { + "epoch": 0.06183733497503249, + "grad_norm": 2.8295275861524805, + "learning_rate": 9.973490041150024e-06, + "loss": 0.463, + "step": 904 + }, + { + "epoch": 0.061905739106642044, + "grad_norm": 3.5169794313427505, + "learning_rate": 9.973375998624811e-06, + "loss": 0.7422, + "step": 905 + }, + { + "epoch": 0.06197414323825159, + "grad_norm": 2.0305928340981363, + "learning_rate": 9.973261711982287e-06, + "loss": 0.3628, + "step": 906 + }, + { + "epoch": 0.06204254736986114, + "grad_norm": 1.9382609172955425, + "learning_rate": 9.973147181228062e-06, + "loss": 0.3576, + "step": 907 + }, + { + "epoch": 0.06211095150147069, + "grad_norm": 2.715048901607533, + "learning_rate": 9.973032406367758e-06, + "loss": 0.7963, + "step": 908 + }, + { + "epoch": 0.06217935563308024, + "grad_norm": 2.5128756953939964, + "learning_rate": 9.97291738740701e-06, + "loss": 0.5051, + "step": 909 + }, + { + "epoch": 0.06224775976468979, + "grad_norm": 2.7498763013242127, + "learning_rate": 9.972802124351461e-06, + "loss": 0.7066, + "step": 910 + }, + { + "epoch": 0.062316163896299334, + "grad_norm": 2.511064073307623, + "learning_rate": 9.972686617206773e-06, + "loss": 0.5362, + "step": 911 + }, + { + "epoch": 0.062384568027908885, + "grad_norm": 2.6836682049724327, + "learning_rate": 9.972570865978612e-06, + "loss": 0.5801, + "step": 912 + }, + { + "epoch": 0.062452972159518436, + "grad_norm": 1.7320070597981576, + "learning_rate": 9.97245487067266e-06, + "loss": 0.1939, + "step": 913 + }, + { + "epoch": 0.06252137629112799, + "grad_norm": 2.069980767501643, + "learning_rate": 9.972338631294611e-06, + "loss": 0.4395, + "step": 914 + }, + { + "epoch": 0.06258978042273754, + "grad_norm": 2.08477922491963, + "learning_rate": 9.972222147850173e-06, + "loss": 0.425, + "step": 915 + }, + { + "epoch": 0.06265818455434709, + "grad_norm": 2.2976105349187947, + "learning_rate": 9.972105420345058e-06, + "loss": 0.5508, + "step": 916 + }, + { + "epoch": 0.06272658868595664, + "grad_norm": 2.2376307096342947, + "learning_rate": 9.971988448785002e-06, + "loss": 0.4197, + "step": 917 + }, + { + "epoch": 0.06279499281756618, + "grad_norm": 1.785020860675655, + "learning_rate": 9.97187123317574e-06, + "loss": 0.1627, + "step": 918 + }, + { + "epoch": 0.06286339694917573, + "grad_norm": 2.547862616179099, + "learning_rate": 9.971753773523032e-06, + "loss": 0.5211, + "step": 919 + }, + { + "epoch": 0.06293180108078528, + "grad_norm": 2.2846397207380402, + "learning_rate": 9.97163606983264e-06, + "loss": 0.4611, + "step": 920 + }, + { + "epoch": 0.06300020521239483, + "grad_norm": 2.280033982833032, + "learning_rate": 9.97151812211034e-06, + "loss": 0.5492, + "step": 921 + }, + { + "epoch": 0.06306860934400438, + "grad_norm": 2.8694729700149804, + "learning_rate": 9.971399930361924e-06, + "loss": 0.4057, + "step": 922 + }, + { + "epoch": 0.06313701347561393, + "grad_norm": 2.09159545569829, + "learning_rate": 9.971281494593194e-06, + "loss": 0.3997, + "step": 923 + }, + { + "epoch": 0.06320541760722348, + "grad_norm": 1.6017822591690358, + "learning_rate": 9.971162814809962e-06, + "loss": 0.2121, + "step": 924 + }, + { + "epoch": 0.06327382173883303, + "grad_norm": 1.5821487006399673, + "learning_rate": 9.971043891018053e-06, + "loss": 0.251, + "step": 925 + }, + { + "epoch": 0.06334222587044258, + "grad_norm": 4.267253816271588, + "learning_rate": 9.970924723223306e-06, + "loss": 0.3508, + "step": 926 + }, + { + "epoch": 0.06341063000205212, + "grad_norm": 12.948371926260517, + "learning_rate": 9.970805311431566e-06, + "loss": 0.4605, + "step": 927 + }, + { + "epoch": 0.06347903413366167, + "grad_norm": 2.251327394630374, + "learning_rate": 9.970685655648702e-06, + "loss": 0.4338, + "step": 928 + }, + { + "epoch": 0.06354743826527122, + "grad_norm": 3.2704055601568, + "learning_rate": 9.970565755880579e-06, + "loss": 0.5031, + "step": 929 + }, + { + "epoch": 0.06361584239688077, + "grad_norm": 2.787958920553477, + "learning_rate": 9.970445612133088e-06, + "loss": 0.2409, + "step": 930 + }, + { + "epoch": 0.06368424652849032, + "grad_norm": 3.012501477601255, + "learning_rate": 9.970325224412125e-06, + "loss": 0.5846, + "step": 931 + }, + { + "epoch": 0.06375265066009987, + "grad_norm": 2.5459011123227615, + "learning_rate": 9.970204592723598e-06, + "loss": 0.4442, + "step": 932 + }, + { + "epoch": 0.06382105479170942, + "grad_norm": 7.836329137536698, + "learning_rate": 9.970083717073427e-06, + "loss": 0.2004, + "step": 933 + }, + { + "epoch": 0.06388945892331897, + "grad_norm": 2.741210366722295, + "learning_rate": 9.969962597467548e-06, + "loss": 0.2851, + "step": 934 + }, + { + "epoch": 0.06395786305492852, + "grad_norm": 2.4373830559382834, + "learning_rate": 9.969841233911905e-06, + "loss": 0.3466, + "step": 935 + }, + { + "epoch": 0.06402626718653806, + "grad_norm": 2.863561796050765, + "learning_rate": 9.969719626412456e-06, + "loss": 0.744, + "step": 936 + }, + { + "epoch": 0.06409467131814761, + "grad_norm": 2.6016031554198573, + "learning_rate": 9.969597774975169e-06, + "loss": 0.5556, + "step": 937 + }, + { + "epoch": 0.06416307544975716, + "grad_norm": 2.4142189715886957, + "learning_rate": 9.969475679606025e-06, + "loss": 0.639, + "step": 938 + }, + { + "epoch": 0.06423147958136671, + "grad_norm": 2.493724062313865, + "learning_rate": 9.969353340311017e-06, + "loss": 0.5196, + "step": 939 + }, + { + "epoch": 0.06429988371297626, + "grad_norm": 2.259362375189832, + "learning_rate": 9.96923075709615e-06, + "loss": 0.2889, + "step": 940 + }, + { + "epoch": 0.06436828784458581, + "grad_norm": 2.4593037101757016, + "learning_rate": 9.96910792996744e-06, + "loss": 0.4073, + "step": 941 + }, + { + "epoch": 0.06443669197619536, + "grad_norm": 2.638357904759169, + "learning_rate": 9.96898485893092e-06, + "loss": 0.5279, + "step": 942 + }, + { + "epoch": 0.06450509610780492, + "grad_norm": 2.471612242852739, + "learning_rate": 9.968861543992628e-06, + "loss": 0.3464, + "step": 943 + }, + { + "epoch": 0.06457350023941447, + "grad_norm": 2.5985116138314326, + "learning_rate": 9.968737985158615e-06, + "loss": 0.4986, + "step": 944 + }, + { + "epoch": 0.064641904371024, + "grad_norm": 2.88906585297047, + "learning_rate": 9.968614182434949e-06, + "loss": 0.3231, + "step": 945 + }, + { + "epoch": 0.06471030850263355, + "grad_norm": 2.824241564224556, + "learning_rate": 9.968490135827707e-06, + "loss": 0.6046, + "step": 946 + }, + { + "epoch": 0.0647787126342431, + "grad_norm": 2.788936451860544, + "learning_rate": 9.968365845342973e-06, + "loss": 0.8125, + "step": 947 + }, + { + "epoch": 0.06484711676585266, + "grad_norm": 2.231387428517909, + "learning_rate": 9.968241310986855e-06, + "loss": 0.3335, + "step": 948 + }, + { + "epoch": 0.0649155208974622, + "grad_norm": 2.648629550103403, + "learning_rate": 9.96811653276546e-06, + "loss": 0.5245, + "step": 949 + }, + { + "epoch": 0.06498392502907176, + "grad_norm": 2.1254031906780084, + "learning_rate": 9.967991510684916e-06, + "loss": 0.3916, + "step": 950 + }, + { + "epoch": 0.06505232916068131, + "grad_norm": 2.7888986234845423, + "learning_rate": 9.967866244751356e-06, + "loss": 0.5935, + "step": 951 + }, + { + "epoch": 0.06512073329229086, + "grad_norm": 2.53549216797209, + "learning_rate": 9.967740734970934e-06, + "loss": 0.5389, + "step": 952 + }, + { + "epoch": 0.06518913742390041, + "grad_norm": 2.0505460622441922, + "learning_rate": 9.967614981349807e-06, + "loss": 0.3343, + "step": 953 + }, + { + "epoch": 0.06525754155550996, + "grad_norm": 2.033231614102896, + "learning_rate": 9.967488983894147e-06, + "loss": 0.2591, + "step": 954 + }, + { + "epoch": 0.0653259456871195, + "grad_norm": 2.047461272864372, + "learning_rate": 9.967362742610141e-06, + "loss": 0.4399, + "step": 955 + }, + { + "epoch": 0.06539434981872905, + "grad_norm": 2.4575445309191095, + "learning_rate": 9.967236257503984e-06, + "loss": 0.5772, + "step": 956 + }, + { + "epoch": 0.0654627539503386, + "grad_norm": 2.365222836378887, + "learning_rate": 9.967109528581886e-06, + "loss": 0.4493, + "step": 957 + }, + { + "epoch": 0.06553115808194815, + "grad_norm": 2.673612609604145, + "learning_rate": 9.966982555850066e-06, + "loss": 0.6708, + "step": 958 + }, + { + "epoch": 0.0655995622135577, + "grad_norm": 2.3283003350613534, + "learning_rate": 9.966855339314756e-06, + "loss": 0.317, + "step": 959 + }, + { + "epoch": 0.06566796634516725, + "grad_norm": 2.7970693563349527, + "learning_rate": 9.966727878982203e-06, + "loss": 0.3967, + "step": 960 + }, + { + "epoch": 0.0657363704767768, + "grad_norm": 2.5598827987809907, + "learning_rate": 9.96660017485866e-06, + "loss": 0.6765, + "step": 961 + }, + { + "epoch": 0.06580477460838635, + "grad_norm": 2.1771196806457684, + "learning_rate": 9.966472226950396e-06, + "loss": 0.3974, + "step": 962 + }, + { + "epoch": 0.0658731787399959, + "grad_norm": 1.6718957745963696, + "learning_rate": 9.966344035263696e-06, + "loss": 0.2621, + "step": 963 + }, + { + "epoch": 0.06594158287160544, + "grad_norm": 8.51010910717314, + "learning_rate": 9.966215599804846e-06, + "loss": 0.3781, + "step": 964 + }, + { + "epoch": 0.06600998700321499, + "grad_norm": 2.450196869394683, + "learning_rate": 9.966086920580152e-06, + "loss": 0.2412, + "step": 965 + }, + { + "epoch": 0.06607839113482454, + "grad_norm": 3.412403550374459, + "learning_rate": 9.965957997595932e-06, + "loss": 0.5763, + "step": 966 + }, + { + "epoch": 0.06614679526643409, + "grad_norm": 2.6786944770933063, + "learning_rate": 9.965828830858513e-06, + "loss": 0.4666, + "step": 967 + }, + { + "epoch": 0.06621519939804364, + "grad_norm": 4.206406684213161, + "learning_rate": 9.965699420374235e-06, + "loss": 0.4244, + "step": 968 + }, + { + "epoch": 0.06628360352965319, + "grad_norm": 2.885017130947985, + "learning_rate": 9.96556976614945e-06, + "loss": 0.4521, + "step": 969 + }, + { + "epoch": 0.06635200766126274, + "grad_norm": 2.7045157774166877, + "learning_rate": 9.965439868190524e-06, + "loss": 0.5924, + "step": 970 + }, + { + "epoch": 0.0664204117928723, + "grad_norm": 2.2816320435803936, + "learning_rate": 9.96530972650383e-06, + "loss": 0.2791, + "step": 971 + }, + { + "epoch": 0.06648881592448184, + "grad_norm": 2.8717078142618466, + "learning_rate": 9.965179341095758e-06, + "loss": 0.5337, + "step": 972 + }, + { + "epoch": 0.06655722005609138, + "grad_norm": 5.665693040492871, + "learning_rate": 9.965048711972707e-06, + "loss": 0.5172, + "step": 973 + }, + { + "epoch": 0.06662562418770093, + "grad_norm": 2.5150796464988847, + "learning_rate": 9.964917839141089e-06, + "loss": 0.4313, + "step": 974 + }, + { + "epoch": 0.06669402831931048, + "grad_norm": 1.8042985406154783, + "learning_rate": 9.96478672260733e-06, + "loss": 0.181, + "step": 975 + }, + { + "epoch": 0.06676243245092003, + "grad_norm": 3.081157671723186, + "learning_rate": 9.96465536237786e-06, + "loss": 0.5547, + "step": 976 + }, + { + "epoch": 0.06683083658252958, + "grad_norm": 3.916482782443657, + "learning_rate": 9.964523758459134e-06, + "loss": 0.399, + "step": 977 + }, + { + "epoch": 0.06689924071413914, + "grad_norm": 1.8922081655607064, + "learning_rate": 9.964391910857608e-06, + "loss": 0.3603, + "step": 978 + }, + { + "epoch": 0.06696764484574869, + "grad_norm": 1.8340830750967925, + "learning_rate": 9.964259819579754e-06, + "loss": 0.3779, + "step": 979 + }, + { + "epoch": 0.06703604897735824, + "grad_norm": 2.678309272471299, + "learning_rate": 9.964127484632054e-06, + "loss": 0.7094, + "step": 980 + }, + { + "epoch": 0.06710445310896779, + "grad_norm": 2.1789935179666933, + "learning_rate": 9.963994906021008e-06, + "loss": 0.3597, + "step": 981 + }, + { + "epoch": 0.06717285724057732, + "grad_norm": 2.763689123308795, + "learning_rate": 9.96386208375312e-06, + "loss": 0.5027, + "step": 982 + }, + { + "epoch": 0.06724126137218688, + "grad_norm": 2.949107197047589, + "learning_rate": 9.96372901783491e-06, + "loss": 0.4787, + "step": 983 + }, + { + "epoch": 0.06730966550379643, + "grad_norm": 2.173838602539477, + "learning_rate": 9.963595708272913e-06, + "loss": 0.2186, + "step": 984 + }, + { + "epoch": 0.06737806963540598, + "grad_norm": 1.9422003889078168, + "learning_rate": 9.963462155073667e-06, + "loss": 0.4371, + "step": 985 + }, + { + "epoch": 0.06744647376701553, + "grad_norm": 2.3113350769320085, + "learning_rate": 9.963328358243732e-06, + "loss": 0.4089, + "step": 986 + }, + { + "epoch": 0.06751487789862508, + "grad_norm": 2.4222693306855043, + "learning_rate": 9.963194317789672e-06, + "loss": 0.3948, + "step": 987 + }, + { + "epoch": 0.06758328203023463, + "grad_norm": 2.343704305839156, + "learning_rate": 9.963060033718068e-06, + "loss": 0.3073, + "step": 988 + }, + { + "epoch": 0.06765168616184418, + "grad_norm": 2.962808073499399, + "learning_rate": 9.962925506035512e-06, + "loss": 0.242, + "step": 989 + }, + { + "epoch": 0.06772009029345373, + "grad_norm": 3.6694872701834464, + "learning_rate": 9.962790734748606e-06, + "loss": 1.0186, + "step": 990 + }, + { + "epoch": 0.06778849442506327, + "grad_norm": 2.6994021038764564, + "learning_rate": 9.962655719863965e-06, + "loss": 0.2553, + "step": 991 + }, + { + "epoch": 0.06785689855667282, + "grad_norm": 2.4496465002211254, + "learning_rate": 9.96252046138822e-06, + "loss": 0.3159, + "step": 992 + }, + { + "epoch": 0.06792530268828237, + "grad_norm": 2.604339148849341, + "learning_rate": 9.962384959328002e-06, + "loss": 0.541, + "step": 993 + }, + { + "epoch": 0.06799370681989192, + "grad_norm": 2.571094243325737, + "learning_rate": 9.96224921368997e-06, + "loss": 0.4937, + "step": 994 + }, + { + "epoch": 0.06806211095150147, + "grad_norm": 3.0423169594890385, + "learning_rate": 9.962113224480783e-06, + "loss": 0.5872, + "step": 995 + }, + { + "epoch": 0.06813051508311102, + "grad_norm": 2.4621359858647622, + "learning_rate": 9.961976991707118e-06, + "loss": 0.4698, + "step": 996 + }, + { + "epoch": 0.06819891921472057, + "grad_norm": 2.6407397092925495, + "learning_rate": 9.961840515375663e-06, + "loss": 0.5666, + "step": 997 + }, + { + "epoch": 0.06826732334633012, + "grad_norm": 2.783615863379243, + "learning_rate": 9.961703795493111e-06, + "loss": 0.8431, + "step": 998 + }, + { + "epoch": 0.06833572747793967, + "grad_norm": 2.1074493554898246, + "learning_rate": 9.96156683206618e-06, + "loss": 0.3992, + "step": 999 + }, + { + "epoch": 0.06840413160954922, + "grad_norm": 2.4335955746859694, + "learning_rate": 9.961429625101589e-06, + "loss": 0.4933, + "step": 1000 + }, + { + "epoch": 0.06847253574115876, + "grad_norm": 2.4003982357590457, + "learning_rate": 9.961292174606073e-06, + "loss": 0.5583, + "step": 1001 + }, + { + "epoch": 0.06854093987276831, + "grad_norm": 8.03069175828441, + "learning_rate": 9.961154480586379e-06, + "loss": 0.3758, + "step": 1002 + }, + { + "epoch": 0.06860934400437786, + "grad_norm": 2.024991727299923, + "learning_rate": 9.961016543049267e-06, + "loss": 0.3516, + "step": 1003 + }, + { + "epoch": 0.06867774813598741, + "grad_norm": 2.98784312109352, + "learning_rate": 9.960878362001505e-06, + "loss": 0.5376, + "step": 1004 + }, + { + "epoch": 0.06874615226759696, + "grad_norm": 3.263873803439799, + "learning_rate": 9.960739937449877e-06, + "loss": 0.8229, + "step": 1005 + }, + { + "epoch": 0.06881455639920651, + "grad_norm": 4.904585319006385, + "learning_rate": 9.96060126940118e-06, + "loss": 0.3604, + "step": 1006 + }, + { + "epoch": 0.06888296053081606, + "grad_norm": 2.316986493776772, + "learning_rate": 9.960462357862216e-06, + "loss": 0.2947, + "step": 1007 + }, + { + "epoch": 0.06895136466242562, + "grad_norm": 1.917835611667519, + "learning_rate": 9.960323202839806e-06, + "loss": 0.2573, + "step": 1008 + }, + { + "epoch": 0.06901976879403517, + "grad_norm": 2.505825773922189, + "learning_rate": 9.960183804340781e-06, + "loss": 0.3942, + "step": 1009 + }, + { + "epoch": 0.0690881729256447, + "grad_norm": 2.1819883187315248, + "learning_rate": 9.960044162371982e-06, + "loss": 0.3733, + "step": 1010 + }, + { + "epoch": 0.06915657705725425, + "grad_norm": 2.826285327672023, + "learning_rate": 9.959904276940265e-06, + "loss": 0.3789, + "step": 1011 + }, + { + "epoch": 0.0692249811888638, + "grad_norm": 2.1301830260321, + "learning_rate": 9.959764148052493e-06, + "loss": 0.2425, + "step": 1012 + }, + { + "epoch": 0.06929338532047336, + "grad_norm": 2.5464661509924853, + "learning_rate": 9.959623775715546e-06, + "loss": 0.2416, + "step": 1013 + }, + { + "epoch": 0.0693617894520829, + "grad_norm": 2.439948637140326, + "learning_rate": 9.959483159936316e-06, + "loss": 0.5277, + "step": 1014 + }, + { + "epoch": 0.06943019358369246, + "grad_norm": 2.602979669257779, + "learning_rate": 9.959342300721702e-06, + "loss": 0.6769, + "step": 1015 + }, + { + "epoch": 0.06949859771530201, + "grad_norm": 2.1303887341381227, + "learning_rate": 9.959201198078621e-06, + "loss": 0.4614, + "step": 1016 + }, + { + "epoch": 0.06956700184691156, + "grad_norm": 2.0838843717597166, + "learning_rate": 9.959059852013994e-06, + "loss": 0.4435, + "step": 1017 + }, + { + "epoch": 0.06963540597852111, + "grad_norm": 2.736639730721296, + "learning_rate": 9.958918262534767e-06, + "loss": 0.6352, + "step": 1018 + }, + { + "epoch": 0.06970381011013065, + "grad_norm": 1.6325679805864135, + "learning_rate": 9.958776429647882e-06, + "loss": 0.3318, + "step": 1019 + }, + { + "epoch": 0.0697722142417402, + "grad_norm": 2.938476075679172, + "learning_rate": 9.958634353360305e-06, + "loss": 0.7283, + "step": 1020 + }, + { + "epoch": 0.06984061837334975, + "grad_norm": 2.714180482486242, + "learning_rate": 9.95849203367901e-06, + "loss": 0.6226, + "step": 1021 + }, + { + "epoch": 0.0699090225049593, + "grad_norm": 2.7502208913606885, + "learning_rate": 9.95834947061098e-06, + "loss": 0.6839, + "step": 1022 + }, + { + "epoch": 0.06997742663656885, + "grad_norm": 2.1840569920920023, + "learning_rate": 9.958206664163215e-06, + "loss": 0.4763, + "step": 1023 + }, + { + "epoch": 0.0700458307681784, + "grad_norm": 2.9430433380825884, + "learning_rate": 9.958063614342726e-06, + "loss": 0.7998, + "step": 1024 + }, + { + "epoch": 0.07011423489978795, + "grad_norm": 2.7031278940968098, + "learning_rate": 9.95792032115653e-06, + "loss": 0.6107, + "step": 1025 + }, + { + "epoch": 0.0701826390313975, + "grad_norm": 2.245311209320568, + "learning_rate": 9.957776784611663e-06, + "loss": 0.6327, + "step": 1026 + }, + { + "epoch": 0.07025104316300705, + "grad_norm": 1.813649824058935, + "learning_rate": 9.957633004715173e-06, + "loss": 0.3409, + "step": 1027 + }, + { + "epoch": 0.07031944729461659, + "grad_norm": 6.212961234273394, + "learning_rate": 9.957488981474113e-06, + "loss": 0.2331, + "step": 1028 + }, + { + "epoch": 0.07038785142622614, + "grad_norm": 2.282731692002078, + "learning_rate": 9.957344714895554e-06, + "loss": 0.3352, + "step": 1029 + }, + { + "epoch": 0.07045625555783569, + "grad_norm": 2.3870921927576387, + "learning_rate": 9.95720020498658e-06, + "loss": 0.5065, + "step": 1030 + }, + { + "epoch": 0.07052465968944524, + "grad_norm": 2.07707728598992, + "learning_rate": 9.957055451754279e-06, + "loss": 0.2718, + "step": 1031 + }, + { + "epoch": 0.07059306382105479, + "grad_norm": 2.174911845821369, + "learning_rate": 9.95691045520576e-06, + "loss": 0.4036, + "step": 1032 + }, + { + "epoch": 0.07066146795266434, + "grad_norm": 2.7400428014391616, + "learning_rate": 9.95676521534814e-06, + "loss": 0.577, + "step": 1033 + }, + { + "epoch": 0.07072987208427389, + "grad_norm": 1.9644052383979276, + "learning_rate": 9.956619732188547e-06, + "loss": 0.4244, + "step": 1034 + }, + { + "epoch": 0.07079827621588344, + "grad_norm": 2.243348787878243, + "learning_rate": 9.95647400573412e-06, + "loss": 0.1936, + "step": 1035 + }, + { + "epoch": 0.070866680347493, + "grad_norm": 2.820697314355854, + "learning_rate": 9.956328035992017e-06, + "loss": 0.5918, + "step": 1036 + }, + { + "epoch": 0.07093508447910253, + "grad_norm": 2.5232409232371387, + "learning_rate": 9.9561818229694e-06, + "loss": 0.3069, + "step": 1037 + }, + { + "epoch": 0.07100348861071208, + "grad_norm": 2.5483300902689257, + "learning_rate": 9.956035366673444e-06, + "loss": 0.5191, + "step": 1038 + }, + { + "epoch": 0.07107189274232163, + "grad_norm": 1.9759432459646187, + "learning_rate": 9.955888667111341e-06, + "loss": 0.3643, + "step": 1039 + }, + { + "epoch": 0.07114029687393118, + "grad_norm": 2.734980826702942, + "learning_rate": 9.955741724290291e-06, + "loss": 0.6579, + "step": 1040 + }, + { + "epoch": 0.07120870100554073, + "grad_norm": 2.2807755107458547, + "learning_rate": 9.955594538217505e-06, + "loss": 0.3344, + "step": 1041 + }, + { + "epoch": 0.07127710513715028, + "grad_norm": 2.2859704058945534, + "learning_rate": 9.955447108900212e-06, + "loss": 0.3504, + "step": 1042 + }, + { + "epoch": 0.07134550926875984, + "grad_norm": 2.592662382172054, + "learning_rate": 9.955299436345641e-06, + "loss": 0.5526, + "step": 1043 + }, + { + "epoch": 0.07141391340036939, + "grad_norm": 2.1806541336728698, + "learning_rate": 9.955151520561045e-06, + "loss": 0.415, + "step": 1044 + }, + { + "epoch": 0.07148231753197894, + "grad_norm": 2.0974863168232605, + "learning_rate": 9.955003361553688e-06, + "loss": 0.2659, + "step": 1045 + }, + { + "epoch": 0.07155072166358849, + "grad_norm": 2.4965238961684943, + "learning_rate": 9.954854959330835e-06, + "loss": 0.5773, + "step": 1046 + }, + { + "epoch": 0.07161912579519802, + "grad_norm": 2.4924343279879095, + "learning_rate": 9.954706313899776e-06, + "loss": 0.2072, + "step": 1047 + }, + { + "epoch": 0.07168752992680757, + "grad_norm": 1.7107451866446768, + "learning_rate": 9.954557425267804e-06, + "loss": 0.2268, + "step": 1048 + }, + { + "epoch": 0.07175593405841713, + "grad_norm": 1.8427143825333239, + "learning_rate": 9.954408293442228e-06, + "loss": 0.311, + "step": 1049 + }, + { + "epoch": 0.07182433819002668, + "grad_norm": 3.0714400067861094, + "learning_rate": 9.954258918430368e-06, + "loss": 0.5399, + "step": 1050 + }, + { + "epoch": 0.07189274232163623, + "grad_norm": 2.4892377706673443, + "learning_rate": 9.954109300239559e-06, + "loss": 0.5529, + "step": 1051 + }, + { + "epoch": 0.07196114645324578, + "grad_norm": 2.458524895871505, + "learning_rate": 9.95395943887714e-06, + "loss": 0.6119, + "step": 1052 + }, + { + "epoch": 0.07202955058485533, + "grad_norm": 3.601897482695106, + "learning_rate": 9.95380933435047e-06, + "loss": 0.7938, + "step": 1053 + }, + { + "epoch": 0.07209795471646488, + "grad_norm": 2.863765436219965, + "learning_rate": 9.953658986666916e-06, + "loss": 0.4471, + "step": 1054 + }, + { + "epoch": 0.07216635884807443, + "grad_norm": 2.1140092452531896, + "learning_rate": 9.953508395833857e-06, + "loss": 0.3621, + "step": 1055 + }, + { + "epoch": 0.07223476297968397, + "grad_norm": 3.507976366731207, + "learning_rate": 9.953357561858688e-06, + "loss": 0.3872, + "step": 1056 + }, + { + "epoch": 0.07230316711129352, + "grad_norm": 12.269420255288304, + "learning_rate": 9.953206484748807e-06, + "loss": 0.5578, + "step": 1057 + }, + { + "epoch": 0.07237157124290307, + "grad_norm": 1.955268900096957, + "learning_rate": 9.953055164511635e-06, + "loss": 0.2682, + "step": 1058 + }, + { + "epoch": 0.07243997537451262, + "grad_norm": 2.6371850343184837, + "learning_rate": 9.952903601154598e-06, + "loss": 0.4443, + "step": 1059 + }, + { + "epoch": 0.07250837950612217, + "grad_norm": 2.674310951303961, + "learning_rate": 9.952751794685135e-06, + "loss": 0.5505, + "step": 1060 + }, + { + "epoch": 0.07257678363773172, + "grad_norm": 2.184814350820009, + "learning_rate": 9.952599745110697e-06, + "loss": 0.321, + "step": 1061 + }, + { + "epoch": 0.07264518776934127, + "grad_norm": 2.4112677955465154, + "learning_rate": 9.952447452438747e-06, + "loss": 0.6462, + "step": 1062 + }, + { + "epoch": 0.07271359190095082, + "grad_norm": 2.7291597658657354, + "learning_rate": 9.952294916676762e-06, + "loss": 0.3574, + "step": 1063 + }, + { + "epoch": 0.07278199603256037, + "grad_norm": 2.7133733431155154, + "learning_rate": 9.952142137832226e-06, + "loss": 0.511, + "step": 1064 + }, + { + "epoch": 0.07285040016416991, + "grad_norm": 2.2399771358928167, + "learning_rate": 9.951989115912641e-06, + "loss": 0.5471, + "step": 1065 + }, + { + "epoch": 0.07291880429577946, + "grad_norm": 2.3908808141232716, + "learning_rate": 9.951835850925519e-06, + "loss": 0.3687, + "step": 1066 + }, + { + "epoch": 0.07298720842738901, + "grad_norm": 2.4220736637466045, + "learning_rate": 9.951682342878379e-06, + "loss": 0.4914, + "step": 1067 + }, + { + "epoch": 0.07305561255899856, + "grad_norm": 2.501763794500433, + "learning_rate": 9.95152859177876e-06, + "loss": 0.3954, + "step": 1068 + }, + { + "epoch": 0.07312401669060811, + "grad_norm": 2.130433030313395, + "learning_rate": 9.951374597634208e-06, + "loss": 0.3349, + "step": 1069 + }, + { + "epoch": 0.07319242082221766, + "grad_norm": 27.39849931825789, + "learning_rate": 9.951220360452277e-06, + "loss": 0.2009, + "step": 1070 + }, + { + "epoch": 0.07326082495382721, + "grad_norm": 2.3277777278401177, + "learning_rate": 9.951065880240545e-06, + "loss": 0.2969, + "step": 1071 + }, + { + "epoch": 0.07332922908543676, + "grad_norm": 4.239793421564506, + "learning_rate": 9.950911157006588e-06, + "loss": 0.3955, + "step": 1072 + }, + { + "epoch": 0.07339763321704632, + "grad_norm": 2.2500974845960666, + "learning_rate": 9.950756190758006e-06, + "loss": 0.4414, + "step": 1073 + }, + { + "epoch": 0.07346603734865585, + "grad_norm": 2.3382791500674784, + "learning_rate": 9.950600981502403e-06, + "loss": 0.3794, + "step": 1074 + }, + { + "epoch": 0.0735344414802654, + "grad_norm": 2.149605149367718, + "learning_rate": 9.950445529247397e-06, + "loss": 0.2963, + "step": 1075 + }, + { + "epoch": 0.07360284561187495, + "grad_norm": 2.5783069488592876, + "learning_rate": 9.950289834000619e-06, + "loss": 0.4034, + "step": 1076 + }, + { + "epoch": 0.0736712497434845, + "grad_norm": 3.9158428353790145, + "learning_rate": 9.950133895769712e-06, + "loss": 0.3434, + "step": 1077 + }, + { + "epoch": 0.07373965387509406, + "grad_norm": 3.2613222532211483, + "learning_rate": 9.949977714562329e-06, + "loss": 0.8255, + "step": 1078 + }, + { + "epoch": 0.0738080580067036, + "grad_norm": 2.6011985031577507, + "learning_rate": 9.949821290386137e-06, + "loss": 0.3436, + "step": 1079 + }, + { + "epoch": 0.07387646213831316, + "grad_norm": 2.568121112171609, + "learning_rate": 9.949664623248813e-06, + "loss": 0.5479, + "step": 1080 + }, + { + "epoch": 0.07394486626992271, + "grad_norm": 2.083469384439942, + "learning_rate": 9.949507713158048e-06, + "loss": 0.362, + "step": 1081 + }, + { + "epoch": 0.07401327040153226, + "grad_norm": 3.0667755730656485, + "learning_rate": 9.949350560121543e-06, + "loss": 0.531, + "step": 1082 + }, + { + "epoch": 0.07408167453314181, + "grad_norm": 2.2359452898594907, + "learning_rate": 9.949193164147012e-06, + "loss": 0.4589, + "step": 1083 + }, + { + "epoch": 0.07415007866475135, + "grad_norm": 2.754907666742986, + "learning_rate": 9.949035525242181e-06, + "loss": 0.5378, + "step": 1084 + }, + { + "epoch": 0.0742184827963609, + "grad_norm": 2.7109988092970756, + "learning_rate": 9.94887764341479e-06, + "loss": 0.5811, + "step": 1085 + }, + { + "epoch": 0.07428688692797045, + "grad_norm": 1.8869144188174087, + "learning_rate": 9.948719518672584e-06, + "loss": 0.2918, + "step": 1086 + }, + { + "epoch": 0.07435529105958, + "grad_norm": 2.7516804892410183, + "learning_rate": 9.948561151023328e-06, + "loss": 0.6971, + "step": 1087 + }, + { + "epoch": 0.07442369519118955, + "grad_norm": 2.689578749438299, + "learning_rate": 9.948402540474793e-06, + "loss": 0.6808, + "step": 1088 + }, + { + "epoch": 0.0744920993227991, + "grad_norm": 2.0837339306173415, + "learning_rate": 9.948243687034768e-06, + "loss": 0.4334, + "step": 1089 + }, + { + "epoch": 0.07456050345440865, + "grad_norm": 2.958319359509056, + "learning_rate": 9.948084590711047e-06, + "loss": 0.7501, + "step": 1090 + }, + { + "epoch": 0.0746289075860182, + "grad_norm": 3.26161155296106, + "learning_rate": 9.94792525151144e-06, + "loss": 0.5213, + "step": 1091 + }, + { + "epoch": 0.07469731171762775, + "grad_norm": 2.811260692089301, + "learning_rate": 9.947765669443768e-06, + "loss": 0.3806, + "step": 1092 + }, + { + "epoch": 0.07476571584923729, + "grad_norm": 2.0494170801548948, + "learning_rate": 9.947605844515863e-06, + "loss": 0.3226, + "step": 1093 + }, + { + "epoch": 0.07483411998084684, + "grad_norm": 2.31212117984347, + "learning_rate": 9.947445776735574e-06, + "loss": 0.349, + "step": 1094 + }, + { + "epoch": 0.07490252411245639, + "grad_norm": 8.48779733667929, + "learning_rate": 9.947285466110755e-06, + "loss": 0.42, + "step": 1095 + }, + { + "epoch": 0.07497092824406594, + "grad_norm": 2.286329845099989, + "learning_rate": 9.947124912649275e-06, + "loss": 0.3795, + "step": 1096 + }, + { + "epoch": 0.07503933237567549, + "grad_norm": 2.706773170400206, + "learning_rate": 9.946964116359015e-06, + "loss": 0.6394, + "step": 1097 + }, + { + "epoch": 0.07510773650728504, + "grad_norm": 2.5888915346800183, + "learning_rate": 9.946803077247866e-06, + "loss": 0.4358, + "step": 1098 + }, + { + "epoch": 0.07517614063889459, + "grad_norm": 3.4481266837490185, + "learning_rate": 9.946641795323737e-06, + "loss": 0.5366, + "step": 1099 + }, + { + "epoch": 0.07524454477050414, + "grad_norm": 1.3880381190945505, + "learning_rate": 9.946480270594539e-06, + "loss": 0.1792, + "step": 1100 + }, + { + "epoch": 0.0753129489021137, + "grad_norm": 1.388793035981185, + "learning_rate": 9.946318503068205e-06, + "loss": 0.195, + "step": 1101 + }, + { + "epoch": 0.07538135303372323, + "grad_norm": 2.5190241695204647, + "learning_rate": 9.94615649275267e-06, + "loss": 0.5724, + "step": 1102 + }, + { + "epoch": 0.07544975716533278, + "grad_norm": 2.635088878848173, + "learning_rate": 9.945994239655892e-06, + "loss": 0.649, + "step": 1103 + }, + { + "epoch": 0.07551816129694233, + "grad_norm": 2.514263329506905, + "learning_rate": 9.945831743785832e-06, + "loss": 0.6236, + "step": 1104 + }, + { + "epoch": 0.07558656542855188, + "grad_norm": 2.2233146000858826, + "learning_rate": 9.945669005150467e-06, + "loss": 0.45, + "step": 1105 + }, + { + "epoch": 0.07565496956016143, + "grad_norm": 2.9574775510179423, + "learning_rate": 9.945506023757786e-06, + "loss": 0.1938, + "step": 1106 + }, + { + "epoch": 0.07572337369177098, + "grad_norm": 2.773254549681466, + "learning_rate": 9.945342799615786e-06, + "loss": 0.5639, + "step": 1107 + }, + { + "epoch": 0.07579177782338054, + "grad_norm": 2.6933960887904616, + "learning_rate": 9.945179332732482e-06, + "loss": 0.7536, + "step": 1108 + }, + { + "epoch": 0.07586018195499009, + "grad_norm": 1.8153213214183834, + "learning_rate": 9.945015623115896e-06, + "loss": 0.1637, + "step": 1109 + }, + { + "epoch": 0.07592858608659964, + "grad_norm": 2.596127741565586, + "learning_rate": 9.944851670774064e-06, + "loss": 0.5732, + "step": 1110 + }, + { + "epoch": 0.07599699021820917, + "grad_norm": 2.298214589894453, + "learning_rate": 9.944687475715034e-06, + "loss": 0.4043, + "step": 1111 + }, + { + "epoch": 0.07606539434981872, + "grad_norm": 2.5908068999149263, + "learning_rate": 9.944523037946863e-06, + "loss": 0.6742, + "step": 1112 + }, + { + "epoch": 0.07613379848142827, + "grad_norm": 3.6503811219215687, + "learning_rate": 9.944358357477625e-06, + "loss": 0.437, + "step": 1113 + }, + { + "epoch": 0.07620220261303783, + "grad_norm": 1.7422212588884096, + "learning_rate": 9.944193434315406e-06, + "loss": 0.2249, + "step": 1114 + }, + { + "epoch": 0.07627060674464738, + "grad_norm": 2.121067742456988, + "learning_rate": 9.944028268468294e-06, + "loss": 0.3245, + "step": 1115 + }, + { + "epoch": 0.07633901087625693, + "grad_norm": 1.8364369950665915, + "learning_rate": 9.943862859944402e-06, + "loss": 0.3279, + "step": 1116 + }, + { + "epoch": 0.07640741500786648, + "grad_norm": 2.50753928862092, + "learning_rate": 9.943697208751847e-06, + "loss": 0.6031, + "step": 1117 + }, + { + "epoch": 0.07647581913947603, + "grad_norm": 2.5395109514201977, + "learning_rate": 9.94353131489876e-06, + "loss": 0.6491, + "step": 1118 + }, + { + "epoch": 0.07654422327108558, + "grad_norm": 1.6592630204889256, + "learning_rate": 9.943365178393283e-06, + "loss": 0.1443, + "step": 1119 + }, + { + "epoch": 0.07661262740269512, + "grad_norm": 2.974407908078639, + "learning_rate": 9.943198799243575e-06, + "loss": 0.2536, + "step": 1120 + }, + { + "epoch": 0.07668103153430467, + "grad_norm": 2.2385296176167797, + "learning_rate": 9.943032177457796e-06, + "loss": 0.5557, + "step": 1121 + }, + { + "epoch": 0.07674943566591422, + "grad_norm": 2.5435955136600756, + "learning_rate": 9.94286531304413e-06, + "loss": 0.5373, + "step": 1122 + }, + { + "epoch": 0.07681783979752377, + "grad_norm": 2.4961924466244785, + "learning_rate": 9.942698206010764e-06, + "loss": 0.3286, + "step": 1123 + }, + { + "epoch": 0.07688624392913332, + "grad_norm": 2.0812113151162603, + "learning_rate": 9.942530856365902e-06, + "loss": 0.478, + "step": 1124 + }, + { + "epoch": 0.07695464806074287, + "grad_norm": 2.114569590356325, + "learning_rate": 9.94236326411776e-06, + "loss": 0.3924, + "step": 1125 + }, + { + "epoch": 0.07702305219235242, + "grad_norm": 2.628700554383146, + "learning_rate": 9.942195429274562e-06, + "loss": 0.4322, + "step": 1126 + }, + { + "epoch": 0.07709145632396197, + "grad_norm": 2.180790386307906, + "learning_rate": 9.942027351844545e-06, + "loss": 0.5081, + "step": 1127 + }, + { + "epoch": 0.07715986045557152, + "grad_norm": 3.0540891485711485, + "learning_rate": 9.941859031835964e-06, + "loss": 0.4666, + "step": 1128 + }, + { + "epoch": 0.07722826458718107, + "grad_norm": 2.4452764733923837, + "learning_rate": 9.941690469257075e-06, + "loss": 0.5147, + "step": 1129 + }, + { + "epoch": 0.07729666871879061, + "grad_norm": 3.618004197927457, + "learning_rate": 9.941521664116156e-06, + "loss": 0.3539, + "step": 1130 + }, + { + "epoch": 0.07736507285040016, + "grad_norm": 2.2260470580067113, + "learning_rate": 9.94135261642149e-06, + "loss": 0.4078, + "step": 1131 + }, + { + "epoch": 0.07743347698200971, + "grad_norm": 2.5456373682993036, + "learning_rate": 9.941183326181377e-06, + "loss": 0.5591, + "step": 1132 + }, + { + "epoch": 0.07750188111361926, + "grad_norm": 1.7454834069744825, + "learning_rate": 9.941013793404126e-06, + "loss": 0.2913, + "step": 1133 + }, + { + "epoch": 0.07757028524522881, + "grad_norm": 2.465971662956434, + "learning_rate": 9.940844018098057e-06, + "loss": 0.5041, + "step": 1134 + }, + { + "epoch": 0.07763868937683836, + "grad_norm": 2.2909086743925693, + "learning_rate": 9.940674000271505e-06, + "loss": 0.5478, + "step": 1135 + }, + { + "epoch": 0.07770709350844791, + "grad_norm": 1.8096186450776313, + "learning_rate": 9.940503739932815e-06, + "loss": 0.2513, + "step": 1136 + }, + { + "epoch": 0.07777549764005746, + "grad_norm": 2.4581696703683384, + "learning_rate": 9.940333237090345e-06, + "loss": 0.3763, + "step": 1137 + }, + { + "epoch": 0.07784390177166702, + "grad_norm": 2.9147272860188993, + "learning_rate": 9.940162491752461e-06, + "loss": 0.5572, + "step": 1138 + }, + { + "epoch": 0.07791230590327655, + "grad_norm": 1.891122843085123, + "learning_rate": 9.939991503927548e-06, + "loss": 0.287, + "step": 1139 + }, + { + "epoch": 0.0779807100348861, + "grad_norm": 2.484433419362462, + "learning_rate": 9.939820273623996e-06, + "loss": 0.5293, + "step": 1140 + }, + { + "epoch": 0.07804911416649565, + "grad_norm": 2.270294730739799, + "learning_rate": 9.93964880085021e-06, + "loss": 0.4939, + "step": 1141 + }, + { + "epoch": 0.0781175182981052, + "grad_norm": 2.8729565033745255, + "learning_rate": 9.939477085614608e-06, + "loss": 0.4051, + "step": 1142 + }, + { + "epoch": 0.07818592242971475, + "grad_norm": 2.3275850681202273, + "learning_rate": 9.939305127925621e-06, + "loss": 0.5406, + "step": 1143 + }, + { + "epoch": 0.0782543265613243, + "grad_norm": 2.0758393237540407, + "learning_rate": 9.939132927791684e-06, + "loss": 0.4152, + "step": 1144 + }, + { + "epoch": 0.07832273069293386, + "grad_norm": 2.177385256029136, + "learning_rate": 9.938960485221252e-06, + "loss": 0.4319, + "step": 1145 + }, + { + "epoch": 0.07839113482454341, + "grad_norm": 2.684058825132479, + "learning_rate": 9.93878780022279e-06, + "loss": 0.4993, + "step": 1146 + }, + { + "epoch": 0.07845953895615296, + "grad_norm": 3.8437912481312253, + "learning_rate": 9.938614872804773e-06, + "loss": 0.5068, + "step": 1147 + }, + { + "epoch": 0.0785279430877625, + "grad_norm": 2.342023951423456, + "learning_rate": 9.938441702975689e-06, + "loss": 0.5398, + "step": 1148 + }, + { + "epoch": 0.07859634721937205, + "grad_norm": 2.4354689462487533, + "learning_rate": 9.938268290744039e-06, + "loss": 0.5742, + "step": 1149 + }, + { + "epoch": 0.0786647513509816, + "grad_norm": 2.3135216621199324, + "learning_rate": 9.938094636118336e-06, + "loss": 0.492, + "step": 1150 + }, + { + "epoch": 0.07873315548259115, + "grad_norm": 2.743440434191089, + "learning_rate": 9.937920739107102e-06, + "loss": 0.7444, + "step": 1151 + }, + { + "epoch": 0.0788015596142007, + "grad_norm": 2.7621622228604146, + "learning_rate": 9.937746599718873e-06, + "loss": 0.6435, + "step": 1152 + }, + { + "epoch": 0.07886996374581025, + "grad_norm": 2.5624796354833728, + "learning_rate": 9.937572217962197e-06, + "loss": 0.2986, + "step": 1153 + }, + { + "epoch": 0.0789383678774198, + "grad_norm": 2.328004437083509, + "learning_rate": 9.937397593845634e-06, + "loss": 0.4305, + "step": 1154 + }, + { + "epoch": 0.07900677200902935, + "grad_norm": 1.7041421355306077, + "learning_rate": 9.937222727377753e-06, + "loss": 0.2749, + "step": 1155 + }, + { + "epoch": 0.0790751761406389, + "grad_norm": 1.8449907169263573, + "learning_rate": 9.93704761856714e-06, + "loss": 0.3106, + "step": 1156 + }, + { + "epoch": 0.07914358027224844, + "grad_norm": 2.5131739647152576, + "learning_rate": 9.936872267422389e-06, + "loss": 0.4545, + "step": 1157 + }, + { + "epoch": 0.07921198440385799, + "grad_norm": 2.3696901669965786, + "learning_rate": 9.936696673952107e-06, + "loss": 0.471, + "step": 1158 + }, + { + "epoch": 0.07928038853546754, + "grad_norm": 2.4846723186675392, + "learning_rate": 9.936520838164912e-06, + "loss": 0.5902, + "step": 1159 + }, + { + "epoch": 0.07934879266707709, + "grad_norm": 1.9369280263111415, + "learning_rate": 9.936344760069437e-06, + "loss": 0.3738, + "step": 1160 + }, + { + "epoch": 0.07941719679868664, + "grad_norm": 1.9991798954624456, + "learning_rate": 9.936168439674326e-06, + "loss": 0.2936, + "step": 1161 + }, + { + "epoch": 0.07948560093029619, + "grad_norm": 1.891088030996455, + "learning_rate": 9.935991876988228e-06, + "loss": 0.3636, + "step": 1162 + }, + { + "epoch": 0.07955400506190574, + "grad_norm": 2.5316459852296376, + "learning_rate": 9.935815072019815e-06, + "loss": 0.5082, + "step": 1163 + }, + { + "epoch": 0.07962240919351529, + "grad_norm": 1.969236941896376, + "learning_rate": 9.935638024777763e-06, + "loss": 0.3074, + "step": 1164 + }, + { + "epoch": 0.07969081332512484, + "grad_norm": 2.20447743616131, + "learning_rate": 9.935460735270762e-06, + "loss": 0.2146, + "step": 1165 + }, + { + "epoch": 0.0797592174567344, + "grad_norm": 2.0573897052695713, + "learning_rate": 9.935283203507516e-06, + "loss": 0.3431, + "step": 1166 + }, + { + "epoch": 0.07982762158834393, + "grad_norm": 2.3085920136988998, + "learning_rate": 9.935105429496737e-06, + "loss": 0.2898, + "step": 1167 + }, + { + "epoch": 0.07989602571995348, + "grad_norm": 2.049938028550261, + "learning_rate": 9.934927413247152e-06, + "loss": 0.433, + "step": 1168 + }, + { + "epoch": 0.07996442985156303, + "grad_norm": 2.3120828909629205, + "learning_rate": 9.934749154767503e-06, + "loss": 0.6091, + "step": 1169 + }, + { + "epoch": 0.08003283398317258, + "grad_norm": 3.1100037288950064, + "learning_rate": 9.93457065406653e-06, + "loss": 0.7005, + "step": 1170 + }, + { + "epoch": 0.08010123811478213, + "grad_norm": 2.256315931548306, + "learning_rate": 9.934391911153006e-06, + "loss": 0.3124, + "step": 1171 + }, + { + "epoch": 0.08016964224639168, + "grad_norm": 1.4971727509111505, + "learning_rate": 9.934212926035695e-06, + "loss": 0.2245, + "step": 1172 + }, + { + "epoch": 0.08023804637800123, + "grad_norm": 2.499054926815104, + "learning_rate": 9.93403369872339e-06, + "loss": 0.6353, + "step": 1173 + }, + { + "epoch": 0.08030645050961079, + "grad_norm": 2.2198163277449408, + "learning_rate": 9.93385422922488e-06, + "loss": 0.4164, + "step": 1174 + }, + { + "epoch": 0.08037485464122034, + "grad_norm": 1.6845318599126964, + "learning_rate": 9.933674517548984e-06, + "loss": 0.3459, + "step": 1175 + }, + { + "epoch": 0.08044325877282987, + "grad_norm": 1.2994357645231358, + "learning_rate": 9.933494563704515e-06, + "loss": 0.212, + "step": 1176 + }, + { + "epoch": 0.08051166290443942, + "grad_norm": 2.511728627851412, + "learning_rate": 9.933314367700311e-06, + "loss": 0.4755, + "step": 1177 + }, + { + "epoch": 0.08058006703604897, + "grad_norm": 2.682785900963887, + "learning_rate": 9.933133929545213e-06, + "loss": 0.645, + "step": 1178 + }, + { + "epoch": 0.08064847116765853, + "grad_norm": 1.9793492751622508, + "learning_rate": 9.932953249248082e-06, + "loss": 0.2986, + "step": 1179 + }, + { + "epoch": 0.08071687529926808, + "grad_norm": 2.425069375864637, + "learning_rate": 9.932772326817784e-06, + "loss": 0.5512, + "step": 1180 + }, + { + "epoch": 0.08078527943087763, + "grad_norm": 2.037467754948447, + "learning_rate": 9.932591162263199e-06, + "loss": 0.3237, + "step": 1181 + }, + { + "epoch": 0.08085368356248718, + "grad_norm": 3.5078248361485302, + "learning_rate": 9.932409755593223e-06, + "loss": 0.8034, + "step": 1182 + }, + { + "epoch": 0.08092208769409673, + "grad_norm": 2.3764149598089888, + "learning_rate": 9.932228106816755e-06, + "loss": 0.4298, + "step": 1183 + }, + { + "epoch": 0.08099049182570628, + "grad_norm": 2.9670193094692983, + "learning_rate": 9.932046215942715e-06, + "loss": 0.5693, + "step": 1184 + }, + { + "epoch": 0.08105889595731582, + "grad_norm": 2.245805871178687, + "learning_rate": 9.93186408298003e-06, + "loss": 0.4186, + "step": 1185 + }, + { + "epoch": 0.08112730008892537, + "grad_norm": 2.3706973277090397, + "learning_rate": 9.931681707937642e-06, + "loss": 0.3128, + "step": 1186 + }, + { + "epoch": 0.08119570422053492, + "grad_norm": 2.0274892765801105, + "learning_rate": 9.931499090824499e-06, + "loss": 0.4016, + "step": 1187 + }, + { + "epoch": 0.08126410835214447, + "grad_norm": 2.7615910545082403, + "learning_rate": 9.931316231649565e-06, + "loss": 0.7062, + "step": 1188 + }, + { + "epoch": 0.08133251248375402, + "grad_norm": 2.2111081101333543, + "learning_rate": 9.931133130421821e-06, + "loss": 0.5111, + "step": 1189 + }, + { + "epoch": 0.08140091661536357, + "grad_norm": 2.912089295960864, + "learning_rate": 9.93094978715025e-06, + "loss": 0.4021, + "step": 1190 + }, + { + "epoch": 0.08146932074697312, + "grad_norm": 2.0671120071178115, + "learning_rate": 9.930766201843852e-06, + "loss": 0.4596, + "step": 1191 + }, + { + "epoch": 0.08153772487858267, + "grad_norm": 2.896782117010925, + "learning_rate": 9.930582374511635e-06, + "loss": 0.5599, + "step": 1192 + }, + { + "epoch": 0.08160612901019222, + "grad_norm": 2.6253631147644056, + "learning_rate": 9.93039830516263e-06, + "loss": 0.6979, + "step": 1193 + }, + { + "epoch": 0.08167453314180176, + "grad_norm": 2.7746273228697085, + "learning_rate": 9.930213993805863e-06, + "loss": 0.4984, + "step": 1194 + }, + { + "epoch": 0.08174293727341131, + "grad_norm": 2.3765101588176747, + "learning_rate": 9.930029440450391e-06, + "loss": 0.4718, + "step": 1195 + }, + { + "epoch": 0.08181134140502086, + "grad_norm": 2.9412957888831444, + "learning_rate": 9.929844645105263e-06, + "loss": 0.7151, + "step": 1196 + }, + { + "epoch": 0.08187974553663041, + "grad_norm": 3.0822157520529805, + "learning_rate": 9.929659607779555e-06, + "loss": 0.7752, + "step": 1197 + }, + { + "epoch": 0.08194814966823996, + "grad_norm": 2.3074188502343382, + "learning_rate": 9.92947432848235e-06, + "loss": 0.3132, + "step": 1198 + }, + { + "epoch": 0.08201655379984951, + "grad_norm": 2.032577150782066, + "learning_rate": 9.929288807222738e-06, + "loss": 0.4465, + "step": 1199 + }, + { + "epoch": 0.08208495793145906, + "grad_norm": 2.847540425272367, + "learning_rate": 9.929103044009829e-06, + "loss": 0.7622, + "step": 1200 + }, + { + "epoch": 0.08215336206306861, + "grad_norm": 2.785384865643226, + "learning_rate": 9.928917038852741e-06, + "loss": 0.4816, + "step": 1201 + }, + { + "epoch": 0.08222176619467816, + "grad_norm": 2.356525133648848, + "learning_rate": 9.928730791760602e-06, + "loss": 0.23, + "step": 1202 + }, + { + "epoch": 0.0822901703262877, + "grad_norm": 2.885556447720842, + "learning_rate": 9.928544302742556e-06, + "loss": 0.4198, + "step": 1203 + }, + { + "epoch": 0.08235857445789725, + "grad_norm": 2.231306376399473, + "learning_rate": 9.928357571807757e-06, + "loss": 0.3524, + "step": 1204 + }, + { + "epoch": 0.0824269785895068, + "grad_norm": 1.9000675989347786, + "learning_rate": 9.928170598965368e-06, + "loss": 0.3559, + "step": 1205 + }, + { + "epoch": 0.08249538272111635, + "grad_norm": 2.7173924619423433, + "learning_rate": 9.927983384224567e-06, + "loss": 0.4564, + "step": 1206 + }, + { + "epoch": 0.0825637868527259, + "grad_norm": 2.3977217926632926, + "learning_rate": 9.927795927594547e-06, + "loss": 0.5139, + "step": 1207 + }, + { + "epoch": 0.08263219098433545, + "grad_norm": 2.479264473121129, + "learning_rate": 9.927608229084506e-06, + "loss": 0.5508, + "step": 1208 + }, + { + "epoch": 0.082700595115945, + "grad_norm": 2.3820052483610534, + "learning_rate": 9.927420288703657e-06, + "loss": 0.3908, + "step": 1209 + }, + { + "epoch": 0.08276899924755456, + "grad_norm": 2.1720227527445335, + "learning_rate": 9.927232106461226e-06, + "loss": 0.516, + "step": 1210 + }, + { + "epoch": 0.0828374033791641, + "grad_norm": 2.182789649379566, + "learning_rate": 9.927043682366452e-06, + "loss": 0.5835, + "step": 1211 + }, + { + "epoch": 0.08290580751077366, + "grad_norm": 2.1291937537688073, + "learning_rate": 9.92685501642858e-06, + "loss": 0.3252, + "step": 1212 + }, + { + "epoch": 0.0829742116423832, + "grad_norm": 2.1271078720757175, + "learning_rate": 9.926666108656873e-06, + "loss": 0.3884, + "step": 1213 + }, + { + "epoch": 0.08304261577399275, + "grad_norm": 2.382339042778516, + "learning_rate": 9.926476959060602e-06, + "loss": 0.601, + "step": 1214 + }, + { + "epoch": 0.0831110199056023, + "grad_norm": 3.357852537574129, + "learning_rate": 9.926287567649051e-06, + "loss": 0.4132, + "step": 1215 + }, + { + "epoch": 0.08317942403721185, + "grad_norm": 2.2704269952644704, + "learning_rate": 9.92609793443152e-06, + "loss": 0.5033, + "step": 1216 + }, + { + "epoch": 0.0832478281688214, + "grad_norm": 1.6078726297558432, + "learning_rate": 9.925908059417313e-06, + "loss": 0.2466, + "step": 1217 + }, + { + "epoch": 0.08331623230043095, + "grad_norm": 1.73668458881835, + "learning_rate": 9.925717942615752e-06, + "loss": 0.3716, + "step": 1218 + }, + { + "epoch": 0.0833846364320405, + "grad_norm": 2.1369033852259816, + "learning_rate": 9.925527584036167e-06, + "loss": 0.5188, + "step": 1219 + }, + { + "epoch": 0.08345304056365005, + "grad_norm": 2.1321121015975417, + "learning_rate": 9.925336983687905e-06, + "loss": 0.4737, + "step": 1220 + }, + { + "epoch": 0.0835214446952596, + "grad_norm": 2.5964791606760613, + "learning_rate": 9.925146141580317e-06, + "loss": 0.5737, + "step": 1221 + }, + { + "epoch": 0.08358984882686914, + "grad_norm": 2.0857230872779895, + "learning_rate": 9.924955057722776e-06, + "loss": 0.1974, + "step": 1222 + }, + { + "epoch": 0.08365825295847869, + "grad_norm": 3.4476777173516497, + "learning_rate": 9.924763732124657e-06, + "loss": 0.4962, + "step": 1223 + }, + { + "epoch": 0.08372665709008824, + "grad_norm": 2.292440051290243, + "learning_rate": 9.924572164795352e-06, + "loss": 0.701, + "step": 1224 + }, + { + "epoch": 0.08379506122169779, + "grad_norm": 2.463595876977694, + "learning_rate": 9.924380355744266e-06, + "loss": 0.6776, + "step": 1225 + }, + { + "epoch": 0.08386346535330734, + "grad_norm": 1.5062981484560714, + "learning_rate": 9.924188304980812e-06, + "loss": 0.1355, + "step": 1226 + }, + { + "epoch": 0.08393186948491689, + "grad_norm": 2.4242742406092046, + "learning_rate": 9.923996012514416e-06, + "loss": 0.4888, + "step": 1227 + }, + { + "epoch": 0.08400027361652644, + "grad_norm": 2.2892694054270906, + "learning_rate": 9.92380347835452e-06, + "loss": 0.2165, + "step": 1228 + }, + { + "epoch": 0.08406867774813599, + "grad_norm": 2.2873782676605896, + "learning_rate": 9.923610702510571e-06, + "loss": 0.4236, + "step": 1229 + }, + { + "epoch": 0.08413708187974554, + "grad_norm": 2.8465178402781026, + "learning_rate": 9.923417684992033e-06, + "loss": 0.7389, + "step": 1230 + }, + { + "epoch": 0.08420548601135508, + "grad_norm": 2.4612023831668264, + "learning_rate": 9.923224425808381e-06, + "loss": 0.5255, + "step": 1231 + }, + { + "epoch": 0.08427389014296463, + "grad_norm": 2.4401566809530317, + "learning_rate": 9.923030924969101e-06, + "loss": 0.4386, + "step": 1232 + }, + { + "epoch": 0.08434229427457418, + "grad_norm": 1.3130130588155557, + "learning_rate": 9.92283718248369e-06, + "loss": 0.1456, + "step": 1233 + }, + { + "epoch": 0.08441069840618373, + "grad_norm": 2.451556913370705, + "learning_rate": 9.922643198361655e-06, + "loss": 0.5747, + "step": 1234 + }, + { + "epoch": 0.08447910253779328, + "grad_norm": 2.509890779573401, + "learning_rate": 9.922448972612524e-06, + "loss": 0.3913, + "step": 1235 + }, + { + "epoch": 0.08454750666940283, + "grad_norm": 1.7773275269826443, + "learning_rate": 9.922254505245825e-06, + "loss": 0.2326, + "step": 1236 + }, + { + "epoch": 0.08461591080101238, + "grad_norm": 2.308197942841891, + "learning_rate": 9.922059796271107e-06, + "loss": 0.5138, + "step": 1237 + }, + { + "epoch": 0.08468431493262193, + "grad_norm": 1.0403646684757177, + "learning_rate": 9.921864845697924e-06, + "loss": 0.2023, + "step": 1238 + }, + { + "epoch": 0.08475271906423149, + "grad_norm": 2.8221619794880963, + "learning_rate": 9.921669653535848e-06, + "loss": 0.6408, + "step": 1239 + }, + { + "epoch": 0.08482112319584102, + "grad_norm": 2.3098647012659033, + "learning_rate": 9.921474219794459e-06, + "loss": 0.4226, + "step": 1240 + }, + { + "epoch": 0.08488952732745057, + "grad_norm": 2.3641779372829923, + "learning_rate": 9.921278544483349e-06, + "loss": 0.2653, + "step": 1241 + }, + { + "epoch": 0.08495793145906012, + "grad_norm": 3.1148827907468353, + "learning_rate": 9.921082627612125e-06, + "loss": 0.9064, + "step": 1242 + }, + { + "epoch": 0.08502633559066967, + "grad_norm": 2.410126636453032, + "learning_rate": 9.920886469190401e-06, + "loss": 0.2768, + "step": 1243 + }, + { + "epoch": 0.08509473972227923, + "grad_norm": 2.1607114585381693, + "learning_rate": 9.920690069227805e-06, + "loss": 0.3975, + "step": 1244 + }, + { + "epoch": 0.08516314385388878, + "grad_norm": 2.4252998132303825, + "learning_rate": 9.920493427733981e-06, + "loss": 0.4116, + "step": 1245 + }, + { + "epoch": 0.08523154798549833, + "grad_norm": 1.9140296777519055, + "learning_rate": 9.920296544718577e-06, + "loss": 0.2258, + "step": 1246 + }, + { + "epoch": 0.08529995211710788, + "grad_norm": 2.715995796621719, + "learning_rate": 9.92009942019126e-06, + "loss": 0.5232, + "step": 1247 + }, + { + "epoch": 0.08536835624871743, + "grad_norm": 1.601810605905163, + "learning_rate": 9.919902054161704e-06, + "loss": 0.2684, + "step": 1248 + }, + { + "epoch": 0.08543676038032696, + "grad_norm": 2.866745044132113, + "learning_rate": 9.919704446639597e-06, + "loss": 0.7081, + "step": 1249 + }, + { + "epoch": 0.08550516451193652, + "grad_norm": 2.3306326258323105, + "learning_rate": 9.919506597634638e-06, + "loss": 0.3879, + "step": 1250 + }, + { + "epoch": 0.08557356864354607, + "grad_norm": 2.4299796931474003, + "learning_rate": 9.919308507156542e-06, + "loss": 0.4378, + "step": 1251 + }, + { + "epoch": 0.08564197277515562, + "grad_norm": 2.5959552903608745, + "learning_rate": 9.919110175215027e-06, + "loss": 0.6486, + "step": 1252 + }, + { + "epoch": 0.08571037690676517, + "grad_norm": 1.7353128998871183, + "learning_rate": 9.91891160181983e-06, + "loss": 0.2408, + "step": 1253 + }, + { + "epoch": 0.08577878103837472, + "grad_norm": 2.3309134611155202, + "learning_rate": 9.9187127869807e-06, + "loss": 0.3765, + "step": 1254 + }, + { + "epoch": 0.08584718516998427, + "grad_norm": 2.0657124355122645, + "learning_rate": 9.918513730707392e-06, + "loss": 0.4397, + "step": 1255 + }, + { + "epoch": 0.08591558930159382, + "grad_norm": 2.033041853284862, + "learning_rate": 9.918314433009681e-06, + "loss": 0.4264, + "step": 1256 + }, + { + "epoch": 0.08598399343320337, + "grad_norm": 2.1899111674420806, + "learning_rate": 9.918114893897348e-06, + "loss": 0.4381, + "step": 1257 + }, + { + "epoch": 0.08605239756481292, + "grad_norm": 2.890762449553603, + "learning_rate": 9.917915113380186e-06, + "loss": 0.6049, + "step": 1258 + }, + { + "epoch": 0.08612080169642246, + "grad_norm": 2.4769211698912406, + "learning_rate": 9.917715091467999e-06, + "loss": 0.7902, + "step": 1259 + }, + { + "epoch": 0.08618920582803201, + "grad_norm": 2.3827851841095664, + "learning_rate": 9.91751482817061e-06, + "loss": 0.4084, + "step": 1260 + }, + { + "epoch": 0.08625760995964156, + "grad_norm": 3.000357412862898, + "learning_rate": 9.917314323497848e-06, + "loss": 0.4498, + "step": 1261 + }, + { + "epoch": 0.08632601409125111, + "grad_norm": 2.0098191041939626, + "learning_rate": 9.917113577459553e-06, + "loss": 0.3717, + "step": 1262 + }, + { + "epoch": 0.08639441822286066, + "grad_norm": 2.8234454160662725, + "learning_rate": 9.916912590065578e-06, + "loss": 0.616, + "step": 1263 + }, + { + "epoch": 0.08646282235447021, + "grad_norm": 2.0882720705880247, + "learning_rate": 9.916711361325791e-06, + "loss": 0.2943, + "step": 1264 + }, + { + "epoch": 0.08653122648607976, + "grad_norm": 2.2037292845521055, + "learning_rate": 9.916509891250066e-06, + "loss": 0.3675, + "step": 1265 + }, + { + "epoch": 0.08659963061768931, + "grad_norm": 2.424494587937646, + "learning_rate": 9.916308179848295e-06, + "loss": 0.3401, + "step": 1266 + }, + { + "epoch": 0.08666803474929886, + "grad_norm": 2.3309909602492755, + "learning_rate": 9.916106227130378e-06, + "loss": 0.6571, + "step": 1267 + }, + { + "epoch": 0.0867364388809084, + "grad_norm": 1.9745226915540333, + "learning_rate": 9.915904033106227e-06, + "loss": 0.4967, + "step": 1268 + }, + { + "epoch": 0.08680484301251795, + "grad_norm": 2.527029775605986, + "learning_rate": 9.915701597785769e-06, + "loss": 0.3122, + "step": 1269 + }, + { + "epoch": 0.0868732471441275, + "grad_norm": 2.276098617623925, + "learning_rate": 9.915498921178937e-06, + "loss": 0.4438, + "step": 1270 + }, + { + "epoch": 0.08694165127573705, + "grad_norm": 2.1380344015628725, + "learning_rate": 9.915296003295682e-06, + "loss": 0.2944, + "step": 1271 + }, + { + "epoch": 0.0870100554073466, + "grad_norm": 2.044012266357404, + "learning_rate": 9.915092844145963e-06, + "loss": 0.2948, + "step": 1272 + }, + { + "epoch": 0.08707845953895615, + "grad_norm": 2.2790167799056142, + "learning_rate": 9.914889443739755e-06, + "loss": 0.4051, + "step": 1273 + }, + { + "epoch": 0.0871468636705657, + "grad_norm": 1.7923895039412203, + "learning_rate": 9.914685802087037e-06, + "loss": 0.3446, + "step": 1274 + }, + { + "epoch": 0.08721526780217526, + "grad_norm": 1.6026432717128722, + "learning_rate": 9.914481919197807e-06, + "loss": 0.2855, + "step": 1275 + }, + { + "epoch": 0.0872836719337848, + "grad_norm": 2.64503052814656, + "learning_rate": 9.914277795082072e-06, + "loss": 0.4811, + "step": 1276 + }, + { + "epoch": 0.08735207606539434, + "grad_norm": 2.4401725154221507, + "learning_rate": 9.914073429749853e-06, + "loss": 0.5482, + "step": 1277 + }, + { + "epoch": 0.0874204801970039, + "grad_norm": 2.1378326964375796, + "learning_rate": 9.913868823211182e-06, + "loss": 0.3952, + "step": 1278 + }, + { + "epoch": 0.08748888432861344, + "grad_norm": 2.578506181454309, + "learning_rate": 9.913663975476099e-06, + "loss": 0.5322, + "step": 1279 + }, + { + "epoch": 0.087557288460223, + "grad_norm": 2.5426626328832604, + "learning_rate": 9.913458886554659e-06, + "loss": 0.432, + "step": 1280 + }, + { + "epoch": 0.08762569259183255, + "grad_norm": 2.6392363672253976, + "learning_rate": 9.91325355645693e-06, + "loss": 0.6995, + "step": 1281 + }, + { + "epoch": 0.0876940967234421, + "grad_norm": 2.518423123600879, + "learning_rate": 9.913047985192991e-06, + "loss": 0.4143, + "step": 1282 + }, + { + "epoch": 0.08776250085505165, + "grad_norm": 2.320930616975427, + "learning_rate": 9.912842172772935e-06, + "loss": 0.3663, + "step": 1283 + }, + { + "epoch": 0.0878309049866612, + "grad_norm": 1.7606301507410682, + "learning_rate": 9.912636119206857e-06, + "loss": 0.3513, + "step": 1284 + }, + { + "epoch": 0.08789930911827075, + "grad_norm": 2.352000807155419, + "learning_rate": 9.912429824504877e-06, + "loss": 0.4475, + "step": 1285 + }, + { + "epoch": 0.08796771324988029, + "grad_norm": 1.8527176026420162, + "learning_rate": 9.91222328867712e-06, + "loss": 0.385, + "step": 1286 + }, + { + "epoch": 0.08803611738148984, + "grad_norm": 2.1003351030749666, + "learning_rate": 9.912016511733722e-06, + "loss": 0.4993, + "step": 1287 + }, + { + "epoch": 0.08810452151309939, + "grad_norm": 2.2478175506153213, + "learning_rate": 9.911809493684834e-06, + "loss": 0.4184, + "step": 1288 + }, + { + "epoch": 0.08817292564470894, + "grad_norm": 2.380086652927344, + "learning_rate": 9.911602234540618e-06, + "loss": 0.5502, + "step": 1289 + }, + { + "epoch": 0.08824132977631849, + "grad_norm": 1.288646185723821, + "learning_rate": 9.911394734311246e-06, + "loss": 0.1993, + "step": 1290 + }, + { + "epoch": 0.08830973390792804, + "grad_norm": 3.1583295005374152, + "learning_rate": 9.911186993006903e-06, + "loss": 0.6863, + "step": 1291 + }, + { + "epoch": 0.08837813803953759, + "grad_norm": 2.282709543150597, + "learning_rate": 9.910979010637787e-06, + "loss": 0.5198, + "step": 1292 + }, + { + "epoch": 0.08844654217114714, + "grad_norm": 2.3445086936030872, + "learning_rate": 9.910770787214106e-06, + "loss": 0.3424, + "step": 1293 + }, + { + "epoch": 0.08851494630275669, + "grad_norm": 2.550500559296179, + "learning_rate": 9.910562322746083e-06, + "loss": 0.5032, + "step": 1294 + }, + { + "epoch": 0.08858335043436624, + "grad_norm": 2.451789565356712, + "learning_rate": 9.910353617243944e-06, + "loss": 0.6054, + "step": 1295 + }, + { + "epoch": 0.08865175456597578, + "grad_norm": 1.56020691454614, + "learning_rate": 9.910144670717942e-06, + "loss": 0.1841, + "step": 1296 + }, + { + "epoch": 0.08872015869758533, + "grad_norm": 2.9921403487750844, + "learning_rate": 9.909935483178326e-06, + "loss": 0.7376, + "step": 1297 + }, + { + "epoch": 0.08878856282919488, + "grad_norm": 3.053345680656066, + "learning_rate": 9.909726054635369e-06, + "loss": 0.7177, + "step": 1298 + }, + { + "epoch": 0.08885696696080443, + "grad_norm": 2.3000197777727043, + "learning_rate": 9.909516385099346e-06, + "loss": 0.5793, + "step": 1299 + }, + { + "epoch": 0.08892537109241398, + "grad_norm": 2.1552811533034038, + "learning_rate": 9.909306474580551e-06, + "loss": 0.4253, + "step": 1300 + }, + { + "epoch": 0.08899377522402353, + "grad_norm": 2.218106948991913, + "learning_rate": 9.909096323089288e-06, + "loss": 0.3621, + "step": 1301 + }, + { + "epoch": 0.08906217935563308, + "grad_norm": 2.8055134622298112, + "learning_rate": 9.908885930635872e-06, + "loss": 0.5759, + "step": 1302 + }, + { + "epoch": 0.08913058348724263, + "grad_norm": 1.8458644836777451, + "learning_rate": 9.90867529723063e-06, + "loss": 0.2672, + "step": 1303 + }, + { + "epoch": 0.08919898761885219, + "grad_norm": 2.5636649972595835, + "learning_rate": 9.908464422883901e-06, + "loss": 0.6844, + "step": 1304 + }, + { + "epoch": 0.08926739175046172, + "grad_norm": 2.3566916218858993, + "learning_rate": 9.908253307606035e-06, + "loss": 0.5097, + "step": 1305 + }, + { + "epoch": 0.08933579588207127, + "grad_norm": 2.1653574639297193, + "learning_rate": 9.908041951407394e-06, + "loss": 0.4159, + "step": 1306 + }, + { + "epoch": 0.08940420001368082, + "grad_norm": 1.8298672622800063, + "learning_rate": 9.907830354298355e-06, + "loss": 0.3097, + "step": 1307 + }, + { + "epoch": 0.08947260414529037, + "grad_norm": 2.2774044604576362, + "learning_rate": 9.907618516289301e-06, + "loss": 0.3361, + "step": 1308 + }, + { + "epoch": 0.08954100827689992, + "grad_norm": 2.4647810454939516, + "learning_rate": 9.907406437390632e-06, + "loss": 0.6572, + "step": 1309 + }, + { + "epoch": 0.08960941240850948, + "grad_norm": 2.439012608332497, + "learning_rate": 9.907194117612757e-06, + "loss": 0.2679, + "step": 1310 + }, + { + "epoch": 0.08967781654011903, + "grad_norm": 2.25757856378057, + "learning_rate": 9.9069815569661e-06, + "loss": 0.2744, + "step": 1311 + }, + { + "epoch": 0.08974622067172858, + "grad_norm": 2.7387758425978044, + "learning_rate": 9.906768755461092e-06, + "loss": 0.7606, + "step": 1312 + }, + { + "epoch": 0.08981462480333813, + "grad_norm": 2.015129882416384, + "learning_rate": 9.906555713108176e-06, + "loss": 0.3065, + "step": 1313 + }, + { + "epoch": 0.08988302893494766, + "grad_norm": 1.6721628623099793, + "learning_rate": 9.906342429917815e-06, + "loss": 0.2603, + "step": 1314 + }, + { + "epoch": 0.08995143306655722, + "grad_norm": 1.9142606360199912, + "learning_rate": 9.906128905900475e-06, + "loss": 0.1653, + "step": 1315 + }, + { + "epoch": 0.09001983719816677, + "grad_norm": 3.6653392007151027, + "learning_rate": 9.905915141066636e-06, + "loss": 0.2638, + "step": 1316 + }, + { + "epoch": 0.09008824132977632, + "grad_norm": 2.307658223785356, + "learning_rate": 9.905701135426792e-06, + "loss": 0.2956, + "step": 1317 + }, + { + "epoch": 0.09015664546138587, + "grad_norm": 1.923352425721254, + "learning_rate": 9.905486888991446e-06, + "loss": 0.2678, + "step": 1318 + }, + { + "epoch": 0.09022504959299542, + "grad_norm": 2.359593747589941, + "learning_rate": 9.905272401771115e-06, + "loss": 0.4605, + "step": 1319 + }, + { + "epoch": 0.09029345372460497, + "grad_norm": 2.2182309658463915, + "learning_rate": 9.905057673776329e-06, + "loss": 0.3914, + "step": 1320 + }, + { + "epoch": 0.09036185785621452, + "grad_norm": 2.5099586264175393, + "learning_rate": 9.904842705017625e-06, + "loss": 0.5, + "step": 1321 + }, + { + "epoch": 0.09043026198782407, + "grad_norm": 2.2938992173722585, + "learning_rate": 9.904627495505557e-06, + "loss": 0.4293, + "step": 1322 + }, + { + "epoch": 0.09049866611943361, + "grad_norm": 2.3230813296649315, + "learning_rate": 9.904412045250686e-06, + "loss": 0.3574, + "step": 1323 + }, + { + "epoch": 0.09056707025104316, + "grad_norm": 2.232738729370104, + "learning_rate": 9.904196354263588e-06, + "loss": 0.372, + "step": 1324 + }, + { + "epoch": 0.09063547438265271, + "grad_norm": 2.4477371041744544, + "learning_rate": 9.903980422554851e-06, + "loss": 0.4428, + "step": 1325 + }, + { + "epoch": 0.09070387851426226, + "grad_norm": 3.199474557377205, + "learning_rate": 9.903764250135075e-06, + "loss": 0.6211, + "step": 1326 + }, + { + "epoch": 0.09077228264587181, + "grad_norm": 2.166368564094966, + "learning_rate": 9.903547837014868e-06, + "loss": 0.3557, + "step": 1327 + }, + { + "epoch": 0.09084068677748136, + "grad_norm": 4.335576429024888, + "learning_rate": 9.903331183204856e-06, + "loss": 0.3968, + "step": 1328 + }, + { + "epoch": 0.09090909090909091, + "grad_norm": 2.1703483924616687, + "learning_rate": 9.903114288715671e-06, + "loss": 0.3874, + "step": 1329 + }, + { + "epoch": 0.09097749504070046, + "grad_norm": 1.9481936494139673, + "learning_rate": 9.902897153557959e-06, + "loss": 0.3615, + "step": 1330 + }, + { + "epoch": 0.09104589917231001, + "grad_norm": 2.0647675302779533, + "learning_rate": 9.90267977774238e-06, + "loss": 0.3043, + "step": 1331 + }, + { + "epoch": 0.09111430330391955, + "grad_norm": 2.6350716539698364, + "learning_rate": 9.902462161279601e-06, + "loss": 0.5025, + "step": 1332 + }, + { + "epoch": 0.0911827074355291, + "grad_norm": 3.738807673046127, + "learning_rate": 9.902244304180306e-06, + "loss": 0.5365, + "step": 1333 + }, + { + "epoch": 0.09125111156713865, + "grad_norm": 2.2616950249621732, + "learning_rate": 9.90202620645519e-06, + "loss": 0.5189, + "step": 1334 + }, + { + "epoch": 0.0913195156987482, + "grad_norm": 1.6386319202194326, + "learning_rate": 9.901807868114955e-06, + "loss": 0.2507, + "step": 1335 + }, + { + "epoch": 0.09138791983035775, + "grad_norm": 2.4742760557976813, + "learning_rate": 9.901589289170316e-06, + "loss": 0.3606, + "step": 1336 + }, + { + "epoch": 0.0914563239619673, + "grad_norm": 1.4728610670397417, + "learning_rate": 9.90137046963201e-06, + "loss": 0.2756, + "step": 1337 + }, + { + "epoch": 0.09152472809357685, + "grad_norm": 2.73158406194022, + "learning_rate": 9.901151409510769e-06, + "loss": 0.5913, + "step": 1338 + }, + { + "epoch": 0.0915931322251864, + "grad_norm": 2.27181330076284, + "learning_rate": 9.900932108817352e-06, + "loss": 0.3542, + "step": 1339 + }, + { + "epoch": 0.09166153635679596, + "grad_norm": 2.1018312204425613, + "learning_rate": 9.90071256756252e-06, + "loss": 0.5466, + "step": 1340 + }, + { + "epoch": 0.0917299404884055, + "grad_norm": 2.4421588714200504, + "learning_rate": 9.90049278575705e-06, + "loss": 0.5123, + "step": 1341 + }, + { + "epoch": 0.09179834462001504, + "grad_norm": 2.3501190114829384, + "learning_rate": 9.90027276341173e-06, + "loss": 0.3098, + "step": 1342 + }, + { + "epoch": 0.0918667487516246, + "grad_norm": 4.192025984056581, + "learning_rate": 9.900052500537358e-06, + "loss": 0.4786, + "step": 1343 + }, + { + "epoch": 0.09193515288323414, + "grad_norm": 2.5088999045517992, + "learning_rate": 9.899831997144749e-06, + "loss": 0.6745, + "step": 1344 + }, + { + "epoch": 0.0920035570148437, + "grad_norm": 1.821937834112541, + "learning_rate": 9.899611253244722e-06, + "loss": 0.2788, + "step": 1345 + }, + { + "epoch": 0.09207196114645325, + "grad_norm": 2.0612109086058386, + "learning_rate": 9.899390268848118e-06, + "loss": 0.3823, + "step": 1346 + }, + { + "epoch": 0.0921403652780628, + "grad_norm": 2.2311689410191535, + "learning_rate": 9.899169043965779e-06, + "loss": 0.5482, + "step": 1347 + }, + { + "epoch": 0.09220876940967235, + "grad_norm": 2.1503028817080203, + "learning_rate": 9.898947578608567e-06, + "loss": 0.5428, + "step": 1348 + }, + { + "epoch": 0.0922771735412819, + "grad_norm": 1.8751873479069472, + "learning_rate": 9.898725872787349e-06, + "loss": 0.3721, + "step": 1349 + }, + { + "epoch": 0.09234557767289145, + "grad_norm": 2.793416287505617, + "learning_rate": 9.89850392651301e-06, + "loss": 0.4974, + "step": 1350 + }, + { + "epoch": 0.09241398180450099, + "grad_norm": 3.1683075861224985, + "learning_rate": 9.898281739796444e-06, + "loss": 0.6832, + "step": 1351 + }, + { + "epoch": 0.09248238593611054, + "grad_norm": 1.6571871966944933, + "learning_rate": 9.898059312648557e-06, + "loss": 0.2803, + "step": 1352 + }, + { + "epoch": 0.09255079006772009, + "grad_norm": 1.9949742367283172, + "learning_rate": 9.897836645080266e-06, + "loss": 0.2813, + "step": 1353 + }, + { + "epoch": 0.09261919419932964, + "grad_norm": 2.563502859798151, + "learning_rate": 9.897613737102501e-06, + "loss": 0.5297, + "step": 1354 + }, + { + "epoch": 0.09268759833093919, + "grad_norm": 2.6071111715410664, + "learning_rate": 9.897390588726203e-06, + "loss": 0.7275, + "step": 1355 + }, + { + "epoch": 0.09275600246254874, + "grad_norm": 2.1503570754979395, + "learning_rate": 9.897167199962326e-06, + "loss": 0.298, + "step": 1356 + }, + { + "epoch": 0.09282440659415829, + "grad_norm": 1.9607871252511941, + "learning_rate": 9.896943570821835e-06, + "loss": 0.3875, + "step": 1357 + }, + { + "epoch": 0.09289281072576784, + "grad_norm": 2.0881229517600395, + "learning_rate": 9.896719701315708e-06, + "loss": 0.4681, + "step": 1358 + }, + { + "epoch": 0.09296121485737739, + "grad_norm": 2.557255378874864, + "learning_rate": 9.896495591454929e-06, + "loss": 0.5841, + "step": 1359 + }, + { + "epoch": 0.09302961898898693, + "grad_norm": 2.470334287109426, + "learning_rate": 9.896271241250502e-06, + "loss": 0.5539, + "step": 1360 + }, + { + "epoch": 0.09309802312059648, + "grad_norm": 2.3737928685741245, + "learning_rate": 9.89604665071344e-06, + "loss": 0.4936, + "step": 1361 + }, + { + "epoch": 0.09316642725220603, + "grad_norm": 1.6020962756067796, + "learning_rate": 9.895821819854765e-06, + "loss": 0.2783, + "step": 1362 + }, + { + "epoch": 0.09323483138381558, + "grad_norm": 2.0172261995626326, + "learning_rate": 9.895596748685513e-06, + "loss": 0.2171, + "step": 1363 + }, + { + "epoch": 0.09330323551542513, + "grad_norm": 1.8782829273807584, + "learning_rate": 9.895371437216733e-06, + "loss": 0.1697, + "step": 1364 + }, + { + "epoch": 0.09337163964703468, + "grad_norm": 2.085996334089165, + "learning_rate": 9.895145885459482e-06, + "loss": 0.3198, + "step": 1365 + }, + { + "epoch": 0.09344004377864423, + "grad_norm": 2.07455160412145, + "learning_rate": 9.894920093424834e-06, + "loss": 0.3331, + "step": 1366 + }, + { + "epoch": 0.09350844791025378, + "grad_norm": 2.3232370819895167, + "learning_rate": 9.894694061123868e-06, + "loss": 0.4321, + "step": 1367 + }, + { + "epoch": 0.09357685204186333, + "grad_norm": 2.791268472717152, + "learning_rate": 9.894467788567684e-06, + "loss": 0.508, + "step": 1368 + }, + { + "epoch": 0.09364525617347287, + "grad_norm": 2.0117452953031654, + "learning_rate": 9.894241275767385e-06, + "loss": 0.3367, + "step": 1369 + }, + { + "epoch": 0.09371366030508242, + "grad_norm": 1.9932229660877372, + "learning_rate": 9.89401452273409e-06, + "loss": 0.384, + "step": 1370 + }, + { + "epoch": 0.09378206443669197, + "grad_norm": 2.1056660611017817, + "learning_rate": 9.89378752947893e-06, + "loss": 0.3717, + "step": 1371 + }, + { + "epoch": 0.09385046856830152, + "grad_norm": 2.0505531256815863, + "learning_rate": 9.893560296013045e-06, + "loss": 0.5726, + "step": 1372 + }, + { + "epoch": 0.09391887269991107, + "grad_norm": 2.263799533749674, + "learning_rate": 9.89333282234759e-06, + "loss": 0.4061, + "step": 1373 + }, + { + "epoch": 0.09398727683152062, + "grad_norm": 2.286928216941061, + "learning_rate": 9.89310510849373e-06, + "loss": 0.4437, + "step": 1374 + }, + { + "epoch": 0.09405568096313018, + "grad_norm": 2.245037082406216, + "learning_rate": 9.892877154462643e-06, + "loss": 0.4217, + "step": 1375 + }, + { + "epoch": 0.09412408509473973, + "grad_norm": 2.1087190809170897, + "learning_rate": 9.892648960265518e-06, + "loss": 0.4713, + "step": 1376 + }, + { + "epoch": 0.09419248922634928, + "grad_norm": 2.183330984120403, + "learning_rate": 9.892420525913557e-06, + "loss": 0.5301, + "step": 1377 + }, + { + "epoch": 0.09426089335795883, + "grad_norm": 1.7784116261734508, + "learning_rate": 9.89219185141797e-06, + "loss": 0.2165, + "step": 1378 + }, + { + "epoch": 0.09432929748956836, + "grad_norm": 2.338540862379948, + "learning_rate": 9.891962936789983e-06, + "loss": 0.35, + "step": 1379 + }, + { + "epoch": 0.09439770162117792, + "grad_norm": 3.22002456947395, + "learning_rate": 9.891733782040832e-06, + "loss": 0.5321, + "step": 1380 + }, + { + "epoch": 0.09446610575278747, + "grad_norm": 2.14327029033867, + "learning_rate": 9.891504387181765e-06, + "loss": 0.2843, + "step": 1381 + }, + { + "epoch": 0.09453450988439702, + "grad_norm": 2.5299616790543755, + "learning_rate": 9.891274752224041e-06, + "loss": 0.4522, + "step": 1382 + }, + { + "epoch": 0.09460291401600657, + "grad_norm": 2.053620543692243, + "learning_rate": 9.891044877178934e-06, + "loss": 0.3935, + "step": 1383 + }, + { + "epoch": 0.09467131814761612, + "grad_norm": 2.28561150344231, + "learning_rate": 9.890814762057725e-06, + "loss": 0.5753, + "step": 1384 + }, + { + "epoch": 0.09473972227922567, + "grad_norm": 2.5584457587619855, + "learning_rate": 9.89058440687171e-06, + "loss": 0.7715, + "step": 1385 + }, + { + "epoch": 0.09480812641083522, + "grad_norm": 2.769382607327539, + "learning_rate": 9.890353811632195e-06, + "loss": 0.6922, + "step": 1386 + }, + { + "epoch": 0.09487653054244477, + "grad_norm": 2.0688895175598034, + "learning_rate": 9.890122976350501e-06, + "loss": 0.4702, + "step": 1387 + }, + { + "epoch": 0.09494493467405431, + "grad_norm": 2.077134729115061, + "learning_rate": 9.889891901037957e-06, + "loss": 0.1961, + "step": 1388 + }, + { + "epoch": 0.09501333880566386, + "grad_norm": 2.2881193105662847, + "learning_rate": 9.889660585705905e-06, + "loss": 0.2391, + "step": 1389 + }, + { + "epoch": 0.09508174293727341, + "grad_norm": 1.9254030123169168, + "learning_rate": 9.8894290303657e-06, + "loss": 0.4332, + "step": 1390 + }, + { + "epoch": 0.09515014706888296, + "grad_norm": 2.7069980129386844, + "learning_rate": 9.889197235028707e-06, + "loss": 0.5862, + "step": 1391 + }, + { + "epoch": 0.09521855120049251, + "grad_norm": 1.7929503163572342, + "learning_rate": 9.888965199706304e-06, + "loss": 0.3413, + "step": 1392 + }, + { + "epoch": 0.09528695533210206, + "grad_norm": 2.4751257233524955, + "learning_rate": 9.888732924409881e-06, + "loss": 0.7001, + "step": 1393 + }, + { + "epoch": 0.09535535946371161, + "grad_norm": 2.449768394110744, + "learning_rate": 9.88850040915084e-06, + "loss": 0.5703, + "step": 1394 + }, + { + "epoch": 0.09542376359532116, + "grad_norm": 2.137603973904948, + "learning_rate": 9.888267653940589e-06, + "loss": 0.3145, + "step": 1395 + }, + { + "epoch": 0.09549216772693071, + "grad_norm": 2.3826287433248035, + "learning_rate": 9.888034658790558e-06, + "loss": 0.4986, + "step": 1396 + }, + { + "epoch": 0.09556057185854025, + "grad_norm": 2.3909039116513084, + "learning_rate": 9.887801423712182e-06, + "loss": 0.4985, + "step": 1397 + }, + { + "epoch": 0.0956289759901498, + "grad_norm": 2.3344532579431085, + "learning_rate": 9.88756794871691e-06, + "loss": 0.4693, + "step": 1398 + }, + { + "epoch": 0.09569738012175935, + "grad_norm": 1.9864608210198187, + "learning_rate": 9.887334233816199e-06, + "loss": 0.3817, + "step": 1399 + }, + { + "epoch": 0.0957657842533689, + "grad_norm": 1.782483142516261, + "learning_rate": 9.887100279021525e-06, + "loss": 0.2023, + "step": 1400 + }, + { + "epoch": 0.09583418838497845, + "grad_norm": 2.374454021439254, + "learning_rate": 9.88686608434437e-06, + "loss": 0.4458, + "step": 1401 + }, + { + "epoch": 0.095902592516588, + "grad_norm": 2.2180264596126187, + "learning_rate": 9.886631649796227e-06, + "loss": 0.4697, + "step": 1402 + }, + { + "epoch": 0.09597099664819755, + "grad_norm": 2.3892548817329953, + "learning_rate": 9.886396975388608e-06, + "loss": 0.3303, + "step": 1403 + }, + { + "epoch": 0.0960394007798071, + "grad_norm": 2.2817361653287724, + "learning_rate": 9.886162061133027e-06, + "loss": 0.3705, + "step": 1404 + }, + { + "epoch": 0.09610780491141666, + "grad_norm": 2.1252003603330927, + "learning_rate": 9.88592690704102e-06, + "loss": 0.4748, + "step": 1405 + }, + { + "epoch": 0.09617620904302619, + "grad_norm": 2.3190153587041618, + "learning_rate": 9.885691513124123e-06, + "loss": 0.4763, + "step": 1406 + }, + { + "epoch": 0.09624461317463574, + "grad_norm": 2.098329450249835, + "learning_rate": 9.885455879393896e-06, + "loss": 0.3347, + "step": 1407 + }, + { + "epoch": 0.0963130173062453, + "grad_norm": 2.2966958319921122, + "learning_rate": 9.885220005861901e-06, + "loss": 0.492, + "step": 1408 + }, + { + "epoch": 0.09638142143785484, + "grad_norm": 2.648408343849497, + "learning_rate": 9.884983892539719e-06, + "loss": 0.5811, + "step": 1409 + }, + { + "epoch": 0.0964498255694644, + "grad_norm": 1.8474800235509696, + "learning_rate": 9.884747539438939e-06, + "loss": 0.3621, + "step": 1410 + }, + { + "epoch": 0.09651822970107395, + "grad_norm": 2.5041456420532087, + "learning_rate": 9.884510946571159e-06, + "loss": 0.5254, + "step": 1411 + }, + { + "epoch": 0.0965866338326835, + "grad_norm": 1.7498699889254017, + "learning_rate": 9.884274113947997e-06, + "loss": 0.3544, + "step": 1412 + }, + { + "epoch": 0.09665503796429305, + "grad_norm": 2.2106687564418164, + "learning_rate": 9.884037041581074e-06, + "loss": 0.4624, + "step": 1413 + }, + { + "epoch": 0.0967234420959026, + "grad_norm": 2.3066077557062195, + "learning_rate": 9.883799729482029e-06, + "loss": 0.595, + "step": 1414 + }, + { + "epoch": 0.09679184622751213, + "grad_norm": 1.919557934685412, + "learning_rate": 9.883562177662512e-06, + "loss": 0.2066, + "step": 1415 + }, + { + "epoch": 0.09686025035912169, + "grad_norm": 2.427854011761132, + "learning_rate": 9.883324386134177e-06, + "loss": 0.2919, + "step": 1416 + }, + { + "epoch": 0.09692865449073124, + "grad_norm": 2.3661746511538473, + "learning_rate": 9.883086354908703e-06, + "loss": 0.3589, + "step": 1417 + }, + { + "epoch": 0.09699705862234079, + "grad_norm": 2.039778996421973, + "learning_rate": 9.882848083997766e-06, + "loss": 0.5176, + "step": 1418 + }, + { + "epoch": 0.09706546275395034, + "grad_norm": 3.1419896392138487, + "learning_rate": 9.88260957341307e-06, + "loss": 0.6523, + "step": 1419 + }, + { + "epoch": 0.09713386688555989, + "grad_norm": 1.7700366378483698, + "learning_rate": 9.882370823166317e-06, + "loss": 0.3068, + "step": 1420 + }, + { + "epoch": 0.09720227101716944, + "grad_norm": 1.7318911164083504, + "learning_rate": 9.882131833269227e-06, + "loss": 0.3188, + "step": 1421 + }, + { + "epoch": 0.09727067514877899, + "grad_norm": 2.197635221974013, + "learning_rate": 9.881892603733532e-06, + "loss": 0.2715, + "step": 1422 + }, + { + "epoch": 0.09733907928038854, + "grad_norm": 2.56292599533133, + "learning_rate": 9.881653134570972e-06, + "loss": 0.2877, + "step": 1423 + }, + { + "epoch": 0.09740748341199809, + "grad_norm": 2.0892981949079448, + "learning_rate": 9.881413425793305e-06, + "loss": 0.5293, + "step": 1424 + }, + { + "epoch": 0.09747588754360763, + "grad_norm": 2.3304941424713803, + "learning_rate": 9.881173477412293e-06, + "loss": 0.6877, + "step": 1425 + }, + { + "epoch": 0.09754429167521718, + "grad_norm": 2.2711029322871292, + "learning_rate": 9.880933289439717e-06, + "loss": 0.3724, + "step": 1426 + }, + { + "epoch": 0.09761269580682673, + "grad_norm": 2.1026440854205073, + "learning_rate": 9.880692861887364e-06, + "loss": 0.4124, + "step": 1427 + }, + { + "epoch": 0.09768109993843628, + "grad_norm": 2.159219911391296, + "learning_rate": 9.880452194767039e-06, + "loss": 0.304, + "step": 1428 + }, + { + "epoch": 0.09774950407004583, + "grad_norm": 2.649426278502004, + "learning_rate": 9.880211288090551e-06, + "loss": 0.6371, + "step": 1429 + }, + { + "epoch": 0.09781790820165538, + "grad_norm": 2.188143860202586, + "learning_rate": 9.879970141869726e-06, + "loss": 0.4189, + "step": 1430 + }, + { + "epoch": 0.09788631233326493, + "grad_norm": 1.4768101116233592, + "learning_rate": 9.879728756116403e-06, + "loss": 0.2027, + "step": 1431 + }, + { + "epoch": 0.09795471646487448, + "grad_norm": 2.059262993911311, + "learning_rate": 9.879487130842428e-06, + "loss": 0.3849, + "step": 1432 + }, + { + "epoch": 0.09802312059648403, + "grad_norm": 1.9961369877441608, + "learning_rate": 9.879245266059662e-06, + "loss": 0.5059, + "step": 1433 + }, + { + "epoch": 0.09809152472809357, + "grad_norm": 2.3385311514455895, + "learning_rate": 9.879003161779976e-06, + "loss": 0.4525, + "step": 1434 + }, + { + "epoch": 0.09815992885970312, + "grad_norm": 2.3002372124395993, + "learning_rate": 9.878760818015257e-06, + "loss": 0.4463, + "step": 1435 + }, + { + "epoch": 0.09822833299131267, + "grad_norm": 2.1420107310154974, + "learning_rate": 9.878518234777395e-06, + "loss": 0.2475, + "step": 1436 + }, + { + "epoch": 0.09829673712292222, + "grad_norm": 2.2946317708509403, + "learning_rate": 9.8782754120783e-06, + "loss": 0.5311, + "step": 1437 + }, + { + "epoch": 0.09836514125453177, + "grad_norm": 1.5844192895586515, + "learning_rate": 9.878032349929892e-06, + "loss": 0.3524, + "step": 1438 + }, + { + "epoch": 0.09843354538614132, + "grad_norm": 3.0439087538471212, + "learning_rate": 9.8777890483441e-06, + "loss": 0.6501, + "step": 1439 + }, + { + "epoch": 0.09850194951775088, + "grad_norm": 2.0527226768036915, + "learning_rate": 9.877545507332867e-06, + "loss": 0.3911, + "step": 1440 + }, + { + "epoch": 0.09857035364936043, + "grad_norm": 4.718384179088707, + "learning_rate": 9.877301726908146e-06, + "loss": 0.413, + "step": 1441 + }, + { + "epoch": 0.09863875778096998, + "grad_norm": 1.6838746776879254, + "learning_rate": 9.877057707081905e-06, + "loss": 0.4392, + "step": 1442 + }, + { + "epoch": 0.09870716191257951, + "grad_norm": 2.2789844668215964, + "learning_rate": 9.876813447866122e-06, + "loss": 0.2252, + "step": 1443 + }, + { + "epoch": 0.09877556604418906, + "grad_norm": 2.4278264139517893, + "learning_rate": 9.876568949272784e-06, + "loss": 0.4149, + "step": 1444 + }, + { + "epoch": 0.09884397017579861, + "grad_norm": 2.795938958538268, + "learning_rate": 9.876324211313893e-06, + "loss": 0.4561, + "step": 1445 + }, + { + "epoch": 0.09891237430740817, + "grad_norm": 2.0418663186786965, + "learning_rate": 9.87607923400146e-06, + "loss": 0.4302, + "step": 1446 + }, + { + "epoch": 0.09898077843901772, + "grad_norm": 2.551044047489408, + "learning_rate": 9.875834017347515e-06, + "loss": 0.6083, + "step": 1447 + }, + { + "epoch": 0.09904918257062727, + "grad_norm": 1.8510740134747994, + "learning_rate": 9.87558856136409e-06, + "loss": 0.4429, + "step": 1448 + }, + { + "epoch": 0.09911758670223682, + "grad_norm": 2.127196701716963, + "learning_rate": 9.875342866063235e-06, + "loss": 0.2942, + "step": 1449 + }, + { + "epoch": 0.09918599083384637, + "grad_norm": 2.3576169885327647, + "learning_rate": 9.87509693145701e-06, + "loss": 0.5223, + "step": 1450 + }, + { + "epoch": 0.09925439496545592, + "grad_norm": 3.734870909639304, + "learning_rate": 9.874850757557483e-06, + "loss": 0.4812, + "step": 1451 + }, + { + "epoch": 0.09932279909706546, + "grad_norm": 2.2790177998983787, + "learning_rate": 9.874604344376743e-06, + "loss": 0.5075, + "step": 1452 + }, + { + "epoch": 0.099391203228675, + "grad_norm": 2.5195437231420454, + "learning_rate": 9.874357691926883e-06, + "loss": 0.6485, + "step": 1453 + }, + { + "epoch": 0.09945960736028456, + "grad_norm": 2.439184328270941, + "learning_rate": 9.874110800220008e-06, + "loss": 0.5913, + "step": 1454 + }, + { + "epoch": 0.09952801149189411, + "grad_norm": 2.0330309762916854, + "learning_rate": 9.873863669268239e-06, + "loss": 0.3461, + "step": 1455 + }, + { + "epoch": 0.09959641562350366, + "grad_norm": 1.674432376150482, + "learning_rate": 9.873616299083705e-06, + "loss": 0.2457, + "step": 1456 + }, + { + "epoch": 0.09966481975511321, + "grad_norm": 1.633517893229935, + "learning_rate": 9.87336868967855e-06, + "loss": 0.3234, + "step": 1457 + }, + { + "epoch": 0.09973322388672276, + "grad_norm": 4.823930615920926, + "learning_rate": 9.873120841064924e-06, + "loss": 0.3307, + "step": 1458 + }, + { + "epoch": 0.09980162801833231, + "grad_norm": 2.2280510828616555, + "learning_rate": 9.872872753254996e-06, + "loss": 0.4589, + "step": 1459 + }, + { + "epoch": 0.09987003214994186, + "grad_norm": 2.3998207458718843, + "learning_rate": 9.872624426260942e-06, + "loss": 0.4665, + "step": 1460 + }, + { + "epoch": 0.0999384362815514, + "grad_norm": 1.9555848157685283, + "learning_rate": 9.872375860094951e-06, + "loss": 0.3268, + "step": 1461 + }, + { + "epoch": 0.10000684041316095, + "grad_norm": 2.5956397927898593, + "learning_rate": 9.872127054769227e-06, + "loss": 0.5275, + "step": 1462 + }, + { + "epoch": 0.1000752445447705, + "grad_norm": 2.029738094618409, + "learning_rate": 9.871878010295978e-06, + "loss": 0.3171, + "step": 1463 + }, + { + "epoch": 0.10014364867638005, + "grad_norm": 3.133158771660856, + "learning_rate": 9.87162872668743e-06, + "loss": 0.8681, + "step": 1464 + }, + { + "epoch": 0.1002120528079896, + "grad_norm": 2.402182820209993, + "learning_rate": 9.87137920395582e-06, + "loss": 0.5139, + "step": 1465 + }, + { + "epoch": 0.10028045693959915, + "grad_norm": 2.2205357954047273, + "learning_rate": 9.871129442113394e-06, + "loss": 0.5019, + "step": 1466 + }, + { + "epoch": 0.1003488610712087, + "grad_norm": 2.26197678406798, + "learning_rate": 9.870879441172414e-06, + "loss": 0.6009, + "step": 1467 + }, + { + "epoch": 0.10041726520281825, + "grad_norm": 1.9419147355158268, + "learning_rate": 9.870629201145149e-06, + "loss": 0.3085, + "step": 1468 + }, + { + "epoch": 0.1004856693344278, + "grad_norm": 3.9191658724124943, + "learning_rate": 9.870378722043882e-06, + "loss": 0.2614, + "step": 1469 + }, + { + "epoch": 0.10055407346603736, + "grad_norm": 1.689324906615917, + "learning_rate": 9.87012800388091e-06, + "loss": 0.3248, + "step": 1470 + }, + { + "epoch": 0.10062247759764689, + "grad_norm": 1.9273391558125035, + "learning_rate": 9.869877046668536e-06, + "loss": 0.4139, + "step": 1471 + }, + { + "epoch": 0.10069088172925644, + "grad_norm": 2.002638737578967, + "learning_rate": 9.869625850419083e-06, + "loss": 0.3316, + "step": 1472 + }, + { + "epoch": 0.100759285860866, + "grad_norm": 1.6024487346179612, + "learning_rate": 9.869374415144876e-06, + "loss": 0.2615, + "step": 1473 + }, + { + "epoch": 0.10082768999247554, + "grad_norm": 2.310727316355756, + "learning_rate": 9.869122740858259e-06, + "loss": 0.3379, + "step": 1474 + }, + { + "epoch": 0.1008960941240851, + "grad_norm": 2.394373193101393, + "learning_rate": 9.868870827571585e-06, + "loss": 0.4258, + "step": 1475 + }, + { + "epoch": 0.10096449825569465, + "grad_norm": 1.891604339329735, + "learning_rate": 9.868618675297219e-06, + "loss": 0.3564, + "step": 1476 + }, + { + "epoch": 0.1010329023873042, + "grad_norm": 2.1729381415444933, + "learning_rate": 9.86836628404754e-06, + "loss": 0.3913, + "step": 1477 + }, + { + "epoch": 0.10110130651891375, + "grad_norm": 2.225832361000684, + "learning_rate": 9.868113653834934e-06, + "loss": 0.563, + "step": 1478 + }, + { + "epoch": 0.1011697106505233, + "grad_norm": 3.857386319926131, + "learning_rate": 9.8678607846718e-06, + "loss": 0.4169, + "step": 1479 + }, + { + "epoch": 0.10123811478213283, + "grad_norm": 2.177718185569903, + "learning_rate": 9.867607676570554e-06, + "loss": 0.4746, + "step": 1480 + }, + { + "epoch": 0.10130651891374239, + "grad_norm": 2.1612335850405016, + "learning_rate": 9.867354329543617e-06, + "loss": 0.4657, + "step": 1481 + }, + { + "epoch": 0.10137492304535194, + "grad_norm": 2.679750135929771, + "learning_rate": 9.867100743603425e-06, + "loss": 0.7071, + "step": 1482 + }, + { + "epoch": 0.10144332717696149, + "grad_norm": 2.152305699616163, + "learning_rate": 9.866846918762426e-06, + "loss": 0.3735, + "step": 1483 + }, + { + "epoch": 0.10151173130857104, + "grad_norm": 2.4057632145877967, + "learning_rate": 9.86659285503308e-06, + "loss": 0.2301, + "step": 1484 + }, + { + "epoch": 0.10158013544018059, + "grad_norm": 2.0627471096664447, + "learning_rate": 9.866338552427852e-06, + "loss": 0.2513, + "step": 1485 + }, + { + "epoch": 0.10164853957179014, + "grad_norm": 2.2625099866867955, + "learning_rate": 9.866084010959232e-06, + "loss": 0.4654, + "step": 1486 + }, + { + "epoch": 0.10171694370339969, + "grad_norm": 2.51169853142985, + "learning_rate": 9.86582923063971e-06, + "loss": 0.6371, + "step": 1487 + }, + { + "epoch": 0.10178534783500924, + "grad_norm": 1.6279114949308828, + "learning_rate": 9.865574211481792e-06, + "loss": 0.3199, + "step": 1488 + }, + { + "epoch": 0.10185375196661878, + "grad_norm": 1.7902938236934396, + "learning_rate": 9.865318953497997e-06, + "loss": 0.2681, + "step": 1489 + }, + { + "epoch": 0.10192215609822833, + "grad_norm": 1.538821274108402, + "learning_rate": 9.865063456700851e-06, + "loss": 0.2271, + "step": 1490 + }, + { + "epoch": 0.10199056022983788, + "grad_norm": 2.2297536552121016, + "learning_rate": 9.8648077211029e-06, + "loss": 0.3975, + "step": 1491 + }, + { + "epoch": 0.10205896436144743, + "grad_norm": 2.3887129005783008, + "learning_rate": 9.864551746716694e-06, + "loss": 0.2127, + "step": 1492 + }, + { + "epoch": 0.10212736849305698, + "grad_norm": 2.656975663861877, + "learning_rate": 9.864295533554798e-06, + "loss": 0.6215, + "step": 1493 + }, + { + "epoch": 0.10219577262466653, + "grad_norm": 2.127116929989824, + "learning_rate": 9.864039081629786e-06, + "loss": 0.3034, + "step": 1494 + }, + { + "epoch": 0.10226417675627608, + "grad_norm": 2.3206288927961514, + "learning_rate": 9.863782390954248e-06, + "loss": 0.5487, + "step": 1495 + }, + { + "epoch": 0.10233258088788563, + "grad_norm": 1.9611045530676792, + "learning_rate": 9.863525461540784e-06, + "loss": 0.3678, + "step": 1496 + }, + { + "epoch": 0.10240098501949518, + "grad_norm": 1.8890158686155625, + "learning_rate": 9.863268293402005e-06, + "loss": 0.4011, + "step": 1497 + }, + { + "epoch": 0.10246938915110472, + "grad_norm": 2.4053385853898606, + "learning_rate": 9.863010886550532e-06, + "loss": 0.4334, + "step": 1498 + }, + { + "epoch": 0.10253779328271427, + "grad_norm": 2.19606970221219, + "learning_rate": 9.862753240999001e-06, + "loss": 0.3785, + "step": 1499 + }, + { + "epoch": 0.10260619741432382, + "grad_norm": 2.5034898837262083, + "learning_rate": 9.86249535676006e-06, + "loss": 0.4143, + "step": 1500 + }, + { + "epoch": 0.10267460154593337, + "grad_norm": 2.0738638594022754, + "learning_rate": 9.862237233846367e-06, + "loss": 0.4562, + "step": 1501 + }, + { + "epoch": 0.10274300567754292, + "grad_norm": 5.056913354502133, + "learning_rate": 9.86197887227059e-06, + "loss": 0.8505, + "step": 1502 + }, + { + "epoch": 0.10281140980915247, + "grad_norm": 1.3471274305869285, + "learning_rate": 9.861720272045413e-06, + "loss": 0.1232, + "step": 1503 + }, + { + "epoch": 0.10287981394076202, + "grad_norm": 2.23049084974788, + "learning_rate": 9.861461433183527e-06, + "loss": 0.2925, + "step": 1504 + }, + { + "epoch": 0.10294821807237158, + "grad_norm": 1.6391914121518798, + "learning_rate": 9.861202355697639e-06, + "loss": 0.1774, + "step": 1505 + }, + { + "epoch": 0.10301662220398113, + "grad_norm": 2.5034620393877227, + "learning_rate": 9.860943039600464e-06, + "loss": 0.4814, + "step": 1506 + }, + { + "epoch": 0.10308502633559068, + "grad_norm": 2.857418985945518, + "learning_rate": 9.860683484904731e-06, + "loss": 0.4242, + "step": 1507 + }, + { + "epoch": 0.10315343046720021, + "grad_norm": 1.75090669236185, + "learning_rate": 9.860423691623182e-06, + "loss": 0.2069, + "step": 1508 + }, + { + "epoch": 0.10322183459880976, + "grad_norm": 1.4828688760537903, + "learning_rate": 9.860163659768565e-06, + "loss": 0.2011, + "step": 1509 + }, + { + "epoch": 0.10329023873041931, + "grad_norm": 1.985051584168061, + "learning_rate": 9.85990338935365e-06, + "loss": 0.321, + "step": 1510 + }, + { + "epoch": 0.10335864286202887, + "grad_norm": 2.510788206778915, + "learning_rate": 9.859642880391205e-06, + "loss": 0.5876, + "step": 1511 + }, + { + "epoch": 0.10342704699363842, + "grad_norm": 2.1428070954455904, + "learning_rate": 9.859382132894021e-06, + "loss": 0.4522, + "step": 1512 + }, + { + "epoch": 0.10349545112524797, + "grad_norm": 2.235619285219997, + "learning_rate": 9.859121146874897e-06, + "loss": 0.464, + "step": 1513 + }, + { + "epoch": 0.10356385525685752, + "grad_norm": 2.033927920213521, + "learning_rate": 9.858859922346642e-06, + "loss": 0.3481, + "step": 1514 + }, + { + "epoch": 0.10363225938846707, + "grad_norm": 2.8218869149895407, + "learning_rate": 9.85859845932208e-06, + "loss": 0.5252, + "step": 1515 + }, + { + "epoch": 0.10370066352007662, + "grad_norm": 1.9979625396978058, + "learning_rate": 9.858336757814042e-06, + "loss": 0.4973, + "step": 1516 + }, + { + "epoch": 0.10376906765168616, + "grad_norm": 2.1501045625787714, + "learning_rate": 9.858074817835376e-06, + "loss": 0.4196, + "step": 1517 + }, + { + "epoch": 0.1038374717832957, + "grad_norm": 2.469420130835498, + "learning_rate": 9.857812639398938e-06, + "loss": 0.5456, + "step": 1518 + }, + { + "epoch": 0.10390587591490526, + "grad_norm": 1.653202983300422, + "learning_rate": 9.857550222517598e-06, + "loss": 0.234, + "step": 1519 + }, + { + "epoch": 0.10397428004651481, + "grad_norm": 2.608457666750786, + "learning_rate": 9.857287567204237e-06, + "loss": 0.6565, + "step": 1520 + }, + { + "epoch": 0.10404268417812436, + "grad_norm": 2.8595819867369143, + "learning_rate": 9.857024673471747e-06, + "loss": 0.758, + "step": 1521 + }, + { + "epoch": 0.10411108830973391, + "grad_norm": 2.144872554864362, + "learning_rate": 9.856761541333031e-06, + "loss": 0.3834, + "step": 1522 + }, + { + "epoch": 0.10417949244134346, + "grad_norm": 2.189787669829777, + "learning_rate": 9.856498170801005e-06, + "loss": 0.4126, + "step": 1523 + }, + { + "epoch": 0.10424789657295301, + "grad_norm": 2.5421918291618937, + "learning_rate": 9.856234561888596e-06, + "loss": 0.2994, + "step": 1524 + }, + { + "epoch": 0.10431630070456256, + "grad_norm": 2.2481527295283557, + "learning_rate": 9.855970714608746e-06, + "loss": 0.4185, + "step": 1525 + }, + { + "epoch": 0.1043847048361721, + "grad_norm": 2.792616313759118, + "learning_rate": 9.855706628974402e-06, + "loss": 0.6226, + "step": 1526 + }, + { + "epoch": 0.10445310896778165, + "grad_norm": 2.612182805435557, + "learning_rate": 9.855442304998531e-06, + "loss": 0.2935, + "step": 1527 + }, + { + "epoch": 0.1045215130993912, + "grad_norm": 2.7053751317657704, + "learning_rate": 9.855177742694105e-06, + "loss": 0.6862, + "step": 1528 + }, + { + "epoch": 0.10458991723100075, + "grad_norm": 3.8119667024178887, + "learning_rate": 9.854912942074109e-06, + "loss": 0.7868, + "step": 1529 + }, + { + "epoch": 0.1046583213626103, + "grad_norm": 1.8273511781252043, + "learning_rate": 9.854647903151542e-06, + "loss": 0.3353, + "step": 1530 + }, + { + "epoch": 0.10472672549421985, + "grad_norm": 3.152494008620496, + "learning_rate": 9.854382625939412e-06, + "loss": 0.3686, + "step": 1531 + }, + { + "epoch": 0.1047951296258294, + "grad_norm": 2.6048700825864337, + "learning_rate": 9.854117110450742e-06, + "loss": 0.5224, + "step": 1532 + }, + { + "epoch": 0.10486353375743895, + "grad_norm": 2.6917630195841333, + "learning_rate": 9.853851356698564e-06, + "loss": 0.5859, + "step": 1533 + }, + { + "epoch": 0.1049319378890485, + "grad_norm": 2.3022319643592306, + "learning_rate": 9.853585364695921e-06, + "loss": 0.3941, + "step": 1534 + }, + { + "epoch": 0.10500034202065804, + "grad_norm": 2.452250737920388, + "learning_rate": 9.853319134455872e-06, + "loss": 0.1923, + "step": 1535 + }, + { + "epoch": 0.10506874615226759, + "grad_norm": 2.142912310928951, + "learning_rate": 9.853052665991485e-06, + "loss": 0.4999, + "step": 1536 + }, + { + "epoch": 0.10513715028387714, + "grad_norm": 2.4458066544602195, + "learning_rate": 9.852785959315835e-06, + "loss": 0.5952, + "step": 1537 + }, + { + "epoch": 0.1052055544154867, + "grad_norm": 1.8980675796098652, + "learning_rate": 9.852519014442018e-06, + "loss": 0.3314, + "step": 1538 + }, + { + "epoch": 0.10527395854709624, + "grad_norm": 2.0829172736198376, + "learning_rate": 9.852251831383136e-06, + "loss": 0.4319, + "step": 1539 + }, + { + "epoch": 0.1053423626787058, + "grad_norm": 2.4621929296750062, + "learning_rate": 9.8519844101523e-06, + "loss": 0.5704, + "step": 1540 + }, + { + "epoch": 0.10541076681031535, + "grad_norm": 1.8854673136886395, + "learning_rate": 9.851716750762641e-06, + "loss": 0.2299, + "step": 1541 + }, + { + "epoch": 0.1054791709419249, + "grad_norm": 5.145969499750406, + "learning_rate": 9.851448853227294e-06, + "loss": 0.5829, + "step": 1542 + }, + { + "epoch": 0.10554757507353445, + "grad_norm": 2.667450811286529, + "learning_rate": 9.851180717559413e-06, + "loss": 0.5084, + "step": 1543 + }, + { + "epoch": 0.10561597920514398, + "grad_norm": 1.9717379935014994, + "learning_rate": 9.850912343772152e-06, + "loss": 0.3049, + "step": 1544 + }, + { + "epoch": 0.10568438333675353, + "grad_norm": 2.0994043507388183, + "learning_rate": 9.85064373187869e-06, + "loss": 0.4056, + "step": 1545 + }, + { + "epoch": 0.10575278746836309, + "grad_norm": 1.6276153462343919, + "learning_rate": 9.850374881892212e-06, + "loss": 0.3052, + "step": 1546 + }, + { + "epoch": 0.10582119159997264, + "grad_norm": 2.4907787371276315, + "learning_rate": 9.85010579382591e-06, + "loss": 0.3961, + "step": 1547 + }, + { + "epoch": 0.10588959573158219, + "grad_norm": 1.8528934163986406, + "learning_rate": 9.849836467692996e-06, + "loss": 0.2542, + "step": 1548 + }, + { + "epoch": 0.10595799986319174, + "grad_norm": 2.727752731809943, + "learning_rate": 9.849566903506689e-06, + "loss": 0.913, + "step": 1549 + }, + { + "epoch": 0.10602640399480129, + "grad_norm": 1.768355663271228, + "learning_rate": 9.84929710128022e-06, + "loss": 0.2668, + "step": 1550 + }, + { + "epoch": 0.10609480812641084, + "grad_norm": 2.2098125891763285, + "learning_rate": 9.84902706102683e-06, + "loss": 0.4574, + "step": 1551 + }, + { + "epoch": 0.10616321225802039, + "grad_norm": 2.145677887083918, + "learning_rate": 9.848756782759778e-06, + "loss": 0.3362, + "step": 1552 + }, + { + "epoch": 0.10623161638962994, + "grad_norm": 1.7951899007730654, + "learning_rate": 9.84848626649233e-06, + "loss": 0.2828, + "step": 1553 + }, + { + "epoch": 0.10630002052123948, + "grad_norm": 2.457152310589652, + "learning_rate": 9.84821551223776e-06, + "loss": 0.3987, + "step": 1554 + }, + { + "epoch": 0.10636842465284903, + "grad_norm": 2.445457015474612, + "learning_rate": 9.847944520009362e-06, + "loss": 0.2452, + "step": 1555 + }, + { + "epoch": 0.10643682878445858, + "grad_norm": 2.0275658869826914, + "learning_rate": 9.847673289820438e-06, + "loss": 0.2917, + "step": 1556 + }, + { + "epoch": 0.10650523291606813, + "grad_norm": 2.0386263793850827, + "learning_rate": 9.847401821684298e-06, + "loss": 0.3179, + "step": 1557 + }, + { + "epoch": 0.10657363704767768, + "grad_norm": 1.544515170398112, + "learning_rate": 9.84713011561427e-06, + "loss": 0.2191, + "step": 1558 + }, + { + "epoch": 0.10664204117928723, + "grad_norm": 1.3209064941972783, + "learning_rate": 9.846858171623687e-06, + "loss": 0.2143, + "step": 1559 + }, + { + "epoch": 0.10671044531089678, + "grad_norm": 1.7043294125794644, + "learning_rate": 9.846585989725902e-06, + "loss": 0.2134, + "step": 1560 + }, + { + "epoch": 0.10677884944250633, + "grad_norm": 2.5334976601530554, + "learning_rate": 9.846313569934272e-06, + "loss": 0.4718, + "step": 1561 + }, + { + "epoch": 0.10684725357411588, + "grad_norm": 2.307010488549124, + "learning_rate": 9.84604091226217e-06, + "loss": 0.3506, + "step": 1562 + }, + { + "epoch": 0.10691565770572542, + "grad_norm": 1.978210237363039, + "learning_rate": 9.845768016722979e-06, + "loss": 0.2587, + "step": 1563 + }, + { + "epoch": 0.10698406183733497, + "grad_norm": 2.100552504099181, + "learning_rate": 9.84549488333009e-06, + "loss": 0.3838, + "step": 1564 + }, + { + "epoch": 0.10705246596894452, + "grad_norm": 2.4221235393849123, + "learning_rate": 9.845221512096919e-06, + "loss": 0.4369, + "step": 1565 + }, + { + "epoch": 0.10712087010055407, + "grad_norm": 1.958981349482534, + "learning_rate": 9.844947903036875e-06, + "loss": 0.2085, + "step": 1566 + }, + { + "epoch": 0.10718927423216362, + "grad_norm": 2.581363360753668, + "learning_rate": 9.844674056163393e-06, + "loss": 0.4856, + "step": 1567 + }, + { + "epoch": 0.10725767836377317, + "grad_norm": 2.3731299491752353, + "learning_rate": 9.844399971489915e-06, + "loss": 0.4187, + "step": 1568 + }, + { + "epoch": 0.10732608249538272, + "grad_norm": 2.962026602106582, + "learning_rate": 9.844125649029891e-06, + "loss": 0.803, + "step": 1569 + }, + { + "epoch": 0.10739448662699227, + "grad_norm": 2.6734635889685507, + "learning_rate": 9.843851088796788e-06, + "loss": 0.3767, + "step": 1570 + }, + { + "epoch": 0.10746289075860183, + "grad_norm": 2.1575439551197855, + "learning_rate": 9.843576290804084e-06, + "loss": 0.1759, + "step": 1571 + }, + { + "epoch": 0.10753129489021136, + "grad_norm": 2.1479170637071787, + "learning_rate": 9.843301255065267e-06, + "loss": 0.3494, + "step": 1572 + }, + { + "epoch": 0.10759969902182091, + "grad_norm": 2.094377409462039, + "learning_rate": 9.843025981593835e-06, + "loss": 0.4827, + "step": 1573 + }, + { + "epoch": 0.10766810315343046, + "grad_norm": 1.7743239560969963, + "learning_rate": 9.842750470403301e-06, + "loss": 0.3575, + "step": 1574 + }, + { + "epoch": 0.10773650728504001, + "grad_norm": 3.4330438818064644, + "learning_rate": 9.84247472150719e-06, + "loss": 0.3866, + "step": 1575 + }, + { + "epoch": 0.10780491141664957, + "grad_norm": 2.8110372872317186, + "learning_rate": 9.842198734919033e-06, + "loss": 0.533, + "step": 1576 + }, + { + "epoch": 0.10787331554825912, + "grad_norm": 2.0057541430541206, + "learning_rate": 9.841922510652383e-06, + "loss": 0.4185, + "step": 1577 + }, + { + "epoch": 0.10794171967986867, + "grad_norm": 2.373541434286981, + "learning_rate": 9.841646048720791e-06, + "loss": 0.5401, + "step": 1578 + }, + { + "epoch": 0.10801012381147822, + "grad_norm": 2.0204056446173126, + "learning_rate": 9.841369349137832e-06, + "loss": 0.2818, + "step": 1579 + }, + { + "epoch": 0.10807852794308777, + "grad_norm": 1.5221570600521466, + "learning_rate": 9.841092411917089e-06, + "loss": 0.2908, + "step": 1580 + }, + { + "epoch": 0.1081469320746973, + "grad_norm": 2.5083077556683198, + "learning_rate": 9.84081523707215e-06, + "loss": 0.614, + "step": 1581 + }, + { + "epoch": 0.10821533620630686, + "grad_norm": 2.071900589958949, + "learning_rate": 9.840537824616623e-06, + "loss": 0.3442, + "step": 1582 + }, + { + "epoch": 0.1082837403379164, + "grad_norm": 2.6459180613090147, + "learning_rate": 9.840260174564128e-06, + "loss": 0.7893, + "step": 1583 + }, + { + "epoch": 0.10835214446952596, + "grad_norm": 2.534134428175068, + "learning_rate": 9.839982286928287e-06, + "loss": 0.4244, + "step": 1584 + }, + { + "epoch": 0.10842054860113551, + "grad_norm": 2.169970747897757, + "learning_rate": 9.839704161722745e-06, + "loss": 0.4631, + "step": 1585 + }, + { + "epoch": 0.10848895273274506, + "grad_norm": 2.789877150139894, + "learning_rate": 9.839425798961152e-06, + "loss": 0.1902, + "step": 1586 + }, + { + "epoch": 0.10855735686435461, + "grad_norm": 2.453387197829872, + "learning_rate": 9.83914719865717e-06, + "loss": 0.5626, + "step": 1587 + }, + { + "epoch": 0.10862576099596416, + "grad_norm": 1.9562835787977744, + "learning_rate": 9.838868360824478e-06, + "loss": 0.1898, + "step": 1588 + }, + { + "epoch": 0.10869416512757371, + "grad_norm": 1.9245444731069536, + "learning_rate": 9.838589285476759e-06, + "loss": 0.2971, + "step": 1589 + }, + { + "epoch": 0.10876256925918326, + "grad_norm": 2.145212419652392, + "learning_rate": 9.838309972627711e-06, + "loss": 0.3154, + "step": 1590 + }, + { + "epoch": 0.1088309733907928, + "grad_norm": 3.586784146493143, + "learning_rate": 9.838030422291047e-06, + "loss": 0.8206, + "step": 1591 + }, + { + "epoch": 0.10889937752240235, + "grad_norm": 2.6283983133832964, + "learning_rate": 9.837750634480487e-06, + "loss": 0.6194, + "step": 1592 + }, + { + "epoch": 0.1089677816540119, + "grad_norm": 2.086446228569926, + "learning_rate": 9.837470609209765e-06, + "loss": 0.4056, + "step": 1593 + }, + { + "epoch": 0.10903618578562145, + "grad_norm": 2.4267964626742096, + "learning_rate": 9.837190346492626e-06, + "loss": 0.5023, + "step": 1594 + }, + { + "epoch": 0.109104589917231, + "grad_norm": 2.180582259369662, + "learning_rate": 9.836909846342826e-06, + "loss": 0.3814, + "step": 1595 + }, + { + "epoch": 0.10917299404884055, + "grad_norm": 2.913455189069457, + "learning_rate": 9.836629108774132e-06, + "loss": 0.5542, + "step": 1596 + }, + { + "epoch": 0.1092413981804501, + "grad_norm": 2.401473286174761, + "learning_rate": 9.836348133800327e-06, + "loss": 0.2284, + "step": 1597 + }, + { + "epoch": 0.10930980231205965, + "grad_norm": 3.2336257859251614, + "learning_rate": 9.8360669214352e-06, + "loss": 0.3631, + "step": 1598 + }, + { + "epoch": 0.1093782064436692, + "grad_norm": 1.951005953434572, + "learning_rate": 9.835785471692559e-06, + "loss": 0.244, + "step": 1599 + }, + { + "epoch": 0.10944661057527874, + "grad_norm": 1.9546677066218585, + "learning_rate": 9.835503784586211e-06, + "loss": 0.3818, + "step": 1600 + }, + { + "epoch": 0.10951501470688829, + "grad_norm": 2.6321140857138183, + "learning_rate": 9.835221860129989e-06, + "loss": 0.849, + "step": 1601 + }, + { + "epoch": 0.10958341883849784, + "grad_norm": 2.806099950415127, + "learning_rate": 9.834939698337728e-06, + "loss": 0.1882, + "step": 1602 + }, + { + "epoch": 0.1096518229701074, + "grad_norm": 2.121259482637339, + "learning_rate": 9.834657299223279e-06, + "loss": 0.3059, + "step": 1603 + }, + { + "epoch": 0.10972022710171694, + "grad_norm": 1.873378575124528, + "learning_rate": 9.834374662800504e-06, + "loss": 0.3548, + "step": 1604 + }, + { + "epoch": 0.1097886312333265, + "grad_norm": 2.4630275388774128, + "learning_rate": 9.834091789083276e-06, + "loss": 0.6196, + "step": 1605 + }, + { + "epoch": 0.10985703536493605, + "grad_norm": 1.9081966513267197, + "learning_rate": 9.833808678085478e-06, + "loss": 0.2863, + "step": 1606 + }, + { + "epoch": 0.1099254394965456, + "grad_norm": 2.539489894988604, + "learning_rate": 9.833525329821008e-06, + "loss": 0.5827, + "step": 1607 + }, + { + "epoch": 0.10999384362815515, + "grad_norm": 2.1749111538312444, + "learning_rate": 9.833241744303775e-06, + "loss": 0.3605, + "step": 1608 + }, + { + "epoch": 0.11006224775976468, + "grad_norm": 2.0765956614812264, + "learning_rate": 9.832957921547697e-06, + "loss": 0.4388, + "step": 1609 + }, + { + "epoch": 0.11013065189137423, + "grad_norm": 2.7678804468017675, + "learning_rate": 9.832673861566706e-06, + "loss": 0.8241, + "step": 1610 + }, + { + "epoch": 0.11019905602298379, + "grad_norm": 2.528471348783112, + "learning_rate": 9.832389564374745e-06, + "loss": 0.4448, + "step": 1611 + }, + { + "epoch": 0.11026746015459334, + "grad_norm": 1.3655549465691437, + "learning_rate": 9.83210502998577e-06, + "loss": 0.205, + "step": 1612 + }, + { + "epoch": 0.11033586428620289, + "grad_norm": 2.102857910161409, + "learning_rate": 9.831820258413744e-06, + "loss": 0.2416, + "step": 1613 + }, + { + "epoch": 0.11040426841781244, + "grad_norm": 2.093392701051883, + "learning_rate": 9.831535249672651e-06, + "loss": 0.52, + "step": 1614 + }, + { + "epoch": 0.11047267254942199, + "grad_norm": 2.1037084584942702, + "learning_rate": 9.831250003776473e-06, + "loss": 0.4265, + "step": 1615 + }, + { + "epoch": 0.11054107668103154, + "grad_norm": 2.255203191268263, + "learning_rate": 9.830964520739216e-06, + "loss": 0.4355, + "step": 1616 + }, + { + "epoch": 0.11060948081264109, + "grad_norm": 2.309419248773874, + "learning_rate": 9.830678800574891e-06, + "loss": 0.5338, + "step": 1617 + }, + { + "epoch": 0.11067788494425063, + "grad_norm": 1.658276820258206, + "learning_rate": 9.830392843297525e-06, + "loss": 0.3878, + "step": 1618 + }, + { + "epoch": 0.11074628907586018, + "grad_norm": 2.0822365947610884, + "learning_rate": 9.830106648921152e-06, + "loss": 0.3906, + "step": 1619 + }, + { + "epoch": 0.11081469320746973, + "grad_norm": 2.1945547547376862, + "learning_rate": 9.82982021745982e-06, + "loss": 0.2675, + "step": 1620 + }, + { + "epoch": 0.11088309733907928, + "grad_norm": 2.267679151441602, + "learning_rate": 9.82953354892759e-06, + "loss": 0.6074, + "step": 1621 + }, + { + "epoch": 0.11095150147068883, + "grad_norm": 2.030566439786606, + "learning_rate": 9.82924664333853e-06, + "loss": 0.3783, + "step": 1622 + }, + { + "epoch": 0.11101990560229838, + "grad_norm": 1.7114296187984959, + "learning_rate": 9.828959500706725e-06, + "loss": 0.2884, + "step": 1623 + }, + { + "epoch": 0.11108830973390793, + "grad_norm": 2.527280461788037, + "learning_rate": 9.82867212104627e-06, + "loss": 0.6023, + "step": 1624 + }, + { + "epoch": 0.11115671386551748, + "grad_norm": 3.0032661311542714, + "learning_rate": 9.82838450437127e-06, + "loss": 0.3831, + "step": 1625 + }, + { + "epoch": 0.11122511799712703, + "grad_norm": 2.4512675840378355, + "learning_rate": 9.82809665069584e-06, + "loss": 0.3929, + "step": 1626 + }, + { + "epoch": 0.11129352212873657, + "grad_norm": 1.9783954828657422, + "learning_rate": 9.827808560034113e-06, + "loss": 0.3396, + "step": 1627 + }, + { + "epoch": 0.11136192626034612, + "grad_norm": 2.4477850445742058, + "learning_rate": 9.82752023240023e-06, + "loss": 0.5224, + "step": 1628 + }, + { + "epoch": 0.11143033039195567, + "grad_norm": 2.315474198845485, + "learning_rate": 9.82723166780834e-06, + "loss": 0.533, + "step": 1629 + }, + { + "epoch": 0.11149873452356522, + "grad_norm": 2.3347678046801104, + "learning_rate": 9.82694286627261e-06, + "loss": 0.3988, + "step": 1630 + }, + { + "epoch": 0.11156713865517477, + "grad_norm": 1.913443414646884, + "learning_rate": 9.826653827807215e-06, + "loss": 0.4146, + "step": 1631 + }, + { + "epoch": 0.11163554278678432, + "grad_norm": 2.13630517447499, + "learning_rate": 9.826364552426345e-06, + "loss": 0.5623, + "step": 1632 + }, + { + "epoch": 0.11170394691839387, + "grad_norm": 2.2440771852561796, + "learning_rate": 9.826075040144194e-06, + "loss": 0.578, + "step": 1633 + }, + { + "epoch": 0.11177235105000342, + "grad_norm": 2.340041647361128, + "learning_rate": 9.825785290974975e-06, + "loss": 0.568, + "step": 1634 + }, + { + "epoch": 0.11184075518161297, + "grad_norm": 2.0620074261948793, + "learning_rate": 9.82549530493291e-06, + "loss": 0.4915, + "step": 1635 + }, + { + "epoch": 0.11190915931322253, + "grad_norm": 2.797958185584247, + "learning_rate": 9.825205082032235e-06, + "loss": 0.7703, + "step": 1636 + }, + { + "epoch": 0.11197756344483206, + "grad_norm": 3.8979122518979628, + "learning_rate": 9.824914622287195e-06, + "loss": 0.2177, + "step": 1637 + }, + { + "epoch": 0.11204596757644161, + "grad_norm": 2.6592752563472852, + "learning_rate": 9.824623925712042e-06, + "loss": 0.3804, + "step": 1638 + }, + { + "epoch": 0.11211437170805116, + "grad_norm": 1.7092489680774894, + "learning_rate": 9.824332992321052e-06, + "loss": 0.2083, + "step": 1639 + }, + { + "epoch": 0.11218277583966071, + "grad_norm": 2.2255393906764525, + "learning_rate": 9.8240418221285e-06, + "loss": 0.5422, + "step": 1640 + }, + { + "epoch": 0.11225117997127027, + "grad_norm": 1.830266566228542, + "learning_rate": 9.823750415148683e-06, + "loss": 0.2469, + "step": 1641 + }, + { + "epoch": 0.11231958410287982, + "grad_norm": 2.4081317306293806, + "learning_rate": 9.8234587713959e-06, + "loss": 0.6026, + "step": 1642 + }, + { + "epoch": 0.11238798823448937, + "grad_norm": 2.9697902864533026, + "learning_rate": 9.82316689088447e-06, + "loss": 0.932, + "step": 1643 + }, + { + "epoch": 0.11245639236609892, + "grad_norm": 2.345277406174529, + "learning_rate": 9.822874773628717e-06, + "loss": 0.5104, + "step": 1644 + }, + { + "epoch": 0.11252479649770847, + "grad_norm": 2.043501166410559, + "learning_rate": 9.82258241964298e-06, + "loss": 0.4529, + "step": 1645 + }, + { + "epoch": 0.112593200629318, + "grad_norm": 2.1390236134195444, + "learning_rate": 9.822289828941612e-06, + "loss": 0.5291, + "step": 1646 + }, + { + "epoch": 0.11266160476092756, + "grad_norm": 2.254748727439008, + "learning_rate": 9.82199700153897e-06, + "loss": 0.5042, + "step": 1647 + }, + { + "epoch": 0.1127300088925371, + "grad_norm": 1.9469919117562153, + "learning_rate": 9.821703937449431e-06, + "loss": 0.4333, + "step": 1648 + }, + { + "epoch": 0.11279841302414666, + "grad_norm": 2.418546863276165, + "learning_rate": 9.82141063668738e-06, + "loss": 0.6881, + "step": 1649 + }, + { + "epoch": 0.11286681715575621, + "grad_norm": 1.5779074197853638, + "learning_rate": 9.821117099267211e-06, + "loss": 0.2242, + "step": 1650 + }, + { + "epoch": 0.11293522128736576, + "grad_norm": 2.000053781292932, + "learning_rate": 9.820823325203335e-06, + "loss": 0.5178, + "step": 1651 + }, + { + "epoch": 0.11300362541897531, + "grad_norm": 2.791138493454162, + "learning_rate": 9.82052931451017e-06, + "loss": 0.6862, + "step": 1652 + }, + { + "epoch": 0.11307202955058486, + "grad_norm": 2.4241648523877504, + "learning_rate": 9.820235067202148e-06, + "loss": 0.4801, + "step": 1653 + }, + { + "epoch": 0.11314043368219441, + "grad_norm": 1.4390413505375177, + "learning_rate": 9.819940583293713e-06, + "loss": 0.1785, + "step": 1654 + }, + { + "epoch": 0.11320883781380395, + "grad_norm": 2.1650781761287528, + "learning_rate": 9.819645862799318e-06, + "loss": 0.4248, + "step": 1655 + }, + { + "epoch": 0.1132772419454135, + "grad_norm": 2.043243364217847, + "learning_rate": 9.819350905733431e-06, + "loss": 0.387, + "step": 1656 + }, + { + "epoch": 0.11334564607702305, + "grad_norm": 1.7849907078354195, + "learning_rate": 9.81905571211053e-06, + "loss": 0.228, + "step": 1657 + }, + { + "epoch": 0.1134140502086326, + "grad_norm": 2.601295216894652, + "learning_rate": 9.818760281945102e-06, + "loss": 0.6696, + "step": 1658 + }, + { + "epoch": 0.11348245434024215, + "grad_norm": 3.6842698319222262, + "learning_rate": 9.81846461525165e-06, + "loss": 0.2927, + "step": 1659 + }, + { + "epoch": 0.1135508584718517, + "grad_norm": 2.3141133892019643, + "learning_rate": 9.818168712044687e-06, + "loss": 0.6132, + "step": 1660 + }, + { + "epoch": 0.11361926260346125, + "grad_norm": 1.7471921269535158, + "learning_rate": 9.817872572338737e-06, + "loss": 0.2321, + "step": 1661 + }, + { + "epoch": 0.1136876667350708, + "grad_norm": 1.9032919124639385, + "learning_rate": 9.817576196148334e-06, + "loss": 0.2749, + "step": 1662 + }, + { + "epoch": 0.11375607086668035, + "grad_norm": 1.5821402394463524, + "learning_rate": 9.81727958348803e-06, + "loss": 0.1685, + "step": 1663 + }, + { + "epoch": 0.11382447499828989, + "grad_norm": 2.0897035498250167, + "learning_rate": 9.816982734372382e-06, + "loss": 0.3918, + "step": 1664 + }, + { + "epoch": 0.11389287912989944, + "grad_norm": 1.9499756408050257, + "learning_rate": 9.81668564881596e-06, + "loss": 0.425, + "step": 1665 + }, + { + "epoch": 0.11396128326150899, + "grad_norm": 1.991125908533325, + "learning_rate": 9.816388326833347e-06, + "loss": 0.3657, + "step": 1666 + }, + { + "epoch": 0.11402968739311854, + "grad_norm": 23.431496138563237, + "learning_rate": 9.816090768439135e-06, + "loss": 0.4279, + "step": 1667 + }, + { + "epoch": 0.11409809152472809, + "grad_norm": 1.9053882738286823, + "learning_rate": 9.81579297364793e-06, + "loss": 0.3353, + "step": 1668 + }, + { + "epoch": 0.11416649565633764, + "grad_norm": 2.6282644522882896, + "learning_rate": 9.815494942474354e-06, + "loss": 0.3517, + "step": 1669 + }, + { + "epoch": 0.1142348997879472, + "grad_norm": 1.9057924823784955, + "learning_rate": 9.815196674933031e-06, + "loss": 0.3299, + "step": 1670 + }, + { + "epoch": 0.11430330391955675, + "grad_norm": 2.339522350826572, + "learning_rate": 9.814898171038604e-06, + "loss": 0.4534, + "step": 1671 + }, + { + "epoch": 0.1143717080511663, + "grad_norm": 3.0907840696502964, + "learning_rate": 9.814599430805722e-06, + "loss": 0.4096, + "step": 1672 + }, + { + "epoch": 0.11444011218277583, + "grad_norm": 2.0853490428046504, + "learning_rate": 9.814300454249053e-06, + "loss": 0.3868, + "step": 1673 + }, + { + "epoch": 0.11450851631438538, + "grad_norm": 2.8641138362497967, + "learning_rate": 9.814001241383267e-06, + "loss": 0.7177, + "step": 1674 + }, + { + "epoch": 0.11457692044599493, + "grad_norm": 1.9261626819521958, + "learning_rate": 9.813701792223056e-06, + "loss": 0.4124, + "step": 1675 + }, + { + "epoch": 0.11464532457760448, + "grad_norm": 1.5547266313046735, + "learning_rate": 9.813402106783115e-06, + "loss": 0.1787, + "step": 1676 + }, + { + "epoch": 0.11471372870921404, + "grad_norm": 2.2214740565787974, + "learning_rate": 9.813102185078154e-06, + "loss": 0.6182, + "step": 1677 + }, + { + "epoch": 0.11478213284082359, + "grad_norm": 1.8385601587599305, + "learning_rate": 9.812802027122897e-06, + "loss": 0.2369, + "step": 1678 + }, + { + "epoch": 0.11485053697243314, + "grad_norm": 2.5012173788068135, + "learning_rate": 9.812501632932074e-06, + "loss": 0.6116, + "step": 1679 + }, + { + "epoch": 0.11491894110404269, + "grad_norm": 1.9209674828614183, + "learning_rate": 9.812201002520431e-06, + "loss": 0.255, + "step": 1680 + }, + { + "epoch": 0.11498734523565224, + "grad_norm": 1.8976970963461754, + "learning_rate": 9.811900135902728e-06, + "loss": 0.3529, + "step": 1681 + }, + { + "epoch": 0.11505574936726179, + "grad_norm": 1.433152065786788, + "learning_rate": 9.811599033093728e-06, + "loss": 0.2407, + "step": 1682 + }, + { + "epoch": 0.11512415349887133, + "grad_norm": 2.333138029121911, + "learning_rate": 9.811297694108212e-06, + "loss": 0.5837, + "step": 1683 + }, + { + "epoch": 0.11519255763048088, + "grad_norm": 2.080994004108149, + "learning_rate": 9.810996118960972e-06, + "loss": 0.2852, + "step": 1684 + }, + { + "epoch": 0.11526096176209043, + "grad_norm": 2.519288114896319, + "learning_rate": 9.810694307666811e-06, + "loss": 0.4851, + "step": 1685 + }, + { + "epoch": 0.11532936589369998, + "grad_norm": 2.2744058008563957, + "learning_rate": 9.810392260240543e-06, + "loss": 0.411, + "step": 1686 + }, + { + "epoch": 0.11539777002530953, + "grad_norm": 2.27705772789196, + "learning_rate": 9.810089976696992e-06, + "loss": 0.5342, + "step": 1687 + }, + { + "epoch": 0.11546617415691908, + "grad_norm": 1.6914519090194131, + "learning_rate": 9.809787457051e-06, + "loss": 0.2866, + "step": 1688 + }, + { + "epoch": 0.11553457828852863, + "grad_norm": 2.535233494973369, + "learning_rate": 9.809484701317411e-06, + "loss": 0.6733, + "step": 1689 + }, + { + "epoch": 0.11560298242013818, + "grad_norm": 2.4876315857238605, + "learning_rate": 9.80918170951109e-06, + "loss": 0.5148, + "step": 1690 + }, + { + "epoch": 0.11567138655174773, + "grad_norm": 1.894256882611363, + "learning_rate": 9.808878481646905e-06, + "loss": 0.3596, + "step": 1691 + }, + { + "epoch": 0.11573979068335727, + "grad_norm": 2.1438819896058483, + "learning_rate": 9.808575017739743e-06, + "loss": 0.4783, + "step": 1692 + }, + { + "epoch": 0.11580819481496682, + "grad_norm": 2.02942219439942, + "learning_rate": 9.8082713178045e-06, + "loss": 0.3403, + "step": 1693 + }, + { + "epoch": 0.11587659894657637, + "grad_norm": 2.4078562162566, + "learning_rate": 9.807967381856081e-06, + "loss": 0.3911, + "step": 1694 + }, + { + "epoch": 0.11594500307818592, + "grad_norm": 1.9409140206542737, + "learning_rate": 9.807663209909406e-06, + "loss": 0.2235, + "step": 1695 + }, + { + "epoch": 0.11601340720979547, + "grad_norm": 2.1277899238866684, + "learning_rate": 9.807358801979404e-06, + "loss": 0.4806, + "step": 1696 + }, + { + "epoch": 0.11608181134140502, + "grad_norm": 2.2772490152902765, + "learning_rate": 9.807054158081017e-06, + "loss": 0.4707, + "step": 1697 + }, + { + "epoch": 0.11615021547301457, + "grad_norm": 2.8530489561511594, + "learning_rate": 9.806749278229201e-06, + "loss": 0.5371, + "step": 1698 + }, + { + "epoch": 0.11621861960462412, + "grad_norm": 1.8073370647958118, + "learning_rate": 9.806444162438917e-06, + "loss": 0.1848, + "step": 1699 + }, + { + "epoch": 0.11628702373623367, + "grad_norm": 2.36512502831759, + "learning_rate": 9.806138810725144e-06, + "loss": 0.6134, + "step": 1700 + }, + { + "epoch": 0.11635542786784321, + "grad_norm": 2.2938706250656486, + "learning_rate": 9.80583322310287e-06, + "loss": 0.3959, + "step": 1701 + }, + { + "epoch": 0.11642383199945276, + "grad_norm": 1.5478858703276868, + "learning_rate": 9.805527399587093e-06, + "loss": 0.213, + "step": 1702 + }, + { + "epoch": 0.11649223613106231, + "grad_norm": 2.190439418953084, + "learning_rate": 9.805221340192828e-06, + "loss": 0.5303, + "step": 1703 + }, + { + "epoch": 0.11656064026267186, + "grad_norm": 2.3430635845838443, + "learning_rate": 9.804915044935092e-06, + "loss": 0.4443, + "step": 1704 + }, + { + "epoch": 0.11662904439428141, + "grad_norm": 2.1372385386566726, + "learning_rate": 9.804608513828927e-06, + "loss": 0.3219, + "step": 1705 + }, + { + "epoch": 0.11669744852589096, + "grad_norm": 2.4888815163274214, + "learning_rate": 9.804301746889372e-06, + "loss": 0.4967, + "step": 1706 + }, + { + "epoch": 0.11676585265750052, + "grad_norm": 2.37066576141466, + "learning_rate": 9.803994744131488e-06, + "loss": 0.3443, + "step": 1707 + }, + { + "epoch": 0.11683425678911007, + "grad_norm": 2.496703513681642, + "learning_rate": 9.803687505570343e-06, + "loss": 0.6241, + "step": 1708 + }, + { + "epoch": 0.11690266092071962, + "grad_norm": 3.915655227020817, + "learning_rate": 9.80338003122102e-06, + "loss": 0.5906, + "step": 1709 + }, + { + "epoch": 0.11697106505232915, + "grad_norm": 2.600585685383308, + "learning_rate": 9.803072321098609e-06, + "loss": 0.6715, + "step": 1710 + }, + { + "epoch": 0.1170394691839387, + "grad_norm": 2.118868670614508, + "learning_rate": 9.802764375218215e-06, + "loss": 0.5929, + "step": 1711 + }, + { + "epoch": 0.11710787331554826, + "grad_norm": 2.2519605657978983, + "learning_rate": 9.802456193594953e-06, + "loss": 0.4869, + "step": 1712 + }, + { + "epoch": 0.1171762774471578, + "grad_norm": 2.7464412453066855, + "learning_rate": 9.80214777624395e-06, + "loss": 0.4547, + "step": 1713 + }, + { + "epoch": 0.11724468157876736, + "grad_norm": 2.397192342905208, + "learning_rate": 9.801839123180346e-06, + "loss": 0.6543, + "step": 1714 + }, + { + "epoch": 0.11731308571037691, + "grad_norm": 2.6837934672349557, + "learning_rate": 9.801530234419288e-06, + "loss": 0.8123, + "step": 1715 + }, + { + "epoch": 0.11738148984198646, + "grad_norm": 1.9832942527705997, + "learning_rate": 9.80122110997594e-06, + "loss": 0.3899, + "step": 1716 + }, + { + "epoch": 0.11744989397359601, + "grad_norm": 1.7796538546549305, + "learning_rate": 9.800911749865476e-06, + "loss": 0.3511, + "step": 1717 + }, + { + "epoch": 0.11751829810520556, + "grad_norm": 1.9854728072522685, + "learning_rate": 9.80060215410308e-06, + "loss": 0.3445, + "step": 1718 + }, + { + "epoch": 0.11758670223681511, + "grad_norm": 2.184673636000553, + "learning_rate": 9.800292322703949e-06, + "loss": 0.6327, + "step": 1719 + }, + { + "epoch": 0.11765510636842465, + "grad_norm": 2.3844987499676127, + "learning_rate": 9.799982255683289e-06, + "loss": 0.4964, + "step": 1720 + }, + { + "epoch": 0.1177235105000342, + "grad_norm": 1.7959190355025108, + "learning_rate": 9.799671953056322e-06, + "loss": 0.2555, + "step": 1721 + }, + { + "epoch": 0.11779191463164375, + "grad_norm": 1.9754835056030533, + "learning_rate": 9.799361414838277e-06, + "loss": 0.3434, + "step": 1722 + }, + { + "epoch": 0.1178603187632533, + "grad_norm": 1.9175639764540056, + "learning_rate": 9.799050641044398e-06, + "loss": 0.3121, + "step": 1723 + }, + { + "epoch": 0.11792872289486285, + "grad_norm": 2.078255369643166, + "learning_rate": 9.79873963168994e-06, + "loss": 0.4852, + "step": 1724 + }, + { + "epoch": 0.1179971270264724, + "grad_norm": 2.3280549454869974, + "learning_rate": 9.79842838679017e-06, + "loss": 0.7114, + "step": 1725 + }, + { + "epoch": 0.11806553115808195, + "grad_norm": 2.0587246037105564, + "learning_rate": 9.79811690636036e-06, + "loss": 0.5125, + "step": 1726 + }, + { + "epoch": 0.1181339352896915, + "grad_norm": 2.4104357855628478, + "learning_rate": 9.797805190415803e-06, + "loss": 0.4539, + "step": 1727 + }, + { + "epoch": 0.11820233942130105, + "grad_norm": 2.4686744412798443, + "learning_rate": 9.7974932389718e-06, + "loss": 0.5542, + "step": 1728 + }, + { + "epoch": 0.11827074355291059, + "grad_norm": 2.355549350706986, + "learning_rate": 9.79718105204366e-06, + "loss": 0.5929, + "step": 1729 + }, + { + "epoch": 0.11833914768452014, + "grad_norm": 1.684169644238789, + "learning_rate": 9.79686862964671e-06, + "loss": 0.335, + "step": 1730 + }, + { + "epoch": 0.11840755181612969, + "grad_norm": 0.9768790399616731, + "learning_rate": 9.796555971796285e-06, + "loss": 0.1433, + "step": 1731 + }, + { + "epoch": 0.11847595594773924, + "grad_norm": 2.1859242690410574, + "learning_rate": 9.796243078507729e-06, + "loss": 0.6255, + "step": 1732 + }, + { + "epoch": 0.11854436007934879, + "grad_norm": 2.0089005331323175, + "learning_rate": 9.795929949796402e-06, + "loss": 0.5419, + "step": 1733 + }, + { + "epoch": 0.11861276421095834, + "grad_norm": 1.5152375178195099, + "learning_rate": 9.795616585677674e-06, + "loss": 0.2619, + "step": 1734 + }, + { + "epoch": 0.1186811683425679, + "grad_norm": 1.7628480755239522, + "learning_rate": 9.795302986166926e-06, + "loss": 0.2653, + "step": 1735 + }, + { + "epoch": 0.11874957247417744, + "grad_norm": 2.0094387022098306, + "learning_rate": 9.794989151279553e-06, + "loss": 0.3386, + "step": 1736 + }, + { + "epoch": 0.118817976605787, + "grad_norm": 2.0234618829856115, + "learning_rate": 9.794675081030956e-06, + "loss": 0.4995, + "step": 1737 + }, + { + "epoch": 0.11888638073739653, + "grad_norm": 1.9845657932743246, + "learning_rate": 9.794360775436552e-06, + "loss": 0.4268, + "step": 1738 + }, + { + "epoch": 0.11895478486900608, + "grad_norm": 1.6586448027290883, + "learning_rate": 9.79404623451177e-06, + "loss": 0.4002, + "step": 1739 + }, + { + "epoch": 0.11902318900061563, + "grad_norm": 2.2667965162093857, + "learning_rate": 9.793731458272049e-06, + "loss": 0.5886, + "step": 1740 + }, + { + "epoch": 0.11909159313222518, + "grad_norm": 2.7992274870591087, + "learning_rate": 9.79341644673284e-06, + "loss": 0.9156, + "step": 1741 + }, + { + "epoch": 0.11915999726383474, + "grad_norm": 2.2225615174497926, + "learning_rate": 9.793101199909603e-06, + "loss": 0.6246, + "step": 1742 + }, + { + "epoch": 0.11922840139544429, + "grad_norm": 2.5844211511976996, + "learning_rate": 9.792785717817815e-06, + "loss": 0.4368, + "step": 1743 + }, + { + "epoch": 0.11929680552705384, + "grad_norm": 2.518983358501225, + "learning_rate": 9.792470000472957e-06, + "loss": 0.583, + "step": 1744 + }, + { + "epoch": 0.11936520965866339, + "grad_norm": 3.306673855285047, + "learning_rate": 9.792154047890531e-06, + "loss": 0.4327, + "step": 1745 + }, + { + "epoch": 0.11943361379027294, + "grad_norm": 2.863120138259554, + "learning_rate": 9.791837860086045e-06, + "loss": 0.3093, + "step": 1746 + }, + { + "epoch": 0.11950201792188248, + "grad_norm": 2.402345815518119, + "learning_rate": 9.791521437075015e-06, + "loss": 0.5829, + "step": 1747 + }, + { + "epoch": 0.11957042205349203, + "grad_norm": 2.1330317080685437, + "learning_rate": 9.791204778872976e-06, + "loss": 0.421, + "step": 1748 + }, + { + "epoch": 0.11963882618510158, + "grad_norm": 1.813491369298944, + "learning_rate": 9.790887885495467e-06, + "loss": 0.3755, + "step": 1749 + }, + { + "epoch": 0.11970723031671113, + "grad_norm": 2.495311342811197, + "learning_rate": 9.79057075695805e-06, + "loss": 0.4408, + "step": 1750 + }, + { + "epoch": 0.11977563444832068, + "grad_norm": 3.4206838559545316, + "learning_rate": 9.790253393276283e-06, + "loss": 0.3239, + "step": 1751 + }, + { + "epoch": 0.11984403857993023, + "grad_norm": 1.7960691718319302, + "learning_rate": 9.78993579446575e-06, + "loss": 0.4056, + "step": 1752 + }, + { + "epoch": 0.11991244271153978, + "grad_norm": 2.313429719668884, + "learning_rate": 9.789617960542037e-06, + "loss": 0.2841, + "step": 1753 + }, + { + "epoch": 0.11998084684314933, + "grad_norm": 1.7678841847529967, + "learning_rate": 9.789299891520746e-06, + "loss": 0.4583, + "step": 1754 + }, + { + "epoch": 0.12004925097475888, + "grad_norm": 6.822404363681085, + "learning_rate": 9.788981587417488e-06, + "loss": 0.5134, + "step": 1755 + }, + { + "epoch": 0.12011765510636842, + "grad_norm": 2.6467782058041642, + "learning_rate": 9.788663048247888e-06, + "loss": 0.7722, + "step": 1756 + }, + { + "epoch": 0.12018605923797797, + "grad_norm": 2.250850688865531, + "learning_rate": 9.788344274027584e-06, + "loss": 0.6018, + "step": 1757 + }, + { + "epoch": 0.12025446336958752, + "grad_norm": 2.6035055459379026, + "learning_rate": 9.788025264772218e-06, + "loss": 0.5506, + "step": 1758 + }, + { + "epoch": 0.12032286750119707, + "grad_norm": 2.6178645923491226, + "learning_rate": 9.787706020497451e-06, + "loss": 0.5048, + "step": 1759 + }, + { + "epoch": 0.12039127163280662, + "grad_norm": 2.886214257370465, + "learning_rate": 9.787386541218953e-06, + "loss": 0.8323, + "step": 1760 + }, + { + "epoch": 0.12045967576441617, + "grad_norm": 2.241947373880213, + "learning_rate": 9.787066826952407e-06, + "loss": 0.4842, + "step": 1761 + }, + { + "epoch": 0.12052807989602572, + "grad_norm": 2.0668983594614287, + "learning_rate": 9.786746877713502e-06, + "loss": 0.4695, + "step": 1762 + }, + { + "epoch": 0.12059648402763527, + "grad_norm": 2.9244265100092823, + "learning_rate": 9.786426693517947e-06, + "loss": 0.8329, + "step": 1763 + }, + { + "epoch": 0.12066488815924482, + "grad_norm": 1.9998872546365283, + "learning_rate": 9.786106274381457e-06, + "loss": 0.4922, + "step": 1764 + }, + { + "epoch": 0.12073329229085437, + "grad_norm": 1.6181650662462232, + "learning_rate": 9.785785620319757e-06, + "loss": 0.202, + "step": 1765 + }, + { + "epoch": 0.12080169642246391, + "grad_norm": 2.481171374792831, + "learning_rate": 9.78546473134859e-06, + "loss": 0.4245, + "step": 1766 + }, + { + "epoch": 0.12087010055407346, + "grad_norm": 2.2162378959196136, + "learning_rate": 9.785143607483705e-06, + "loss": 0.3422, + "step": 1767 + }, + { + "epoch": 0.12093850468568301, + "grad_norm": 2.2038801671066497, + "learning_rate": 9.784822248740864e-06, + "loss": 0.5456, + "step": 1768 + }, + { + "epoch": 0.12100690881729256, + "grad_norm": 2.3740356520319326, + "learning_rate": 9.784500655135842e-06, + "loss": 0.6282, + "step": 1769 + }, + { + "epoch": 0.12107531294890211, + "grad_norm": 2.042023814032604, + "learning_rate": 9.784178826684422e-06, + "loss": 0.2763, + "step": 1770 + }, + { + "epoch": 0.12114371708051166, + "grad_norm": 1.7394728101999544, + "learning_rate": 9.783856763402405e-06, + "loss": 0.4007, + "step": 1771 + }, + { + "epoch": 0.12121212121212122, + "grad_norm": 2.3005327566567786, + "learning_rate": 9.783534465305598e-06, + "loss": 0.4674, + "step": 1772 + }, + { + "epoch": 0.12128052534373077, + "grad_norm": 2.589964062642088, + "learning_rate": 9.783211932409818e-06, + "loss": 0.6372, + "step": 1773 + }, + { + "epoch": 0.12134892947534032, + "grad_norm": 2.3807105064008516, + "learning_rate": 9.782889164730897e-06, + "loss": 0.624, + "step": 1774 + }, + { + "epoch": 0.12141733360694985, + "grad_norm": 2.4351230916472564, + "learning_rate": 9.782566162284681e-06, + "loss": 0.5733, + "step": 1775 + }, + { + "epoch": 0.1214857377385594, + "grad_norm": 1.8233835812191066, + "learning_rate": 9.782242925087024e-06, + "loss": 0.3928, + "step": 1776 + }, + { + "epoch": 0.12155414187016896, + "grad_norm": 2.4006553361163263, + "learning_rate": 9.78191945315379e-06, + "loss": 0.6362, + "step": 1777 + }, + { + "epoch": 0.1216225460017785, + "grad_norm": 2.752909443413117, + "learning_rate": 9.781595746500857e-06, + "loss": 0.5138, + "step": 1778 + }, + { + "epoch": 0.12169095013338806, + "grad_norm": 1.7663924566352103, + "learning_rate": 9.781271805144115e-06, + "loss": 0.1713, + "step": 1779 + }, + { + "epoch": 0.12175935426499761, + "grad_norm": 1.7707394902290245, + "learning_rate": 9.780947629099466e-06, + "loss": 0.3899, + "step": 1780 + }, + { + "epoch": 0.12182775839660716, + "grad_norm": 2.492836559763443, + "learning_rate": 9.78062321838282e-06, + "loss": 0.3878, + "step": 1781 + }, + { + "epoch": 0.12189616252821671, + "grad_norm": 2.155463185705875, + "learning_rate": 9.7802985730101e-06, + "loss": 0.5471, + "step": 1782 + }, + { + "epoch": 0.12196456665982626, + "grad_norm": 2.207539558952624, + "learning_rate": 9.77997369299724e-06, + "loss": 0.4865, + "step": 1783 + }, + { + "epoch": 0.1220329707914358, + "grad_norm": 2.0783071940011353, + "learning_rate": 9.779648578360192e-06, + "loss": 0.5116, + "step": 1784 + }, + { + "epoch": 0.12210137492304535, + "grad_norm": 1.5668197832736765, + "learning_rate": 9.77932322911491e-06, + "loss": 0.2819, + "step": 1785 + }, + { + "epoch": 0.1221697790546549, + "grad_norm": 3.433998474640638, + "learning_rate": 9.778997645277366e-06, + "loss": 0.335, + "step": 1786 + }, + { + "epoch": 0.12223818318626445, + "grad_norm": 2.5051422401216827, + "learning_rate": 9.778671826863538e-06, + "loss": 0.6743, + "step": 1787 + }, + { + "epoch": 0.122306587317874, + "grad_norm": 2.349419672562344, + "learning_rate": 9.778345773889422e-06, + "loss": 0.4124, + "step": 1788 + }, + { + "epoch": 0.12237499144948355, + "grad_norm": 1.990845871743225, + "learning_rate": 9.778019486371022e-06, + "loss": 0.4132, + "step": 1789 + }, + { + "epoch": 0.1224433955810931, + "grad_norm": 2.455006576381662, + "learning_rate": 9.77769296432435e-06, + "loss": 0.7719, + "step": 1790 + }, + { + "epoch": 0.12251179971270265, + "grad_norm": 3.1525412761155085, + "learning_rate": 9.777366207765438e-06, + "loss": 0.6021, + "step": 1791 + }, + { + "epoch": 0.1225802038443122, + "grad_norm": 1.9231927858579843, + "learning_rate": 9.777039216710322e-06, + "loss": 0.2783, + "step": 1792 + }, + { + "epoch": 0.12264860797592174, + "grad_norm": 2.736840831092922, + "learning_rate": 9.776711991175054e-06, + "loss": 0.803, + "step": 1793 + }, + { + "epoch": 0.12271701210753129, + "grad_norm": 2.5923053785933954, + "learning_rate": 9.776384531175696e-06, + "loss": 0.5786, + "step": 1794 + }, + { + "epoch": 0.12278541623914084, + "grad_norm": 1.5955139711064785, + "learning_rate": 9.776056836728318e-06, + "loss": 0.2884, + "step": 1795 + }, + { + "epoch": 0.12285382037075039, + "grad_norm": 2.6712857260254026, + "learning_rate": 9.775728907849008e-06, + "loss": 0.5484, + "step": 1796 + }, + { + "epoch": 0.12292222450235994, + "grad_norm": 1.7740108764633804, + "learning_rate": 9.775400744553863e-06, + "loss": 0.2463, + "step": 1797 + }, + { + "epoch": 0.12299062863396949, + "grad_norm": 2.115089334039658, + "learning_rate": 9.775072346858986e-06, + "loss": 0.5128, + "step": 1798 + }, + { + "epoch": 0.12305903276557904, + "grad_norm": 2.2157971249372097, + "learning_rate": 9.774743714780502e-06, + "loss": 0.5534, + "step": 1799 + }, + { + "epoch": 0.1231274368971886, + "grad_norm": 3.213693475563733, + "learning_rate": 9.77441484833454e-06, + "loss": 0.7504, + "step": 1800 + }, + { + "epoch": 0.12319584102879814, + "grad_norm": 2.5304850906075305, + "learning_rate": 9.774085747537241e-06, + "loss": 0.6191, + "step": 1801 + }, + { + "epoch": 0.1232642451604077, + "grad_norm": 2.599927646290457, + "learning_rate": 9.773756412404757e-06, + "loss": 0.309, + "step": 1802 + }, + { + "epoch": 0.12333264929201723, + "grad_norm": 2.8188844748094146, + "learning_rate": 9.77342684295326e-06, + "loss": 0.2909, + "step": 1803 + }, + { + "epoch": 0.12340105342362678, + "grad_norm": 1.9900130653311872, + "learning_rate": 9.773097039198921e-06, + "loss": 0.3283, + "step": 1804 + }, + { + "epoch": 0.12346945755523633, + "grad_norm": 2.380963424148949, + "learning_rate": 9.772767001157932e-06, + "loss": 0.4588, + "step": 1805 + }, + { + "epoch": 0.12353786168684588, + "grad_norm": 1.9563996634042284, + "learning_rate": 9.772436728846488e-06, + "loss": 0.3247, + "step": 1806 + }, + { + "epoch": 0.12360626581845544, + "grad_norm": 3.1993491131092386, + "learning_rate": 9.772106222280806e-06, + "loss": 0.4369, + "step": 1807 + }, + { + "epoch": 0.12367466995006499, + "grad_norm": 2.506287928363212, + "learning_rate": 9.771775481477105e-06, + "loss": 0.206, + "step": 1808 + }, + { + "epoch": 0.12374307408167454, + "grad_norm": 2.2130445744967813, + "learning_rate": 9.771444506451622e-06, + "loss": 0.4588, + "step": 1809 + }, + { + "epoch": 0.12381147821328409, + "grad_norm": 1.4051948827613976, + "learning_rate": 9.771113297220601e-06, + "loss": 0.1764, + "step": 1810 + }, + { + "epoch": 0.12387988234489364, + "grad_norm": 2.5125227096821834, + "learning_rate": 9.7707818538003e-06, + "loss": 0.6564, + "step": 1811 + }, + { + "epoch": 0.12394828647650317, + "grad_norm": 2.140308666523795, + "learning_rate": 9.770450176206987e-06, + "loss": 0.3341, + "step": 1812 + }, + { + "epoch": 0.12401669060811273, + "grad_norm": 1.848024053438744, + "learning_rate": 9.770118264456943e-06, + "loss": 0.392, + "step": 1813 + }, + { + "epoch": 0.12408509473972228, + "grad_norm": 2.7450611783247223, + "learning_rate": 9.769786118566462e-06, + "loss": 0.7273, + "step": 1814 + }, + { + "epoch": 0.12415349887133183, + "grad_norm": 1.7826372232624603, + "learning_rate": 9.769453738551844e-06, + "loss": 0.1831, + "step": 1815 + }, + { + "epoch": 0.12422190300294138, + "grad_norm": 2.3875976230475136, + "learning_rate": 9.769121124429404e-06, + "loss": 0.4759, + "step": 1816 + }, + { + "epoch": 0.12429030713455093, + "grad_norm": 3.0032918770557773, + "learning_rate": 9.768788276215471e-06, + "loss": 0.4111, + "step": 1817 + }, + { + "epoch": 0.12435871126616048, + "grad_norm": 2.0668308276176206, + "learning_rate": 9.768455193926379e-06, + "loss": 0.494, + "step": 1818 + }, + { + "epoch": 0.12442711539777003, + "grad_norm": 2.0340126324282917, + "learning_rate": 9.76812187757848e-06, + "loss": 0.324, + "step": 1819 + }, + { + "epoch": 0.12449551952937958, + "grad_norm": 2.4038698140960655, + "learning_rate": 9.767788327188135e-06, + "loss": 0.3543, + "step": 1820 + }, + { + "epoch": 0.12456392366098912, + "grad_norm": 2.2581309717618323, + "learning_rate": 9.767454542771718e-06, + "loss": 0.5045, + "step": 1821 + }, + { + "epoch": 0.12463232779259867, + "grad_norm": 2.0937012552878613, + "learning_rate": 9.767120524345605e-06, + "loss": 0.4655, + "step": 1822 + }, + { + "epoch": 0.12470073192420822, + "grad_norm": 2.175059522439141, + "learning_rate": 9.7667862719262e-06, + "loss": 0.3723, + "step": 1823 + }, + { + "epoch": 0.12476913605581777, + "grad_norm": 1.9344684741556426, + "learning_rate": 9.766451785529905e-06, + "loss": 0.2474, + "step": 1824 + }, + { + "epoch": 0.12483754018742732, + "grad_norm": 2.3617027843716802, + "learning_rate": 9.76611706517314e-06, + "loss": 0.5173, + "step": 1825 + }, + { + "epoch": 0.12490594431903687, + "grad_norm": 2.6321942879875633, + "learning_rate": 9.765782110872335e-06, + "loss": 0.4946, + "step": 1826 + }, + { + "epoch": 0.12497434845064642, + "grad_norm": 2.662596453007605, + "learning_rate": 9.765446922643928e-06, + "loss": 0.2868, + "step": 1827 + }, + { + "epoch": 0.12504275258225597, + "grad_norm": 2.5315565082623706, + "learning_rate": 9.765111500504376e-06, + "loss": 0.5956, + "step": 1828 + }, + { + "epoch": 0.12511115671386552, + "grad_norm": 2.158120163233529, + "learning_rate": 9.76477584447014e-06, + "loss": 0.4644, + "step": 1829 + }, + { + "epoch": 0.12517956084547507, + "grad_norm": 2.2277006151750656, + "learning_rate": 9.764439954557696e-06, + "loss": 0.6209, + "step": 1830 + }, + { + "epoch": 0.12524796497708462, + "grad_norm": 2.2016629311888214, + "learning_rate": 9.764103830783533e-06, + "loss": 0.4987, + "step": 1831 + }, + { + "epoch": 0.12531636910869418, + "grad_norm": 2.431903803421225, + "learning_rate": 9.763767473164149e-06, + "loss": 0.5969, + "step": 1832 + }, + { + "epoch": 0.12538477324030373, + "grad_norm": 1.760554087276849, + "learning_rate": 9.763430881716052e-06, + "loss": 0.3654, + "step": 1833 + }, + { + "epoch": 0.12545317737191328, + "grad_norm": 2.585634213782241, + "learning_rate": 9.763094056455766e-06, + "loss": 0.6755, + "step": 1834 + }, + { + "epoch": 0.1255215815035228, + "grad_norm": 2.2167614770155297, + "learning_rate": 9.762756997399822e-06, + "loss": 0.6892, + "step": 1835 + }, + { + "epoch": 0.12558998563513235, + "grad_norm": 1.8395315095665907, + "learning_rate": 9.762419704564765e-06, + "loss": 0.3268, + "step": 1836 + }, + { + "epoch": 0.1256583897667419, + "grad_norm": 2.091230983750278, + "learning_rate": 9.762082177967153e-06, + "loss": 0.474, + "step": 1837 + }, + { + "epoch": 0.12572679389835145, + "grad_norm": 1.8749126135807466, + "learning_rate": 9.76174441762355e-06, + "loss": 0.3706, + "step": 1838 + }, + { + "epoch": 0.125795198029961, + "grad_norm": 2.0483169136082453, + "learning_rate": 9.761406423550539e-06, + "loss": 0.3263, + "step": 1839 + }, + { + "epoch": 0.12586360216157055, + "grad_norm": 2.027138363351625, + "learning_rate": 9.761068195764706e-06, + "loss": 0.538, + "step": 1840 + }, + { + "epoch": 0.1259320062931801, + "grad_norm": 1.7365828028026298, + "learning_rate": 9.760729734282656e-06, + "loss": 0.361, + "step": 1841 + }, + { + "epoch": 0.12600041042478966, + "grad_norm": 2.070516561851875, + "learning_rate": 9.760391039121e-06, + "loss": 0.336, + "step": 1842 + }, + { + "epoch": 0.1260688145563992, + "grad_norm": 2.0762486484001856, + "learning_rate": 9.760052110296367e-06, + "loss": 0.5886, + "step": 1843 + }, + { + "epoch": 0.12613721868800876, + "grad_norm": 2.0258225299588206, + "learning_rate": 9.759712947825387e-06, + "loss": 0.2904, + "step": 1844 + }, + { + "epoch": 0.1262056228196183, + "grad_norm": 1.8588204398337242, + "learning_rate": 9.759373551724713e-06, + "loss": 0.2773, + "step": 1845 + }, + { + "epoch": 0.12627402695122786, + "grad_norm": 1.8140355503272019, + "learning_rate": 9.759033922011004e-06, + "loss": 0.4796, + "step": 1846 + }, + { + "epoch": 0.1263424310828374, + "grad_norm": 1.7893304093258204, + "learning_rate": 9.758694058700925e-06, + "loss": 0.2521, + "step": 1847 + }, + { + "epoch": 0.12641083521444696, + "grad_norm": 2.3324403451914315, + "learning_rate": 9.758353961811164e-06, + "loss": 0.4404, + "step": 1848 + }, + { + "epoch": 0.1264792393460565, + "grad_norm": 2.341256315504885, + "learning_rate": 9.758013631358414e-06, + "loss": 0.5682, + "step": 1849 + }, + { + "epoch": 0.12654764347766606, + "grad_norm": 1.8243848124688127, + "learning_rate": 9.757673067359378e-06, + "loss": 0.3374, + "step": 1850 + }, + { + "epoch": 0.1266160476092756, + "grad_norm": 2.4941029497256952, + "learning_rate": 9.757332269830773e-06, + "loss": 0.526, + "step": 1851 + }, + { + "epoch": 0.12668445174088516, + "grad_norm": 2.355502209022456, + "learning_rate": 9.756991238789327e-06, + "loss": 0.7073, + "step": 1852 + }, + { + "epoch": 0.12675285587249469, + "grad_norm": 2.4946425850643092, + "learning_rate": 9.75664997425178e-06, + "loss": 0.5503, + "step": 1853 + }, + { + "epoch": 0.12682126000410424, + "grad_norm": 2.81227480728571, + "learning_rate": 9.756308476234881e-06, + "loss": 0.2349, + "step": 1854 + }, + { + "epoch": 0.1268896641357138, + "grad_norm": 2.2885594645244947, + "learning_rate": 9.755966744755396e-06, + "loss": 0.4933, + "step": 1855 + }, + { + "epoch": 0.12695806826732334, + "grad_norm": 2.269667320965354, + "learning_rate": 9.755624779830098e-06, + "loss": 0.4479, + "step": 1856 + }, + { + "epoch": 0.1270264723989329, + "grad_norm": 1.8633220756104811, + "learning_rate": 9.755282581475769e-06, + "loss": 0.4048, + "step": 1857 + }, + { + "epoch": 0.12709487653054244, + "grad_norm": 2.5560625587004644, + "learning_rate": 9.754940149709209e-06, + "loss": 0.2047, + "step": 1858 + }, + { + "epoch": 0.127163280662152, + "grad_norm": 2.3383285316303333, + "learning_rate": 9.754597484547223e-06, + "loss": 0.2567, + "step": 1859 + }, + { + "epoch": 0.12723168479376154, + "grad_norm": 1.5052739612226513, + "learning_rate": 9.754254586006636e-06, + "loss": 0.1461, + "step": 1860 + }, + { + "epoch": 0.1273000889253711, + "grad_norm": 1.9052990511331511, + "learning_rate": 9.753911454104272e-06, + "loss": 0.2759, + "step": 1861 + }, + { + "epoch": 0.12736849305698064, + "grad_norm": 1.6358282539157967, + "learning_rate": 9.75356808885698e-06, + "loss": 0.2396, + "step": 1862 + }, + { + "epoch": 0.1274368971885902, + "grad_norm": 2.150163551904746, + "learning_rate": 9.753224490281612e-06, + "loss": 0.3399, + "step": 1863 + }, + { + "epoch": 0.12750530132019974, + "grad_norm": 2.1868945578185457, + "learning_rate": 9.75288065839503e-06, + "loss": 0.373, + "step": 1864 + }, + { + "epoch": 0.1275737054518093, + "grad_norm": 1.9812227578200603, + "learning_rate": 9.752536593214118e-06, + "loss": 0.4527, + "step": 1865 + }, + { + "epoch": 0.12764210958341884, + "grad_norm": 1.7417562062079137, + "learning_rate": 9.752192294755757e-06, + "loss": 0.3648, + "step": 1866 + }, + { + "epoch": 0.1277105137150284, + "grad_norm": 3.1999626272969595, + "learning_rate": 9.75184776303685e-06, + "loss": 0.6352, + "step": 1867 + }, + { + "epoch": 0.12777891784663795, + "grad_norm": 2.22096870531945, + "learning_rate": 9.751502998074311e-06, + "loss": 0.3223, + "step": 1868 + }, + { + "epoch": 0.1278473219782475, + "grad_norm": 1.9738558436051141, + "learning_rate": 9.751157999885057e-06, + "loss": 0.3826, + "step": 1869 + }, + { + "epoch": 0.12791572610985705, + "grad_norm": 1.865106440445715, + "learning_rate": 9.750812768486028e-06, + "loss": 0.3312, + "step": 1870 + }, + { + "epoch": 0.1279841302414666, + "grad_norm": 2.8243735179994314, + "learning_rate": 9.750467303894165e-06, + "loss": 0.7745, + "step": 1871 + }, + { + "epoch": 0.12805253437307612, + "grad_norm": 2.0161902056611756, + "learning_rate": 9.750121606126429e-06, + "loss": 0.3783, + "step": 1872 + }, + { + "epoch": 0.12812093850468567, + "grad_norm": 2.542477632084732, + "learning_rate": 9.749775675199784e-06, + "loss": 0.5436, + "step": 1873 + }, + { + "epoch": 0.12818934263629522, + "grad_norm": 1.8415977516017559, + "learning_rate": 9.749429511131213e-06, + "loss": 0.3901, + "step": 1874 + }, + { + "epoch": 0.12825774676790477, + "grad_norm": 1.8175993743710297, + "learning_rate": 9.749083113937707e-06, + "loss": 0.4232, + "step": 1875 + }, + { + "epoch": 0.12832615089951432, + "grad_norm": 2.6459047421758792, + "learning_rate": 9.748736483636269e-06, + "loss": 0.6906, + "step": 1876 + }, + { + "epoch": 0.12839455503112387, + "grad_norm": 1.8352452210692913, + "learning_rate": 9.748389620243914e-06, + "loss": 0.3298, + "step": 1877 + }, + { + "epoch": 0.12846295916273343, + "grad_norm": 1.9235807037035044, + "learning_rate": 9.748042523777665e-06, + "loss": 0.3096, + "step": 1878 + }, + { + "epoch": 0.12853136329434298, + "grad_norm": 2.662125831603516, + "learning_rate": 9.747695194254561e-06, + "loss": 0.5599, + "step": 1879 + }, + { + "epoch": 0.12859976742595253, + "grad_norm": 2.276098653631322, + "learning_rate": 9.747347631691652e-06, + "loss": 0.4067, + "step": 1880 + }, + { + "epoch": 0.12866817155756208, + "grad_norm": 1.5319105981046934, + "learning_rate": 9.746999836105994e-06, + "loss": 0.2812, + "step": 1881 + }, + { + "epoch": 0.12873657568917163, + "grad_norm": 1.7676532539713876, + "learning_rate": 9.746651807514663e-06, + "loss": 0.3288, + "step": 1882 + }, + { + "epoch": 0.12880497982078118, + "grad_norm": 2.233886098392588, + "learning_rate": 9.74630354593474e-06, + "loss": 0.4605, + "step": 1883 + }, + { + "epoch": 0.12887338395239073, + "grad_norm": 1.9780863138908604, + "learning_rate": 9.745955051383319e-06, + "loss": 0.2954, + "step": 1884 + }, + { + "epoch": 0.12894178808400028, + "grad_norm": 2.0523757984344067, + "learning_rate": 9.745606323877506e-06, + "loss": 0.2475, + "step": 1885 + }, + { + "epoch": 0.12901019221560983, + "grad_norm": 1.429874303735376, + "learning_rate": 9.745257363434418e-06, + "loss": 0.2901, + "step": 1886 + }, + { + "epoch": 0.12907859634721938, + "grad_norm": 2.2056505661388264, + "learning_rate": 9.744908170071184e-06, + "loss": 0.5005, + "step": 1887 + }, + { + "epoch": 0.12914700047882893, + "grad_norm": 2.2107958057411574, + "learning_rate": 9.744558743804944e-06, + "loss": 0.4793, + "step": 1888 + }, + { + "epoch": 0.12921540461043848, + "grad_norm": 2.304212249634942, + "learning_rate": 9.74420908465285e-06, + "loss": 0.5769, + "step": 1889 + }, + { + "epoch": 0.129283808742048, + "grad_norm": 2.8498312933663326, + "learning_rate": 9.743859192632065e-06, + "loss": 0.384, + "step": 1890 + }, + { + "epoch": 0.12935221287365756, + "grad_norm": 3.0780212413901995, + "learning_rate": 9.743509067759761e-06, + "loss": 0.7963, + "step": 1891 + }, + { + "epoch": 0.1294206170052671, + "grad_norm": 1.9931385911494284, + "learning_rate": 9.743158710053129e-06, + "loss": 0.5036, + "step": 1892 + }, + { + "epoch": 0.12948902113687666, + "grad_norm": 2.2675367149495043, + "learning_rate": 9.74280811952936e-06, + "loss": 0.4048, + "step": 1893 + }, + { + "epoch": 0.1295574252684862, + "grad_norm": 2.431743683986515, + "learning_rate": 9.742457296205668e-06, + "loss": 0.6561, + "step": 1894 + }, + { + "epoch": 0.12962582940009576, + "grad_norm": 2.6057319336614126, + "learning_rate": 9.742106240099269e-06, + "loss": 0.5389, + "step": 1895 + }, + { + "epoch": 0.1296942335317053, + "grad_norm": 1.8663564692134216, + "learning_rate": 9.741754951227398e-06, + "loss": 0.1985, + "step": 1896 + }, + { + "epoch": 0.12976263766331486, + "grad_norm": 2.404526192135938, + "learning_rate": 9.741403429607294e-06, + "loss": 0.3844, + "step": 1897 + }, + { + "epoch": 0.1298310417949244, + "grad_norm": 2.9663447173002995, + "learning_rate": 9.741051675256216e-06, + "loss": 0.6152, + "step": 1898 + }, + { + "epoch": 0.12989944592653396, + "grad_norm": 1.9287645511177196, + "learning_rate": 9.740699688191426e-06, + "loss": 0.3397, + "step": 1899 + }, + { + "epoch": 0.1299678500581435, + "grad_norm": 2.4569039486169673, + "learning_rate": 9.740347468430203e-06, + "loss": 0.5622, + "step": 1900 + }, + { + "epoch": 0.13003625418975306, + "grad_norm": 2.2024143777370364, + "learning_rate": 9.739995015989835e-06, + "loss": 0.5637, + "step": 1901 + }, + { + "epoch": 0.13010465832136262, + "grad_norm": 2.239843063301165, + "learning_rate": 9.739642330887624e-06, + "loss": 0.5527, + "step": 1902 + }, + { + "epoch": 0.13017306245297217, + "grad_norm": 2.0597818843866103, + "learning_rate": 9.739289413140879e-06, + "loss": 0.325, + "step": 1903 + }, + { + "epoch": 0.13024146658458172, + "grad_norm": 2.257212701081403, + "learning_rate": 9.738936262766923e-06, + "loss": 0.5326, + "step": 1904 + }, + { + "epoch": 0.13030987071619127, + "grad_norm": 2.308533243332008, + "learning_rate": 9.738582879783091e-06, + "loss": 0.6161, + "step": 1905 + }, + { + "epoch": 0.13037827484780082, + "grad_norm": 2.75830867939752, + "learning_rate": 9.73822926420673e-06, + "loss": 0.8568, + "step": 1906 + }, + { + "epoch": 0.13044667897941037, + "grad_norm": 2.136607528107145, + "learning_rate": 9.737875416055196e-06, + "loss": 0.3595, + "step": 1907 + }, + { + "epoch": 0.13051508311101992, + "grad_norm": 2.5720944879904755, + "learning_rate": 9.737521335345858e-06, + "loss": 0.6937, + "step": 1908 + }, + { + "epoch": 0.13058348724262944, + "grad_norm": 2.2053817864304803, + "learning_rate": 9.737167022096095e-06, + "loss": 0.4071, + "step": 1909 + }, + { + "epoch": 0.130651891374239, + "grad_norm": 2.0807597122687587, + "learning_rate": 9.736812476323299e-06, + "loss": 0.44, + "step": 1910 + }, + { + "epoch": 0.13072029550584854, + "grad_norm": 2.6938795067441172, + "learning_rate": 9.736457698044873e-06, + "loss": 0.8803, + "step": 1911 + }, + { + "epoch": 0.1307886996374581, + "grad_norm": 1.7596945287841506, + "learning_rate": 9.73610268727823e-06, + "loss": 0.268, + "step": 1912 + }, + { + "epoch": 0.13085710376906765, + "grad_norm": 2.5085837464681093, + "learning_rate": 9.735747444040798e-06, + "loss": 0.3404, + "step": 1913 + }, + { + "epoch": 0.1309255079006772, + "grad_norm": 2.146217378180681, + "learning_rate": 9.735391968350012e-06, + "loss": 0.3723, + "step": 1914 + }, + { + "epoch": 0.13099391203228675, + "grad_norm": 2.492980424309117, + "learning_rate": 9.735036260223322e-06, + "loss": 0.4639, + "step": 1915 + }, + { + "epoch": 0.1310623161638963, + "grad_norm": 1.832371385310156, + "learning_rate": 9.734680319678187e-06, + "loss": 0.2616, + "step": 1916 + }, + { + "epoch": 0.13113072029550585, + "grad_norm": 2.332137528750146, + "learning_rate": 9.734324146732076e-06, + "loss": 0.2967, + "step": 1917 + }, + { + "epoch": 0.1311991244271154, + "grad_norm": 1.7693153232763632, + "learning_rate": 9.733967741402477e-06, + "loss": 0.3323, + "step": 1918 + }, + { + "epoch": 0.13126752855872495, + "grad_norm": 2.2382068255502197, + "learning_rate": 9.733611103706882e-06, + "loss": 0.4987, + "step": 1919 + }, + { + "epoch": 0.1313359326903345, + "grad_norm": 2.857701538855891, + "learning_rate": 9.733254233662791e-06, + "loss": 0.6157, + "step": 1920 + }, + { + "epoch": 0.13140433682194405, + "grad_norm": 2.2837035949786313, + "learning_rate": 9.732897131287731e-06, + "loss": 0.3372, + "step": 1921 + }, + { + "epoch": 0.1314727409535536, + "grad_norm": 2.089166684858882, + "learning_rate": 9.73253979659922e-06, + "loss": 0.4576, + "step": 1922 + }, + { + "epoch": 0.13154114508516315, + "grad_norm": 2.3138175059362514, + "learning_rate": 9.732182229614807e-06, + "loss": 0.3858, + "step": 1923 + }, + { + "epoch": 0.1316095492167727, + "grad_norm": 2.215998326796716, + "learning_rate": 9.731824430352036e-06, + "loss": 0.4477, + "step": 1924 + }, + { + "epoch": 0.13167795334838225, + "grad_norm": 1.7751059728875564, + "learning_rate": 9.731466398828474e-06, + "loss": 0.2558, + "step": 1925 + }, + { + "epoch": 0.1317463574799918, + "grad_norm": 1.7489935096552958, + "learning_rate": 9.731108135061691e-06, + "loss": 0.2561, + "step": 1926 + }, + { + "epoch": 0.13181476161160133, + "grad_norm": 2.445314974067582, + "learning_rate": 9.730749639069275e-06, + "loss": 0.5215, + "step": 1927 + }, + { + "epoch": 0.13188316574321088, + "grad_norm": 2.4728384402801957, + "learning_rate": 9.730390910868821e-06, + "loss": 0.3142, + "step": 1928 + }, + { + "epoch": 0.13195156987482043, + "grad_norm": 2.170906317418152, + "learning_rate": 9.73003195047794e-06, + "loss": 0.4633, + "step": 1929 + }, + { + "epoch": 0.13201997400642998, + "grad_norm": 2.8114799345095207, + "learning_rate": 9.72967275791425e-06, + "loss": 0.6864, + "step": 1930 + }, + { + "epoch": 0.13208837813803953, + "grad_norm": 2.1956127205822127, + "learning_rate": 9.729313333195383e-06, + "loss": 0.411, + "step": 1931 + }, + { + "epoch": 0.13215678226964908, + "grad_norm": 2.151224766266855, + "learning_rate": 9.728953676338977e-06, + "loss": 0.4941, + "step": 1932 + }, + { + "epoch": 0.13222518640125863, + "grad_norm": 2.0507096995633023, + "learning_rate": 9.72859378736269e-06, + "loss": 0.3809, + "step": 1933 + }, + { + "epoch": 0.13229359053286818, + "grad_norm": 1.7439124266712704, + "learning_rate": 9.728233666284187e-06, + "loss": 0.32, + "step": 1934 + }, + { + "epoch": 0.13236199466447773, + "grad_norm": 2.1244189716242756, + "learning_rate": 9.727873313121142e-06, + "loss": 0.4596, + "step": 1935 + }, + { + "epoch": 0.13243039879608728, + "grad_norm": 2.4204300292251406, + "learning_rate": 9.727512727891242e-06, + "loss": 0.7471, + "step": 1936 + }, + { + "epoch": 0.13249880292769683, + "grad_norm": 1.9817450584777343, + "learning_rate": 9.727151910612192e-06, + "loss": 0.5003, + "step": 1937 + }, + { + "epoch": 0.13256720705930639, + "grad_norm": 2.5543831693566723, + "learning_rate": 9.726790861301696e-06, + "loss": 0.3878, + "step": 1938 + }, + { + "epoch": 0.13263561119091594, + "grad_norm": 2.1271780926678665, + "learning_rate": 9.72642957997748e-06, + "loss": 0.4599, + "step": 1939 + }, + { + "epoch": 0.1327040153225255, + "grad_norm": 2.6586863844362583, + "learning_rate": 9.726068066657279e-06, + "loss": 0.6774, + "step": 1940 + }, + { + "epoch": 0.13277241945413504, + "grad_norm": 2.635576229517165, + "learning_rate": 9.725706321358832e-06, + "loss": 0.6603, + "step": 1941 + }, + { + "epoch": 0.1328408235857446, + "grad_norm": 2.5229031584625075, + "learning_rate": 9.725344344099901e-06, + "loss": 0.4033, + "step": 1942 + }, + { + "epoch": 0.13290922771735414, + "grad_norm": 1.552052968536777, + "learning_rate": 9.72498213489825e-06, + "loss": 0.1749, + "step": 1943 + }, + { + "epoch": 0.1329776318489637, + "grad_norm": 1.5690256107486313, + "learning_rate": 9.724619693771658e-06, + "loss": 0.1958, + "step": 1944 + }, + { + "epoch": 0.1330460359805732, + "grad_norm": 2.281599880194962, + "learning_rate": 9.724257020737918e-06, + "loss": 0.3821, + "step": 1945 + }, + { + "epoch": 0.13311444011218276, + "grad_norm": 1.9155817697510373, + "learning_rate": 9.72389411581483e-06, + "loss": 0.1854, + "step": 1946 + }, + { + "epoch": 0.13318284424379231, + "grad_norm": 2.4884821273230107, + "learning_rate": 9.723530979020208e-06, + "loss": 0.5781, + "step": 1947 + }, + { + "epoch": 0.13325124837540187, + "grad_norm": 2.901475573723365, + "learning_rate": 9.723167610371876e-06, + "loss": 0.742, + "step": 1948 + }, + { + "epoch": 0.13331965250701142, + "grad_norm": 1.649407090329871, + "learning_rate": 9.722804009887667e-06, + "loss": 0.3278, + "step": 1949 + }, + { + "epoch": 0.13338805663862097, + "grad_norm": 1.8498835738057033, + "learning_rate": 9.722440177585433e-06, + "loss": 0.3068, + "step": 1950 + }, + { + "epoch": 0.13345646077023052, + "grad_norm": 2.4421589171823097, + "learning_rate": 9.72207611348303e-06, + "loss": 0.4016, + "step": 1951 + }, + { + "epoch": 0.13352486490184007, + "grad_norm": 2.124816451840917, + "learning_rate": 9.72171181759833e-06, + "loss": 0.269, + "step": 1952 + }, + { + "epoch": 0.13359326903344962, + "grad_norm": 2.269298410207959, + "learning_rate": 9.721347289949212e-06, + "loss": 0.32, + "step": 1953 + }, + { + "epoch": 0.13366167316505917, + "grad_norm": 2.5444219248929967, + "learning_rate": 9.72098253055357e-06, + "loss": 0.2749, + "step": 1954 + }, + { + "epoch": 0.13373007729666872, + "grad_norm": 2.4479708192124705, + "learning_rate": 9.720617539429308e-06, + "loss": 0.5894, + "step": 1955 + }, + { + "epoch": 0.13379848142827827, + "grad_norm": 2.7307663565710447, + "learning_rate": 9.720252316594344e-06, + "loss": 0.6695, + "step": 1956 + }, + { + "epoch": 0.13386688555988782, + "grad_norm": 2.519429002210049, + "learning_rate": 9.7198868620666e-06, + "loss": 0.6163, + "step": 1957 + }, + { + "epoch": 0.13393528969149737, + "grad_norm": 2.3713024462337984, + "learning_rate": 9.719521175864014e-06, + "loss": 0.55, + "step": 1958 + }, + { + "epoch": 0.13400369382310692, + "grad_norm": 2.1448449841641186, + "learning_rate": 9.719155258004542e-06, + "loss": 0.3758, + "step": 1959 + }, + { + "epoch": 0.13407209795471647, + "grad_norm": 1.5366152947203988, + "learning_rate": 9.718789108506141e-06, + "loss": 0.2458, + "step": 1960 + }, + { + "epoch": 0.13414050208632602, + "grad_norm": 3.0339636797915595, + "learning_rate": 9.718422727386782e-06, + "loss": 0.6408, + "step": 1961 + }, + { + "epoch": 0.13420890621793558, + "grad_norm": 4.671436914425687, + "learning_rate": 9.718056114664453e-06, + "loss": 0.504, + "step": 1962 + }, + { + "epoch": 0.13427731034954513, + "grad_norm": 2.458926247901266, + "learning_rate": 9.717689270357145e-06, + "loss": 0.5717, + "step": 1963 + }, + { + "epoch": 0.13434571448115465, + "grad_norm": 1.8838378441141885, + "learning_rate": 9.717322194482867e-06, + "loss": 0.4188, + "step": 1964 + }, + { + "epoch": 0.1344141186127642, + "grad_norm": 3.3449066514587917, + "learning_rate": 9.716954887059636e-06, + "loss": 1.1848, + "step": 1965 + }, + { + "epoch": 0.13448252274437375, + "grad_norm": 2.439953645007891, + "learning_rate": 9.71658734810548e-06, + "loss": 0.334, + "step": 1966 + }, + { + "epoch": 0.1345509268759833, + "grad_norm": 2.38381167240508, + "learning_rate": 9.716219577638442e-06, + "loss": 0.5066, + "step": 1967 + }, + { + "epoch": 0.13461933100759285, + "grad_norm": 2.294347553218899, + "learning_rate": 9.715851575676572e-06, + "loss": 0.4672, + "step": 1968 + }, + { + "epoch": 0.1346877351392024, + "grad_norm": 2.4364995890273953, + "learning_rate": 9.715483342237935e-06, + "loss": 0.6473, + "step": 1969 + }, + { + "epoch": 0.13475613927081195, + "grad_norm": 2.793581029830044, + "learning_rate": 9.715114877340604e-06, + "loss": 0.7333, + "step": 1970 + }, + { + "epoch": 0.1348245434024215, + "grad_norm": 2.1767510593287303, + "learning_rate": 9.714746181002665e-06, + "loss": 0.5939, + "step": 1971 + }, + { + "epoch": 0.13489294753403105, + "grad_norm": 2.166915142654646, + "learning_rate": 9.714377253242218e-06, + "loss": 0.4489, + "step": 1972 + }, + { + "epoch": 0.1349613516656406, + "grad_norm": 2.2530040356654504, + "learning_rate": 9.714008094077367e-06, + "loss": 0.4041, + "step": 1973 + }, + { + "epoch": 0.13502975579725016, + "grad_norm": 1.5230437821471559, + "learning_rate": 9.713638703526239e-06, + "loss": 0.2935, + "step": 1974 + }, + { + "epoch": 0.1350981599288597, + "grad_norm": 2.018569126824943, + "learning_rate": 9.713269081606958e-06, + "loss": 0.452, + "step": 1975 + }, + { + "epoch": 0.13516656406046926, + "grad_norm": 1.8878803941957387, + "learning_rate": 9.71289922833767e-06, + "loss": 0.3135, + "step": 1976 + }, + { + "epoch": 0.1352349681920788, + "grad_norm": 2.2328536115936504, + "learning_rate": 9.712529143736533e-06, + "loss": 0.4808, + "step": 1977 + }, + { + "epoch": 0.13530337232368836, + "grad_norm": 2.1790928832856853, + "learning_rate": 9.712158827821706e-06, + "loss": 0.4548, + "step": 1978 + }, + { + "epoch": 0.1353717764552979, + "grad_norm": 2.3841556578173893, + "learning_rate": 9.711788280611371e-06, + "loss": 0.5887, + "step": 1979 + }, + { + "epoch": 0.13544018058690746, + "grad_norm": 2.084968516851938, + "learning_rate": 9.711417502123714e-06, + "loss": 0.3197, + "step": 1980 + }, + { + "epoch": 0.135508584718517, + "grad_norm": 1.861134185774496, + "learning_rate": 9.711046492376935e-06, + "loss": 0.4027, + "step": 1981 + }, + { + "epoch": 0.13557698885012653, + "grad_norm": 1.6430285263796418, + "learning_rate": 9.710675251389243e-06, + "loss": 0.3301, + "step": 1982 + }, + { + "epoch": 0.13564539298173608, + "grad_norm": 1.3373422710162273, + "learning_rate": 9.710303779178863e-06, + "loss": 0.1939, + "step": 1983 + }, + { + "epoch": 0.13571379711334564, + "grad_norm": 2.3767987799078076, + "learning_rate": 9.709932075764028e-06, + "loss": 0.5444, + "step": 1984 + }, + { + "epoch": 0.1357822012449552, + "grad_norm": 2.0689039909178177, + "learning_rate": 9.709560141162982e-06, + "loss": 0.4748, + "step": 1985 + }, + { + "epoch": 0.13585060537656474, + "grad_norm": 2.2642558992605317, + "learning_rate": 9.709187975393983e-06, + "loss": 0.6362, + "step": 1986 + }, + { + "epoch": 0.1359190095081743, + "grad_norm": 2.258320319013879, + "learning_rate": 9.708815578475297e-06, + "loss": 0.5387, + "step": 1987 + }, + { + "epoch": 0.13598741363978384, + "grad_norm": 1.9249574589982845, + "learning_rate": 9.708442950425202e-06, + "loss": 0.4808, + "step": 1988 + }, + { + "epoch": 0.1360558177713934, + "grad_norm": 1.6881285132746202, + "learning_rate": 9.708070091261994e-06, + "loss": 0.3892, + "step": 1989 + }, + { + "epoch": 0.13612422190300294, + "grad_norm": 2.118644482241303, + "learning_rate": 9.707697001003968e-06, + "loss": 0.3839, + "step": 1990 + }, + { + "epoch": 0.1361926260346125, + "grad_norm": 2.364233622689171, + "learning_rate": 9.707323679669441e-06, + "loss": 0.4965, + "step": 1991 + }, + { + "epoch": 0.13626103016622204, + "grad_norm": 3.199802359423151, + "learning_rate": 9.706950127276738e-06, + "loss": 0.4802, + "step": 1992 + }, + { + "epoch": 0.1363294342978316, + "grad_norm": 2.204600563741413, + "learning_rate": 9.70657634384419e-06, + "loss": 0.5112, + "step": 1993 + }, + { + "epoch": 0.13639783842944114, + "grad_norm": 1.782736392323246, + "learning_rate": 9.706202329390148e-06, + "loss": 0.2756, + "step": 1994 + }, + { + "epoch": 0.1364662425610507, + "grad_norm": 2.2001299012370503, + "learning_rate": 9.705828083932972e-06, + "loss": 0.4285, + "step": 1995 + }, + { + "epoch": 0.13653464669266024, + "grad_norm": 2.2148195732990446, + "learning_rate": 9.705453607491027e-06, + "loss": 0.6457, + "step": 1996 + }, + { + "epoch": 0.1366030508242698, + "grad_norm": 2.369052643176135, + "learning_rate": 9.705078900082696e-06, + "loss": 0.3413, + "step": 1997 + }, + { + "epoch": 0.13667145495587935, + "grad_norm": 2.362352551294313, + "learning_rate": 9.704703961726374e-06, + "loss": 0.5719, + "step": 1998 + }, + { + "epoch": 0.1367398590874889, + "grad_norm": 1.5217825777318414, + "learning_rate": 9.704328792440462e-06, + "loss": 0.236, + "step": 1999 + }, + { + "epoch": 0.13680826321909845, + "grad_norm": 2.036180856993218, + "learning_rate": 9.703953392243377e-06, + "loss": 0.4581, + "step": 2000 + }, + { + "epoch": 0.13687666735070797, + "grad_norm": 1.8486653550271668, + "learning_rate": 9.703577761153542e-06, + "loss": 0.3318, + "step": 2001 + }, + { + "epoch": 0.13694507148231752, + "grad_norm": 1.2987867162313211, + "learning_rate": 9.703201899189399e-06, + "loss": 0.2184, + "step": 2002 + }, + { + "epoch": 0.13701347561392707, + "grad_norm": 1.7897481062738274, + "learning_rate": 9.702825806369394e-06, + "loss": 0.3741, + "step": 2003 + }, + { + "epoch": 0.13708187974553662, + "grad_norm": 2.5418801827069535, + "learning_rate": 9.702449482711989e-06, + "loss": 0.1836, + "step": 2004 + }, + { + "epoch": 0.13715028387714617, + "grad_norm": 1.7452305531196306, + "learning_rate": 9.702072928235658e-06, + "loss": 0.2937, + "step": 2005 + }, + { + "epoch": 0.13721868800875572, + "grad_norm": 1.8067937177797997, + "learning_rate": 9.701696142958878e-06, + "loss": 0.2085, + "step": 2006 + }, + { + "epoch": 0.13728709214036527, + "grad_norm": 2.0812669893665348, + "learning_rate": 9.701319126900149e-06, + "loss": 0.4172, + "step": 2007 + }, + { + "epoch": 0.13735549627197483, + "grad_norm": 1.5174669644799275, + "learning_rate": 9.700941880077974e-06, + "loss": 0.2435, + "step": 2008 + }, + { + "epoch": 0.13742390040358438, + "grad_norm": 2.10649241119347, + "learning_rate": 9.700564402510871e-06, + "loss": 0.4457, + "step": 2009 + }, + { + "epoch": 0.13749230453519393, + "grad_norm": 2.206506717432214, + "learning_rate": 9.700186694217369e-06, + "loss": 0.49, + "step": 2010 + }, + { + "epoch": 0.13756070866680348, + "grad_norm": 3.0885717423739956, + "learning_rate": 9.699808755216007e-06, + "loss": 0.5836, + "step": 2011 + }, + { + "epoch": 0.13762911279841303, + "grad_norm": 2.3946775979752424, + "learning_rate": 9.699430585525336e-06, + "loss": 0.6148, + "step": 2012 + }, + { + "epoch": 0.13769751693002258, + "grad_norm": 2.5847640011156665, + "learning_rate": 9.699052185163917e-06, + "loss": 0.738, + "step": 2013 + }, + { + "epoch": 0.13776592106163213, + "grad_norm": 2.5480936563382537, + "learning_rate": 9.698673554150327e-06, + "loss": 0.5849, + "step": 2014 + }, + { + "epoch": 0.13783432519324168, + "grad_norm": 1.3610430053563436, + "learning_rate": 9.69829469250315e-06, + "loss": 0.228, + "step": 2015 + }, + { + "epoch": 0.13790272932485123, + "grad_norm": 2.2483209729567717, + "learning_rate": 9.697915600240979e-06, + "loss": 0.3754, + "step": 2016 + }, + { + "epoch": 0.13797113345646078, + "grad_norm": 2.0270920452062704, + "learning_rate": 9.697536277382426e-06, + "loss": 0.3626, + "step": 2017 + }, + { + "epoch": 0.13803953758807033, + "grad_norm": 2.453272438478107, + "learning_rate": 9.697156723946108e-06, + "loss": 0.4761, + "step": 2018 + }, + { + "epoch": 0.13810794171967986, + "grad_norm": 1.5917288736433541, + "learning_rate": 9.696776939950657e-06, + "loss": 0.2658, + "step": 2019 + }, + { + "epoch": 0.1381763458512894, + "grad_norm": 1.3937693151923212, + "learning_rate": 9.696396925414711e-06, + "loss": 0.1962, + "step": 2020 + }, + { + "epoch": 0.13824474998289896, + "grad_norm": 1.7328042293278971, + "learning_rate": 9.696016680356928e-06, + "loss": 0.2696, + "step": 2021 + }, + { + "epoch": 0.1383131541145085, + "grad_norm": 2.295154900590374, + "learning_rate": 9.695636204795966e-06, + "loss": 0.5574, + "step": 2022 + }, + { + "epoch": 0.13838155824611806, + "grad_norm": 2.1042599578672565, + "learning_rate": 9.695255498750506e-06, + "loss": 0.3678, + "step": 2023 + }, + { + "epoch": 0.1384499623777276, + "grad_norm": 2.3633904676184936, + "learning_rate": 9.694874562239233e-06, + "loss": 0.2503, + "step": 2024 + }, + { + "epoch": 0.13851836650933716, + "grad_norm": 1.9122351896296228, + "learning_rate": 9.694493395280846e-06, + "loss": 0.3889, + "step": 2025 + }, + { + "epoch": 0.1385867706409467, + "grad_norm": 1.7650504823251731, + "learning_rate": 9.694111997894053e-06, + "loss": 0.3249, + "step": 2026 + }, + { + "epoch": 0.13865517477255626, + "grad_norm": 2.8098964932747044, + "learning_rate": 9.693730370097576e-06, + "loss": 0.7769, + "step": 2027 + }, + { + "epoch": 0.1387235789041658, + "grad_norm": 1.9536058062977935, + "learning_rate": 9.693348511910146e-06, + "loss": 0.3124, + "step": 2028 + }, + { + "epoch": 0.13879198303577536, + "grad_norm": 1.5576379613178652, + "learning_rate": 9.692966423350505e-06, + "loss": 0.2021, + "step": 2029 + }, + { + "epoch": 0.1388603871673849, + "grad_norm": 2.1321474862065557, + "learning_rate": 9.692584104437413e-06, + "loss": 0.4331, + "step": 2030 + }, + { + "epoch": 0.13892879129899446, + "grad_norm": 2.048851383975424, + "learning_rate": 9.69220155518963e-06, + "loss": 0.3186, + "step": 2031 + }, + { + "epoch": 0.13899719543060401, + "grad_norm": 1.4580967972436072, + "learning_rate": 9.691818775625939e-06, + "loss": 0.2412, + "step": 2032 + }, + { + "epoch": 0.13906559956221357, + "grad_norm": 2.4178104841853267, + "learning_rate": 9.691435765765124e-06, + "loss": 0.3115, + "step": 2033 + }, + { + "epoch": 0.13913400369382312, + "grad_norm": 2.1553079717277974, + "learning_rate": 9.691052525625989e-06, + "loss": 0.4676, + "step": 2034 + }, + { + "epoch": 0.13920240782543267, + "grad_norm": 2.3440548634935534, + "learning_rate": 9.690669055227341e-06, + "loss": 0.2184, + "step": 2035 + }, + { + "epoch": 0.13927081195704222, + "grad_norm": 2.1908591160309707, + "learning_rate": 9.690285354588007e-06, + "loss": 0.4734, + "step": 2036 + }, + { + "epoch": 0.13933921608865177, + "grad_norm": 1.8857450421004023, + "learning_rate": 9.689901423726817e-06, + "loss": 0.1736, + "step": 2037 + }, + { + "epoch": 0.1394076202202613, + "grad_norm": 2.31130465981862, + "learning_rate": 9.689517262662617e-06, + "loss": 0.7169, + "step": 2038 + }, + { + "epoch": 0.13947602435187084, + "grad_norm": 2.260899337678477, + "learning_rate": 9.689132871414266e-06, + "loss": 0.4768, + "step": 2039 + }, + { + "epoch": 0.1395444284834804, + "grad_norm": 2.1717523567645367, + "learning_rate": 9.68874825000063e-06, + "loss": 0.5489, + "step": 2040 + }, + { + "epoch": 0.13961283261508994, + "grad_norm": 1.8490050878568252, + "learning_rate": 9.688363398440587e-06, + "loss": 0.4488, + "step": 2041 + }, + { + "epoch": 0.1396812367466995, + "grad_norm": 2.2785154276271755, + "learning_rate": 9.68797831675303e-06, + "loss": 0.2408, + "step": 2042 + }, + { + "epoch": 0.13974964087830904, + "grad_norm": 1.7196518699286447, + "learning_rate": 9.687593004956858e-06, + "loss": 0.3034, + "step": 2043 + }, + { + "epoch": 0.1398180450099186, + "grad_norm": 2.560873422525091, + "learning_rate": 9.687207463070986e-06, + "loss": 0.3931, + "step": 2044 + }, + { + "epoch": 0.13988644914152815, + "grad_norm": 1.8359587850757195, + "learning_rate": 9.686821691114339e-06, + "loss": 0.2424, + "step": 2045 + }, + { + "epoch": 0.1399548532731377, + "grad_norm": 2.2499690252397238, + "learning_rate": 9.686435689105848e-06, + "loss": 0.4995, + "step": 2046 + }, + { + "epoch": 0.14002325740474725, + "grad_norm": 2.2015709773293692, + "learning_rate": 9.686049457064463e-06, + "loss": 0.6154, + "step": 2047 + }, + { + "epoch": 0.1400916615363568, + "grad_norm": 2.218750151110362, + "learning_rate": 9.685662995009145e-06, + "loss": 0.5267, + "step": 2048 + }, + { + "epoch": 0.14016006566796635, + "grad_norm": 1.7414583296794262, + "learning_rate": 9.685276302958857e-06, + "loss": 0.3711, + "step": 2049 + }, + { + "epoch": 0.1402284697995759, + "grad_norm": 1.9223658179202698, + "learning_rate": 9.684889380932584e-06, + "loss": 0.4077, + "step": 2050 + }, + { + "epoch": 0.14029687393118545, + "grad_norm": 2.265957354141773, + "learning_rate": 9.684502228949317e-06, + "loss": 0.5655, + "step": 2051 + }, + { + "epoch": 0.140365278062795, + "grad_norm": 2.8575033461767245, + "learning_rate": 9.68411484702806e-06, + "loss": 0.8683, + "step": 2052 + }, + { + "epoch": 0.14043368219440455, + "grad_norm": 2.037327623974184, + "learning_rate": 9.683727235187826e-06, + "loss": 0.3915, + "step": 2053 + }, + { + "epoch": 0.1405020863260141, + "grad_norm": 2.059346367285291, + "learning_rate": 9.683339393447644e-06, + "loss": 0.3918, + "step": 2054 + }, + { + "epoch": 0.14057049045762365, + "grad_norm": 1.888587606575812, + "learning_rate": 9.682951321826546e-06, + "loss": 0.3621, + "step": 2055 + }, + { + "epoch": 0.14063889458923318, + "grad_norm": 2.347529193658053, + "learning_rate": 9.682563020343585e-06, + "loss": 0.4032, + "step": 2056 + }, + { + "epoch": 0.14070729872084273, + "grad_norm": 2.183145695475487, + "learning_rate": 9.682174489017819e-06, + "loss": 0.1978, + "step": 2057 + }, + { + "epoch": 0.14077570285245228, + "grad_norm": 1.9728351625191949, + "learning_rate": 9.681785727868319e-06, + "loss": 0.4631, + "step": 2058 + }, + { + "epoch": 0.14084410698406183, + "grad_norm": 1.8511381065313721, + "learning_rate": 9.681396736914169e-06, + "loss": 0.3733, + "step": 2059 + }, + { + "epoch": 0.14091251111567138, + "grad_norm": 2.1022022400984715, + "learning_rate": 9.681007516174458e-06, + "loss": 0.4377, + "step": 2060 + }, + { + "epoch": 0.14098091524728093, + "grad_norm": 2.262492915400588, + "learning_rate": 9.680618065668295e-06, + "loss": 0.5854, + "step": 2061 + }, + { + "epoch": 0.14104931937889048, + "grad_norm": 1.9233313369272318, + "learning_rate": 9.680228385414795e-06, + "loss": 0.4265, + "step": 2062 + }, + { + "epoch": 0.14111772351050003, + "grad_norm": 2.0133095904699236, + "learning_rate": 9.679838475433085e-06, + "loss": 0.2846, + "step": 2063 + }, + { + "epoch": 0.14118612764210958, + "grad_norm": 2.1438478864559247, + "learning_rate": 9.679448335742304e-06, + "loss": 0.416, + "step": 2064 + }, + { + "epoch": 0.14125453177371913, + "grad_norm": 2.44221460006429, + "learning_rate": 9.679057966361603e-06, + "loss": 0.6043, + "step": 2065 + }, + { + "epoch": 0.14132293590532868, + "grad_norm": 2.0719061745589675, + "learning_rate": 9.67866736731014e-06, + "loss": 0.4274, + "step": 2066 + }, + { + "epoch": 0.14139134003693823, + "grad_norm": 1.727802355742209, + "learning_rate": 9.67827653860709e-06, + "loss": 0.3115, + "step": 2067 + }, + { + "epoch": 0.14145974416854779, + "grad_norm": 2.010531302792559, + "learning_rate": 9.677885480271637e-06, + "loss": 0.4693, + "step": 2068 + }, + { + "epoch": 0.14152814830015734, + "grad_norm": 2.0419783589475453, + "learning_rate": 9.677494192322976e-06, + "loss": 0.4363, + "step": 2069 + }, + { + "epoch": 0.1415965524317669, + "grad_norm": 2.6348880318046533, + "learning_rate": 9.677102674780312e-06, + "loss": 0.68, + "step": 2070 + }, + { + "epoch": 0.14166495656337644, + "grad_norm": 2.1912592398279496, + "learning_rate": 9.676710927662863e-06, + "loss": 0.4933, + "step": 2071 + }, + { + "epoch": 0.141733360694986, + "grad_norm": 2.5309823124238133, + "learning_rate": 9.676318950989858e-06, + "loss": 0.3394, + "step": 2072 + }, + { + "epoch": 0.14180176482659554, + "grad_norm": 2.3589964840902393, + "learning_rate": 9.675926744780537e-06, + "loss": 0.5221, + "step": 2073 + }, + { + "epoch": 0.14187016895820506, + "grad_norm": 1.6297485588829246, + "learning_rate": 9.67553430905415e-06, + "loss": 0.3542, + "step": 2074 + }, + { + "epoch": 0.1419385730898146, + "grad_norm": 2.361794664662599, + "learning_rate": 9.675141643829964e-06, + "loss": 0.7145, + "step": 2075 + }, + { + "epoch": 0.14200697722142416, + "grad_norm": 2.046888183958968, + "learning_rate": 9.67474874912725e-06, + "loss": 0.4165, + "step": 2076 + }, + { + "epoch": 0.14207538135303371, + "grad_norm": 1.8712954999747085, + "learning_rate": 9.674355624965293e-06, + "loss": 0.2801, + "step": 2077 + }, + { + "epoch": 0.14214378548464326, + "grad_norm": 2.128694856280006, + "learning_rate": 9.673962271363388e-06, + "loss": 0.5386, + "step": 2078 + }, + { + "epoch": 0.14221218961625282, + "grad_norm": 1.9483767890802042, + "learning_rate": 9.673568688340846e-06, + "loss": 0.4396, + "step": 2079 + }, + { + "epoch": 0.14228059374786237, + "grad_norm": 2.611172490985139, + "learning_rate": 9.673174875916984e-06, + "loss": 0.5465, + "step": 2080 + }, + { + "epoch": 0.14234899787947192, + "grad_norm": 2.346219732488669, + "learning_rate": 9.672780834111131e-06, + "loss": 0.5871, + "step": 2081 + }, + { + "epoch": 0.14241740201108147, + "grad_norm": 2.5704039815374404, + "learning_rate": 9.672386562942633e-06, + "loss": 0.3931, + "step": 2082 + }, + { + "epoch": 0.14248580614269102, + "grad_norm": 2.155405065598809, + "learning_rate": 9.671992062430837e-06, + "loss": 0.536, + "step": 2083 + }, + { + "epoch": 0.14255421027430057, + "grad_norm": 2.697498966098407, + "learning_rate": 9.671597332595111e-06, + "loss": 0.8295, + "step": 2084 + }, + { + "epoch": 0.14262261440591012, + "grad_norm": 2.3084173570815083, + "learning_rate": 9.67120237345483e-06, + "loss": 0.5222, + "step": 2085 + }, + { + "epoch": 0.14269101853751967, + "grad_norm": 1.60351037455546, + "learning_rate": 9.670807185029378e-06, + "loss": 0.2718, + "step": 2086 + }, + { + "epoch": 0.14275942266912922, + "grad_norm": 1.8581985309985896, + "learning_rate": 9.670411767338155e-06, + "loss": 0.4675, + "step": 2087 + }, + { + "epoch": 0.14282782680073877, + "grad_norm": 2.802539193936775, + "learning_rate": 9.670016120400568e-06, + "loss": 0.2734, + "step": 2088 + }, + { + "epoch": 0.14289623093234832, + "grad_norm": 1.9119739737804298, + "learning_rate": 9.66962024423604e-06, + "loss": 0.4351, + "step": 2089 + }, + { + "epoch": 0.14296463506395787, + "grad_norm": 2.618243656278975, + "learning_rate": 9.669224138863999e-06, + "loss": 0.735, + "step": 2090 + }, + { + "epoch": 0.14303303919556742, + "grad_norm": 1.5616221393741294, + "learning_rate": 9.668827804303893e-06, + "loss": 0.1733, + "step": 2091 + }, + { + "epoch": 0.14310144332717697, + "grad_norm": 4.130159287163491, + "learning_rate": 9.668431240575172e-06, + "loss": 0.5376, + "step": 2092 + }, + { + "epoch": 0.1431698474587865, + "grad_norm": 2.32791578389671, + "learning_rate": 9.6680344476973e-06, + "loss": 0.6292, + "step": 2093 + }, + { + "epoch": 0.14323825159039605, + "grad_norm": 1.99425557073327, + "learning_rate": 9.667637425689757e-06, + "loss": 0.1918, + "step": 2094 + }, + { + "epoch": 0.1433066557220056, + "grad_norm": 2.157650009324152, + "learning_rate": 9.66724017457203e-06, + "loss": 0.2503, + "step": 2095 + }, + { + "epoch": 0.14337505985361515, + "grad_norm": 2.815847322871972, + "learning_rate": 9.666842694363616e-06, + "loss": 0.6139, + "step": 2096 + }, + { + "epoch": 0.1434434639852247, + "grad_norm": 2.9217075565114943, + "learning_rate": 9.666444985084027e-06, + "loss": 0.5545, + "step": 2097 + }, + { + "epoch": 0.14351186811683425, + "grad_norm": 2.3132466837662267, + "learning_rate": 9.666047046752784e-06, + "loss": 0.5213, + "step": 2098 + }, + { + "epoch": 0.1435802722484438, + "grad_norm": 3.428514001764872, + "learning_rate": 9.66564887938942e-06, + "loss": 0.5929, + "step": 2099 + }, + { + "epoch": 0.14364867638005335, + "grad_norm": 1.3041459518631873, + "learning_rate": 9.66525048301348e-06, + "loss": 0.1909, + "step": 2100 + }, + { + "epoch": 0.1437170805116629, + "grad_norm": 1.7972157569870015, + "learning_rate": 9.664851857644517e-06, + "loss": 0.393, + "step": 2101 + }, + { + "epoch": 0.14378548464327245, + "grad_norm": 2.3667401309655305, + "learning_rate": 9.664453003302097e-06, + "loss": 0.5968, + "step": 2102 + }, + { + "epoch": 0.143853888774882, + "grad_norm": 2.714916546620888, + "learning_rate": 9.664053920005802e-06, + "loss": 0.6704, + "step": 2103 + }, + { + "epoch": 0.14392229290649156, + "grad_norm": 2.3538029792078334, + "learning_rate": 9.663654607775216e-06, + "loss": 0.4812, + "step": 2104 + }, + { + "epoch": 0.1439906970381011, + "grad_norm": 2.094655502999177, + "learning_rate": 9.663255066629942e-06, + "loss": 0.5781, + "step": 2105 + }, + { + "epoch": 0.14405910116971066, + "grad_norm": 2.152260538143857, + "learning_rate": 9.66285529658959e-06, + "loss": 0.3499, + "step": 2106 + }, + { + "epoch": 0.1441275053013202, + "grad_norm": 1.242663332833704, + "learning_rate": 9.662455297673784e-06, + "loss": 0.1808, + "step": 2107 + }, + { + "epoch": 0.14419590943292976, + "grad_norm": 2.2664187915222254, + "learning_rate": 9.662055069902158e-06, + "loss": 0.478, + "step": 2108 + }, + { + "epoch": 0.1442643135645393, + "grad_norm": 2.468725361278571, + "learning_rate": 9.661654613294356e-06, + "loss": 0.6861, + "step": 2109 + }, + { + "epoch": 0.14433271769614886, + "grad_norm": 2.178979280054231, + "learning_rate": 9.661253927870035e-06, + "loss": 0.4723, + "step": 2110 + }, + { + "epoch": 0.14440112182775838, + "grad_norm": 2.9327453016204963, + "learning_rate": 9.660853013648861e-06, + "loss": 0.2665, + "step": 2111 + }, + { + "epoch": 0.14446952595936793, + "grad_norm": 2.639791012722927, + "learning_rate": 9.660451870650514e-06, + "loss": 0.7062, + "step": 2112 + }, + { + "epoch": 0.14453793009097748, + "grad_norm": 2.315541483908433, + "learning_rate": 9.660050498894685e-06, + "loss": 0.2095, + "step": 2113 + }, + { + "epoch": 0.14460633422258704, + "grad_norm": 2.092029099169495, + "learning_rate": 9.659648898401075e-06, + "loss": 0.34, + "step": 2114 + }, + { + "epoch": 0.14467473835419659, + "grad_norm": 1.624568368367163, + "learning_rate": 9.659247069189394e-06, + "loss": 0.2361, + "step": 2115 + }, + { + "epoch": 0.14474314248580614, + "grad_norm": 1.877842000976201, + "learning_rate": 9.65884501127937e-06, + "loss": 0.3009, + "step": 2116 + }, + { + "epoch": 0.1448115466174157, + "grad_norm": 2.360279913916867, + "learning_rate": 9.658442724690733e-06, + "loss": 0.5544, + "step": 2117 + }, + { + "epoch": 0.14487995074902524, + "grad_norm": 2.811039375804274, + "learning_rate": 9.658040209443234e-06, + "loss": 0.8833, + "step": 2118 + }, + { + "epoch": 0.1449483548806348, + "grad_norm": 2.2720945527665832, + "learning_rate": 9.657637465556626e-06, + "loss": 0.427, + "step": 2119 + }, + { + "epoch": 0.14501675901224434, + "grad_norm": 2.400597348479758, + "learning_rate": 9.65723449305068e-06, + "loss": 0.4437, + "step": 2120 + }, + { + "epoch": 0.1450851631438539, + "grad_norm": 2.250059848890117, + "learning_rate": 9.656831291945178e-06, + "loss": 0.4431, + "step": 2121 + }, + { + "epoch": 0.14515356727546344, + "grad_norm": 1.8622162692399853, + "learning_rate": 9.656427862259904e-06, + "loss": 0.519, + "step": 2122 + }, + { + "epoch": 0.145221971407073, + "grad_norm": 1.7048549833811304, + "learning_rate": 9.656024204014668e-06, + "loss": 0.3212, + "step": 2123 + }, + { + "epoch": 0.14529037553868254, + "grad_norm": 2.1115211518560835, + "learning_rate": 9.65562031722928e-06, + "loss": 0.6032, + "step": 2124 + }, + { + "epoch": 0.1453587796702921, + "grad_norm": 2.206548246346544, + "learning_rate": 9.655216201923565e-06, + "loss": 0.5149, + "step": 2125 + }, + { + "epoch": 0.14542718380190164, + "grad_norm": 1.857040350044752, + "learning_rate": 9.654811858117359e-06, + "loss": 0.3703, + "step": 2126 + }, + { + "epoch": 0.1454955879335112, + "grad_norm": 1.843000445511775, + "learning_rate": 9.654407285830509e-06, + "loss": 0.3417, + "step": 2127 + }, + { + "epoch": 0.14556399206512075, + "grad_norm": 2.09737785807819, + "learning_rate": 9.654002485082874e-06, + "loss": 0.4641, + "step": 2128 + }, + { + "epoch": 0.1456323961967303, + "grad_norm": 2.2886580477712073, + "learning_rate": 9.653597455894322e-06, + "loss": 0.5964, + "step": 2129 + }, + { + "epoch": 0.14570080032833982, + "grad_norm": 2.0262273993017104, + "learning_rate": 9.653192198284735e-06, + "loss": 0.3009, + "step": 2130 + }, + { + "epoch": 0.14576920445994937, + "grad_norm": 1.848085289110202, + "learning_rate": 9.652786712274005e-06, + "loss": 0.3815, + "step": 2131 + }, + { + "epoch": 0.14583760859155892, + "grad_norm": 2.290918860379821, + "learning_rate": 9.652380997882035e-06, + "loss": 0.6519, + "step": 2132 + }, + { + "epoch": 0.14590601272316847, + "grad_norm": 2.373339875896237, + "learning_rate": 9.651975055128741e-06, + "loss": 0.5203, + "step": 2133 + }, + { + "epoch": 0.14597441685477802, + "grad_norm": 2.0882915686378474, + "learning_rate": 9.651568884034047e-06, + "loss": 0.5337, + "step": 2134 + }, + { + "epoch": 0.14604282098638757, + "grad_norm": 5.011674127060455, + "learning_rate": 9.65116248461789e-06, + "loss": 0.5411, + "step": 2135 + }, + { + "epoch": 0.14611122511799712, + "grad_norm": 1.7487823239196032, + "learning_rate": 9.650755856900216e-06, + "loss": 0.3692, + "step": 2136 + }, + { + "epoch": 0.14617962924960667, + "grad_norm": 1.8117213056315187, + "learning_rate": 9.650349000900988e-06, + "loss": 0.3813, + "step": 2137 + }, + { + "epoch": 0.14624803338121622, + "grad_norm": 3.760885148264482, + "learning_rate": 9.649941916640173e-06, + "loss": 0.3994, + "step": 2138 + }, + { + "epoch": 0.14631643751282578, + "grad_norm": 2.6071690183183596, + "learning_rate": 9.649534604137755e-06, + "loss": 0.3758, + "step": 2139 + }, + { + "epoch": 0.14638484164443533, + "grad_norm": 1.7063622309380455, + "learning_rate": 9.649127063413727e-06, + "loss": 0.2426, + "step": 2140 + }, + { + "epoch": 0.14645324577604488, + "grad_norm": 5.095246608545552, + "learning_rate": 9.648719294488092e-06, + "loss": 0.4455, + "step": 2141 + }, + { + "epoch": 0.14652164990765443, + "grad_norm": 2.99158413931698, + "learning_rate": 9.648311297380866e-06, + "loss": 0.3983, + "step": 2142 + }, + { + "epoch": 0.14659005403926398, + "grad_norm": 2.12374637243647, + "learning_rate": 9.647903072112075e-06, + "loss": 0.3937, + "step": 2143 + }, + { + "epoch": 0.14665845817087353, + "grad_norm": 2.03596140919931, + "learning_rate": 9.647494618701756e-06, + "loss": 0.5472, + "step": 2144 + }, + { + "epoch": 0.14672686230248308, + "grad_norm": 3.3936520086371, + "learning_rate": 9.647085937169958e-06, + "loss": 0.3384, + "step": 2145 + }, + { + "epoch": 0.14679526643409263, + "grad_norm": 2.3151586752883535, + "learning_rate": 9.646677027536744e-06, + "loss": 0.6207, + "step": 2146 + }, + { + "epoch": 0.14686367056570218, + "grad_norm": 1.9369731886743167, + "learning_rate": 9.646267889822181e-06, + "loss": 0.3349, + "step": 2147 + }, + { + "epoch": 0.1469320746973117, + "grad_norm": 1.592436749618523, + "learning_rate": 9.645858524046354e-06, + "loss": 0.3292, + "step": 2148 + }, + { + "epoch": 0.14700047882892125, + "grad_norm": 1.4246699457076548, + "learning_rate": 9.645448930229356e-06, + "loss": 0.2533, + "step": 2149 + }, + { + "epoch": 0.1470688829605308, + "grad_norm": 1.7819039332336521, + "learning_rate": 9.645039108391293e-06, + "loss": 0.3733, + "step": 2150 + }, + { + "epoch": 0.14713728709214036, + "grad_norm": 2.7206713694453364, + "learning_rate": 9.644629058552277e-06, + "loss": 0.7635, + "step": 2151 + }, + { + "epoch": 0.1472056912237499, + "grad_norm": 1.618024093661319, + "learning_rate": 9.64421878073244e-06, + "loss": 0.392, + "step": 2152 + }, + { + "epoch": 0.14727409535535946, + "grad_norm": 2.715292156703912, + "learning_rate": 9.643808274951919e-06, + "loss": 0.7555, + "step": 2153 + }, + { + "epoch": 0.147342499486969, + "grad_norm": 7.082879295008917, + "learning_rate": 9.643397541230862e-06, + "loss": 0.4448, + "step": 2154 + }, + { + "epoch": 0.14741090361857856, + "grad_norm": 2.253156603434126, + "learning_rate": 9.64298657958943e-06, + "loss": 0.2316, + "step": 2155 + }, + { + "epoch": 0.1474793077501881, + "grad_norm": 1.9391994521708684, + "learning_rate": 9.6425753900478e-06, + "loss": 0.3919, + "step": 2156 + }, + { + "epoch": 0.14754771188179766, + "grad_norm": 1.6838366074437578, + "learning_rate": 9.642163972626147e-06, + "loss": 0.4545, + "step": 2157 + }, + { + "epoch": 0.1476161160134072, + "grad_norm": 2.084603722589718, + "learning_rate": 9.641752327344671e-06, + "loss": 0.4437, + "step": 2158 + }, + { + "epoch": 0.14768452014501676, + "grad_norm": 2.263859886675678, + "learning_rate": 9.641340454223576e-06, + "loss": 0.3457, + "step": 2159 + }, + { + "epoch": 0.1477529242766263, + "grad_norm": 2.836889312163441, + "learning_rate": 9.64092835328308e-06, + "loss": 0.3643, + "step": 2160 + }, + { + "epoch": 0.14782132840823586, + "grad_norm": 1.5820725612191606, + "learning_rate": 9.640516024543407e-06, + "loss": 0.3457, + "step": 2161 + }, + { + "epoch": 0.14788973253984541, + "grad_norm": 22.70215379353321, + "learning_rate": 9.6401034680248e-06, + "loss": 0.3901, + "step": 2162 + }, + { + "epoch": 0.14795813667145497, + "grad_norm": 1.5613114294782426, + "learning_rate": 9.639690683747507e-06, + "loss": 0.32, + "step": 2163 + }, + { + "epoch": 0.14802654080306452, + "grad_norm": 2.525169631363602, + "learning_rate": 9.639277671731791e-06, + "loss": 0.3585, + "step": 2164 + }, + { + "epoch": 0.14809494493467407, + "grad_norm": 2.136405691753452, + "learning_rate": 9.638864431997925e-06, + "loss": 0.4952, + "step": 2165 + }, + { + "epoch": 0.14816334906628362, + "grad_norm": 1.8061633165144269, + "learning_rate": 9.638450964566191e-06, + "loss": 0.3947, + "step": 2166 + }, + { + "epoch": 0.14823175319789314, + "grad_norm": 1.9832775355218388, + "learning_rate": 9.638037269456886e-06, + "loss": 0.5227, + "step": 2167 + }, + { + "epoch": 0.1483001573295027, + "grad_norm": 1.7617485525989502, + "learning_rate": 9.637623346690313e-06, + "loss": 0.284, + "step": 2168 + }, + { + "epoch": 0.14836856146111224, + "grad_norm": 2.1330758026579177, + "learning_rate": 9.637209196286792e-06, + "loss": 0.2789, + "step": 2169 + }, + { + "epoch": 0.1484369655927218, + "grad_norm": 1.8648550556608852, + "learning_rate": 9.63679481826665e-06, + "loss": 0.3374, + "step": 2170 + }, + { + "epoch": 0.14850536972433134, + "grad_norm": 2.6223371714470995, + "learning_rate": 9.636380212650229e-06, + "loss": 0.5167, + "step": 2171 + }, + { + "epoch": 0.1485737738559409, + "grad_norm": 1.8671368528224697, + "learning_rate": 9.635965379457879e-06, + "loss": 0.3823, + "step": 2172 + }, + { + "epoch": 0.14864217798755044, + "grad_norm": 2.2630747154736546, + "learning_rate": 9.63555031870996e-06, + "loss": 0.508, + "step": 2173 + }, + { + "epoch": 0.14871058211916, + "grad_norm": 1.8274827546298815, + "learning_rate": 9.635135030426846e-06, + "loss": 0.4184, + "step": 2174 + }, + { + "epoch": 0.14877898625076955, + "grad_norm": 2.16853642635083, + "learning_rate": 9.634719514628923e-06, + "loss": 0.6382, + "step": 2175 + }, + { + "epoch": 0.1488473903823791, + "grad_norm": 1.9474643234760636, + "learning_rate": 9.634303771336584e-06, + "loss": 0.4594, + "step": 2176 + }, + { + "epoch": 0.14891579451398865, + "grad_norm": 2.47310099888228, + "learning_rate": 9.633887800570238e-06, + "loss": 0.5918, + "step": 2177 + }, + { + "epoch": 0.1489841986455982, + "grad_norm": 4.263248829773394, + "learning_rate": 9.633471602350302e-06, + "loss": 0.4219, + "step": 2178 + }, + { + "epoch": 0.14905260277720775, + "grad_norm": 2.3039101016897536, + "learning_rate": 9.633055176697205e-06, + "loss": 0.5851, + "step": 2179 + }, + { + "epoch": 0.1491210069088173, + "grad_norm": 2.0195490321662204, + "learning_rate": 9.632638523631388e-06, + "loss": 0.2149, + "step": 2180 + }, + { + "epoch": 0.14918941104042685, + "grad_norm": 1.6285313625403366, + "learning_rate": 9.6322216431733e-06, + "loss": 0.2023, + "step": 2181 + }, + { + "epoch": 0.1492578151720364, + "grad_norm": 1.8032869915779504, + "learning_rate": 9.631804535343404e-06, + "loss": 0.3478, + "step": 2182 + }, + { + "epoch": 0.14932621930364595, + "grad_norm": 2.1568979588947244, + "learning_rate": 9.631387200162176e-06, + "loss": 0.3087, + "step": 2183 + }, + { + "epoch": 0.1493946234352555, + "grad_norm": 2.216515645072842, + "learning_rate": 9.6309696376501e-06, + "loss": 0.7174, + "step": 2184 + }, + { + "epoch": 0.14946302756686503, + "grad_norm": 1.9056469870300112, + "learning_rate": 9.630551847827672e-06, + "loss": 0.4544, + "step": 2185 + }, + { + "epoch": 0.14953143169847458, + "grad_norm": 1.6600109575309236, + "learning_rate": 9.630133830715396e-06, + "loss": 0.3257, + "step": 2186 + }, + { + "epoch": 0.14959983583008413, + "grad_norm": 1.534251863953843, + "learning_rate": 9.629715586333794e-06, + "loss": 0.1324, + "step": 2187 + }, + { + "epoch": 0.14966823996169368, + "grad_norm": 1.7625448928857979, + "learning_rate": 9.629297114703395e-06, + "loss": 0.4194, + "step": 2188 + }, + { + "epoch": 0.14973664409330323, + "grad_norm": 1.7442019516948344, + "learning_rate": 9.62887841584474e-06, + "loss": 0.4148, + "step": 2189 + }, + { + "epoch": 0.14980504822491278, + "grad_norm": 1.794573603766733, + "learning_rate": 9.628459489778378e-06, + "loss": 0.1837, + "step": 2190 + }, + { + "epoch": 0.14987345235652233, + "grad_norm": 2.2385970553620482, + "learning_rate": 9.628040336524875e-06, + "loss": 0.3078, + "step": 2191 + }, + { + "epoch": 0.14994185648813188, + "grad_norm": 2.105578255150083, + "learning_rate": 9.627620956104801e-06, + "loss": 0.3739, + "step": 2192 + }, + { + "epoch": 0.15001026061974143, + "grad_norm": 1.968251057252657, + "learning_rate": 9.627201348538746e-06, + "loss": 0.3903, + "step": 2193 + }, + { + "epoch": 0.15007866475135098, + "grad_norm": 2.335850041005697, + "learning_rate": 9.626781513847305e-06, + "loss": 0.5805, + "step": 2194 + }, + { + "epoch": 0.15014706888296053, + "grad_norm": 1.972957323723561, + "learning_rate": 9.626361452051084e-06, + "loss": 0.55, + "step": 2195 + }, + { + "epoch": 0.15021547301457008, + "grad_norm": 3.1207445543714267, + "learning_rate": 9.625941163170702e-06, + "loss": 0.2969, + "step": 2196 + }, + { + "epoch": 0.15028387714617963, + "grad_norm": 2.647672701717264, + "learning_rate": 9.625520647226788e-06, + "loss": 0.4223, + "step": 2197 + }, + { + "epoch": 0.15035228127778918, + "grad_norm": 2.8542149265877867, + "learning_rate": 9.625099904239987e-06, + "loss": 0.4703, + "step": 2198 + }, + { + "epoch": 0.15042068540939874, + "grad_norm": 2.502228703088731, + "learning_rate": 9.624678934230948e-06, + "loss": 0.5446, + "step": 2199 + }, + { + "epoch": 0.1504890895410083, + "grad_norm": 1.7222758627898944, + "learning_rate": 9.624257737220333e-06, + "loss": 0.4304, + "step": 2200 + }, + { + "epoch": 0.15055749367261784, + "grad_norm": 1.7536489514613114, + "learning_rate": 9.62383631322882e-06, + "loss": 0.3473, + "step": 2201 + }, + { + "epoch": 0.1506258978042274, + "grad_norm": 2.526928509379891, + "learning_rate": 9.62341466227709e-06, + "loss": 0.6633, + "step": 2202 + }, + { + "epoch": 0.1506943019358369, + "grad_norm": 2.528530968133775, + "learning_rate": 9.622992784385842e-06, + "loss": 0.3827, + "step": 2203 + }, + { + "epoch": 0.15076270606744646, + "grad_norm": 2.4309838936318067, + "learning_rate": 9.622570679575787e-06, + "loss": 0.6125, + "step": 2204 + }, + { + "epoch": 0.150831110199056, + "grad_norm": 2.239587602315575, + "learning_rate": 9.62214834786764e-06, + "loss": 0.4868, + "step": 2205 + }, + { + "epoch": 0.15089951433066556, + "grad_norm": 1.8578631932743188, + "learning_rate": 9.62172578928213e-06, + "loss": 0.1995, + "step": 2206 + }, + { + "epoch": 0.1509679184622751, + "grad_norm": 1.914463114172423, + "learning_rate": 9.62130300384e-06, + "loss": 0.4998, + "step": 2207 + }, + { + "epoch": 0.15103632259388466, + "grad_norm": 2.669081203268765, + "learning_rate": 9.620879991562005e-06, + "loss": 0.3088, + "step": 2208 + }, + { + "epoch": 0.15110472672549422, + "grad_norm": 2.208402902853859, + "learning_rate": 9.620456752468903e-06, + "loss": 0.4568, + "step": 2209 + }, + { + "epoch": 0.15117313085710377, + "grad_norm": 2.4060644102980047, + "learning_rate": 9.620033286581475e-06, + "loss": 0.4741, + "step": 2210 + }, + { + "epoch": 0.15124153498871332, + "grad_norm": 1.7922254821091648, + "learning_rate": 9.619609593920502e-06, + "loss": 0.4006, + "step": 2211 + }, + { + "epoch": 0.15130993912032287, + "grad_norm": 2.642167571910239, + "learning_rate": 9.619185674506781e-06, + "loss": 0.7166, + "step": 2212 + }, + { + "epoch": 0.15137834325193242, + "grad_norm": 1.7134450452361185, + "learning_rate": 9.618761528361123e-06, + "loss": 0.3288, + "step": 2213 + }, + { + "epoch": 0.15144674738354197, + "grad_norm": 1.8769682368564289, + "learning_rate": 9.618337155504345e-06, + "loss": 0.3854, + "step": 2214 + }, + { + "epoch": 0.15151515151515152, + "grad_norm": 1.470104709720277, + "learning_rate": 9.617912555957277e-06, + "loss": 0.1681, + "step": 2215 + }, + { + "epoch": 0.15158355564676107, + "grad_norm": 1.9507823627852205, + "learning_rate": 9.617487729740761e-06, + "loss": 0.3976, + "step": 2216 + }, + { + "epoch": 0.15165195977837062, + "grad_norm": 2.4223899232383146, + "learning_rate": 9.61706267687565e-06, + "loss": 0.3877, + "step": 2217 + }, + { + "epoch": 0.15172036390998017, + "grad_norm": 2.2735582070570577, + "learning_rate": 9.616637397382807e-06, + "loss": 0.3468, + "step": 2218 + }, + { + "epoch": 0.15178876804158972, + "grad_norm": 0.9744117513519485, + "learning_rate": 9.616211891283108e-06, + "loss": 0.1729, + "step": 2219 + }, + { + "epoch": 0.15185717217319927, + "grad_norm": 1.992801407114219, + "learning_rate": 9.615786158597438e-06, + "loss": 0.3074, + "step": 2220 + }, + { + "epoch": 0.15192557630480882, + "grad_norm": 2.180812927986842, + "learning_rate": 9.615360199346694e-06, + "loss": 0.3792, + "step": 2221 + }, + { + "epoch": 0.15199398043641835, + "grad_norm": 1.8358836957474027, + "learning_rate": 9.614934013551781e-06, + "loss": 0.4021, + "step": 2222 + }, + { + "epoch": 0.1520623845680279, + "grad_norm": 1.673780192998661, + "learning_rate": 9.614507601233625e-06, + "loss": 0.2235, + "step": 2223 + }, + { + "epoch": 0.15213078869963745, + "grad_norm": 2.763411433480792, + "learning_rate": 9.61408096241315e-06, + "loss": 0.7753, + "step": 2224 + }, + { + "epoch": 0.152199192831247, + "grad_norm": 2.000374352914679, + "learning_rate": 9.613654097111302e-06, + "loss": 0.5419, + "step": 2225 + }, + { + "epoch": 0.15226759696285655, + "grad_norm": 2.4074130158243574, + "learning_rate": 9.61322700534903e-06, + "loss": 0.4524, + "step": 2226 + }, + { + "epoch": 0.1523360010944661, + "grad_norm": 2.1501464635840986, + "learning_rate": 9.612799687147302e-06, + "loss": 0.504, + "step": 2227 + }, + { + "epoch": 0.15240440522607565, + "grad_norm": 1.8560079212834109, + "learning_rate": 9.61237214252709e-06, + "loss": 0.3014, + "step": 2228 + }, + { + "epoch": 0.1524728093576852, + "grad_norm": 2.0250019208139562, + "learning_rate": 9.611944371509378e-06, + "loss": 0.3005, + "step": 2229 + }, + { + "epoch": 0.15254121348929475, + "grad_norm": 2.3393706223887425, + "learning_rate": 9.611516374115169e-06, + "loss": 0.6339, + "step": 2230 + }, + { + "epoch": 0.1526096176209043, + "grad_norm": 2.3403865649473383, + "learning_rate": 9.611088150365464e-06, + "loss": 0.6797, + "step": 2231 + }, + { + "epoch": 0.15267802175251385, + "grad_norm": 2.157881169880393, + "learning_rate": 9.610659700281288e-06, + "loss": 0.274, + "step": 2232 + }, + { + "epoch": 0.1527464258841234, + "grad_norm": 2.476575442259272, + "learning_rate": 9.610231023883667e-06, + "loss": 0.5521, + "step": 2233 + }, + { + "epoch": 0.15281483001573296, + "grad_norm": 2.194535358842023, + "learning_rate": 9.609802121193648e-06, + "loss": 0.442, + "step": 2234 + }, + { + "epoch": 0.1528832341473425, + "grad_norm": 2.3548807640728024, + "learning_rate": 9.609372992232279e-06, + "loss": 0.553, + "step": 2235 + }, + { + "epoch": 0.15295163827895206, + "grad_norm": 2.503073650600163, + "learning_rate": 9.608943637020625e-06, + "loss": 0.6567, + "step": 2236 + }, + { + "epoch": 0.1530200424105616, + "grad_norm": 1.920232815559692, + "learning_rate": 9.60851405557976e-06, + "loss": 0.4005, + "step": 2237 + }, + { + "epoch": 0.15308844654217116, + "grad_norm": 1.7672893773064502, + "learning_rate": 9.608084247930772e-06, + "loss": 0.3429, + "step": 2238 + }, + { + "epoch": 0.1531568506737807, + "grad_norm": 2.733506582505147, + "learning_rate": 9.607654214094757e-06, + "loss": 0.6171, + "step": 2239 + }, + { + "epoch": 0.15322525480539023, + "grad_norm": 2.0436873548863397, + "learning_rate": 9.60722395409282e-06, + "loss": 0.3362, + "step": 2240 + }, + { + "epoch": 0.15329365893699978, + "grad_norm": 1.723510435260913, + "learning_rate": 9.606793467946088e-06, + "loss": 0.297, + "step": 2241 + }, + { + "epoch": 0.15336206306860933, + "grad_norm": 1.98640593279768, + "learning_rate": 9.606362755675685e-06, + "loss": 0.4841, + "step": 2242 + }, + { + "epoch": 0.15343046720021888, + "grad_norm": 2.34697764041422, + "learning_rate": 9.605931817302752e-06, + "loss": 0.524, + "step": 2243 + }, + { + "epoch": 0.15349887133182843, + "grad_norm": 2.223025996824526, + "learning_rate": 9.605500652848447e-06, + "loss": 0.5387, + "step": 2244 + }, + { + "epoch": 0.15356727546343799, + "grad_norm": 3.1624064314800227, + "learning_rate": 9.605069262333928e-06, + "loss": 0.6898, + "step": 2245 + }, + { + "epoch": 0.15363567959504754, + "grad_norm": 2.2424072131392707, + "learning_rate": 9.604637645780373e-06, + "loss": 0.5918, + "step": 2246 + }, + { + "epoch": 0.1537040837266571, + "grad_norm": 1.9160883789559597, + "learning_rate": 9.604205803208966e-06, + "loss": 0.3825, + "step": 2247 + }, + { + "epoch": 0.15377248785826664, + "grad_norm": 2.0649004535488333, + "learning_rate": 9.603773734640905e-06, + "loss": 0.3781, + "step": 2248 + }, + { + "epoch": 0.1538408919898762, + "grad_norm": 1.9225805759069405, + "learning_rate": 9.603341440097397e-06, + "loss": 0.3683, + "step": 2249 + }, + { + "epoch": 0.15390929612148574, + "grad_norm": 2.0897603633323367, + "learning_rate": 9.602908919599662e-06, + "loss": 0.3659, + "step": 2250 + }, + { + "epoch": 0.1539777002530953, + "grad_norm": 2.099975355321597, + "learning_rate": 9.60247617316893e-06, + "loss": 0.3321, + "step": 2251 + }, + { + "epoch": 0.15404610438470484, + "grad_norm": 1.644992392212224, + "learning_rate": 9.602043200826443e-06, + "loss": 0.2275, + "step": 2252 + }, + { + "epoch": 0.1541145085163144, + "grad_norm": 2.805530198352986, + "learning_rate": 9.601610002593451e-06, + "loss": 0.9473, + "step": 2253 + }, + { + "epoch": 0.15418291264792394, + "grad_norm": 1.6428647670046936, + "learning_rate": 9.60117657849122e-06, + "loss": 0.2986, + "step": 2254 + }, + { + "epoch": 0.1542513167795335, + "grad_norm": 2.2833765778193893, + "learning_rate": 9.600742928541025e-06, + "loss": 0.4241, + "step": 2255 + }, + { + "epoch": 0.15431972091114304, + "grad_norm": 2.4765881076776437, + "learning_rate": 9.60030905276415e-06, + "loss": 0.3666, + "step": 2256 + }, + { + "epoch": 0.1543881250427526, + "grad_norm": 2.236911707531892, + "learning_rate": 9.59987495118189e-06, + "loss": 0.6149, + "step": 2257 + }, + { + "epoch": 0.15445652917436214, + "grad_norm": 2.039517479776048, + "learning_rate": 9.599440623815557e-06, + "loss": 0.3609, + "step": 2258 + }, + { + "epoch": 0.15452493330597167, + "grad_norm": 2.148288374840518, + "learning_rate": 9.599006070686467e-06, + "loss": 0.6327, + "step": 2259 + }, + { + "epoch": 0.15459333743758122, + "grad_norm": 6.575142517285814, + "learning_rate": 9.59857129181595e-06, + "loss": 0.559, + "step": 2260 + }, + { + "epoch": 0.15466174156919077, + "grad_norm": 1.7689855504648018, + "learning_rate": 9.59813628722535e-06, + "loss": 0.4193, + "step": 2261 + }, + { + "epoch": 0.15473014570080032, + "grad_norm": 2.4850982716814736, + "learning_rate": 9.597701056936014e-06, + "loss": 0.683, + "step": 2262 + }, + { + "epoch": 0.15479854983240987, + "grad_norm": 2.0984767893097684, + "learning_rate": 9.59726560096931e-06, + "loss": 0.5553, + "step": 2263 + }, + { + "epoch": 0.15486695396401942, + "grad_norm": 1.6595716703826209, + "learning_rate": 9.596829919346608e-06, + "loss": 0.1753, + "step": 2264 + }, + { + "epoch": 0.15493535809562897, + "grad_norm": 3.0128724180621274, + "learning_rate": 9.596394012089298e-06, + "loss": 0.5919, + "step": 2265 + }, + { + "epoch": 0.15500376222723852, + "grad_norm": 2.2049404300646547, + "learning_rate": 9.595957879218773e-06, + "loss": 0.4598, + "step": 2266 + }, + { + "epoch": 0.15507216635884807, + "grad_norm": 1.826539068993618, + "learning_rate": 9.595521520756443e-06, + "loss": 0.3516, + "step": 2267 + }, + { + "epoch": 0.15514057049045762, + "grad_norm": 1.1965722196577244, + "learning_rate": 9.595084936723725e-06, + "loss": 0.1631, + "step": 2268 + }, + { + "epoch": 0.15520897462206718, + "grad_norm": 3.2580532515598595, + "learning_rate": 9.594648127142048e-06, + "loss": 0.5278, + "step": 2269 + }, + { + "epoch": 0.15527737875367673, + "grad_norm": 2.1551661877377786, + "learning_rate": 9.594211092032853e-06, + "loss": 0.3569, + "step": 2270 + }, + { + "epoch": 0.15534578288528628, + "grad_norm": 1.8529457928751847, + "learning_rate": 9.593773831417592e-06, + "loss": 0.3754, + "step": 2271 + }, + { + "epoch": 0.15541418701689583, + "grad_norm": 1.9414326399270236, + "learning_rate": 9.59333634531773e-06, + "loss": 0.3546, + "step": 2272 + }, + { + "epoch": 0.15548259114850538, + "grad_norm": 1.9020401675210254, + "learning_rate": 9.592898633754739e-06, + "loss": 0.4208, + "step": 2273 + }, + { + "epoch": 0.15555099528011493, + "grad_norm": 2.4554752887440934, + "learning_rate": 9.592460696750103e-06, + "loss": 0.3832, + "step": 2274 + }, + { + "epoch": 0.15561939941172448, + "grad_norm": 1.5799653846496486, + "learning_rate": 9.59202253432532e-06, + "loss": 0.2551, + "step": 2275 + }, + { + "epoch": 0.15568780354333403, + "grad_norm": 2.0430481751212337, + "learning_rate": 9.591584146501896e-06, + "loss": 0.4025, + "step": 2276 + }, + { + "epoch": 0.15575620767494355, + "grad_norm": 1.6833451877505496, + "learning_rate": 9.591145533301351e-06, + "loss": 0.285, + "step": 2277 + }, + { + "epoch": 0.1558246118065531, + "grad_norm": 1.95808179910577, + "learning_rate": 9.590706694745211e-06, + "loss": 0.3805, + "step": 2278 + }, + { + "epoch": 0.15589301593816265, + "grad_norm": 2.0402364238193367, + "learning_rate": 9.59026763085502e-06, + "loss": 0.4091, + "step": 2279 + }, + { + "epoch": 0.1559614200697722, + "grad_norm": 1.4811798240037455, + "learning_rate": 9.589828341652326e-06, + "loss": 0.2511, + "step": 2280 + }, + { + "epoch": 0.15602982420138176, + "grad_norm": 1.7497053835009917, + "learning_rate": 9.589388827158692e-06, + "loss": 0.1967, + "step": 2281 + }, + { + "epoch": 0.1560982283329913, + "grad_norm": 1.4550500393403547, + "learning_rate": 9.588949087395694e-06, + "loss": 0.1784, + "step": 2282 + }, + { + "epoch": 0.15616663246460086, + "grad_norm": 1.6129979815500026, + "learning_rate": 9.588509122384915e-06, + "loss": 0.2009, + "step": 2283 + }, + { + "epoch": 0.1562350365962104, + "grad_norm": 1.664294378932995, + "learning_rate": 9.588068932147949e-06, + "loss": 0.2175, + "step": 2284 + }, + { + "epoch": 0.15630344072781996, + "grad_norm": 2.01247006062338, + "learning_rate": 9.587628516706406e-06, + "loss": 0.4795, + "step": 2285 + }, + { + "epoch": 0.1563718448594295, + "grad_norm": 2.619125662013473, + "learning_rate": 9.587187876081901e-06, + "loss": 0.6372, + "step": 2286 + }, + { + "epoch": 0.15644024899103906, + "grad_norm": 2.498708128928976, + "learning_rate": 9.586747010296064e-06, + "loss": 0.6686, + "step": 2287 + }, + { + "epoch": 0.1565086531226486, + "grad_norm": 2.1056579688829435, + "learning_rate": 9.586305919370535e-06, + "loss": 0.4886, + "step": 2288 + }, + { + "epoch": 0.15657705725425816, + "grad_norm": 3.107924722186384, + "learning_rate": 9.585864603326964e-06, + "loss": 0.4943, + "step": 2289 + }, + { + "epoch": 0.1566454613858677, + "grad_norm": 2.269993260642479, + "learning_rate": 9.585423062187014e-06, + "loss": 0.6212, + "step": 2290 + }, + { + "epoch": 0.15671386551747726, + "grad_norm": 2.2633692791759517, + "learning_rate": 9.584981295972357e-06, + "loss": 0.5507, + "step": 2291 + }, + { + "epoch": 0.15678226964908681, + "grad_norm": 1.5754709351469696, + "learning_rate": 9.584539304704676e-06, + "loss": 0.2198, + "step": 2292 + }, + { + "epoch": 0.15685067378069636, + "grad_norm": 1.923008971893619, + "learning_rate": 9.58409708840567e-06, + "loss": 0.3849, + "step": 2293 + }, + { + "epoch": 0.15691907791230592, + "grad_norm": 2.6941210283250054, + "learning_rate": 9.583654647097041e-06, + "loss": 0.6694, + "step": 2294 + }, + { + "epoch": 0.15698748204391547, + "grad_norm": 2.4652095468836697, + "learning_rate": 9.583211980800507e-06, + "loss": 0.3554, + "step": 2295 + }, + { + "epoch": 0.157055886175525, + "grad_norm": 1.997481242761403, + "learning_rate": 9.582769089537797e-06, + "loss": 0.4544, + "step": 2296 + }, + { + "epoch": 0.15712429030713454, + "grad_norm": 2.398661295486635, + "learning_rate": 9.582325973330652e-06, + "loss": 0.7057, + "step": 2297 + }, + { + "epoch": 0.1571926944387441, + "grad_norm": 1.7524283114475754, + "learning_rate": 9.581882632200819e-06, + "loss": 0.346, + "step": 2298 + }, + { + "epoch": 0.15726109857035364, + "grad_norm": 1.7153352020173556, + "learning_rate": 9.58143906617006e-06, + "loss": 0.2804, + "step": 2299 + }, + { + "epoch": 0.1573295027019632, + "grad_norm": 1.8024636806209808, + "learning_rate": 9.58099527526015e-06, + "loss": 0.344, + "step": 2300 + }, + { + "epoch": 0.15739790683357274, + "grad_norm": 2.6838990194722188, + "learning_rate": 9.58055125949287e-06, + "loss": 0.7231, + "step": 2301 + }, + { + "epoch": 0.1574663109651823, + "grad_norm": 2.163443638694922, + "learning_rate": 9.580107018890013e-06, + "loss": 0.4954, + "step": 2302 + }, + { + "epoch": 0.15753471509679184, + "grad_norm": 2.338582359602779, + "learning_rate": 9.579662553473388e-06, + "loss": 0.4768, + "step": 2303 + }, + { + "epoch": 0.1576031192284014, + "grad_norm": 2.4023451208095206, + "learning_rate": 9.57921786326481e-06, + "loss": 0.5977, + "step": 2304 + }, + { + "epoch": 0.15767152336001095, + "grad_norm": 2.20992033580209, + "learning_rate": 9.578772948286106e-06, + "loss": 0.4679, + "step": 2305 + }, + { + "epoch": 0.1577399274916205, + "grad_norm": 2.2821547791173664, + "learning_rate": 9.578327808559117e-06, + "loss": 0.2702, + "step": 2306 + }, + { + "epoch": 0.15780833162323005, + "grad_norm": 1.8927641266399058, + "learning_rate": 9.577882444105687e-06, + "loss": 0.5343, + "step": 2307 + }, + { + "epoch": 0.1578767357548396, + "grad_norm": 1.976646368788007, + "learning_rate": 9.577436854947683e-06, + "loss": 0.2896, + "step": 2308 + }, + { + "epoch": 0.15794513988644915, + "grad_norm": 2.449545706139162, + "learning_rate": 9.576991041106973e-06, + "loss": 0.3969, + "step": 2309 + }, + { + "epoch": 0.1580135440180587, + "grad_norm": 2.139682896596448, + "learning_rate": 9.57654500260544e-06, + "loss": 0.4411, + "step": 2310 + }, + { + "epoch": 0.15808194814966825, + "grad_norm": 1.9357985093328576, + "learning_rate": 9.57609873946498e-06, + "loss": 0.4675, + "step": 2311 + }, + { + "epoch": 0.1581503522812778, + "grad_norm": 2.0232793422144684, + "learning_rate": 9.575652251707495e-06, + "loss": 0.1984, + "step": 2312 + }, + { + "epoch": 0.15821875641288735, + "grad_norm": 2.2321791028758415, + "learning_rate": 9.575205539354902e-06, + "loss": 0.5097, + "step": 2313 + }, + { + "epoch": 0.15828716054449687, + "grad_norm": 1.7814734720288674, + "learning_rate": 9.574758602429128e-06, + "loss": 0.2178, + "step": 2314 + }, + { + "epoch": 0.15835556467610643, + "grad_norm": 2.0162777250432957, + "learning_rate": 9.574311440952109e-06, + "loss": 0.4358, + "step": 2315 + }, + { + "epoch": 0.15842396880771598, + "grad_norm": 1.670429179449294, + "learning_rate": 9.573864054945797e-06, + "loss": 0.1493, + "step": 2316 + }, + { + "epoch": 0.15849237293932553, + "grad_norm": 2.1439217602900142, + "learning_rate": 9.57341644443215e-06, + "loss": 0.2195, + "step": 2317 + }, + { + "epoch": 0.15856077707093508, + "grad_norm": 1.6469931172859407, + "learning_rate": 9.57296860943314e-06, + "loss": 0.312, + "step": 2318 + }, + { + "epoch": 0.15862918120254463, + "grad_norm": 2.0874384505512085, + "learning_rate": 9.572520549970746e-06, + "loss": 0.3703, + "step": 2319 + }, + { + "epoch": 0.15869758533415418, + "grad_norm": 1.4033317246898838, + "learning_rate": 9.572072266066963e-06, + "loss": 0.207, + "step": 2320 + }, + { + "epoch": 0.15876598946576373, + "grad_norm": 2.542324054757312, + "learning_rate": 9.571623757743797e-06, + "loss": 0.7559, + "step": 2321 + }, + { + "epoch": 0.15883439359737328, + "grad_norm": 2.3778139246779553, + "learning_rate": 9.57117502502326e-06, + "loss": 0.6293, + "step": 2322 + }, + { + "epoch": 0.15890279772898283, + "grad_norm": 1.641395147395906, + "learning_rate": 9.570726067927378e-06, + "loss": 0.3191, + "step": 2323 + }, + { + "epoch": 0.15897120186059238, + "grad_norm": 1.9953690312501864, + "learning_rate": 9.570276886478189e-06, + "loss": 0.3614, + "step": 2324 + }, + { + "epoch": 0.15903960599220193, + "grad_norm": 2.6641996689551037, + "learning_rate": 9.569827480697742e-06, + "loss": 0.5688, + "step": 2325 + }, + { + "epoch": 0.15910801012381148, + "grad_norm": 3.1620125711572373, + "learning_rate": 9.569377850608093e-06, + "loss": 0.4368, + "step": 2326 + }, + { + "epoch": 0.15917641425542103, + "grad_norm": 1.8880136125799196, + "learning_rate": 9.568927996231314e-06, + "loss": 0.3972, + "step": 2327 + }, + { + "epoch": 0.15924481838703058, + "grad_norm": 2.063708671427346, + "learning_rate": 9.568477917589486e-06, + "loss": 0.2248, + "step": 2328 + }, + { + "epoch": 0.15931322251864014, + "grad_norm": 2.166431393977833, + "learning_rate": 9.568027614704701e-06, + "loss": 0.5516, + "step": 2329 + }, + { + "epoch": 0.15938162665024969, + "grad_norm": 1.6733666031865373, + "learning_rate": 9.567577087599063e-06, + "loss": 0.3074, + "step": 2330 + }, + { + "epoch": 0.15945003078185924, + "grad_norm": 2.9594009257291747, + "learning_rate": 9.567126336294684e-06, + "loss": 0.8672, + "step": 2331 + }, + { + "epoch": 0.1595184349134688, + "grad_norm": 1.9435093040714675, + "learning_rate": 9.566675360813688e-06, + "loss": 0.3714, + "step": 2332 + }, + { + "epoch": 0.1595868390450783, + "grad_norm": 2.006427134807766, + "learning_rate": 9.566224161178217e-06, + "loss": 0.3686, + "step": 2333 + }, + { + "epoch": 0.15965524317668786, + "grad_norm": 1.7735650504032372, + "learning_rate": 9.565772737410412e-06, + "loss": 0.2437, + "step": 2334 + }, + { + "epoch": 0.1597236473082974, + "grad_norm": 2.5155066942153566, + "learning_rate": 9.565321089532432e-06, + "loss": 0.6334, + "step": 2335 + }, + { + "epoch": 0.15979205143990696, + "grad_norm": 2.2186479410688387, + "learning_rate": 9.564869217566447e-06, + "loss": 0.6981, + "step": 2336 + }, + { + "epoch": 0.1598604555715165, + "grad_norm": 2.407548931581213, + "learning_rate": 9.56441712153464e-06, + "loss": 0.3317, + "step": 2337 + }, + { + "epoch": 0.15992885970312606, + "grad_norm": 2.1141998142101794, + "learning_rate": 9.563964801459199e-06, + "loss": 0.5906, + "step": 2338 + }, + { + "epoch": 0.15999726383473561, + "grad_norm": 2.3407656138446415, + "learning_rate": 9.563512257362325e-06, + "loss": 0.5347, + "step": 2339 + }, + { + "epoch": 0.16006566796634517, + "grad_norm": 1.4980142632109072, + "learning_rate": 9.563059489266233e-06, + "loss": 0.2865, + "step": 2340 + }, + { + "epoch": 0.16013407209795472, + "grad_norm": 1.63627413170915, + "learning_rate": 9.562606497193146e-06, + "loss": 0.3476, + "step": 2341 + }, + { + "epoch": 0.16020247622956427, + "grad_norm": 2.4663298369379993, + "learning_rate": 9.5621532811653e-06, + "loss": 0.8419, + "step": 2342 + }, + { + "epoch": 0.16027088036117382, + "grad_norm": 1.8727071253072827, + "learning_rate": 9.561699841204941e-06, + "loss": 0.4353, + "step": 2343 + }, + { + "epoch": 0.16033928449278337, + "grad_norm": 2.48384937538015, + "learning_rate": 9.561246177334325e-06, + "loss": 0.2945, + "step": 2344 + }, + { + "epoch": 0.16040768862439292, + "grad_norm": 1.5311848154596057, + "learning_rate": 9.56079228957572e-06, + "loss": 0.1949, + "step": 2345 + }, + { + "epoch": 0.16047609275600247, + "grad_norm": 1.2862759918134516, + "learning_rate": 9.560338177951407e-06, + "loss": 0.1636, + "step": 2346 + }, + { + "epoch": 0.16054449688761202, + "grad_norm": 1.9366225632283065, + "learning_rate": 9.559883842483672e-06, + "loss": 0.3335, + "step": 2347 + }, + { + "epoch": 0.16061290101922157, + "grad_norm": 1.8483025950638312, + "learning_rate": 9.559429283194822e-06, + "loss": 0.4417, + "step": 2348 + }, + { + "epoch": 0.16068130515083112, + "grad_norm": 1.4344117587912415, + "learning_rate": 9.558974500107163e-06, + "loss": 0.3258, + "step": 2349 + }, + { + "epoch": 0.16074970928244067, + "grad_norm": 2.8210296960582393, + "learning_rate": 9.558519493243022e-06, + "loss": 0.4201, + "step": 2350 + }, + { + "epoch": 0.1608181134140502, + "grad_norm": 2.526701850091519, + "learning_rate": 9.558064262624731e-06, + "loss": 0.5455, + "step": 2351 + }, + { + "epoch": 0.16088651754565975, + "grad_norm": 2.1121261457733733, + "learning_rate": 9.557608808274634e-06, + "loss": 0.63, + "step": 2352 + }, + { + "epoch": 0.1609549216772693, + "grad_norm": 2.330376240668212, + "learning_rate": 9.557153130215089e-06, + "loss": 0.4741, + "step": 2353 + }, + { + "epoch": 0.16102332580887885, + "grad_norm": 1.797075699917438, + "learning_rate": 9.556697228468463e-06, + "loss": 0.4226, + "step": 2354 + }, + { + "epoch": 0.1610917299404884, + "grad_norm": 1.8503342188040615, + "learning_rate": 9.556241103057132e-06, + "loss": 0.5055, + "step": 2355 + }, + { + "epoch": 0.16116013407209795, + "grad_norm": 2.0110586398138297, + "learning_rate": 9.555784754003487e-06, + "loss": 0.5633, + "step": 2356 + }, + { + "epoch": 0.1612285382037075, + "grad_norm": 1.966291913969929, + "learning_rate": 9.555328181329925e-06, + "loss": 0.3766, + "step": 2357 + }, + { + "epoch": 0.16129694233531705, + "grad_norm": 2.2537585429909295, + "learning_rate": 9.55487138505886e-06, + "loss": 0.3235, + "step": 2358 + }, + { + "epoch": 0.1613653464669266, + "grad_norm": 1.617330199317452, + "learning_rate": 9.55441436521271e-06, + "loss": 0.2888, + "step": 2359 + }, + { + "epoch": 0.16143375059853615, + "grad_norm": 2.448182015883553, + "learning_rate": 9.553957121813912e-06, + "loss": 0.5513, + "step": 2360 + }, + { + "epoch": 0.1615021547301457, + "grad_norm": 2.0610389953447386, + "learning_rate": 9.553499654884905e-06, + "loss": 0.1725, + "step": 2361 + }, + { + "epoch": 0.16157055886175525, + "grad_norm": 1.9933154484601416, + "learning_rate": 9.553041964448148e-06, + "loss": 0.4787, + "step": 2362 + }, + { + "epoch": 0.1616389629933648, + "grad_norm": 2.545492516668301, + "learning_rate": 9.552584050526105e-06, + "loss": 0.2524, + "step": 2363 + }, + { + "epoch": 0.16170736712497435, + "grad_norm": 2.2939929137116315, + "learning_rate": 9.552125913141254e-06, + "loss": 0.4494, + "step": 2364 + }, + { + "epoch": 0.1617757712565839, + "grad_norm": 1.647577074770403, + "learning_rate": 9.55166755231608e-06, + "loss": 0.3546, + "step": 2365 + }, + { + "epoch": 0.16184417538819346, + "grad_norm": 1.9245403501173601, + "learning_rate": 9.551208968073082e-06, + "loss": 0.3056, + "step": 2366 + }, + { + "epoch": 0.161912579519803, + "grad_norm": 2.281197384005419, + "learning_rate": 9.550750160434772e-06, + "loss": 0.3817, + "step": 2367 + }, + { + "epoch": 0.16198098365141256, + "grad_norm": 1.7931079041745377, + "learning_rate": 9.550291129423667e-06, + "loss": 0.2627, + "step": 2368 + }, + { + "epoch": 0.16204938778302208, + "grad_norm": 2.0192612000764387, + "learning_rate": 9.5498318750623e-06, + "loss": 0.3906, + "step": 2369 + }, + { + "epoch": 0.16211779191463163, + "grad_norm": 2.209391285783671, + "learning_rate": 9.549372397373216e-06, + "loss": 0.4083, + "step": 2370 + }, + { + "epoch": 0.16218619604624118, + "grad_norm": 1.8150662637599762, + "learning_rate": 9.548912696378965e-06, + "loss": 0.3315, + "step": 2371 + }, + { + "epoch": 0.16225460017785073, + "grad_norm": 1.7436489253676937, + "learning_rate": 9.548452772102111e-06, + "loss": 0.3573, + "step": 2372 + }, + { + "epoch": 0.16232300430946028, + "grad_norm": 2.3492645365772824, + "learning_rate": 9.547992624565233e-06, + "loss": 0.5478, + "step": 2373 + }, + { + "epoch": 0.16239140844106983, + "grad_norm": 1.987249610756272, + "learning_rate": 9.547532253790912e-06, + "loss": 0.4034, + "step": 2374 + }, + { + "epoch": 0.16245981257267939, + "grad_norm": 1.7015145500860207, + "learning_rate": 9.547071659801752e-06, + "loss": 0.2706, + "step": 2375 + }, + { + "epoch": 0.16252821670428894, + "grad_norm": 1.5681491671743357, + "learning_rate": 9.546610842620355e-06, + "loss": 0.2161, + "step": 2376 + }, + { + "epoch": 0.1625966208358985, + "grad_norm": 1.8765011123062616, + "learning_rate": 9.546149802269342e-06, + "loss": 0.3336, + "step": 2377 + }, + { + "epoch": 0.16266502496750804, + "grad_norm": 2.424595701296787, + "learning_rate": 9.545688538771343e-06, + "loss": 0.3325, + "step": 2378 + }, + { + "epoch": 0.1627334290991176, + "grad_norm": 2.0868532049635267, + "learning_rate": 9.545227052149002e-06, + "loss": 0.4796, + "step": 2379 + }, + { + "epoch": 0.16280183323072714, + "grad_norm": 2.4014476345342732, + "learning_rate": 9.544765342424968e-06, + "loss": 0.4143, + "step": 2380 + }, + { + "epoch": 0.1628702373623367, + "grad_norm": 1.86004652190736, + "learning_rate": 9.544303409621904e-06, + "loss": 0.3252, + "step": 2381 + }, + { + "epoch": 0.16293864149394624, + "grad_norm": 2.3645546501089263, + "learning_rate": 9.543841253762484e-06, + "loss": 0.5674, + "step": 2382 + }, + { + "epoch": 0.1630070456255558, + "grad_norm": 1.9211944367534108, + "learning_rate": 9.543378874869393e-06, + "loss": 0.3993, + "step": 2383 + }, + { + "epoch": 0.16307544975716534, + "grad_norm": 2.0625647007301717, + "learning_rate": 9.542916272965327e-06, + "loss": 0.4338, + "step": 2384 + }, + { + "epoch": 0.1631438538887749, + "grad_norm": 1.6123548946962152, + "learning_rate": 9.542453448072994e-06, + "loss": 0.1563, + "step": 2385 + }, + { + "epoch": 0.16321225802038444, + "grad_norm": 2.7437378774475074, + "learning_rate": 9.541990400215111e-06, + "loss": 0.236, + "step": 2386 + }, + { + "epoch": 0.163280662151994, + "grad_norm": 1.9189547705963241, + "learning_rate": 9.541527129414405e-06, + "loss": 0.2123, + "step": 2387 + }, + { + "epoch": 0.16334906628360352, + "grad_norm": 2.7449132517186126, + "learning_rate": 9.541063635693617e-06, + "loss": 0.2748, + "step": 2388 + }, + { + "epoch": 0.16341747041521307, + "grad_norm": 2.5483642682207712, + "learning_rate": 9.540599919075497e-06, + "loss": 0.5756, + "step": 2389 + }, + { + "epoch": 0.16348587454682262, + "grad_norm": 2.2391428595615244, + "learning_rate": 9.540135979582807e-06, + "loss": 0.5662, + "step": 2390 + }, + { + "epoch": 0.16355427867843217, + "grad_norm": 2.028466841538812, + "learning_rate": 9.539671817238318e-06, + "loss": 0.343, + "step": 2391 + }, + { + "epoch": 0.16362268281004172, + "grad_norm": 1.9751056911460676, + "learning_rate": 9.539207432064816e-06, + "loss": 0.2801, + "step": 2392 + }, + { + "epoch": 0.16369108694165127, + "grad_norm": 2.078803496632967, + "learning_rate": 9.538742824085092e-06, + "loss": 0.3793, + "step": 2393 + }, + { + "epoch": 0.16375949107326082, + "grad_norm": 2.110430827631485, + "learning_rate": 9.538277993321955e-06, + "loss": 0.5305, + "step": 2394 + }, + { + "epoch": 0.16382789520487037, + "grad_norm": 2.0520242974620944, + "learning_rate": 9.537812939798217e-06, + "loss": 0.3936, + "step": 2395 + }, + { + "epoch": 0.16389629933647992, + "grad_norm": 2.892221329633891, + "learning_rate": 9.537347663536706e-06, + "loss": 0.2144, + "step": 2396 + }, + { + "epoch": 0.16396470346808947, + "grad_norm": 1.9469000110581058, + "learning_rate": 9.536882164560264e-06, + "loss": 0.4684, + "step": 2397 + }, + { + "epoch": 0.16403310759969902, + "grad_norm": 2.631999691201905, + "learning_rate": 9.536416442891735e-06, + "loss": 0.3855, + "step": 2398 + }, + { + "epoch": 0.16410151173130857, + "grad_norm": 2.0822423931651106, + "learning_rate": 9.53595049855398e-06, + "loss": 0.5516, + "step": 2399 + }, + { + "epoch": 0.16416991586291813, + "grad_norm": 1.2372200190983804, + "learning_rate": 9.535484331569873e-06, + "loss": 0.1224, + "step": 2400 + }, + { + "epoch": 0.16423831999452768, + "grad_norm": 2.0981042692508556, + "learning_rate": 9.535017941962292e-06, + "loss": 0.4706, + "step": 2401 + }, + { + "epoch": 0.16430672412613723, + "grad_norm": 2.829481247536939, + "learning_rate": 9.534551329754132e-06, + "loss": 0.5657, + "step": 2402 + }, + { + "epoch": 0.16437512825774678, + "grad_norm": 1.8604610260725976, + "learning_rate": 9.534084494968296e-06, + "loss": 0.2077, + "step": 2403 + }, + { + "epoch": 0.16444353238935633, + "grad_norm": 1.9714365521596702, + "learning_rate": 9.533617437627696e-06, + "loss": 0.4246, + "step": 2404 + }, + { + "epoch": 0.16451193652096588, + "grad_norm": 2.554665719968946, + "learning_rate": 9.533150157755262e-06, + "loss": 0.4292, + "step": 2405 + }, + { + "epoch": 0.1645803406525754, + "grad_norm": 1.9242972639438982, + "learning_rate": 9.532682655373927e-06, + "loss": 0.312, + "step": 2406 + }, + { + "epoch": 0.16464874478418495, + "grad_norm": 2.653985617873916, + "learning_rate": 9.532214930506638e-06, + "loss": 0.4983, + "step": 2407 + }, + { + "epoch": 0.1647171489157945, + "grad_norm": 2.3184031712959547, + "learning_rate": 9.531746983176356e-06, + "loss": 0.3044, + "step": 2408 + }, + { + "epoch": 0.16478555304740405, + "grad_norm": 2.6702601254116485, + "learning_rate": 9.531278813406047e-06, + "loss": 0.5747, + "step": 2409 + }, + { + "epoch": 0.1648539571790136, + "grad_norm": 1.9572272059850013, + "learning_rate": 9.530810421218693e-06, + "loss": 0.2678, + "step": 2410 + }, + { + "epoch": 0.16492236131062316, + "grad_norm": 1.8292286101137656, + "learning_rate": 9.530341806637285e-06, + "loss": 0.3516, + "step": 2411 + }, + { + "epoch": 0.1649907654422327, + "grad_norm": 2.448620110075896, + "learning_rate": 9.529872969684824e-06, + "loss": 0.4529, + "step": 2412 + }, + { + "epoch": 0.16505916957384226, + "grad_norm": 2.1172673326481695, + "learning_rate": 9.529403910384322e-06, + "loss": 0.5051, + "step": 2413 + }, + { + "epoch": 0.1651275737054518, + "grad_norm": 2.7681567139831817, + "learning_rate": 9.528934628758805e-06, + "loss": 0.7439, + "step": 2414 + }, + { + "epoch": 0.16519597783706136, + "grad_norm": 2.0983909016631186, + "learning_rate": 9.528465124831307e-06, + "loss": 0.423, + "step": 2415 + }, + { + "epoch": 0.1652643819686709, + "grad_norm": 2.78371038800111, + "learning_rate": 9.527995398624871e-06, + "loss": 0.4633, + "step": 2416 + }, + { + "epoch": 0.16533278610028046, + "grad_norm": 1.7842738309930775, + "learning_rate": 9.527525450162558e-06, + "loss": 0.3067, + "step": 2417 + }, + { + "epoch": 0.16540119023189, + "grad_norm": 2.507885111889496, + "learning_rate": 9.527055279467432e-06, + "loss": 0.6526, + "step": 2418 + }, + { + "epoch": 0.16546959436349956, + "grad_norm": 2.3842315450687876, + "learning_rate": 9.526584886562571e-06, + "loss": 0.6934, + "step": 2419 + }, + { + "epoch": 0.1655379984951091, + "grad_norm": 2.352250743029446, + "learning_rate": 9.526114271471065e-06, + "loss": 0.5087, + "step": 2420 + }, + { + "epoch": 0.16560640262671866, + "grad_norm": 1.5317565070525072, + "learning_rate": 9.525643434216014e-06, + "loss": 0.2894, + "step": 2421 + }, + { + "epoch": 0.1656748067583282, + "grad_norm": 2.048323802862455, + "learning_rate": 9.52517237482053e-06, + "loss": 0.4677, + "step": 2422 + }, + { + "epoch": 0.16574321088993776, + "grad_norm": 2.645247025870921, + "learning_rate": 9.524701093307735e-06, + "loss": 0.5617, + "step": 2423 + }, + { + "epoch": 0.16581161502154732, + "grad_norm": 1.8705949058622824, + "learning_rate": 9.524229589700759e-06, + "loss": 0.4023, + "step": 2424 + }, + { + "epoch": 0.16588001915315684, + "grad_norm": 1.8061114593028378, + "learning_rate": 9.523757864022747e-06, + "loss": 0.3524, + "step": 2425 + }, + { + "epoch": 0.1659484232847664, + "grad_norm": 2.4195313219335866, + "learning_rate": 9.523285916296856e-06, + "loss": 0.565, + "step": 2426 + }, + { + "epoch": 0.16601682741637594, + "grad_norm": 2.623286306626401, + "learning_rate": 9.52281374654625e-06, + "loss": 0.668, + "step": 2427 + }, + { + "epoch": 0.1660852315479855, + "grad_norm": 1.6504301529162948, + "learning_rate": 9.522341354794101e-06, + "loss": 0.3207, + "step": 2428 + }, + { + "epoch": 0.16615363567959504, + "grad_norm": 1.979509915318122, + "learning_rate": 9.521868741063603e-06, + "loss": 0.4662, + "step": 2429 + }, + { + "epoch": 0.1662220398112046, + "grad_norm": 2.0952999299516484, + "learning_rate": 9.52139590537795e-06, + "loss": 0.2751, + "step": 2430 + }, + { + "epoch": 0.16629044394281414, + "grad_norm": 1.8626841602069584, + "learning_rate": 9.520922847760355e-06, + "loss": 0.2926, + "step": 2431 + }, + { + "epoch": 0.1663588480744237, + "grad_norm": 2.028866084233344, + "learning_rate": 9.520449568234033e-06, + "loss": 0.3993, + "step": 2432 + }, + { + "epoch": 0.16642725220603324, + "grad_norm": 2.174548616651746, + "learning_rate": 9.519976066822217e-06, + "loss": 0.4892, + "step": 2433 + }, + { + "epoch": 0.1664956563376428, + "grad_norm": 1.942945081553618, + "learning_rate": 9.51950234354815e-06, + "loss": 0.3966, + "step": 2434 + }, + { + "epoch": 0.16656406046925235, + "grad_norm": 2.4723930881748317, + "learning_rate": 9.519028398435082e-06, + "loss": 0.6617, + "step": 2435 + }, + { + "epoch": 0.1666324646008619, + "grad_norm": 2.2617287248513374, + "learning_rate": 9.518554231506281e-06, + "loss": 0.4346, + "step": 2436 + }, + { + "epoch": 0.16670086873247145, + "grad_norm": 2.586927198619331, + "learning_rate": 9.518079842785015e-06, + "loss": 0.3499, + "step": 2437 + }, + { + "epoch": 0.166769272864081, + "grad_norm": 2.2884666852224305, + "learning_rate": 9.517605232294575e-06, + "loss": 0.4898, + "step": 2438 + }, + { + "epoch": 0.16683767699569055, + "grad_norm": 2.100169294350989, + "learning_rate": 9.517130400058255e-06, + "loss": 0.2122, + "step": 2439 + }, + { + "epoch": 0.1669060811273001, + "grad_norm": 2.089786026059408, + "learning_rate": 9.51665534609936e-06, + "loss": 0.3381, + "step": 2440 + }, + { + "epoch": 0.16697448525890965, + "grad_norm": 2.279248143110725, + "learning_rate": 9.516180070441212e-06, + "loss": 0.2685, + "step": 2441 + }, + { + "epoch": 0.1670428893905192, + "grad_norm": 1.704540805990932, + "learning_rate": 9.515704573107136e-06, + "loss": 0.3429, + "step": 2442 + }, + { + "epoch": 0.16711129352212872, + "grad_norm": 2.440878730240296, + "learning_rate": 9.515228854120475e-06, + "loss": 0.492, + "step": 2443 + }, + { + "epoch": 0.16717969765373827, + "grad_norm": 2.3252137594760973, + "learning_rate": 9.514752913504577e-06, + "loss": 0.5762, + "step": 2444 + }, + { + "epoch": 0.16724810178534782, + "grad_norm": 2.321391095786421, + "learning_rate": 9.514276751282805e-06, + "loss": 0.4501, + "step": 2445 + }, + { + "epoch": 0.16731650591695738, + "grad_norm": 1.8259268366789392, + "learning_rate": 9.51380036747853e-06, + "loss": 0.2229, + "step": 2446 + }, + { + "epoch": 0.16738491004856693, + "grad_norm": 1.2390630898227708, + "learning_rate": 9.513323762115137e-06, + "loss": 0.1472, + "step": 2447 + }, + { + "epoch": 0.16745331418017648, + "grad_norm": 1.737464515749075, + "learning_rate": 9.512846935216018e-06, + "loss": 0.3724, + "step": 2448 + }, + { + "epoch": 0.16752171831178603, + "grad_norm": 1.9241262718029943, + "learning_rate": 9.51236988680458e-06, + "loss": 0.3734, + "step": 2449 + }, + { + "epoch": 0.16759012244339558, + "grad_norm": 2.1141270797519427, + "learning_rate": 9.511892616904239e-06, + "loss": 0.4086, + "step": 2450 + }, + { + "epoch": 0.16765852657500513, + "grad_norm": 3.185636480534137, + "learning_rate": 9.511415125538419e-06, + "loss": 0.6929, + "step": 2451 + }, + { + "epoch": 0.16772693070661468, + "grad_norm": 2.562562912657153, + "learning_rate": 9.510937412730558e-06, + "loss": 0.6494, + "step": 2452 + }, + { + "epoch": 0.16779533483822423, + "grad_norm": 2.1569030436237755, + "learning_rate": 9.510459478504109e-06, + "loss": 0.4648, + "step": 2453 + }, + { + "epoch": 0.16786373896983378, + "grad_norm": 1.6005115596487263, + "learning_rate": 9.509981322882527e-06, + "loss": 0.2518, + "step": 2454 + }, + { + "epoch": 0.16793214310144333, + "grad_norm": 2.156373341805898, + "learning_rate": 9.509502945889281e-06, + "loss": 0.4645, + "step": 2455 + }, + { + "epoch": 0.16800054723305288, + "grad_norm": 1.7457573436478535, + "learning_rate": 9.509024347547858e-06, + "loss": 0.2637, + "step": 2456 + }, + { + "epoch": 0.16806895136466243, + "grad_norm": 1.8840040812923988, + "learning_rate": 9.508545527881743e-06, + "loss": 0.2743, + "step": 2457 + }, + { + "epoch": 0.16813735549627198, + "grad_norm": 1.9225229899925522, + "learning_rate": 9.508066486914442e-06, + "loss": 0.2541, + "step": 2458 + }, + { + "epoch": 0.16820575962788153, + "grad_norm": 1.5725739928941773, + "learning_rate": 9.50758722466947e-06, + "loss": 0.1702, + "step": 2459 + }, + { + "epoch": 0.16827416375949109, + "grad_norm": 1.8678642606687528, + "learning_rate": 9.50710774117035e-06, + "loss": 0.2017, + "step": 2460 + }, + { + "epoch": 0.16834256789110064, + "grad_norm": 1.7200772839263594, + "learning_rate": 9.506628036440617e-06, + "loss": 0.3316, + "step": 2461 + }, + { + "epoch": 0.16841097202271016, + "grad_norm": 1.4286687666118185, + "learning_rate": 9.506148110503818e-06, + "loss": 0.216, + "step": 2462 + }, + { + "epoch": 0.1684793761543197, + "grad_norm": 2.004203227830548, + "learning_rate": 9.50566796338351e-06, + "loss": 0.3756, + "step": 2463 + }, + { + "epoch": 0.16854778028592926, + "grad_norm": 1.868077764197891, + "learning_rate": 9.505187595103261e-06, + "loss": 0.2952, + "step": 2464 + }, + { + "epoch": 0.1686161844175388, + "grad_norm": 2.9638801788095632, + "learning_rate": 9.504707005686647e-06, + "loss": 0.7451, + "step": 2465 + }, + { + "epoch": 0.16868458854914836, + "grad_norm": 1.8937647478234412, + "learning_rate": 9.504226195157263e-06, + "loss": 0.2771, + "step": 2466 + }, + { + "epoch": 0.1687529926807579, + "grad_norm": 1.9917159401279834, + "learning_rate": 9.503745163538704e-06, + "loss": 0.5777, + "step": 2467 + }, + { + "epoch": 0.16882139681236746, + "grad_norm": 2.3601000600147346, + "learning_rate": 9.503263910854586e-06, + "loss": 0.3709, + "step": 2468 + }, + { + "epoch": 0.16888980094397701, + "grad_norm": 2.1623389510729885, + "learning_rate": 9.50278243712853e-06, + "loss": 0.5019, + "step": 2469 + }, + { + "epoch": 0.16895820507558656, + "grad_norm": 1.800474393107946, + "learning_rate": 9.502300742384167e-06, + "loss": 0.3915, + "step": 2470 + }, + { + "epoch": 0.16902660920719612, + "grad_norm": 6.113233591692226, + "learning_rate": 9.501818826645142e-06, + "loss": 0.6142, + "step": 2471 + }, + { + "epoch": 0.16909501333880567, + "grad_norm": 2.3455535085005104, + "learning_rate": 9.50133668993511e-06, + "loss": 0.4625, + "step": 2472 + }, + { + "epoch": 0.16916341747041522, + "grad_norm": 1.769304050402877, + "learning_rate": 9.500854332277737e-06, + "loss": 0.3327, + "step": 2473 + }, + { + "epoch": 0.16923182160202477, + "grad_norm": 2.2128350889773913, + "learning_rate": 9.500371753696698e-06, + "loss": 0.3145, + "step": 2474 + }, + { + "epoch": 0.16930022573363432, + "grad_norm": 1.2268235481577463, + "learning_rate": 9.499888954215682e-06, + "loss": 0.1788, + "step": 2475 + }, + { + "epoch": 0.16936862986524387, + "grad_norm": 2.1129580760083377, + "learning_rate": 9.499405933858386e-06, + "loss": 0.4716, + "step": 2476 + }, + { + "epoch": 0.16943703399685342, + "grad_norm": 2.2944479075470605, + "learning_rate": 9.49892269264852e-06, + "loss": 0.4414, + "step": 2477 + }, + { + "epoch": 0.16950543812846297, + "grad_norm": 1.5526277301561016, + "learning_rate": 9.498439230609804e-06, + "loss": 0.1982, + "step": 2478 + }, + { + "epoch": 0.16957384226007252, + "grad_norm": 2.3117744235405633, + "learning_rate": 9.497955547765966e-06, + "loss": 0.4474, + "step": 2479 + }, + { + "epoch": 0.16964224639168204, + "grad_norm": 2.2075545745477947, + "learning_rate": 9.49747164414075e-06, + "loss": 0.5726, + "step": 2480 + }, + { + "epoch": 0.1697106505232916, + "grad_norm": 1.9551611872205006, + "learning_rate": 9.496987519757907e-06, + "loss": 0.3344, + "step": 2481 + }, + { + "epoch": 0.16977905465490115, + "grad_norm": 2.639064734160413, + "learning_rate": 9.496503174641202e-06, + "loss": 0.725, + "step": 2482 + }, + { + "epoch": 0.1698474587865107, + "grad_norm": 2.7052682527455665, + "learning_rate": 9.496018608814407e-06, + "loss": 0.9301, + "step": 2483 + }, + { + "epoch": 0.16991586291812025, + "grad_norm": 1.7680609617910599, + "learning_rate": 9.495533822301307e-06, + "loss": 0.3508, + "step": 2484 + }, + { + "epoch": 0.1699842670497298, + "grad_norm": 1.4813247460607966, + "learning_rate": 9.4950488151257e-06, + "loss": 0.1903, + "step": 2485 + }, + { + "epoch": 0.17005267118133935, + "grad_norm": 1.5887003003839881, + "learning_rate": 9.49456358731139e-06, + "loss": 0.1971, + "step": 2486 + }, + { + "epoch": 0.1701210753129489, + "grad_norm": 2.2403566149192593, + "learning_rate": 9.494078138882195e-06, + "loss": 0.407, + "step": 2487 + }, + { + "epoch": 0.17018947944455845, + "grad_norm": 2.11986157507441, + "learning_rate": 9.493592469861944e-06, + "loss": 0.6203, + "step": 2488 + }, + { + "epoch": 0.170257883576168, + "grad_norm": 2.626024772794484, + "learning_rate": 9.493106580274473e-06, + "loss": 0.8577, + "step": 2489 + }, + { + "epoch": 0.17032628770777755, + "grad_norm": 1.9747096299435958, + "learning_rate": 9.492620470143636e-06, + "loss": 0.4243, + "step": 2490 + }, + { + "epoch": 0.1703946918393871, + "grad_norm": 1.9862412515283148, + "learning_rate": 9.492134139493292e-06, + "loss": 0.511, + "step": 2491 + }, + { + "epoch": 0.17046309597099665, + "grad_norm": 1.2141350767419874, + "learning_rate": 9.49164758834731e-06, + "loss": 0.21, + "step": 2492 + }, + { + "epoch": 0.1705315001026062, + "grad_norm": 2.7831326659803954, + "learning_rate": 9.491160816729577e-06, + "loss": 0.5177, + "step": 2493 + }, + { + "epoch": 0.17059990423421575, + "grad_norm": 3.7160778542894692, + "learning_rate": 9.490673824663983e-06, + "loss": 0.3173, + "step": 2494 + }, + { + "epoch": 0.1706683083658253, + "grad_norm": 2.337453235576621, + "learning_rate": 9.49018661217443e-06, + "loss": 0.6656, + "step": 2495 + }, + { + "epoch": 0.17073671249743486, + "grad_norm": 1.897741716139916, + "learning_rate": 9.489699179284837e-06, + "loss": 0.1442, + "step": 2496 + }, + { + "epoch": 0.1708051166290444, + "grad_norm": 1.8310427550880732, + "learning_rate": 9.489211526019128e-06, + "loss": 0.2239, + "step": 2497 + }, + { + "epoch": 0.17087352076065393, + "grad_norm": 2.1492218111849097, + "learning_rate": 9.48872365240124e-06, + "loss": 0.4285, + "step": 2498 + }, + { + "epoch": 0.17094192489226348, + "grad_norm": 2.412661445236072, + "learning_rate": 9.488235558455118e-06, + "loss": 0.5381, + "step": 2499 + }, + { + "epoch": 0.17101032902387303, + "grad_norm": 1.8516450453115956, + "learning_rate": 9.48774724420472e-06, + "loss": 0.3513, + "step": 2500 + }, + { + "epoch": 0.17107873315548258, + "grad_norm": 2.9030219064891885, + "learning_rate": 9.487258709674019e-06, + "loss": 0.6403, + "step": 2501 + }, + { + "epoch": 0.17114713728709213, + "grad_norm": 2.4539933398998914, + "learning_rate": 9.48676995488699e-06, + "loss": 0.684, + "step": 2502 + }, + { + "epoch": 0.17121554141870168, + "grad_norm": 1.6820211124070374, + "learning_rate": 9.486280979867627e-06, + "loss": 0.3042, + "step": 2503 + }, + { + "epoch": 0.17128394555031123, + "grad_norm": 1.7195106687098334, + "learning_rate": 9.485791784639926e-06, + "loss": 0.2151, + "step": 2504 + }, + { + "epoch": 0.17135234968192078, + "grad_norm": 1.8171936830463, + "learning_rate": 9.485302369227905e-06, + "loss": 0.3796, + "step": 2505 + }, + { + "epoch": 0.17142075381353034, + "grad_norm": 2.3627070399293912, + "learning_rate": 9.484812733655584e-06, + "loss": 0.6758, + "step": 2506 + }, + { + "epoch": 0.1714891579451399, + "grad_norm": 2.1694361145568117, + "learning_rate": 9.484322877946997e-06, + "loss": 0.5503, + "step": 2507 + }, + { + "epoch": 0.17155756207674944, + "grad_norm": 1.626729457542612, + "learning_rate": 9.48383280212619e-06, + "loss": 0.2417, + "step": 2508 + }, + { + "epoch": 0.171625966208359, + "grad_norm": 1.5907113053715851, + "learning_rate": 9.483342506217214e-06, + "loss": 0.2115, + "step": 2509 + }, + { + "epoch": 0.17169437033996854, + "grad_norm": 1.9803001061139414, + "learning_rate": 9.48285199024414e-06, + "loss": 0.2641, + "step": 2510 + }, + { + "epoch": 0.1717627744715781, + "grad_norm": 4.656182512651914, + "learning_rate": 9.482361254231043e-06, + "loss": 0.1662, + "step": 2511 + }, + { + "epoch": 0.17183117860318764, + "grad_norm": 1.756038402048669, + "learning_rate": 9.481870298202009e-06, + "loss": 0.3139, + "step": 2512 + }, + { + "epoch": 0.1718995827347972, + "grad_norm": 1.9517427665094622, + "learning_rate": 9.48137912218114e-06, + "loss": 0.391, + "step": 2513 + }, + { + "epoch": 0.17196798686640674, + "grad_norm": 1.7231416429304658, + "learning_rate": 9.480887726192542e-06, + "loss": 0.4141, + "step": 2514 + }, + { + "epoch": 0.1720363909980163, + "grad_norm": 2.3139312799663427, + "learning_rate": 9.480396110260338e-06, + "loss": 0.5129, + "step": 2515 + }, + { + "epoch": 0.17210479512962584, + "grad_norm": 2.050116302226635, + "learning_rate": 9.479904274408656e-06, + "loss": 0.4525, + "step": 2516 + }, + { + "epoch": 0.17217319926123537, + "grad_norm": 1.7706512207933613, + "learning_rate": 9.479412218661637e-06, + "loss": 0.3481, + "step": 2517 + }, + { + "epoch": 0.17224160339284492, + "grad_norm": 2.1991801859860622, + "learning_rate": 9.47891994304344e-06, + "loss": 0.4405, + "step": 2518 + }, + { + "epoch": 0.17231000752445447, + "grad_norm": 2.7440173572805966, + "learning_rate": 9.47842744757822e-06, + "loss": 0.5497, + "step": 2519 + }, + { + "epoch": 0.17237841165606402, + "grad_norm": 2.1485403521800794, + "learning_rate": 9.477934732290155e-06, + "loss": 0.3394, + "step": 2520 + }, + { + "epoch": 0.17244681578767357, + "grad_norm": 2.101562698534421, + "learning_rate": 9.477441797203431e-06, + "loss": 0.4417, + "step": 2521 + }, + { + "epoch": 0.17251521991928312, + "grad_norm": 2.403959782115622, + "learning_rate": 9.47694864234224e-06, + "loss": 0.379, + "step": 2522 + }, + { + "epoch": 0.17258362405089267, + "grad_norm": 2.036461764688032, + "learning_rate": 9.476455267730792e-06, + "loss": 0.4978, + "step": 2523 + }, + { + "epoch": 0.17265202818250222, + "grad_norm": 1.2456307163767706, + "learning_rate": 9.475961673393306e-06, + "loss": 0.2375, + "step": 2524 + }, + { + "epoch": 0.17272043231411177, + "grad_norm": 2.2720637546084936, + "learning_rate": 9.475467859354004e-06, + "loss": 0.5643, + "step": 2525 + }, + { + "epoch": 0.17278883644572132, + "grad_norm": 2.0333509876964824, + "learning_rate": 9.474973825637126e-06, + "loss": 0.3608, + "step": 2526 + }, + { + "epoch": 0.17285724057733087, + "grad_norm": 2.5291592947167976, + "learning_rate": 9.474479572266926e-06, + "loss": 0.5658, + "step": 2527 + }, + { + "epoch": 0.17292564470894042, + "grad_norm": 2.0216360578895585, + "learning_rate": 9.47398509926766e-06, + "loss": 0.5599, + "step": 2528 + }, + { + "epoch": 0.17299404884054997, + "grad_norm": 2.0694414753665007, + "learning_rate": 9.4734904066636e-06, + "loss": 0.3405, + "step": 2529 + }, + { + "epoch": 0.17306245297215953, + "grad_norm": 2.4419839770543574, + "learning_rate": 9.47299549447903e-06, + "loss": 0.7923, + "step": 2530 + }, + { + "epoch": 0.17313085710376908, + "grad_norm": 1.311383533187853, + "learning_rate": 9.472500362738242e-06, + "loss": 0.1904, + "step": 2531 + }, + { + "epoch": 0.17319926123537863, + "grad_norm": 2.0175927337974326, + "learning_rate": 9.47200501146554e-06, + "loss": 0.4764, + "step": 2532 + }, + { + "epoch": 0.17326766536698818, + "grad_norm": 1.6293927479323522, + "learning_rate": 9.471509440685234e-06, + "loss": 0.2858, + "step": 2533 + }, + { + "epoch": 0.17333606949859773, + "grad_norm": 2.076528409128379, + "learning_rate": 9.471013650421652e-06, + "loss": 0.5763, + "step": 2534 + }, + { + "epoch": 0.17340447363020725, + "grad_norm": 2.72865251376611, + "learning_rate": 9.47051764069913e-06, + "loss": 0.6735, + "step": 2535 + }, + { + "epoch": 0.1734728777618168, + "grad_norm": 2.2206856518060514, + "learning_rate": 9.470021411542014e-06, + "loss": 0.5233, + "step": 2536 + }, + { + "epoch": 0.17354128189342635, + "grad_norm": 1.771260080154046, + "learning_rate": 9.469524962974663e-06, + "loss": 0.3432, + "step": 2537 + }, + { + "epoch": 0.1736096860250359, + "grad_norm": 2.0810280578427487, + "learning_rate": 9.469028295021444e-06, + "loss": 0.4928, + "step": 2538 + }, + { + "epoch": 0.17367809015664545, + "grad_norm": 1.5458783541812877, + "learning_rate": 9.468531407706733e-06, + "loss": 0.2241, + "step": 2539 + }, + { + "epoch": 0.173746494288255, + "grad_norm": 5.578477266018291, + "learning_rate": 9.468034301054925e-06, + "loss": 0.352, + "step": 2540 + }, + { + "epoch": 0.17381489841986456, + "grad_norm": 2.0124032679998245, + "learning_rate": 9.467536975090416e-06, + "loss": 0.3303, + "step": 2541 + }, + { + "epoch": 0.1738833025514741, + "grad_norm": 2.3779971636024544, + "learning_rate": 9.467039429837619e-06, + "loss": 0.6691, + "step": 2542 + }, + { + "epoch": 0.17395170668308366, + "grad_norm": 2.1238078040799104, + "learning_rate": 9.466541665320957e-06, + "loss": 0.5858, + "step": 2543 + }, + { + "epoch": 0.1740201108146932, + "grad_norm": 2.572178012917071, + "learning_rate": 9.46604368156486e-06, + "loss": 0.536, + "step": 2544 + }, + { + "epoch": 0.17408851494630276, + "grad_norm": 2.410890878410778, + "learning_rate": 9.465545478593772e-06, + "loss": 0.4276, + "step": 2545 + }, + { + "epoch": 0.1741569190779123, + "grad_norm": 1.5260298069915126, + "learning_rate": 9.465047056432149e-06, + "loss": 0.3202, + "step": 2546 + }, + { + "epoch": 0.17422532320952186, + "grad_norm": 3.434660814105427, + "learning_rate": 9.464548415104456e-06, + "loss": 0.6383, + "step": 2547 + }, + { + "epoch": 0.1742937273411314, + "grad_norm": 1.5806876234181442, + "learning_rate": 9.464049554635167e-06, + "loss": 0.2701, + "step": 2548 + }, + { + "epoch": 0.17436213147274096, + "grad_norm": 2.0448984693118644, + "learning_rate": 9.463550475048769e-06, + "loss": 0.4608, + "step": 2549 + }, + { + "epoch": 0.1744305356043505, + "grad_norm": 1.788478056217681, + "learning_rate": 9.463051176369758e-06, + "loss": 0.2794, + "step": 2550 + }, + { + "epoch": 0.17449893973596006, + "grad_norm": 2.2804640696400114, + "learning_rate": 9.462551658622646e-06, + "loss": 0.6566, + "step": 2551 + }, + { + "epoch": 0.1745673438675696, + "grad_norm": 1.6524258902043174, + "learning_rate": 9.462051921831946e-06, + "loss": 0.2407, + "step": 2552 + }, + { + "epoch": 0.17463574799917916, + "grad_norm": 2.849774536931241, + "learning_rate": 9.461551966022193e-06, + "loss": 0.617, + "step": 2553 + }, + { + "epoch": 0.1747041521307887, + "grad_norm": 1.8472323506021047, + "learning_rate": 9.461051791217924e-06, + "loss": 0.3133, + "step": 2554 + }, + { + "epoch": 0.17477255626239824, + "grad_norm": 2.116415354560203, + "learning_rate": 9.460551397443691e-06, + "loss": 0.4196, + "step": 2555 + }, + { + "epoch": 0.1748409603940078, + "grad_norm": 2.0101947456051135, + "learning_rate": 9.460050784724056e-06, + "loss": 0.3398, + "step": 2556 + }, + { + "epoch": 0.17490936452561734, + "grad_norm": 2.195814095750056, + "learning_rate": 9.45954995308359e-06, + "loss": 0.6937, + "step": 2557 + }, + { + "epoch": 0.1749777686572269, + "grad_norm": 1.6345580577675844, + "learning_rate": 9.459048902546878e-06, + "loss": 0.1803, + "step": 2558 + }, + { + "epoch": 0.17504617278883644, + "grad_norm": 1.9667792220019518, + "learning_rate": 9.458547633138515e-06, + "loss": 0.4558, + "step": 2559 + }, + { + "epoch": 0.175114576920446, + "grad_norm": 1.9265152729739854, + "learning_rate": 9.458046144883103e-06, + "loss": 0.4637, + "step": 2560 + }, + { + "epoch": 0.17518298105205554, + "grad_norm": 2.123560993233004, + "learning_rate": 9.457544437805258e-06, + "loss": 0.485, + "step": 2561 + }, + { + "epoch": 0.1752513851836651, + "grad_norm": 1.5568392015489219, + "learning_rate": 9.457042511929606e-06, + "loss": 0.3531, + "step": 2562 + }, + { + "epoch": 0.17531978931527464, + "grad_norm": 2.021740566365618, + "learning_rate": 9.456540367280787e-06, + "loss": 0.4594, + "step": 2563 + }, + { + "epoch": 0.1753881934468842, + "grad_norm": 1.8913869898562714, + "learning_rate": 9.456038003883446e-06, + "loss": 0.4432, + "step": 2564 + }, + { + "epoch": 0.17545659757849374, + "grad_norm": 2.0868751904263885, + "learning_rate": 9.455535421762238e-06, + "loss": 0.4439, + "step": 2565 + }, + { + "epoch": 0.1755250017101033, + "grad_norm": 1.9231754958125342, + "learning_rate": 9.45503262094184e-06, + "loss": 0.4866, + "step": 2566 + }, + { + "epoch": 0.17559340584171285, + "grad_norm": 2.422965112044937, + "learning_rate": 9.454529601446927e-06, + "loss": 0.3627, + "step": 2567 + }, + { + "epoch": 0.1756618099733224, + "grad_norm": 1.7313277994846443, + "learning_rate": 9.45402636330219e-06, + "loss": 0.3639, + "step": 2568 + }, + { + "epoch": 0.17573021410493195, + "grad_norm": 1.6029303175965757, + "learning_rate": 9.453522906532331e-06, + "loss": 0.1939, + "step": 2569 + }, + { + "epoch": 0.1757986182365415, + "grad_norm": 1.7768655259365875, + "learning_rate": 9.453019231162062e-06, + "loss": 0.3485, + "step": 2570 + }, + { + "epoch": 0.17586702236815105, + "grad_norm": 1.6441574124824858, + "learning_rate": 9.452515337216104e-06, + "loss": 0.3132, + "step": 2571 + }, + { + "epoch": 0.17593542649976057, + "grad_norm": 1.9156452874856484, + "learning_rate": 9.452011224719195e-06, + "loss": 0.4747, + "step": 2572 + }, + { + "epoch": 0.17600383063137012, + "grad_norm": 2.0033688819295805, + "learning_rate": 9.451506893696074e-06, + "loss": 0.371, + "step": 2573 + }, + { + "epoch": 0.17607223476297967, + "grad_norm": 1.5430154720910203, + "learning_rate": 9.451002344171501e-06, + "loss": 0.2259, + "step": 2574 + }, + { + "epoch": 0.17614063889458922, + "grad_norm": 2.241670708463587, + "learning_rate": 9.450497576170238e-06, + "loss": 0.5827, + "step": 2575 + }, + { + "epoch": 0.17620904302619878, + "grad_norm": 2.1539672571729107, + "learning_rate": 9.449992589717063e-06, + "loss": 0.5303, + "step": 2576 + }, + { + "epoch": 0.17627744715780833, + "grad_norm": 2.3772666931200086, + "learning_rate": 9.449487384836763e-06, + "loss": 0.4722, + "step": 2577 + }, + { + "epoch": 0.17634585128941788, + "grad_norm": 2.031791039084187, + "learning_rate": 9.448981961554136e-06, + "loss": 0.4236, + "step": 2578 + }, + { + "epoch": 0.17641425542102743, + "grad_norm": 2.011463363880865, + "learning_rate": 9.448476319893989e-06, + "loss": 0.4773, + "step": 2579 + }, + { + "epoch": 0.17648265955263698, + "grad_norm": 2.1483123326687026, + "learning_rate": 9.447970459881144e-06, + "loss": 0.3381, + "step": 2580 + }, + { + "epoch": 0.17655106368424653, + "grad_norm": 1.4090478641133144, + "learning_rate": 9.447464381540428e-06, + "loss": 0.1535, + "step": 2581 + }, + { + "epoch": 0.17661946781585608, + "grad_norm": 2.0075682072138967, + "learning_rate": 9.446958084896686e-06, + "loss": 0.4754, + "step": 2582 + }, + { + "epoch": 0.17668787194746563, + "grad_norm": 2.485298494922841, + "learning_rate": 9.446451569974766e-06, + "loss": 0.7572, + "step": 2583 + }, + { + "epoch": 0.17675627607907518, + "grad_norm": 1.4665888144771826, + "learning_rate": 9.445944836799531e-06, + "loss": 0.3349, + "step": 2584 + }, + { + "epoch": 0.17682468021068473, + "grad_norm": 2.295769341010253, + "learning_rate": 9.445437885395854e-06, + "loss": 0.6102, + "step": 2585 + }, + { + "epoch": 0.17689308434229428, + "grad_norm": 2.185206839364895, + "learning_rate": 9.444930715788621e-06, + "loss": 0.6371, + "step": 2586 + }, + { + "epoch": 0.17696148847390383, + "grad_norm": 1.943700888832051, + "learning_rate": 9.444423328002721e-06, + "loss": 0.3997, + "step": 2587 + }, + { + "epoch": 0.17702989260551338, + "grad_norm": 2.0606712267627856, + "learning_rate": 9.443915722063064e-06, + "loss": 0.4479, + "step": 2588 + }, + { + "epoch": 0.17709829673712293, + "grad_norm": 2.310609876300989, + "learning_rate": 9.443407897994563e-06, + "loss": 0.4744, + "step": 2589 + }, + { + "epoch": 0.17716670086873249, + "grad_norm": 1.7001491454910054, + "learning_rate": 9.442899855822147e-06, + "loss": 0.3254, + "step": 2590 + }, + { + "epoch": 0.177235105000342, + "grad_norm": 1.9201782226174484, + "learning_rate": 9.44239159557075e-06, + "loss": 0.3939, + "step": 2591 + }, + { + "epoch": 0.17730350913195156, + "grad_norm": 1.9412071502031434, + "learning_rate": 9.441883117265322e-06, + "loss": 0.2838, + "step": 2592 + }, + { + "epoch": 0.1773719132635611, + "grad_norm": 2.2037399563260274, + "learning_rate": 9.44137442093082e-06, + "loss": 0.5071, + "step": 2593 + }, + { + "epoch": 0.17744031739517066, + "grad_norm": 1.6369523994627344, + "learning_rate": 9.440865506592215e-06, + "loss": 0.2819, + "step": 2594 + }, + { + "epoch": 0.1775087215267802, + "grad_norm": 2.1953781566100874, + "learning_rate": 9.440356374274486e-06, + "loss": 0.6081, + "step": 2595 + }, + { + "epoch": 0.17757712565838976, + "grad_norm": 2.1789298193399866, + "learning_rate": 9.439847024002626e-06, + "loss": 0.6323, + "step": 2596 + }, + { + "epoch": 0.1776455297899993, + "grad_norm": 1.767799561682307, + "learning_rate": 9.439337455801632e-06, + "loss": 0.3833, + "step": 2597 + }, + { + "epoch": 0.17771393392160886, + "grad_norm": 2.4419850571208004, + "learning_rate": 9.438827669696517e-06, + "loss": 0.6384, + "step": 2598 + }, + { + "epoch": 0.1777823380532184, + "grad_norm": 1.5570894262046449, + "learning_rate": 9.438317665712308e-06, + "loss": 0.2473, + "step": 2599 + }, + { + "epoch": 0.17785074218482796, + "grad_norm": 1.4885847403000492, + "learning_rate": 9.437807443874033e-06, + "loss": 0.2309, + "step": 2600 + }, + { + "epoch": 0.17791914631643752, + "grad_norm": 1.7817927713129889, + "learning_rate": 9.437297004206742e-06, + "loss": 0.2343, + "step": 2601 + }, + { + "epoch": 0.17798755044804707, + "grad_norm": 2.1425716848633054, + "learning_rate": 9.436786346735484e-06, + "loss": 0.4874, + "step": 2602 + }, + { + "epoch": 0.17805595457965662, + "grad_norm": 1.7912391098846745, + "learning_rate": 9.436275471485328e-06, + "loss": 0.4063, + "step": 2603 + }, + { + "epoch": 0.17812435871126617, + "grad_norm": 10.94428324029346, + "learning_rate": 9.43576437848135e-06, + "loss": 0.4638, + "step": 2604 + }, + { + "epoch": 0.17819276284287572, + "grad_norm": 2.1572156485354435, + "learning_rate": 9.435253067748633e-06, + "loss": 0.3916, + "step": 2605 + }, + { + "epoch": 0.17826116697448527, + "grad_norm": 3.3512928816468217, + "learning_rate": 9.43474153931228e-06, + "loss": 0.4674, + "step": 2606 + }, + { + "epoch": 0.17832957110609482, + "grad_norm": 1.6393456738997951, + "learning_rate": 9.434229793197398e-06, + "loss": 0.2465, + "step": 2607 + }, + { + "epoch": 0.17839797523770437, + "grad_norm": 2.251180974653004, + "learning_rate": 9.433717829429103e-06, + "loss": 0.6196, + "step": 2608 + }, + { + "epoch": 0.1784663793693139, + "grad_norm": 2.113891062301652, + "learning_rate": 9.433205648032528e-06, + "loss": 0.5546, + "step": 2609 + }, + { + "epoch": 0.17853478350092344, + "grad_norm": 2.092413076899209, + "learning_rate": 9.432693249032812e-06, + "loss": 0.4691, + "step": 2610 + }, + { + "epoch": 0.178603187632533, + "grad_norm": 1.8892619918229636, + "learning_rate": 9.432180632455107e-06, + "loss": 0.4336, + "step": 2611 + }, + { + "epoch": 0.17867159176414255, + "grad_norm": 2.0309699085500705, + "learning_rate": 9.431667798324573e-06, + "loss": 0.2592, + "step": 2612 + }, + { + "epoch": 0.1787399958957521, + "grad_norm": 2.26325624548913, + "learning_rate": 9.431154746666382e-06, + "loss": 0.3349, + "step": 2613 + }, + { + "epoch": 0.17880840002736165, + "grad_norm": 2.040997146449788, + "learning_rate": 9.43064147750572e-06, + "loss": 0.38, + "step": 2614 + }, + { + "epoch": 0.1788768041589712, + "grad_norm": 1.73732105481145, + "learning_rate": 9.430127990867778e-06, + "loss": 0.3035, + "step": 2615 + }, + { + "epoch": 0.17894520829058075, + "grad_norm": 2.2242434253402514, + "learning_rate": 9.429614286777762e-06, + "loss": 0.5901, + "step": 2616 + }, + { + "epoch": 0.1790136124221903, + "grad_norm": 2.0351854640745617, + "learning_rate": 9.429100365260886e-06, + "loss": 0.4907, + "step": 2617 + }, + { + "epoch": 0.17908201655379985, + "grad_norm": 2.189723048386382, + "learning_rate": 9.428586226342375e-06, + "loss": 0.4398, + "step": 2618 + }, + { + "epoch": 0.1791504206854094, + "grad_norm": 2.564253602470076, + "learning_rate": 9.428071870047469e-06, + "loss": 0.3835, + "step": 2619 + }, + { + "epoch": 0.17921882481701895, + "grad_norm": 1.7639700594216021, + "learning_rate": 9.427557296401411e-06, + "loss": 0.3607, + "step": 2620 + }, + { + "epoch": 0.1792872289486285, + "grad_norm": 1.9157029342977967, + "learning_rate": 9.42704250542946e-06, + "loss": 0.3761, + "step": 2621 + }, + { + "epoch": 0.17935563308023805, + "grad_norm": 1.8553002692576512, + "learning_rate": 9.426527497156885e-06, + "loss": 0.5257, + "step": 2622 + }, + { + "epoch": 0.1794240372118476, + "grad_norm": 1.684420641455901, + "learning_rate": 9.426012271608967e-06, + "loss": 0.3335, + "step": 2623 + }, + { + "epoch": 0.17949244134345715, + "grad_norm": 2.422596128837807, + "learning_rate": 9.425496828810992e-06, + "loss": 0.5597, + "step": 2624 + }, + { + "epoch": 0.1795608454750667, + "grad_norm": 2.300897345989752, + "learning_rate": 9.42498116878826e-06, + "loss": 0.628, + "step": 2625 + }, + { + "epoch": 0.17962924960667626, + "grad_norm": 2.4071871307424266, + "learning_rate": 9.424465291566088e-06, + "loss": 0.3359, + "step": 2626 + }, + { + "epoch": 0.17969765373828578, + "grad_norm": 2.121221900286262, + "learning_rate": 9.423949197169792e-06, + "loss": 0.2124, + "step": 2627 + }, + { + "epoch": 0.17976605786989533, + "grad_norm": 2.0502487617554523, + "learning_rate": 9.423432885624708e-06, + "loss": 0.531, + "step": 2628 + }, + { + "epoch": 0.17983446200150488, + "grad_norm": 2.5493263079310857, + "learning_rate": 9.422916356956175e-06, + "loss": 0.6223, + "step": 2629 + }, + { + "epoch": 0.17990286613311443, + "grad_norm": 2.9309238520803853, + "learning_rate": 9.422399611189553e-06, + "loss": 0.3322, + "step": 2630 + }, + { + "epoch": 0.17997127026472398, + "grad_norm": 1.8512772046575234, + "learning_rate": 9.421882648350198e-06, + "loss": 0.2724, + "step": 2631 + }, + { + "epoch": 0.18003967439633353, + "grad_norm": 1.9057323386106932, + "learning_rate": 9.421365468463493e-06, + "loss": 0.2639, + "step": 2632 + }, + { + "epoch": 0.18010807852794308, + "grad_norm": 1.7897014018498498, + "learning_rate": 9.420848071554819e-06, + "loss": 0.4349, + "step": 2633 + }, + { + "epoch": 0.18017648265955263, + "grad_norm": 2.0320930052117685, + "learning_rate": 9.420330457649573e-06, + "loss": 0.2799, + "step": 2634 + }, + { + "epoch": 0.18024488679116218, + "grad_norm": 2.0930213371980853, + "learning_rate": 9.419812626773163e-06, + "loss": 0.2096, + "step": 2635 + }, + { + "epoch": 0.18031329092277174, + "grad_norm": 2.6455476010844867, + "learning_rate": 9.419294578951006e-06, + "loss": 0.5554, + "step": 2636 + }, + { + "epoch": 0.18038169505438129, + "grad_norm": 1.7047049713812876, + "learning_rate": 9.418776314208532e-06, + "loss": 0.2687, + "step": 2637 + }, + { + "epoch": 0.18045009918599084, + "grad_norm": 2.2133174730346883, + "learning_rate": 9.418257832571176e-06, + "loss": 0.5936, + "step": 2638 + }, + { + "epoch": 0.1805185033176004, + "grad_norm": 2.0849705215698373, + "learning_rate": 9.417739134064392e-06, + "loss": 0.5457, + "step": 2639 + }, + { + "epoch": 0.18058690744920994, + "grad_norm": 1.8464780704494765, + "learning_rate": 9.417220218713636e-06, + "loss": 0.2589, + "step": 2640 + }, + { + "epoch": 0.1806553115808195, + "grad_norm": 2.2713201173153017, + "learning_rate": 9.416701086544384e-06, + "loss": 0.3117, + "step": 2641 + }, + { + "epoch": 0.18072371571242904, + "grad_norm": 2.3993777441956547, + "learning_rate": 9.416181737582112e-06, + "loss": 0.4758, + "step": 2642 + }, + { + "epoch": 0.1807921198440386, + "grad_norm": 2.16697762318101, + "learning_rate": 9.415662171852317e-06, + "loss": 0.5773, + "step": 2643 + }, + { + "epoch": 0.18086052397564814, + "grad_norm": 1.7404126783448868, + "learning_rate": 9.415142389380498e-06, + "loss": 0.2783, + "step": 2644 + }, + { + "epoch": 0.1809289281072577, + "grad_norm": 1.4949898444424041, + "learning_rate": 9.414622390192171e-06, + "loss": 0.3453, + "step": 2645 + }, + { + "epoch": 0.18099733223886721, + "grad_norm": 1.4438995072075584, + "learning_rate": 9.414102174312859e-06, + "loss": 0.2372, + "step": 2646 + }, + { + "epoch": 0.18106573637047677, + "grad_norm": 1.9267143789849224, + "learning_rate": 9.413581741768096e-06, + "loss": 0.2429, + "step": 2647 + }, + { + "epoch": 0.18113414050208632, + "grad_norm": 2.7904531073520293, + "learning_rate": 9.41306109258343e-06, + "loss": 0.4634, + "step": 2648 + }, + { + "epoch": 0.18120254463369587, + "grad_norm": 1.8370273926254699, + "learning_rate": 9.412540226784413e-06, + "loss": 0.3856, + "step": 2649 + }, + { + "epoch": 0.18127094876530542, + "grad_norm": 1.9189446252719726, + "learning_rate": 9.412019144396616e-06, + "loss": 0.3916, + "step": 2650 + }, + { + "epoch": 0.18133935289691497, + "grad_norm": 2.43709047251783, + "learning_rate": 9.411497845445612e-06, + "loss": 0.5569, + "step": 2651 + }, + { + "epoch": 0.18140775702852452, + "grad_norm": 2.0347969338375016, + "learning_rate": 9.410976329956992e-06, + "loss": 0.1683, + "step": 2652 + }, + { + "epoch": 0.18147616116013407, + "grad_norm": 2.1097724045699464, + "learning_rate": 9.410454597956351e-06, + "loss": 0.6015, + "step": 2653 + }, + { + "epoch": 0.18154456529174362, + "grad_norm": 1.96228984539729, + "learning_rate": 9.409932649469302e-06, + "loss": 0.5205, + "step": 2654 + }, + { + "epoch": 0.18161296942335317, + "grad_norm": 1.602145623838583, + "learning_rate": 9.409410484521464e-06, + "loss": 0.1611, + "step": 2655 + }, + { + "epoch": 0.18168137355496272, + "grad_norm": 1.9375732930994778, + "learning_rate": 9.408888103138467e-06, + "loss": 0.3227, + "step": 2656 + }, + { + "epoch": 0.18174977768657227, + "grad_norm": 9.223901686376882, + "learning_rate": 9.40836550534595e-06, + "loss": 0.2303, + "step": 2657 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 2.0486226265165177, + "learning_rate": 9.407842691169566e-06, + "loss": 0.3358, + "step": 2658 + }, + { + "epoch": 0.18188658594979137, + "grad_norm": 1.686691028460811, + "learning_rate": 9.40731966063498e-06, + "loss": 0.3672, + "step": 2659 + }, + { + "epoch": 0.18195499008140092, + "grad_norm": 2.2619415463358177, + "learning_rate": 9.40679641376786e-06, + "loss": 0.3529, + "step": 2660 + }, + { + "epoch": 0.18202339421301048, + "grad_norm": 1.5860100762433806, + "learning_rate": 9.406272950593892e-06, + "loss": 0.2703, + "step": 2661 + }, + { + "epoch": 0.18209179834462003, + "grad_norm": 2.134360160608494, + "learning_rate": 9.405749271138771e-06, + "loss": 0.4408, + "step": 2662 + }, + { + "epoch": 0.18216020247622958, + "grad_norm": 1.8227092656724593, + "learning_rate": 9.4052253754282e-06, + "loss": 0.4037, + "step": 2663 + }, + { + "epoch": 0.1822286066078391, + "grad_norm": 1.7539424085394437, + "learning_rate": 9.404701263487893e-06, + "loss": 0.2555, + "step": 2664 + }, + { + "epoch": 0.18229701073944865, + "grad_norm": 2.021698613262951, + "learning_rate": 9.40417693534358e-06, + "loss": 0.5047, + "step": 2665 + }, + { + "epoch": 0.1823654148710582, + "grad_norm": 2.060528716237837, + "learning_rate": 9.403652391020996e-06, + "loss": 0.4486, + "step": 2666 + }, + { + "epoch": 0.18243381900266775, + "grad_norm": 2.025721821794134, + "learning_rate": 9.403127630545887e-06, + "loss": 0.5139, + "step": 2667 + }, + { + "epoch": 0.1825022231342773, + "grad_norm": 2.968831071248616, + "learning_rate": 9.402602653944012e-06, + "loss": 0.319, + "step": 2668 + }, + { + "epoch": 0.18257062726588685, + "grad_norm": 2.0459292571083987, + "learning_rate": 9.402077461241139e-06, + "loss": 0.423, + "step": 2669 + }, + { + "epoch": 0.1826390313974964, + "grad_norm": 1.328461315544582, + "learning_rate": 9.401552052463046e-06, + "loss": 0.1062, + "step": 2670 + }, + { + "epoch": 0.18270743552910595, + "grad_norm": 2.3125621684197806, + "learning_rate": 9.401026427635522e-06, + "loss": 0.4412, + "step": 2671 + }, + { + "epoch": 0.1827758396607155, + "grad_norm": 1.6802377730941944, + "learning_rate": 9.40050058678437e-06, + "loss": 0.3354, + "step": 2672 + }, + { + "epoch": 0.18284424379232506, + "grad_norm": 2.218047514373858, + "learning_rate": 9.3999745299354e-06, + "loss": 0.5354, + "step": 2673 + }, + { + "epoch": 0.1829126479239346, + "grad_norm": 2.0886919501848586, + "learning_rate": 9.399448257114432e-06, + "loss": 0.5341, + "step": 2674 + }, + { + "epoch": 0.18298105205554416, + "grad_norm": 2.465689893238432, + "learning_rate": 9.398921768347298e-06, + "loss": 0.4261, + "step": 2675 + }, + { + "epoch": 0.1830494561871537, + "grad_norm": 1.6469516249534444, + "learning_rate": 9.398395063659841e-06, + "loss": 0.2917, + "step": 2676 + }, + { + "epoch": 0.18311786031876326, + "grad_norm": 3.7603775121001215, + "learning_rate": 9.397868143077916e-06, + "loss": 0.3657, + "step": 2677 + }, + { + "epoch": 0.1831862644503728, + "grad_norm": 2.483703990504375, + "learning_rate": 9.397341006627387e-06, + "loss": 0.6388, + "step": 2678 + }, + { + "epoch": 0.18325466858198236, + "grad_norm": 2.078786400541084, + "learning_rate": 9.396813654334124e-06, + "loss": 0.4568, + "step": 2679 + }, + { + "epoch": 0.1833230727135919, + "grad_norm": 2.0348887203636727, + "learning_rate": 9.396286086224017e-06, + "loss": 0.3429, + "step": 2680 + }, + { + "epoch": 0.18339147684520146, + "grad_norm": 2.028480247983894, + "learning_rate": 9.39575830232296e-06, + "loss": 0.6193, + "step": 2681 + }, + { + "epoch": 0.183459880976811, + "grad_norm": 6.6899365416112975, + "learning_rate": 9.395230302656856e-06, + "loss": 0.5096, + "step": 2682 + }, + { + "epoch": 0.18352828510842054, + "grad_norm": 2.9919803039685324, + "learning_rate": 9.394702087251626e-06, + "loss": 0.1604, + "step": 2683 + }, + { + "epoch": 0.1835966892400301, + "grad_norm": 2.1348177667027874, + "learning_rate": 9.394173656133195e-06, + "loss": 0.2709, + "step": 2684 + }, + { + "epoch": 0.18366509337163964, + "grad_norm": 2.1807295724925946, + "learning_rate": 9.393645009327502e-06, + "loss": 0.4724, + "step": 2685 + }, + { + "epoch": 0.1837334975032492, + "grad_norm": 2.535306217830459, + "learning_rate": 9.393116146860496e-06, + "loss": 0.834, + "step": 2686 + }, + { + "epoch": 0.18380190163485874, + "grad_norm": 2.4943891668789586, + "learning_rate": 9.392587068758135e-06, + "loss": 0.2532, + "step": 2687 + }, + { + "epoch": 0.1838703057664683, + "grad_norm": 2.014096086812431, + "learning_rate": 9.392057775046389e-06, + "loss": 0.4795, + "step": 2688 + }, + { + "epoch": 0.18393870989807784, + "grad_norm": 2.104311977379725, + "learning_rate": 9.391528265751237e-06, + "loss": 0.3132, + "step": 2689 + }, + { + "epoch": 0.1840071140296874, + "grad_norm": 2.1848769128379564, + "learning_rate": 9.390998540898672e-06, + "loss": 0.5289, + "step": 2690 + }, + { + "epoch": 0.18407551816129694, + "grad_norm": 1.8739051960452566, + "learning_rate": 9.390468600514694e-06, + "loss": 0.4255, + "step": 2691 + }, + { + "epoch": 0.1841439222929065, + "grad_norm": 2.0196830887962496, + "learning_rate": 9.389938444625318e-06, + "loss": 0.3605, + "step": 2692 + }, + { + "epoch": 0.18421232642451604, + "grad_norm": 2.259159930137765, + "learning_rate": 9.389408073256563e-06, + "loss": 0.4955, + "step": 2693 + }, + { + "epoch": 0.1842807305561256, + "grad_norm": 1.7635570881644909, + "learning_rate": 9.388877486434465e-06, + "loss": 0.2002, + "step": 2694 + }, + { + "epoch": 0.18434913468773514, + "grad_norm": 2.3056598939579236, + "learning_rate": 9.388346684185063e-06, + "loss": 0.6199, + "step": 2695 + }, + { + "epoch": 0.1844175388193447, + "grad_norm": 1.9688055091177472, + "learning_rate": 9.387815666534417e-06, + "loss": 0.1968, + "step": 2696 + }, + { + "epoch": 0.18448594295095425, + "grad_norm": 1.8654534655615094, + "learning_rate": 9.38728443350859e-06, + "loss": 0.4712, + "step": 2697 + }, + { + "epoch": 0.1845543470825638, + "grad_norm": 1.803295573029018, + "learning_rate": 9.386752985133656e-06, + "loss": 0.4261, + "step": 2698 + }, + { + "epoch": 0.18462275121417335, + "grad_norm": 1.885980126824415, + "learning_rate": 9.386221321435702e-06, + "loss": 0.1988, + "step": 2699 + }, + { + "epoch": 0.1846911553457829, + "grad_norm": 1.9998393776588976, + "learning_rate": 9.385689442440825e-06, + "loss": 0.2518, + "step": 2700 + }, + { + "epoch": 0.18475955947739242, + "grad_norm": 1.4033360450515782, + "learning_rate": 9.385157348175131e-06, + "loss": 0.1808, + "step": 2701 + }, + { + "epoch": 0.18482796360900197, + "grad_norm": 1.9159181931713938, + "learning_rate": 9.384625038664742e-06, + "loss": 0.1946, + "step": 2702 + }, + { + "epoch": 0.18489636774061152, + "grad_norm": 1.559179731271297, + "learning_rate": 9.384092513935779e-06, + "loss": 0.3693, + "step": 2703 + }, + { + "epoch": 0.18496477187222107, + "grad_norm": 1.8539234392066126, + "learning_rate": 9.383559774014386e-06, + "loss": 0.3586, + "step": 2704 + }, + { + "epoch": 0.18503317600383062, + "grad_norm": 1.5825201150480546, + "learning_rate": 9.383026818926711e-06, + "loss": 0.283, + "step": 2705 + }, + { + "epoch": 0.18510158013544017, + "grad_norm": 2.5629256842758514, + "learning_rate": 9.382493648698915e-06, + "loss": 0.2935, + "step": 2706 + }, + { + "epoch": 0.18516998426704973, + "grad_norm": 2.2619984112918647, + "learning_rate": 9.381960263357168e-06, + "loss": 0.5045, + "step": 2707 + }, + { + "epoch": 0.18523838839865928, + "grad_norm": 1.9557756366787848, + "learning_rate": 9.38142666292765e-06, + "loss": 0.3973, + "step": 2708 + }, + { + "epoch": 0.18530679253026883, + "grad_norm": 1.3030764975543632, + "learning_rate": 9.380892847436555e-06, + "loss": 0.1463, + "step": 2709 + }, + { + "epoch": 0.18537519666187838, + "grad_norm": 1.8532888717951337, + "learning_rate": 9.380358816910083e-06, + "loss": 0.2939, + "step": 2710 + }, + { + "epoch": 0.18544360079348793, + "grad_norm": 2.156400730559123, + "learning_rate": 9.379824571374449e-06, + "loss": 0.655, + "step": 2711 + }, + { + "epoch": 0.18551200492509748, + "grad_norm": 2.3201358708672175, + "learning_rate": 9.379290110855873e-06, + "loss": 0.3417, + "step": 2712 + }, + { + "epoch": 0.18558040905670703, + "grad_norm": 2.505336947850577, + "learning_rate": 9.378755435380592e-06, + "loss": 0.3376, + "step": 2713 + }, + { + "epoch": 0.18564881318831658, + "grad_norm": 1.984864215920992, + "learning_rate": 9.378220544974851e-06, + "loss": 0.3387, + "step": 2714 + }, + { + "epoch": 0.18571721731992613, + "grad_norm": 1.6413903284866693, + "learning_rate": 9.377685439664902e-06, + "loss": 0.182, + "step": 2715 + }, + { + "epoch": 0.18578562145153568, + "grad_norm": 2.2905154970960933, + "learning_rate": 9.377150119477011e-06, + "loss": 0.5967, + "step": 2716 + }, + { + "epoch": 0.18585402558314523, + "grad_norm": 2.1207214084862263, + "learning_rate": 9.376614584437457e-06, + "loss": 0.4705, + "step": 2717 + }, + { + "epoch": 0.18592242971475478, + "grad_norm": 2.3451170050096035, + "learning_rate": 9.376078834572525e-06, + "loss": 0.637, + "step": 2718 + }, + { + "epoch": 0.18599083384636433, + "grad_norm": 1.5301999169990956, + "learning_rate": 9.375542869908509e-06, + "loss": 0.12, + "step": 2719 + }, + { + "epoch": 0.18605923797797386, + "grad_norm": 2.431268711067119, + "learning_rate": 9.375006690471721e-06, + "loss": 0.4966, + "step": 2720 + }, + { + "epoch": 0.1861276421095834, + "grad_norm": 2.198150954859318, + "learning_rate": 9.374470296288479e-06, + "loss": 0.5321, + "step": 2721 + }, + { + "epoch": 0.18619604624119296, + "grad_norm": 2.2357361055706146, + "learning_rate": 9.373933687385111e-06, + "loss": 0.5953, + "step": 2722 + }, + { + "epoch": 0.1862644503728025, + "grad_norm": 1.707991330049267, + "learning_rate": 9.373396863787955e-06, + "loss": 0.3154, + "step": 2723 + }, + { + "epoch": 0.18633285450441206, + "grad_norm": 2.0511785358530754, + "learning_rate": 9.372859825523361e-06, + "loss": 0.5515, + "step": 2724 + }, + { + "epoch": 0.1864012586360216, + "grad_norm": 2.47257941248681, + "learning_rate": 9.372322572617693e-06, + "loss": 0.6247, + "step": 2725 + }, + { + "epoch": 0.18646966276763116, + "grad_norm": 1.4405606640250384, + "learning_rate": 9.371785105097317e-06, + "loss": 0.2584, + "step": 2726 + }, + { + "epoch": 0.1865380668992407, + "grad_norm": 2.2369100155132204, + "learning_rate": 9.37124742298862e-06, + "loss": 0.2485, + "step": 2727 + }, + { + "epoch": 0.18660647103085026, + "grad_norm": 1.7285642627834383, + "learning_rate": 9.370709526317987e-06, + "loss": 0.3464, + "step": 2728 + }, + { + "epoch": 0.1866748751624598, + "grad_norm": 1.8029906523650516, + "learning_rate": 9.370171415111828e-06, + "loss": 0.3813, + "step": 2729 + }, + { + "epoch": 0.18674327929406936, + "grad_norm": 2.5576942435153014, + "learning_rate": 9.36963308939655e-06, + "loss": 0.825, + "step": 2730 + }, + { + "epoch": 0.18681168342567891, + "grad_norm": 1.8068137544724392, + "learning_rate": 9.36909454919858e-06, + "loss": 0.3298, + "step": 2731 + }, + { + "epoch": 0.18688008755728847, + "grad_norm": 1.839127127679421, + "learning_rate": 9.368555794544352e-06, + "loss": 0.2553, + "step": 2732 + }, + { + "epoch": 0.18694849168889802, + "grad_norm": 1.9470590856861558, + "learning_rate": 9.36801682546031e-06, + "loss": 0.2352, + "step": 2733 + }, + { + "epoch": 0.18701689582050757, + "grad_norm": 2.2558949389275926, + "learning_rate": 9.367477641972908e-06, + "loss": 0.6083, + "step": 2734 + }, + { + "epoch": 0.18708529995211712, + "grad_norm": 1.4462142914705143, + "learning_rate": 9.366938244108616e-06, + "loss": 0.2149, + "step": 2735 + }, + { + "epoch": 0.18715370408372667, + "grad_norm": 1.757293675468064, + "learning_rate": 9.366398631893906e-06, + "loss": 0.4399, + "step": 2736 + }, + { + "epoch": 0.18722210821533622, + "grad_norm": 2.3991621767887383, + "learning_rate": 9.365858805355263e-06, + "loss": 0.61, + "step": 2737 + }, + { + "epoch": 0.18729051234694574, + "grad_norm": 1.5391917924744065, + "learning_rate": 9.36531876451919e-06, + "loss": 0.3237, + "step": 2738 + }, + { + "epoch": 0.1873589164785553, + "grad_norm": 1.8731824052164872, + "learning_rate": 9.364778509412191e-06, + "loss": 0.2336, + "step": 2739 + }, + { + "epoch": 0.18742732061016484, + "grad_norm": 1.9979066547849662, + "learning_rate": 9.364238040060787e-06, + "loss": 0.4286, + "step": 2740 + }, + { + "epoch": 0.1874957247417744, + "grad_norm": 1.9534410945692662, + "learning_rate": 9.363697356491505e-06, + "loss": 0.2588, + "step": 2741 + }, + { + "epoch": 0.18756412887338395, + "grad_norm": 2.257610080977435, + "learning_rate": 9.363156458730884e-06, + "loss": 0.6484, + "step": 2742 + }, + { + "epoch": 0.1876325330049935, + "grad_norm": 1.9274719005483454, + "learning_rate": 9.362615346805475e-06, + "loss": 0.4176, + "step": 2743 + }, + { + "epoch": 0.18770093713660305, + "grad_norm": 2.3821082465869843, + "learning_rate": 9.362074020741836e-06, + "loss": 0.5513, + "step": 2744 + }, + { + "epoch": 0.1877693412682126, + "grad_norm": 1.986921433742885, + "learning_rate": 9.361532480566542e-06, + "loss": 0.2246, + "step": 2745 + }, + { + "epoch": 0.18783774539982215, + "grad_norm": 1.911530965383114, + "learning_rate": 9.360990726306171e-06, + "loss": 0.4572, + "step": 2746 + }, + { + "epoch": 0.1879061495314317, + "grad_norm": 2.3423768789572, + "learning_rate": 9.360448757987318e-06, + "loss": 0.4021, + "step": 2747 + }, + { + "epoch": 0.18797455366304125, + "grad_norm": 1.8092749694818524, + "learning_rate": 9.359906575636582e-06, + "loss": 0.3081, + "step": 2748 + }, + { + "epoch": 0.1880429577946508, + "grad_norm": 1.5931455375171455, + "learning_rate": 9.359364179280578e-06, + "loss": 0.1914, + "step": 2749 + }, + { + "epoch": 0.18811136192626035, + "grad_norm": 1.992327685192273, + "learning_rate": 9.358821568945929e-06, + "loss": 0.3721, + "step": 2750 + }, + { + "epoch": 0.1881797660578699, + "grad_norm": 2.1151531288415324, + "learning_rate": 9.358278744659268e-06, + "loss": 0.1904, + "step": 2751 + }, + { + "epoch": 0.18824817018947945, + "grad_norm": 2.3643426828699257, + "learning_rate": 9.357735706447243e-06, + "loss": 0.5296, + "step": 2752 + }, + { + "epoch": 0.188316574321089, + "grad_norm": 2.1666766878654355, + "learning_rate": 9.357192454336503e-06, + "loss": 0.4244, + "step": 2753 + }, + { + "epoch": 0.18838497845269855, + "grad_norm": 1.759571270048457, + "learning_rate": 9.356648988353717e-06, + "loss": 0.3069, + "step": 2754 + }, + { + "epoch": 0.1884533825843081, + "grad_norm": 2.5349344588540608, + "learning_rate": 9.356105308525563e-06, + "loss": 0.3956, + "step": 2755 + }, + { + "epoch": 0.18852178671591766, + "grad_norm": 1.8715762868027241, + "learning_rate": 9.355561414878722e-06, + "loss": 0.2603, + "step": 2756 + }, + { + "epoch": 0.18859019084752718, + "grad_norm": 2.086976921052409, + "learning_rate": 9.355017307439896e-06, + "loss": 0.3657, + "step": 2757 + }, + { + "epoch": 0.18865859497913673, + "grad_norm": 1.885997411207701, + "learning_rate": 9.35447298623579e-06, + "loss": 0.4178, + "step": 2758 + }, + { + "epoch": 0.18872699911074628, + "grad_norm": 2.517994292127004, + "learning_rate": 9.353928451293122e-06, + "loss": 0.7078, + "step": 2759 + }, + { + "epoch": 0.18879540324235583, + "grad_norm": 2.7836569111628386, + "learning_rate": 9.35338370263862e-06, + "loss": 0.776, + "step": 2760 + }, + { + "epoch": 0.18886380737396538, + "grad_norm": 2.3182791781996315, + "learning_rate": 9.352838740299025e-06, + "loss": 0.6546, + "step": 2761 + }, + { + "epoch": 0.18893221150557493, + "grad_norm": 1.4792742880322505, + "learning_rate": 9.352293564301086e-06, + "loss": 0.2127, + "step": 2762 + }, + { + "epoch": 0.18900061563718448, + "grad_norm": 2.037968444362092, + "learning_rate": 9.35174817467156e-06, + "loss": 0.3936, + "step": 2763 + }, + { + "epoch": 0.18906901976879403, + "grad_norm": 2.8300475756652963, + "learning_rate": 9.351202571437219e-06, + "loss": 0.6752, + "step": 2764 + }, + { + "epoch": 0.18913742390040358, + "grad_norm": 2.1361142755788682, + "learning_rate": 9.350656754624846e-06, + "loss": 0.4845, + "step": 2765 + }, + { + "epoch": 0.18920582803201313, + "grad_norm": 2.102153177458918, + "learning_rate": 9.350110724261228e-06, + "loss": 0.5035, + "step": 2766 + }, + { + "epoch": 0.18927423216362269, + "grad_norm": 1.9778266565747757, + "learning_rate": 9.34956448037317e-06, + "loss": 0.3751, + "step": 2767 + }, + { + "epoch": 0.18934263629523224, + "grad_norm": 1.9466115528368033, + "learning_rate": 9.349018022987484e-06, + "loss": 0.4526, + "step": 2768 + }, + { + "epoch": 0.1894110404268418, + "grad_norm": 2.409056506894815, + "learning_rate": 9.348471352130991e-06, + "loss": 0.3644, + "step": 2769 + }, + { + "epoch": 0.18947944455845134, + "grad_norm": 2.32482446423201, + "learning_rate": 9.347924467830527e-06, + "loss": 0.7108, + "step": 2770 + }, + { + "epoch": 0.1895478486900609, + "grad_norm": 2.08143650528071, + "learning_rate": 9.347377370112934e-06, + "loss": 0.2217, + "step": 2771 + }, + { + "epoch": 0.18961625282167044, + "grad_norm": 2.4070480032848236, + "learning_rate": 9.346830059005066e-06, + "loss": 0.4235, + "step": 2772 + }, + { + "epoch": 0.18968465695328, + "grad_norm": 1.7760794701752656, + "learning_rate": 9.346282534533787e-06, + "loss": 0.428, + "step": 2773 + }, + { + "epoch": 0.18975306108488954, + "grad_norm": 2.180120575516667, + "learning_rate": 9.345734796725974e-06, + "loss": 0.669, + "step": 2774 + }, + { + "epoch": 0.18982146521649906, + "grad_norm": 2.499530581272534, + "learning_rate": 9.345186845608512e-06, + "loss": 0.6027, + "step": 2775 + }, + { + "epoch": 0.18988986934810861, + "grad_norm": 2.06694787684539, + "learning_rate": 9.344638681208298e-06, + "loss": 0.6125, + "step": 2776 + }, + { + "epoch": 0.18995827347971816, + "grad_norm": 2.027922556872958, + "learning_rate": 9.344090303552236e-06, + "loss": 0.245, + "step": 2777 + }, + { + "epoch": 0.19002667761132772, + "grad_norm": 1.8585912511230704, + "learning_rate": 9.343541712667245e-06, + "loss": 0.2105, + "step": 2778 + }, + { + "epoch": 0.19009508174293727, + "grad_norm": 1.766274048077285, + "learning_rate": 9.342992908580252e-06, + "loss": 0.293, + "step": 2779 + }, + { + "epoch": 0.19016348587454682, + "grad_norm": 2.2748212372060004, + "learning_rate": 9.342443891318194e-06, + "loss": 0.369, + "step": 2780 + }, + { + "epoch": 0.19023189000615637, + "grad_norm": 2.630992022301799, + "learning_rate": 9.341894660908023e-06, + "loss": 0.5107, + "step": 2781 + }, + { + "epoch": 0.19030029413776592, + "grad_norm": 1.7456300128198277, + "learning_rate": 9.341345217376692e-06, + "loss": 0.3289, + "step": 2782 + }, + { + "epoch": 0.19036869826937547, + "grad_norm": 2.023146722586807, + "learning_rate": 9.340795560751175e-06, + "loss": 0.3337, + "step": 2783 + }, + { + "epoch": 0.19043710240098502, + "grad_norm": 2.348452948584794, + "learning_rate": 9.340245691058451e-06, + "loss": 0.6574, + "step": 2784 + }, + { + "epoch": 0.19050550653259457, + "grad_norm": 1.8322141104660246, + "learning_rate": 9.33969560832551e-06, + "loss": 0.5132, + "step": 2785 + }, + { + "epoch": 0.19057391066420412, + "grad_norm": 2.858206942927068, + "learning_rate": 9.33914531257935e-06, + "loss": 0.4038, + "step": 2786 + }, + { + "epoch": 0.19064231479581367, + "grad_norm": 1.7616309738283908, + "learning_rate": 9.338594803846987e-06, + "loss": 0.3202, + "step": 2787 + }, + { + "epoch": 0.19071071892742322, + "grad_norm": 2.1345051661520693, + "learning_rate": 9.33804408215544e-06, + "loss": 0.691, + "step": 2788 + }, + { + "epoch": 0.19077912305903277, + "grad_norm": 1.703793805595464, + "learning_rate": 9.337493147531739e-06, + "loss": 0.4528, + "step": 2789 + }, + { + "epoch": 0.19084752719064232, + "grad_norm": 2.231042727959554, + "learning_rate": 9.336942000002931e-06, + "loss": 0.5402, + "step": 2790 + }, + { + "epoch": 0.19091593132225188, + "grad_norm": 1.8041662231236266, + "learning_rate": 9.336390639596066e-06, + "loss": 0.3576, + "step": 2791 + }, + { + "epoch": 0.19098433545386143, + "grad_norm": 2.1632382858049133, + "learning_rate": 9.335839066338209e-06, + "loss": 0.501, + "step": 2792 + }, + { + "epoch": 0.19105273958547095, + "grad_norm": 1.390846674106416, + "learning_rate": 9.33528728025643e-06, + "loss": 0.3011, + "step": 2793 + }, + { + "epoch": 0.1911211437170805, + "grad_norm": 1.512651985882751, + "learning_rate": 9.334735281377821e-06, + "loss": 0.2271, + "step": 2794 + }, + { + "epoch": 0.19118954784869005, + "grad_norm": 2.1431437939264515, + "learning_rate": 9.33418306972947e-06, + "loss": 0.5517, + "step": 2795 + }, + { + "epoch": 0.1912579519802996, + "grad_norm": 2.374684369295619, + "learning_rate": 9.333630645338484e-06, + "loss": 0.5415, + "step": 2796 + }, + { + "epoch": 0.19132635611190915, + "grad_norm": 2.415644420580257, + "learning_rate": 9.333078008231976e-06, + "loss": 0.3941, + "step": 2797 + }, + { + "epoch": 0.1913947602435187, + "grad_norm": 2.6863360601354844, + "learning_rate": 9.33252515843708e-06, + "loss": 0.7233, + "step": 2798 + }, + { + "epoch": 0.19146316437512825, + "grad_norm": 2.1807191383113618, + "learning_rate": 9.331972095980927e-06, + "loss": 0.7219, + "step": 2799 + }, + { + "epoch": 0.1915315685067378, + "grad_norm": 1.8052550162523153, + "learning_rate": 9.331418820890664e-06, + "loss": 0.4292, + "step": 2800 + }, + { + "epoch": 0.19159997263834735, + "grad_norm": 2.3450454946224517, + "learning_rate": 9.330865333193449e-06, + "loss": 0.4656, + "step": 2801 + }, + { + "epoch": 0.1916683767699569, + "grad_norm": 1.8976615293200867, + "learning_rate": 9.330311632916448e-06, + "loss": 0.4705, + "step": 2802 + }, + { + "epoch": 0.19173678090156646, + "grad_norm": 1.6743562948927138, + "learning_rate": 9.329757720086845e-06, + "loss": 0.3054, + "step": 2803 + }, + { + "epoch": 0.191805185033176, + "grad_norm": 2.8188160653841052, + "learning_rate": 9.329203594731821e-06, + "loss": 0.6023, + "step": 2804 + }, + { + "epoch": 0.19187358916478556, + "grad_norm": 2.010901388764432, + "learning_rate": 9.32864925687858e-06, + "loss": 0.3022, + "step": 2805 + }, + { + "epoch": 0.1919419932963951, + "grad_norm": 1.930635124996899, + "learning_rate": 9.328094706554332e-06, + "loss": 0.3382, + "step": 2806 + }, + { + "epoch": 0.19201039742800466, + "grad_norm": 1.9043946422030908, + "learning_rate": 9.327539943786295e-06, + "loss": 0.3799, + "step": 2807 + }, + { + "epoch": 0.1920788015596142, + "grad_norm": 2.3810191112630097, + "learning_rate": 9.326984968601701e-06, + "loss": 0.659, + "step": 2808 + }, + { + "epoch": 0.19214720569122376, + "grad_norm": 1.6802298313483306, + "learning_rate": 9.32642978102779e-06, + "loss": 0.2344, + "step": 2809 + }, + { + "epoch": 0.1922156098228333, + "grad_norm": 2.2736463925982, + "learning_rate": 9.32587438109181e-06, + "loss": 0.3113, + "step": 2810 + }, + { + "epoch": 0.19228401395444286, + "grad_norm": 2.369966392922643, + "learning_rate": 9.32531876882103e-06, + "loss": 0.2564, + "step": 2811 + }, + { + "epoch": 0.19235241808605238, + "grad_norm": 2.4197677674073272, + "learning_rate": 9.324762944242716e-06, + "loss": 0.518, + "step": 2812 + }, + { + "epoch": 0.19242082221766194, + "grad_norm": 1.853384803199068, + "learning_rate": 9.324206907384154e-06, + "loss": 0.5308, + "step": 2813 + }, + { + "epoch": 0.19248922634927149, + "grad_norm": 1.7857418998643764, + "learning_rate": 9.323650658272634e-06, + "loss": 0.3386, + "step": 2814 + }, + { + "epoch": 0.19255763048088104, + "grad_norm": 1.4863086002385795, + "learning_rate": 9.323094196935461e-06, + "loss": 0.1819, + "step": 2815 + }, + { + "epoch": 0.1926260346124906, + "grad_norm": 1.847998089449791, + "learning_rate": 9.322537523399951e-06, + "loss": 0.4434, + "step": 2816 + }, + { + "epoch": 0.19269443874410014, + "grad_norm": 2.0456520897306727, + "learning_rate": 9.321980637693423e-06, + "loss": 0.3396, + "step": 2817 + }, + { + "epoch": 0.1927628428757097, + "grad_norm": 1.576673113644762, + "learning_rate": 9.321423539843218e-06, + "loss": 0.2492, + "step": 2818 + }, + { + "epoch": 0.19283124700731924, + "grad_norm": 2.8483379265000606, + "learning_rate": 9.320866229876674e-06, + "loss": 0.4208, + "step": 2819 + }, + { + "epoch": 0.1928996511389288, + "grad_norm": 1.5722701315489447, + "learning_rate": 9.320308707821153e-06, + "loss": 0.1599, + "step": 2820 + }, + { + "epoch": 0.19296805527053834, + "grad_norm": 1.9693273197383898, + "learning_rate": 9.319750973704017e-06, + "loss": 0.3758, + "step": 2821 + }, + { + "epoch": 0.1930364594021479, + "grad_norm": 1.864919322304031, + "learning_rate": 9.319193027552643e-06, + "loss": 0.3557, + "step": 2822 + }, + { + "epoch": 0.19310486353375744, + "grad_norm": 1.9352750269721977, + "learning_rate": 9.318634869394418e-06, + "loss": 0.4693, + "step": 2823 + }, + { + "epoch": 0.193173267665367, + "grad_norm": 1.9348001820131215, + "learning_rate": 9.318076499256741e-06, + "loss": 0.3704, + "step": 2824 + }, + { + "epoch": 0.19324167179697654, + "grad_norm": 2.2479600632756167, + "learning_rate": 9.317517917167016e-06, + "loss": 0.6186, + "step": 2825 + }, + { + "epoch": 0.1933100759285861, + "grad_norm": 2.285340319661985, + "learning_rate": 9.316959123152662e-06, + "loss": 0.6945, + "step": 2826 + }, + { + "epoch": 0.19337848006019565, + "grad_norm": 2.234802952044442, + "learning_rate": 9.31640011724111e-06, + "loss": 0.3024, + "step": 2827 + }, + { + "epoch": 0.1934468841918052, + "grad_norm": 1.8712342514134752, + "learning_rate": 9.315840899459793e-06, + "loss": 0.4842, + "step": 2828 + }, + { + "epoch": 0.19351528832341475, + "grad_norm": 2.37423693796505, + "learning_rate": 9.315281469836169e-06, + "loss": 0.5998, + "step": 2829 + }, + { + "epoch": 0.19358369245502427, + "grad_norm": 1.9722530406143188, + "learning_rate": 9.314721828397687e-06, + "loss": 0.517, + "step": 2830 + }, + { + "epoch": 0.19365209658663382, + "grad_norm": 2.1890143738384076, + "learning_rate": 9.314161975171825e-06, + "loss": 0.511, + "step": 2831 + }, + { + "epoch": 0.19372050071824337, + "grad_norm": 2.3554234002146037, + "learning_rate": 9.313601910186058e-06, + "loss": 0.5533, + "step": 2832 + }, + { + "epoch": 0.19378890484985292, + "grad_norm": 2.302448267873291, + "learning_rate": 9.31304163346788e-06, + "loss": 0.6312, + "step": 2833 + }, + { + "epoch": 0.19385730898146247, + "grad_norm": 3.0153985407893624, + "learning_rate": 9.312481145044794e-06, + "loss": 0.4811, + "step": 2834 + }, + { + "epoch": 0.19392571311307202, + "grad_norm": 2.133853927080325, + "learning_rate": 9.311920444944306e-06, + "loss": 0.6209, + "step": 2835 + }, + { + "epoch": 0.19399411724468157, + "grad_norm": 2.5789449197157523, + "learning_rate": 9.311359533193942e-06, + "loss": 0.7571, + "step": 2836 + }, + { + "epoch": 0.19406252137629112, + "grad_norm": 1.8556241592775997, + "learning_rate": 9.310798409821232e-06, + "loss": 0.2396, + "step": 2837 + }, + { + "epoch": 0.19413092550790068, + "grad_norm": 2.2205381676059233, + "learning_rate": 9.310237074853717e-06, + "loss": 0.4054, + "step": 2838 + }, + { + "epoch": 0.19419932963951023, + "grad_norm": 1.9466138416573802, + "learning_rate": 9.309675528318955e-06, + "loss": 0.4449, + "step": 2839 + }, + { + "epoch": 0.19426773377111978, + "grad_norm": 2.3098070664464614, + "learning_rate": 9.309113770244507e-06, + "loss": 0.6473, + "step": 2840 + }, + { + "epoch": 0.19433613790272933, + "grad_norm": 2.1066690820391405, + "learning_rate": 9.308551800657946e-06, + "loss": 0.4716, + "step": 2841 + }, + { + "epoch": 0.19440454203433888, + "grad_norm": 1.929339877566727, + "learning_rate": 9.307989619586857e-06, + "loss": 0.3591, + "step": 2842 + }, + { + "epoch": 0.19447294616594843, + "grad_norm": 1.9996890705529906, + "learning_rate": 9.307427227058834e-06, + "loss": 0.3439, + "step": 2843 + }, + { + "epoch": 0.19454135029755798, + "grad_norm": 1.7505193231402663, + "learning_rate": 9.306864623101484e-06, + "loss": 0.2461, + "step": 2844 + }, + { + "epoch": 0.19460975442916753, + "grad_norm": 2.593011503910169, + "learning_rate": 9.306301807742416e-06, + "loss": 0.7841, + "step": 2845 + }, + { + "epoch": 0.19467815856077708, + "grad_norm": 1.830041504354089, + "learning_rate": 9.305738781009265e-06, + "loss": 0.3395, + "step": 2846 + }, + { + "epoch": 0.19474656269238663, + "grad_norm": 2.075937977460157, + "learning_rate": 9.305175542929663e-06, + "loss": 0.5236, + "step": 2847 + }, + { + "epoch": 0.19481496682399618, + "grad_norm": 1.2493788233718077, + "learning_rate": 9.304612093531254e-06, + "loss": 0.1594, + "step": 2848 + }, + { + "epoch": 0.1948833709556057, + "grad_norm": 1.742885569490264, + "learning_rate": 9.304048432841695e-06, + "loss": 0.4008, + "step": 2849 + }, + { + "epoch": 0.19495177508721526, + "grad_norm": 1.7659660364456768, + "learning_rate": 9.303484560888658e-06, + "loss": 0.4122, + "step": 2850 + }, + { + "epoch": 0.1950201792188248, + "grad_norm": 2.2103525607861956, + "learning_rate": 9.302920477699814e-06, + "loss": 0.3202, + "step": 2851 + }, + { + "epoch": 0.19508858335043436, + "grad_norm": 1.8071455420053302, + "learning_rate": 9.302356183302857e-06, + "loss": 0.3428, + "step": 2852 + }, + { + "epoch": 0.1951569874820439, + "grad_norm": 2.383251815696168, + "learning_rate": 9.301791677725483e-06, + "loss": 0.6093, + "step": 2853 + }, + { + "epoch": 0.19522539161365346, + "grad_norm": 2.4010393524710425, + "learning_rate": 9.301226960995399e-06, + "loss": 0.4707, + "step": 2854 + }, + { + "epoch": 0.195293795745263, + "grad_norm": 1.4196726516015996, + "learning_rate": 9.300662033140325e-06, + "loss": 0.1887, + "step": 2855 + }, + { + "epoch": 0.19536219987687256, + "grad_norm": 2.378875243637893, + "learning_rate": 9.300096894187991e-06, + "loss": 0.6725, + "step": 2856 + }, + { + "epoch": 0.1954306040084821, + "grad_norm": 2.05850055201312, + "learning_rate": 9.299531544166138e-06, + "loss": 0.6038, + "step": 2857 + }, + { + "epoch": 0.19549900814009166, + "grad_norm": 2.1729825475409275, + "learning_rate": 9.298965983102513e-06, + "loss": 0.5533, + "step": 2858 + }, + { + "epoch": 0.1955674122717012, + "grad_norm": 1.3717938263367884, + "learning_rate": 9.298400211024878e-06, + "loss": 0.1563, + "step": 2859 + }, + { + "epoch": 0.19563581640331076, + "grad_norm": 2.364548108738984, + "learning_rate": 9.297834227961005e-06, + "loss": 0.6167, + "step": 2860 + }, + { + "epoch": 0.19570422053492031, + "grad_norm": 2.7709574074082233, + "learning_rate": 9.297268033938673e-06, + "loss": 0.3873, + "step": 2861 + }, + { + "epoch": 0.19577262466652987, + "grad_norm": 2.3836053373492128, + "learning_rate": 9.296701628985675e-06, + "loss": 0.7724, + "step": 2862 + }, + { + "epoch": 0.19584102879813942, + "grad_norm": 3.012907760771198, + "learning_rate": 9.296135013129812e-06, + "loss": 0.3326, + "step": 2863 + }, + { + "epoch": 0.19590943292974897, + "grad_norm": 2.8962943284186675, + "learning_rate": 9.295568186398897e-06, + "loss": 0.6621, + "step": 2864 + }, + { + "epoch": 0.19597783706135852, + "grad_norm": 1.7127316269541277, + "learning_rate": 9.295001148820754e-06, + "loss": 0.3609, + "step": 2865 + }, + { + "epoch": 0.19604624119296807, + "grad_norm": 2.0119842946613193, + "learning_rate": 9.294433900423212e-06, + "loss": 0.4952, + "step": 2866 + }, + { + "epoch": 0.1961146453245776, + "grad_norm": 1.8330333506736358, + "learning_rate": 9.293866441234119e-06, + "loss": 0.4294, + "step": 2867 + }, + { + "epoch": 0.19618304945618714, + "grad_norm": 1.812924446360908, + "learning_rate": 9.293298771281325e-06, + "loss": 0.4543, + "step": 2868 + }, + { + "epoch": 0.1962514535877967, + "grad_norm": 1.4783438465605343, + "learning_rate": 9.292730890592694e-06, + "loss": 0.2378, + "step": 2869 + }, + { + "epoch": 0.19631985771940624, + "grad_norm": 2.397620512396756, + "learning_rate": 9.2921627991961e-06, + "loss": 0.6802, + "step": 2870 + }, + { + "epoch": 0.1963882618510158, + "grad_norm": 2.348670721243812, + "learning_rate": 9.291594497119433e-06, + "loss": 0.4468, + "step": 2871 + }, + { + "epoch": 0.19645666598262534, + "grad_norm": 2.1130496125537253, + "learning_rate": 9.291025984390583e-06, + "loss": 0.4886, + "step": 2872 + }, + { + "epoch": 0.1965250701142349, + "grad_norm": 2.0391953889068284, + "learning_rate": 9.290457261037456e-06, + "loss": 0.5356, + "step": 2873 + }, + { + "epoch": 0.19659347424584445, + "grad_norm": 2.6337422503595893, + "learning_rate": 9.289888327087967e-06, + "loss": 0.6995, + "step": 2874 + }, + { + "epoch": 0.196661878377454, + "grad_norm": 3.1032393148251036, + "learning_rate": 9.289319182570047e-06, + "loss": 0.2526, + "step": 2875 + }, + { + "epoch": 0.19673028250906355, + "grad_norm": 2.467416801692611, + "learning_rate": 9.288749827511625e-06, + "loss": 0.2816, + "step": 2876 + }, + { + "epoch": 0.1967986866406731, + "grad_norm": 2.604396959930401, + "learning_rate": 9.288180261940654e-06, + "loss": 0.5435, + "step": 2877 + }, + { + "epoch": 0.19686709077228265, + "grad_norm": 1.542419142559085, + "learning_rate": 9.287610485885086e-06, + "loss": 0.3102, + "step": 2878 + }, + { + "epoch": 0.1969354949038922, + "grad_norm": 2.1266141328842854, + "learning_rate": 9.287040499372893e-06, + "loss": 0.5751, + "step": 2879 + }, + { + "epoch": 0.19700389903550175, + "grad_norm": 1.6497591396252547, + "learning_rate": 9.286470302432046e-06, + "loss": 0.2352, + "step": 2880 + }, + { + "epoch": 0.1970723031671113, + "grad_norm": 1.949646140355533, + "learning_rate": 9.285899895090542e-06, + "loss": 0.4174, + "step": 2881 + }, + { + "epoch": 0.19714070729872085, + "grad_norm": 1.7117823600702071, + "learning_rate": 9.285329277376373e-06, + "loss": 0.4072, + "step": 2882 + }, + { + "epoch": 0.1972091114303304, + "grad_norm": 2.1631006106875703, + "learning_rate": 9.284758449317551e-06, + "loss": 0.52, + "step": 2883 + }, + { + "epoch": 0.19727751556193995, + "grad_norm": 1.6641191598263505, + "learning_rate": 9.284187410942091e-06, + "loss": 0.3987, + "step": 2884 + }, + { + "epoch": 0.1973459196935495, + "grad_norm": 2.7121243146385714, + "learning_rate": 9.283616162278025e-06, + "loss": 0.4441, + "step": 2885 + }, + { + "epoch": 0.19741432382515903, + "grad_norm": 2.574307328756592, + "learning_rate": 9.283044703353394e-06, + "loss": 0.6058, + "step": 2886 + }, + { + "epoch": 0.19748272795676858, + "grad_norm": 2.072577047721965, + "learning_rate": 9.282473034196244e-06, + "loss": 0.5158, + "step": 2887 + }, + { + "epoch": 0.19755113208837813, + "grad_norm": 2.2826358883508577, + "learning_rate": 9.28190115483464e-06, + "loss": 0.6501, + "step": 2888 + }, + { + "epoch": 0.19761953621998768, + "grad_norm": 1.7736382223802272, + "learning_rate": 9.28132906529665e-06, + "loss": 0.4595, + "step": 2889 + }, + { + "epoch": 0.19768794035159723, + "grad_norm": 2.7848510360943504, + "learning_rate": 9.280756765610354e-06, + "loss": 0.1684, + "step": 2890 + }, + { + "epoch": 0.19775634448320678, + "grad_norm": 2.2471007370329357, + "learning_rate": 9.280184255803844e-06, + "loss": 0.5044, + "step": 2891 + }, + { + "epoch": 0.19782474861481633, + "grad_norm": 2.0730190723050645, + "learning_rate": 9.279611535905223e-06, + "loss": 0.2657, + "step": 2892 + }, + { + "epoch": 0.19789315274642588, + "grad_norm": 1.885209192448715, + "learning_rate": 9.279038605942603e-06, + "loss": 0.322, + "step": 2893 + }, + { + "epoch": 0.19796155687803543, + "grad_norm": 2.1130222824695997, + "learning_rate": 9.278465465944102e-06, + "loss": 0.3838, + "step": 2894 + }, + { + "epoch": 0.19802996100964498, + "grad_norm": 2.3933378484799803, + "learning_rate": 9.277892115937856e-06, + "loss": 0.7031, + "step": 2895 + }, + { + "epoch": 0.19809836514125453, + "grad_norm": 1.6117038543057214, + "learning_rate": 9.277318555952009e-06, + "loss": 0.3963, + "step": 2896 + }, + { + "epoch": 0.19816676927286409, + "grad_norm": 2.4875283831505826, + "learning_rate": 9.276744786014712e-06, + "loss": 0.5546, + "step": 2897 + }, + { + "epoch": 0.19823517340447364, + "grad_norm": 2.723145795291404, + "learning_rate": 9.276170806154128e-06, + "loss": 0.4118, + "step": 2898 + }, + { + "epoch": 0.1983035775360832, + "grad_norm": 2.2350931980715916, + "learning_rate": 9.275596616398431e-06, + "loss": 0.5052, + "step": 2899 + }, + { + "epoch": 0.19837198166769274, + "grad_norm": 1.5475104344155626, + "learning_rate": 9.275022216775806e-06, + "loss": 0.1807, + "step": 2900 + }, + { + "epoch": 0.1984403857993023, + "grad_norm": 2.039195111226638, + "learning_rate": 9.274447607314447e-06, + "loss": 0.4505, + "step": 2901 + }, + { + "epoch": 0.19850878993091184, + "grad_norm": 2.4777873647461113, + "learning_rate": 9.273872788042559e-06, + "loss": 0.5598, + "step": 2902 + }, + { + "epoch": 0.1985771940625214, + "grad_norm": 2.0226077142095513, + "learning_rate": 9.273297758988355e-06, + "loss": 0.357, + "step": 2903 + }, + { + "epoch": 0.1986455981941309, + "grad_norm": 1.9986027180286552, + "learning_rate": 9.272722520180062e-06, + "loss": 0.4063, + "step": 2904 + }, + { + "epoch": 0.19871400232574046, + "grad_norm": 1.9518726224637462, + "learning_rate": 9.272147071645913e-06, + "loss": 0.5555, + "step": 2905 + }, + { + "epoch": 0.19878240645735, + "grad_norm": 2.16926528377365, + "learning_rate": 9.271571413414157e-06, + "loss": 0.3826, + "step": 2906 + }, + { + "epoch": 0.19885081058895956, + "grad_norm": 1.9851910347501285, + "learning_rate": 9.270995545513051e-06, + "loss": 0.3946, + "step": 2907 + }, + { + "epoch": 0.19891921472056912, + "grad_norm": 1.761385604925556, + "learning_rate": 9.270419467970857e-06, + "loss": 0.3465, + "step": 2908 + }, + { + "epoch": 0.19898761885217867, + "grad_norm": 1.6629686342221597, + "learning_rate": 9.269843180815854e-06, + "loss": 0.3514, + "step": 2909 + }, + { + "epoch": 0.19905602298378822, + "grad_norm": 1.6648470412695728, + "learning_rate": 9.26926668407633e-06, + "loss": 0.3173, + "step": 2910 + }, + { + "epoch": 0.19912442711539777, + "grad_norm": 2.2247241679675644, + "learning_rate": 9.26868997778058e-06, + "loss": 0.7025, + "step": 2911 + }, + { + "epoch": 0.19919283124700732, + "grad_norm": 1.8819655693947945, + "learning_rate": 9.268113061956914e-06, + "loss": 0.4421, + "step": 2912 + }, + { + "epoch": 0.19926123537861687, + "grad_norm": 2.0694130293135027, + "learning_rate": 9.267535936633646e-06, + "loss": 0.4862, + "step": 2913 + }, + { + "epoch": 0.19932963951022642, + "grad_norm": 1.7451186262828189, + "learning_rate": 9.266958601839108e-06, + "loss": 0.3811, + "step": 2914 + }, + { + "epoch": 0.19939804364183597, + "grad_norm": 1.8788051538372963, + "learning_rate": 9.266381057601637e-06, + "loss": 0.4192, + "step": 2915 + }, + { + "epoch": 0.19946644777344552, + "grad_norm": 2.0124215833420194, + "learning_rate": 9.26580330394958e-06, + "loss": 0.4069, + "step": 2916 + }, + { + "epoch": 0.19953485190505507, + "grad_norm": 1.3116669026046008, + "learning_rate": 9.265225340911299e-06, + "loss": 0.2222, + "step": 2917 + }, + { + "epoch": 0.19960325603666462, + "grad_norm": 1.8528966965688267, + "learning_rate": 9.264647168515161e-06, + "loss": 0.3176, + "step": 2918 + }, + { + "epoch": 0.19967166016827417, + "grad_norm": 2.666681501352254, + "learning_rate": 9.264068786789546e-06, + "loss": 0.4684, + "step": 2919 + }, + { + "epoch": 0.19974006429988372, + "grad_norm": 2.260553827533805, + "learning_rate": 9.263490195762846e-06, + "loss": 0.7131, + "step": 2920 + }, + { + "epoch": 0.19980846843149327, + "grad_norm": 1.870581152510716, + "learning_rate": 9.262911395463456e-06, + "loss": 0.3588, + "step": 2921 + }, + { + "epoch": 0.1998768725631028, + "grad_norm": 1.7651375122912722, + "learning_rate": 9.262332385919792e-06, + "loss": 0.4396, + "step": 2922 + }, + { + "epoch": 0.19994527669471235, + "grad_norm": 1.7168871885116304, + "learning_rate": 9.261753167160271e-06, + "loss": 0.2323, + "step": 2923 + }, + { + "epoch": 0.2000136808263219, + "grad_norm": 2.2607199879950817, + "learning_rate": 9.261173739213325e-06, + "loss": 0.6954, + "step": 2924 + }, + { + "epoch": 0.20008208495793145, + "grad_norm": 1.8686175855240315, + "learning_rate": 9.260594102107394e-06, + "loss": 0.3255, + "step": 2925 + }, + { + "epoch": 0.200150489089541, + "grad_norm": 2.12351272230826, + "learning_rate": 9.260014255870931e-06, + "loss": 0.5383, + "step": 2926 + }, + { + "epoch": 0.20021889322115055, + "grad_norm": 2.4819252603898683, + "learning_rate": 9.259434200532395e-06, + "loss": 0.8143, + "step": 2927 + }, + { + "epoch": 0.2002872973527601, + "grad_norm": 1.68049022212159, + "learning_rate": 9.258853936120263e-06, + "loss": 0.1817, + "step": 2928 + }, + { + "epoch": 0.20035570148436965, + "grad_norm": 1.9945164400141822, + "learning_rate": 9.258273462663013e-06, + "loss": 0.4862, + "step": 2929 + }, + { + "epoch": 0.2004241056159792, + "grad_norm": 2.2333721397863253, + "learning_rate": 9.257692780189136e-06, + "loss": 0.7175, + "step": 2930 + }, + { + "epoch": 0.20049250974758875, + "grad_norm": 2.227812452986491, + "learning_rate": 9.25711188872714e-06, + "loss": 0.7099, + "step": 2931 + }, + { + "epoch": 0.2005609138791983, + "grad_norm": 1.7202361184330544, + "learning_rate": 9.256530788305533e-06, + "loss": 0.3354, + "step": 2932 + }, + { + "epoch": 0.20062931801080786, + "grad_norm": 1.9495370371840344, + "learning_rate": 9.255949478952843e-06, + "loss": 0.5645, + "step": 2933 + }, + { + "epoch": 0.2006977221424174, + "grad_norm": 2.160325131704272, + "learning_rate": 9.255367960697599e-06, + "loss": 0.3561, + "step": 2934 + }, + { + "epoch": 0.20076612627402696, + "grad_norm": 1.6141435660050238, + "learning_rate": 9.254786233568347e-06, + "loss": 0.3315, + "step": 2935 + }, + { + "epoch": 0.2008345304056365, + "grad_norm": 2.076166766600905, + "learning_rate": 9.254204297593641e-06, + "loss": 0.444, + "step": 2936 + }, + { + "epoch": 0.20090293453724606, + "grad_norm": 1.6907199498024197, + "learning_rate": 9.253622152802042e-06, + "loss": 0.2776, + "step": 2937 + }, + { + "epoch": 0.2009713386688556, + "grad_norm": 2.128119615584408, + "learning_rate": 9.253039799222131e-06, + "loss": 0.5485, + "step": 2938 + }, + { + "epoch": 0.20103974280046516, + "grad_norm": 1.6691302581508025, + "learning_rate": 9.252457236882487e-06, + "loss": 0.3003, + "step": 2939 + }, + { + "epoch": 0.2011081469320747, + "grad_norm": 1.9380502457827176, + "learning_rate": 9.251874465811708e-06, + "loss": 0.5378, + "step": 2940 + }, + { + "epoch": 0.20117655106368423, + "grad_norm": 2.280355362982674, + "learning_rate": 9.251291486038396e-06, + "loss": 0.6945, + "step": 2941 + }, + { + "epoch": 0.20124495519529378, + "grad_norm": 1.4566730744237673, + "learning_rate": 9.25070829759117e-06, + "loss": 0.2573, + "step": 2942 + }, + { + "epoch": 0.20131335932690334, + "grad_norm": 1.6611950653539276, + "learning_rate": 9.250124900498653e-06, + "loss": 0.2333, + "step": 2943 + }, + { + "epoch": 0.20138176345851289, + "grad_norm": 2.325620161429264, + "learning_rate": 9.249541294789484e-06, + "loss": 0.6103, + "step": 2944 + }, + { + "epoch": 0.20145016759012244, + "grad_norm": 1.7129282234963175, + "learning_rate": 9.248957480492307e-06, + "loss": 0.3835, + "step": 2945 + }, + { + "epoch": 0.201518571721732, + "grad_norm": 2.2280811284094484, + "learning_rate": 9.24837345763578e-06, + "loss": 0.6837, + "step": 2946 + }, + { + "epoch": 0.20158697585334154, + "grad_norm": 2.380936051057975, + "learning_rate": 9.247789226248566e-06, + "loss": 0.313, + "step": 2947 + }, + { + "epoch": 0.2016553799849511, + "grad_norm": 2.184560496922535, + "learning_rate": 9.247204786359345e-06, + "loss": 0.459, + "step": 2948 + }, + { + "epoch": 0.20172378411656064, + "grad_norm": 2.3556993792282097, + "learning_rate": 9.246620137996804e-06, + "loss": 0.6596, + "step": 2949 + }, + { + "epoch": 0.2017921882481702, + "grad_norm": 1.8532198035771064, + "learning_rate": 9.246035281189641e-06, + "loss": 0.283, + "step": 2950 + }, + { + "epoch": 0.20186059237977974, + "grad_norm": 2.6784756647884085, + "learning_rate": 9.245450215966562e-06, + "loss": 0.838, + "step": 2951 + }, + { + "epoch": 0.2019289965113893, + "grad_norm": 1.9987798783751216, + "learning_rate": 9.244864942356284e-06, + "loss": 0.2992, + "step": 2952 + }, + { + "epoch": 0.20199740064299884, + "grad_norm": 2.4888774391148187, + "learning_rate": 9.244279460387538e-06, + "loss": 0.7798, + "step": 2953 + }, + { + "epoch": 0.2020658047746084, + "grad_norm": 2.378307279651486, + "learning_rate": 9.243693770089059e-06, + "loss": 0.6821, + "step": 2954 + }, + { + "epoch": 0.20213420890621794, + "grad_norm": 2.1336909849903325, + "learning_rate": 9.243107871489598e-06, + "loss": 0.6424, + "step": 2955 + }, + { + "epoch": 0.2022026130378275, + "grad_norm": 1.6128781608281018, + "learning_rate": 9.242521764617913e-06, + "loss": 0.214, + "step": 2956 + }, + { + "epoch": 0.20227101716943705, + "grad_norm": 1.82603019123929, + "learning_rate": 9.241935449502773e-06, + "loss": 0.4584, + "step": 2957 + }, + { + "epoch": 0.2023394213010466, + "grad_norm": 1.7853501462915136, + "learning_rate": 9.241348926172957e-06, + "loss": 0.3921, + "step": 2958 + }, + { + "epoch": 0.20240782543265612, + "grad_norm": 2.0932830887761003, + "learning_rate": 9.240762194657254e-06, + "loss": 0.5496, + "step": 2959 + }, + { + "epoch": 0.20247622956426567, + "grad_norm": 1.687731457660818, + "learning_rate": 9.240175254984464e-06, + "loss": 0.3698, + "step": 2960 + }, + { + "epoch": 0.20254463369587522, + "grad_norm": 1.7521762684529636, + "learning_rate": 9.2395881071834e-06, + "loss": 0.2976, + "step": 2961 + }, + { + "epoch": 0.20261303782748477, + "grad_norm": 1.767078476569114, + "learning_rate": 9.239000751282873e-06, + "loss": 0.3994, + "step": 2962 + }, + { + "epoch": 0.20268144195909432, + "grad_norm": 2.1707497364344093, + "learning_rate": 9.238413187311723e-06, + "loss": 0.4609, + "step": 2963 + }, + { + "epoch": 0.20274984609070387, + "grad_norm": 2.7266303275342265, + "learning_rate": 9.237825415298787e-06, + "loss": 0.586, + "step": 2964 + }, + { + "epoch": 0.20281825022231342, + "grad_norm": 1.8032304157799768, + "learning_rate": 9.237237435272915e-06, + "loss": 0.3222, + "step": 2965 + }, + { + "epoch": 0.20288665435392297, + "grad_norm": 2.035327779676522, + "learning_rate": 9.236649247262968e-06, + "loss": 0.4831, + "step": 2966 + }, + { + "epoch": 0.20295505848553252, + "grad_norm": 2.067287163555737, + "learning_rate": 9.236060851297816e-06, + "loss": 0.3938, + "step": 2967 + }, + { + "epoch": 0.20302346261714208, + "grad_norm": 2.1177937741047677, + "learning_rate": 9.235472247406343e-06, + "loss": 0.4919, + "step": 2968 + }, + { + "epoch": 0.20309186674875163, + "grad_norm": 2.400164799197904, + "learning_rate": 9.23488343561744e-06, + "loss": 0.686, + "step": 2969 + }, + { + "epoch": 0.20316027088036118, + "grad_norm": 2.9124839843669843, + "learning_rate": 9.234294415960006e-06, + "loss": 0.7182, + "step": 2970 + }, + { + "epoch": 0.20322867501197073, + "grad_norm": 2.0701949680003526, + "learning_rate": 9.233705188462955e-06, + "loss": 0.372, + "step": 2971 + }, + { + "epoch": 0.20329707914358028, + "grad_norm": 1.7136817505766926, + "learning_rate": 9.23311575315521e-06, + "loss": 0.396, + "step": 2972 + }, + { + "epoch": 0.20336548327518983, + "grad_norm": 2.439833019720369, + "learning_rate": 9.232526110065702e-06, + "loss": 0.4797, + "step": 2973 + }, + { + "epoch": 0.20343388740679938, + "grad_norm": 1.924186711851936, + "learning_rate": 9.231936259223375e-06, + "loss": 0.286, + "step": 2974 + }, + { + "epoch": 0.20350229153840893, + "grad_norm": 1.9567251365780982, + "learning_rate": 9.231346200657179e-06, + "loss": 0.5512, + "step": 2975 + }, + { + "epoch": 0.20357069567001848, + "grad_norm": 2.4209005795154277, + "learning_rate": 9.230755934396081e-06, + "loss": 0.5032, + "step": 2976 + }, + { + "epoch": 0.20363909980162803, + "grad_norm": 2.8719704152425725, + "learning_rate": 9.230165460469048e-06, + "loss": 0.5425, + "step": 2977 + }, + { + "epoch": 0.20370750393323755, + "grad_norm": 1.4498953679270827, + "learning_rate": 9.229574778905069e-06, + "loss": 0.2059, + "step": 2978 + }, + { + "epoch": 0.2037759080648471, + "grad_norm": 2.0996539763215596, + "learning_rate": 9.228983889733135e-06, + "loss": 0.5679, + "step": 2979 + }, + { + "epoch": 0.20384431219645666, + "grad_norm": 4.498310980784858, + "learning_rate": 9.22839279298225e-06, + "loss": 0.4337, + "step": 2980 + }, + { + "epoch": 0.2039127163280662, + "grad_norm": 2.560195916258231, + "learning_rate": 9.22780148868143e-06, + "loss": 0.8927, + "step": 2981 + }, + { + "epoch": 0.20398112045967576, + "grad_norm": 2.506532800087337, + "learning_rate": 9.227209976859695e-06, + "loss": 0.6037, + "step": 2982 + }, + { + "epoch": 0.2040495245912853, + "grad_norm": 2.10736856755529, + "learning_rate": 9.22661825754608e-06, + "loss": 0.257, + "step": 2983 + }, + { + "epoch": 0.20411792872289486, + "grad_norm": 2.2549871434608595, + "learning_rate": 9.226026330769633e-06, + "loss": 0.5763, + "step": 2984 + }, + { + "epoch": 0.2041863328545044, + "grad_norm": 2.2413510280865556, + "learning_rate": 9.225434196559406e-06, + "loss": 0.4744, + "step": 2985 + }, + { + "epoch": 0.20425473698611396, + "grad_norm": 1.983428793317534, + "learning_rate": 9.224841854944464e-06, + "loss": 0.3935, + "step": 2986 + }, + { + "epoch": 0.2043231411177235, + "grad_norm": 2.096122450886205, + "learning_rate": 9.224249305953882e-06, + "loss": 0.4928, + "step": 2987 + }, + { + "epoch": 0.20439154524933306, + "grad_norm": 2.3161553026615818, + "learning_rate": 9.223656549616746e-06, + "loss": 0.4784, + "step": 2988 + }, + { + "epoch": 0.2044599493809426, + "grad_norm": 2.3876048595558994, + "learning_rate": 9.223063585962152e-06, + "loss": 0.4007, + "step": 2989 + }, + { + "epoch": 0.20452835351255216, + "grad_norm": 2.6002099126662204, + "learning_rate": 9.222470415019202e-06, + "loss": 0.6801, + "step": 2990 + }, + { + "epoch": 0.20459675764416171, + "grad_norm": 1.840316882720162, + "learning_rate": 9.221877036817015e-06, + "loss": 0.2334, + "step": 2991 + }, + { + "epoch": 0.20466516177577126, + "grad_norm": 2.1362324218741073, + "learning_rate": 9.221283451384714e-06, + "loss": 0.4292, + "step": 2992 + }, + { + "epoch": 0.20473356590738082, + "grad_norm": 1.9814821806702496, + "learning_rate": 9.220689658751438e-06, + "loss": 0.4632, + "step": 2993 + }, + { + "epoch": 0.20480197003899037, + "grad_norm": 2.993017776175348, + "learning_rate": 9.220095658946333e-06, + "loss": 0.4053, + "step": 2994 + }, + { + "epoch": 0.20487037417059992, + "grad_norm": 2.3672490662729944, + "learning_rate": 9.219501451998554e-06, + "loss": 0.4039, + "step": 2995 + }, + { + "epoch": 0.20493877830220944, + "grad_norm": 1.808716763624043, + "learning_rate": 9.218907037937267e-06, + "loss": 0.3158, + "step": 2996 + }, + { + "epoch": 0.205007182433819, + "grad_norm": 1.8333774228556285, + "learning_rate": 9.218312416791651e-06, + "loss": 0.3221, + "step": 2997 + }, + { + "epoch": 0.20507558656542854, + "grad_norm": 2.02679619165987, + "learning_rate": 9.21771758859089e-06, + "loss": 0.3653, + "step": 2998 + }, + { + "epoch": 0.2051439906970381, + "grad_norm": 2.300506769721275, + "learning_rate": 9.217122553364184e-06, + "loss": 0.7248, + "step": 2999 + }, + { + "epoch": 0.20521239482864764, + "grad_norm": 2.5388506756969056, + "learning_rate": 9.216527311140737e-06, + "loss": 0.507, + "step": 3000 + }, + { + "epoch": 0.2052807989602572, + "grad_norm": 1.8512960596362662, + "learning_rate": 9.21593186194977e-06, + "loss": 0.346, + "step": 3001 + }, + { + "epoch": 0.20534920309186674, + "grad_norm": 2.142928867615643, + "learning_rate": 9.215336205820509e-06, + "loss": 0.2521, + "step": 3002 + }, + { + "epoch": 0.2054176072234763, + "grad_norm": 2.3088957561588237, + "learning_rate": 9.21474034278219e-06, + "loss": 0.6227, + "step": 3003 + }, + { + "epoch": 0.20548601135508585, + "grad_norm": 2.5942818027538346, + "learning_rate": 9.214144272864063e-06, + "loss": 0.5762, + "step": 3004 + }, + { + "epoch": 0.2055544154866954, + "grad_norm": 3.214077697183883, + "learning_rate": 9.213547996095385e-06, + "loss": 0.4569, + "step": 3005 + }, + { + "epoch": 0.20562281961830495, + "grad_norm": 2.7352218569939555, + "learning_rate": 9.212951512505424e-06, + "loss": 0.7376, + "step": 3006 + }, + { + "epoch": 0.2056912237499145, + "grad_norm": 2.070571843583876, + "learning_rate": 9.21235482212346e-06, + "loss": 0.422, + "step": 3007 + }, + { + "epoch": 0.20575962788152405, + "grad_norm": 2.074144645847086, + "learning_rate": 9.211757924978779e-06, + "loss": 0.4214, + "step": 3008 + }, + { + "epoch": 0.2058280320131336, + "grad_norm": 2.7948263433245675, + "learning_rate": 9.21116082110068e-06, + "loss": 0.6386, + "step": 3009 + }, + { + "epoch": 0.20589643614474315, + "grad_norm": 2.7525786209484946, + "learning_rate": 9.210563510518473e-06, + "loss": 0.9305, + "step": 3010 + }, + { + "epoch": 0.2059648402763527, + "grad_norm": 2.313166090686108, + "learning_rate": 9.209965993261475e-06, + "loss": 0.394, + "step": 3011 + }, + { + "epoch": 0.20603324440796225, + "grad_norm": 1.8752580465132187, + "learning_rate": 9.209368269359018e-06, + "loss": 0.5259, + "step": 3012 + }, + { + "epoch": 0.2061016485395718, + "grad_norm": 1.5324409327609179, + "learning_rate": 9.20877033884044e-06, + "loss": 0.1632, + "step": 3013 + }, + { + "epoch": 0.20617005267118135, + "grad_norm": 1.9113443927084297, + "learning_rate": 9.208172201735089e-06, + "loss": 0.3027, + "step": 3014 + }, + { + "epoch": 0.20623845680279088, + "grad_norm": 1.7009076286607072, + "learning_rate": 9.207573858072325e-06, + "loss": 0.1805, + "step": 3015 + }, + { + "epoch": 0.20630686093440043, + "grad_norm": 2.0172987815632055, + "learning_rate": 9.206975307881518e-06, + "loss": 0.4126, + "step": 3016 + }, + { + "epoch": 0.20637526506600998, + "grad_norm": 2.177531440756425, + "learning_rate": 9.206376551192048e-06, + "loss": 0.4368, + "step": 3017 + }, + { + "epoch": 0.20644366919761953, + "grad_norm": 2.3385616349781624, + "learning_rate": 9.205777588033304e-06, + "loss": 0.5469, + "step": 3018 + }, + { + "epoch": 0.20651207332922908, + "grad_norm": 2.5548537139017045, + "learning_rate": 9.205178418434687e-06, + "loss": 0.6337, + "step": 3019 + }, + { + "epoch": 0.20658047746083863, + "grad_norm": 2.170207776480964, + "learning_rate": 9.204579042425606e-06, + "loss": 0.72, + "step": 3020 + }, + { + "epoch": 0.20664888159244818, + "grad_norm": 1.6057080046498355, + "learning_rate": 9.203979460035484e-06, + "loss": 0.2924, + "step": 3021 + }, + { + "epoch": 0.20671728572405773, + "grad_norm": 1.6252804752532768, + "learning_rate": 9.203379671293746e-06, + "loss": 0.2404, + "step": 3022 + }, + { + "epoch": 0.20678568985566728, + "grad_norm": 1.9248806586676288, + "learning_rate": 9.202779676229838e-06, + "loss": 0.4365, + "step": 3023 + }, + { + "epoch": 0.20685409398727683, + "grad_norm": 2.1071974252101886, + "learning_rate": 9.202179474873207e-06, + "loss": 0.3958, + "step": 3024 + }, + { + "epoch": 0.20692249811888638, + "grad_norm": 3.030647306401471, + "learning_rate": 9.201579067253316e-06, + "loss": 0.4848, + "step": 3025 + }, + { + "epoch": 0.20699090225049593, + "grad_norm": 2.18679916190375, + "learning_rate": 9.200978453399635e-06, + "loss": 0.5527, + "step": 3026 + }, + { + "epoch": 0.20705930638210548, + "grad_norm": 2.043869170234657, + "learning_rate": 9.200377633341645e-06, + "loss": 0.4775, + "step": 3027 + }, + { + "epoch": 0.20712771051371504, + "grad_norm": 1.9256641630903228, + "learning_rate": 9.19977660710884e-06, + "loss": 0.3952, + "step": 3028 + }, + { + "epoch": 0.20719611464532459, + "grad_norm": 1.4222119398496118, + "learning_rate": 9.199175374730715e-06, + "loss": 0.2135, + "step": 3029 + }, + { + "epoch": 0.20726451877693414, + "grad_norm": 1.8226999948797522, + "learning_rate": 9.198573936236787e-06, + "loss": 0.3911, + "step": 3030 + }, + { + "epoch": 0.2073329229085437, + "grad_norm": 1.3688342803325502, + "learning_rate": 9.197972291656575e-06, + "loss": 0.1818, + "step": 3031 + }, + { + "epoch": 0.20740132704015324, + "grad_norm": 2.2201453978913386, + "learning_rate": 9.19737044101961e-06, + "loss": 0.5217, + "step": 3032 + }, + { + "epoch": 0.20746973117176276, + "grad_norm": 2.035369254184431, + "learning_rate": 9.196768384355437e-06, + "loss": 0.5608, + "step": 3033 + }, + { + "epoch": 0.2075381353033723, + "grad_norm": 2.1612477433593864, + "learning_rate": 9.196166121693603e-06, + "loss": 0.378, + "step": 3034 + }, + { + "epoch": 0.20760653943498186, + "grad_norm": 1.8283076887873653, + "learning_rate": 9.195563653063676e-06, + "loss": 0.3954, + "step": 3035 + }, + { + "epoch": 0.2076749435665914, + "grad_norm": 1.6530976277881375, + "learning_rate": 9.194960978495222e-06, + "loss": 0.2025, + "step": 3036 + }, + { + "epoch": 0.20774334769820096, + "grad_norm": 2.2871405666970537, + "learning_rate": 9.194358098017829e-06, + "loss": 0.3532, + "step": 3037 + }, + { + "epoch": 0.20781175182981051, + "grad_norm": 1.9509131534028192, + "learning_rate": 9.193755011661082e-06, + "loss": 0.4244, + "step": 3038 + }, + { + "epoch": 0.20788015596142007, + "grad_norm": 2.2496254463502705, + "learning_rate": 9.193151719454591e-06, + "loss": 0.3753, + "step": 3039 + }, + { + "epoch": 0.20794856009302962, + "grad_norm": 1.7502692560412494, + "learning_rate": 9.192548221427963e-06, + "loss": 0.4039, + "step": 3040 + }, + { + "epoch": 0.20801696422463917, + "grad_norm": 2.1464923237175286, + "learning_rate": 9.191944517610824e-06, + "loss": 0.5115, + "step": 3041 + }, + { + "epoch": 0.20808536835624872, + "grad_norm": 1.7394101109937736, + "learning_rate": 9.191340608032805e-06, + "loss": 0.4283, + "step": 3042 + }, + { + "epoch": 0.20815377248785827, + "grad_norm": 2.4332266908840405, + "learning_rate": 9.190736492723548e-06, + "loss": 0.673, + "step": 3043 + }, + { + "epoch": 0.20822217661946782, + "grad_norm": 2.1539739676295317, + "learning_rate": 9.190132171712709e-06, + "loss": 0.4808, + "step": 3044 + }, + { + "epoch": 0.20829058075107737, + "grad_norm": 2.6458414739698797, + "learning_rate": 9.189527645029948e-06, + "loss": 0.842, + "step": 3045 + }, + { + "epoch": 0.20835898488268692, + "grad_norm": 1.6806476355043019, + "learning_rate": 9.188922912704938e-06, + "loss": 0.3309, + "step": 3046 + }, + { + "epoch": 0.20842738901429647, + "grad_norm": 2.181383983574107, + "learning_rate": 9.188317974767367e-06, + "loss": 0.4042, + "step": 3047 + }, + { + "epoch": 0.20849579314590602, + "grad_norm": 2.0597604236070493, + "learning_rate": 9.187712831246921e-06, + "loss": 0.5169, + "step": 3048 + }, + { + "epoch": 0.20856419727751557, + "grad_norm": 1.9480454041953739, + "learning_rate": 9.187107482173308e-06, + "loss": 0.3867, + "step": 3049 + }, + { + "epoch": 0.20863260140912512, + "grad_norm": 2.1418625393837027, + "learning_rate": 9.186501927576242e-06, + "loss": 0.3969, + "step": 3050 + }, + { + "epoch": 0.20870100554073467, + "grad_norm": 1.6319569212411766, + "learning_rate": 9.185896167485443e-06, + "loss": 0.2058, + "step": 3051 + }, + { + "epoch": 0.2087694096723442, + "grad_norm": 2.2323075913540493, + "learning_rate": 9.185290201930648e-06, + "loss": 0.3278, + "step": 3052 + }, + { + "epoch": 0.20883781380395375, + "grad_norm": 1.2183628537200841, + "learning_rate": 9.184684030941597e-06, + "loss": 0.242, + "step": 3053 + }, + { + "epoch": 0.2089062179355633, + "grad_norm": 2.105583527502779, + "learning_rate": 9.184077654548048e-06, + "loss": 0.479, + "step": 3054 + }, + { + "epoch": 0.20897462206717285, + "grad_norm": 2.1381689586430372, + "learning_rate": 9.183471072779764e-06, + "loss": 0.631, + "step": 3055 + }, + { + "epoch": 0.2090430261987824, + "grad_norm": 1.8579473183483117, + "learning_rate": 9.182864285666516e-06, + "loss": 0.1868, + "step": 3056 + }, + { + "epoch": 0.20911143033039195, + "grad_norm": 2.7558460951714276, + "learning_rate": 9.182257293238091e-06, + "loss": 0.9268, + "step": 3057 + }, + { + "epoch": 0.2091798344620015, + "grad_norm": 1.9372230339514291, + "learning_rate": 9.181650095524282e-06, + "loss": 0.4947, + "step": 3058 + }, + { + "epoch": 0.20924823859361105, + "grad_norm": 1.5962466991483035, + "learning_rate": 9.181042692554894e-06, + "loss": 0.1675, + "step": 3059 + }, + { + "epoch": 0.2093166427252206, + "grad_norm": 1.679217885039268, + "learning_rate": 9.180435084359738e-06, + "loss": 0.3412, + "step": 3060 + }, + { + "epoch": 0.20938504685683015, + "grad_norm": 1.8707028659606106, + "learning_rate": 9.179827270968644e-06, + "loss": 0.4426, + "step": 3061 + }, + { + "epoch": 0.2094534509884397, + "grad_norm": 2.3156134388901455, + "learning_rate": 9.179219252411441e-06, + "loss": 0.605, + "step": 3062 + }, + { + "epoch": 0.20952185512004926, + "grad_norm": 3.319400617256068, + "learning_rate": 9.178611028717978e-06, + "loss": 0.4197, + "step": 3063 + }, + { + "epoch": 0.2095902592516588, + "grad_norm": 1.8245886441953691, + "learning_rate": 9.178002599918106e-06, + "loss": 0.1856, + "step": 3064 + }, + { + "epoch": 0.20965866338326836, + "grad_norm": 2.052563011355685, + "learning_rate": 9.177393966041692e-06, + "loss": 0.3522, + "step": 3065 + }, + { + "epoch": 0.2097270675148779, + "grad_norm": 2.0440893801362376, + "learning_rate": 9.176785127118607e-06, + "loss": 0.4962, + "step": 3066 + }, + { + "epoch": 0.20979547164648746, + "grad_norm": 2.2236477294743193, + "learning_rate": 9.176176083178744e-06, + "loss": 0.6689, + "step": 3067 + }, + { + "epoch": 0.209863875778097, + "grad_norm": 1.9578029534913175, + "learning_rate": 9.175566834251987e-06, + "loss": 0.4973, + "step": 3068 + }, + { + "epoch": 0.20993227990970656, + "grad_norm": 1.8681796881365724, + "learning_rate": 9.174957380368248e-06, + "loss": 0.325, + "step": 3069 + }, + { + "epoch": 0.21000068404131608, + "grad_norm": 1.8645882615875844, + "learning_rate": 9.17434772155744e-06, + "loss": 0.2131, + "step": 3070 + }, + { + "epoch": 0.21006908817292563, + "grad_norm": 1.6681467626845015, + "learning_rate": 9.17373785784949e-06, + "loss": 0.3115, + "step": 3071 + }, + { + "epoch": 0.21013749230453518, + "grad_norm": 1.754106480441087, + "learning_rate": 9.173127789274329e-06, + "loss": 0.3838, + "step": 3072 + }, + { + "epoch": 0.21020589643614473, + "grad_norm": 2.211224143112671, + "learning_rate": 9.172517515861903e-06, + "loss": 0.6015, + "step": 3073 + }, + { + "epoch": 0.21027430056775429, + "grad_norm": 1.9871232420923208, + "learning_rate": 9.171907037642171e-06, + "loss": 0.5284, + "step": 3074 + }, + { + "epoch": 0.21034270469936384, + "grad_norm": 2.64831739057656, + "learning_rate": 9.171296354645094e-06, + "loss": 0.3742, + "step": 3075 + }, + { + "epoch": 0.2104111088309734, + "grad_norm": 1.6130339268970113, + "learning_rate": 9.17068546690065e-06, + "loss": 0.3461, + "step": 3076 + }, + { + "epoch": 0.21047951296258294, + "grad_norm": 2.045842179501346, + "learning_rate": 9.170074374438825e-06, + "loss": 0.3319, + "step": 3077 + }, + { + "epoch": 0.2105479170941925, + "grad_norm": 1.6434614661667653, + "learning_rate": 9.169463077289608e-06, + "loss": 0.2568, + "step": 3078 + }, + { + "epoch": 0.21061632122580204, + "grad_norm": 1.3532141257484709, + "learning_rate": 9.168851575483013e-06, + "loss": 0.1814, + "step": 3079 + }, + { + "epoch": 0.2106847253574116, + "grad_norm": 2.0304262911975983, + "learning_rate": 9.168239869049051e-06, + "loss": 0.4178, + "step": 3080 + }, + { + "epoch": 0.21075312948902114, + "grad_norm": 2.1379917718179895, + "learning_rate": 9.167627958017747e-06, + "loss": 0.6301, + "step": 3081 + }, + { + "epoch": 0.2108215336206307, + "grad_norm": 1.6066897128095825, + "learning_rate": 9.167015842419137e-06, + "loss": 0.3287, + "step": 3082 + }, + { + "epoch": 0.21088993775224024, + "grad_norm": 1.8545084948071338, + "learning_rate": 9.166403522283269e-06, + "loss": 0.1535, + "step": 3083 + }, + { + "epoch": 0.2109583418838498, + "grad_norm": 2.068109841283981, + "learning_rate": 9.165790997640193e-06, + "loss": 0.4031, + "step": 3084 + }, + { + "epoch": 0.21102674601545934, + "grad_norm": 1.7790936794728254, + "learning_rate": 9.165178268519983e-06, + "loss": 0.4245, + "step": 3085 + }, + { + "epoch": 0.2110951501470689, + "grad_norm": 2.2679200748341137, + "learning_rate": 9.164565334952709e-06, + "loss": 0.5391, + "step": 3086 + }, + { + "epoch": 0.21116355427867844, + "grad_norm": 1.7155110357819605, + "learning_rate": 9.163952196968455e-06, + "loss": 0.3251, + "step": 3087 + }, + { + "epoch": 0.21123195841028797, + "grad_norm": 2.340950139105422, + "learning_rate": 9.163338854597324e-06, + "loss": 0.277, + "step": 3088 + }, + { + "epoch": 0.21130036254189752, + "grad_norm": 2.201809632486657, + "learning_rate": 9.162725307869413e-06, + "loss": 0.4729, + "step": 3089 + }, + { + "epoch": 0.21136876667350707, + "grad_norm": 1.301159961088178, + "learning_rate": 9.162111556814845e-06, + "loss": 0.1665, + "step": 3090 + }, + { + "epoch": 0.21143717080511662, + "grad_norm": 1.6152245776099572, + "learning_rate": 9.161497601463742e-06, + "loss": 0.208, + "step": 3091 + }, + { + "epoch": 0.21150557493672617, + "grad_norm": 2.318212310060095, + "learning_rate": 9.160883441846242e-06, + "loss": 0.2448, + "step": 3092 + }, + { + "epoch": 0.21157397906833572, + "grad_norm": 2.777402879860806, + "learning_rate": 9.16026907799249e-06, + "loss": 0.7133, + "step": 3093 + }, + { + "epoch": 0.21164238319994527, + "grad_norm": 2.306898175363184, + "learning_rate": 9.159654509932642e-06, + "loss": 0.2338, + "step": 3094 + }, + { + "epoch": 0.21171078733155482, + "grad_norm": 1.917746793253195, + "learning_rate": 9.159039737696864e-06, + "loss": 0.4111, + "step": 3095 + }, + { + "epoch": 0.21177919146316437, + "grad_norm": 2.6061596847912787, + "learning_rate": 9.15842476131533e-06, + "loss": 0.6197, + "step": 3096 + }, + { + "epoch": 0.21184759559477392, + "grad_norm": 1.668990110596316, + "learning_rate": 9.15780958081823e-06, + "loss": 0.2858, + "step": 3097 + }, + { + "epoch": 0.21191599972638347, + "grad_norm": 2.235357682005979, + "learning_rate": 9.157194196235756e-06, + "loss": 0.4822, + "step": 3098 + }, + { + "epoch": 0.21198440385799303, + "grad_norm": 2.1190930268879535, + "learning_rate": 9.156578607598118e-06, + "loss": 0.4639, + "step": 3099 + }, + { + "epoch": 0.21205280798960258, + "grad_norm": 2.114603087531839, + "learning_rate": 9.155962814935528e-06, + "loss": 0.5573, + "step": 3100 + }, + { + "epoch": 0.21212121212121213, + "grad_norm": 2.473703707926404, + "learning_rate": 9.155346818278216e-06, + "loss": 0.4219, + "step": 3101 + }, + { + "epoch": 0.21218961625282168, + "grad_norm": 1.9985690250318244, + "learning_rate": 9.154730617656416e-06, + "loss": 0.4934, + "step": 3102 + }, + { + "epoch": 0.21225802038443123, + "grad_norm": 2.0656147686795823, + "learning_rate": 9.154114213100373e-06, + "loss": 0.2963, + "step": 3103 + }, + { + "epoch": 0.21232642451604078, + "grad_norm": 2.3827935827786315, + "learning_rate": 9.153497604640344e-06, + "loss": 0.6522, + "step": 3104 + }, + { + "epoch": 0.21239482864765033, + "grad_norm": 2.1503929429210675, + "learning_rate": 9.152880792306598e-06, + "loss": 0.3869, + "step": 3105 + }, + { + "epoch": 0.21246323277925988, + "grad_norm": 2.1809473737741047, + "learning_rate": 9.152263776129407e-06, + "loss": 0.5154, + "step": 3106 + }, + { + "epoch": 0.2125316369108694, + "grad_norm": 2.2334213806094794, + "learning_rate": 9.15164655613906e-06, + "loss": 0.5125, + "step": 3107 + }, + { + "epoch": 0.21260004104247895, + "grad_norm": 2.177214320058535, + "learning_rate": 9.15102913236585e-06, + "loss": 0.3462, + "step": 3108 + }, + { + "epoch": 0.2126684451740885, + "grad_norm": 1.4326237767437908, + "learning_rate": 9.150411504840086e-06, + "loss": 0.1953, + "step": 3109 + }, + { + "epoch": 0.21273684930569806, + "grad_norm": 1.6357683777491696, + "learning_rate": 9.149793673592085e-06, + "loss": 0.3416, + "step": 3110 + }, + { + "epoch": 0.2128052534373076, + "grad_norm": 2.48468911686306, + "learning_rate": 9.149175638652168e-06, + "loss": 0.6558, + "step": 3111 + }, + { + "epoch": 0.21287365756891716, + "grad_norm": 2.561128080444107, + "learning_rate": 9.148557400050678e-06, + "loss": 0.5341, + "step": 3112 + }, + { + "epoch": 0.2129420617005267, + "grad_norm": 1.9782509509059858, + "learning_rate": 9.147938957817955e-06, + "loss": 0.3832, + "step": 3113 + }, + { + "epoch": 0.21301046583213626, + "grad_norm": 1.6207667951749265, + "learning_rate": 9.147320311984358e-06, + "loss": 0.2829, + "step": 3114 + }, + { + "epoch": 0.2130788699637458, + "grad_norm": 1.9404861449661377, + "learning_rate": 9.146701462580254e-06, + "loss": 0.5374, + "step": 3115 + }, + { + "epoch": 0.21314727409535536, + "grad_norm": 2.268515242891765, + "learning_rate": 9.146082409636017e-06, + "loss": 0.4653, + "step": 3116 + }, + { + "epoch": 0.2132156782269649, + "grad_norm": 1.9537376053429771, + "learning_rate": 9.145463153182035e-06, + "loss": 0.3381, + "step": 3117 + }, + { + "epoch": 0.21328408235857446, + "grad_norm": 1.4662638164216464, + "learning_rate": 9.144843693248702e-06, + "loss": 0.2889, + "step": 3118 + }, + { + "epoch": 0.213352486490184, + "grad_norm": 2.9351063878177714, + "learning_rate": 9.144224029866426e-06, + "loss": 0.3104, + "step": 3119 + }, + { + "epoch": 0.21342089062179356, + "grad_norm": 1.584049860850618, + "learning_rate": 9.143604163065622e-06, + "loss": 0.3084, + "step": 3120 + }, + { + "epoch": 0.2134892947534031, + "grad_norm": 1.8236509905837877, + "learning_rate": 9.142984092876717e-06, + "loss": 0.3313, + "step": 3121 + }, + { + "epoch": 0.21355769888501266, + "grad_norm": 2.0146185557913303, + "learning_rate": 9.142363819330146e-06, + "loss": 0.2957, + "step": 3122 + }, + { + "epoch": 0.21362610301662222, + "grad_norm": 1.8131574166058573, + "learning_rate": 9.141743342456358e-06, + "loss": 0.3693, + "step": 3123 + }, + { + "epoch": 0.21369450714823177, + "grad_norm": 1.823296873542392, + "learning_rate": 9.141122662285801e-06, + "loss": 0.3179, + "step": 3124 + }, + { + "epoch": 0.2137629112798413, + "grad_norm": 1.750539251673454, + "learning_rate": 9.140501778848951e-06, + "loss": 0.3468, + "step": 3125 + }, + { + "epoch": 0.21383131541145084, + "grad_norm": 2.2482335123360873, + "learning_rate": 9.139880692176279e-06, + "loss": 0.6477, + "step": 3126 + }, + { + "epoch": 0.2138997195430604, + "grad_norm": 2.1769587482658355, + "learning_rate": 9.13925940229827e-06, + "loss": 0.2946, + "step": 3127 + }, + { + "epoch": 0.21396812367466994, + "grad_norm": 1.9081102815217201, + "learning_rate": 9.138637909245422e-06, + "loss": 0.4809, + "step": 3128 + }, + { + "epoch": 0.2140365278062795, + "grad_norm": 2.05284170274873, + "learning_rate": 9.13801621304824e-06, + "loss": 0.5287, + "step": 3129 + }, + { + "epoch": 0.21410493193788904, + "grad_norm": 2.3374295269955527, + "learning_rate": 9.13739431373724e-06, + "loss": 0.7235, + "step": 3130 + }, + { + "epoch": 0.2141733360694986, + "grad_norm": 1.0760531420555874, + "learning_rate": 9.136772211342947e-06, + "loss": 0.1562, + "step": 3131 + }, + { + "epoch": 0.21424174020110814, + "grad_norm": 2.0676324591820654, + "learning_rate": 9.136149905895897e-06, + "loss": 0.4889, + "step": 3132 + }, + { + "epoch": 0.2143101443327177, + "grad_norm": 2.2852371168907513, + "learning_rate": 9.135527397426637e-06, + "loss": 0.5683, + "step": 3133 + }, + { + "epoch": 0.21437854846432725, + "grad_norm": 2.0335981624700485, + "learning_rate": 9.134904685965725e-06, + "loss": 0.3549, + "step": 3134 + }, + { + "epoch": 0.2144469525959368, + "grad_norm": 2.3580992293649516, + "learning_rate": 9.13428177154372e-06, + "loss": 0.3586, + "step": 3135 + }, + { + "epoch": 0.21451535672754635, + "grad_norm": 2.1302355544598033, + "learning_rate": 9.133658654191205e-06, + "loss": 0.4789, + "step": 3136 + }, + { + "epoch": 0.2145837608591559, + "grad_norm": 1.918219485859641, + "learning_rate": 9.133035333938758e-06, + "loss": 0.2432, + "step": 3137 + }, + { + "epoch": 0.21465216499076545, + "grad_norm": 1.9150031872802487, + "learning_rate": 9.13241181081698e-06, + "loss": 0.3673, + "step": 3138 + }, + { + "epoch": 0.214720569122375, + "grad_norm": 0.998307412438041, + "learning_rate": 9.131788084856477e-06, + "loss": 0.205, + "step": 3139 + }, + { + "epoch": 0.21478897325398455, + "grad_norm": 1.7366381026302071, + "learning_rate": 9.13116415608786e-06, + "loss": 0.361, + "step": 3140 + }, + { + "epoch": 0.2148573773855941, + "grad_norm": 1.080001868027268, + "learning_rate": 9.13054002454176e-06, + "loss": 0.1492, + "step": 3141 + }, + { + "epoch": 0.21492578151720365, + "grad_norm": 1.877914182665412, + "learning_rate": 9.129915690248805e-06, + "loss": 0.402, + "step": 3142 + }, + { + "epoch": 0.2149941856488132, + "grad_norm": 1.846935672053894, + "learning_rate": 9.129291153239647e-06, + "loss": 0.3814, + "step": 3143 + }, + { + "epoch": 0.21506258978042272, + "grad_norm": 1.8427091505423703, + "learning_rate": 9.128666413544939e-06, + "loss": 0.4859, + "step": 3144 + }, + { + "epoch": 0.21513099391203228, + "grad_norm": 1.8304640526900544, + "learning_rate": 9.128041471195349e-06, + "loss": 0.2382, + "step": 3145 + }, + { + "epoch": 0.21519939804364183, + "grad_norm": 1.3314499103857904, + "learning_rate": 9.127416326221547e-06, + "loss": 0.2625, + "step": 3146 + }, + { + "epoch": 0.21526780217525138, + "grad_norm": 1.8430230376465544, + "learning_rate": 9.12679097865422e-06, + "loss": 0.3392, + "step": 3147 + }, + { + "epoch": 0.21533620630686093, + "grad_norm": 2.1927351500682786, + "learning_rate": 9.126165428524065e-06, + "loss": 0.4776, + "step": 3148 + }, + { + "epoch": 0.21540461043847048, + "grad_norm": 2.0060778111694377, + "learning_rate": 9.125539675861786e-06, + "loss": 0.4289, + "step": 3149 + }, + { + "epoch": 0.21547301457008003, + "grad_norm": 1.2361573919491624, + "learning_rate": 9.124913720698097e-06, + "loss": 0.221, + "step": 3150 + }, + { + "epoch": 0.21554141870168958, + "grad_norm": 2.439914253576935, + "learning_rate": 9.124287563063723e-06, + "loss": 0.6324, + "step": 3151 + }, + { + "epoch": 0.21560982283329913, + "grad_norm": 1.1238261579165516, + "learning_rate": 9.1236612029894e-06, + "loss": 0.158, + "step": 3152 + }, + { + "epoch": 0.21567822696490868, + "grad_norm": 2.2943432082423656, + "learning_rate": 9.123034640505871e-06, + "loss": 0.7092, + "step": 3153 + }, + { + "epoch": 0.21574663109651823, + "grad_norm": 2.264817465543784, + "learning_rate": 9.122407875643892e-06, + "loss": 0.4057, + "step": 3154 + }, + { + "epoch": 0.21581503522812778, + "grad_norm": 1.5838533105025214, + "learning_rate": 9.12178090843423e-06, + "loss": 0.2154, + "step": 3155 + }, + { + "epoch": 0.21588343935973733, + "grad_norm": 2.1796357377718625, + "learning_rate": 9.121153738907655e-06, + "loss": 0.6339, + "step": 3156 + }, + { + "epoch": 0.21595184349134688, + "grad_norm": 1.6175124353346466, + "learning_rate": 9.120526367094955e-06, + "loss": 0.3638, + "step": 3157 + }, + { + "epoch": 0.21602024762295644, + "grad_norm": 2.1204534867484846, + "learning_rate": 9.119898793026923e-06, + "loss": 0.451, + "step": 3158 + }, + { + "epoch": 0.21608865175456599, + "grad_norm": 1.4378875904419124, + "learning_rate": 9.11927101673436e-06, + "loss": 0.3217, + "step": 3159 + }, + { + "epoch": 0.21615705588617554, + "grad_norm": 1.9687298592036762, + "learning_rate": 9.118643038248088e-06, + "loss": 0.5371, + "step": 3160 + }, + { + "epoch": 0.2162254600177851, + "grad_norm": 1.242476615608955, + "learning_rate": 9.118014857598927e-06, + "loss": 0.1706, + "step": 3161 + }, + { + "epoch": 0.2162938641493946, + "grad_norm": 1.7529395908086591, + "learning_rate": 9.117386474817709e-06, + "loss": 0.4146, + "step": 3162 + }, + { + "epoch": 0.21636226828100416, + "grad_norm": 2.2995555572666233, + "learning_rate": 9.11675788993528e-06, + "loss": 0.7114, + "step": 3163 + }, + { + "epoch": 0.2164306724126137, + "grad_norm": 2.3813538573737536, + "learning_rate": 9.116129102982496e-06, + "loss": 0.5283, + "step": 3164 + }, + { + "epoch": 0.21649907654422326, + "grad_norm": 1.942251148239741, + "learning_rate": 9.115500113990217e-06, + "loss": 0.4061, + "step": 3165 + }, + { + "epoch": 0.2165674806758328, + "grad_norm": 1.575012801890343, + "learning_rate": 9.11487092298932e-06, + "loss": 0.3609, + "step": 3166 + }, + { + "epoch": 0.21663588480744236, + "grad_norm": 1.9523072438148992, + "learning_rate": 9.114241530010687e-06, + "loss": 0.3058, + "step": 3167 + }, + { + "epoch": 0.21670428893905191, + "grad_norm": 1.7364432872775597, + "learning_rate": 9.113611935085214e-06, + "loss": 0.3832, + "step": 3168 + }, + { + "epoch": 0.21677269307066147, + "grad_norm": 1.5777877055309237, + "learning_rate": 9.112982138243802e-06, + "loss": 0.1703, + "step": 3169 + }, + { + "epoch": 0.21684109720227102, + "grad_norm": 2.5058841481102947, + "learning_rate": 9.112352139517365e-06, + "loss": 0.9084, + "step": 3170 + }, + { + "epoch": 0.21690950133388057, + "grad_norm": 2.0096406999767127, + "learning_rate": 9.111721938936825e-06, + "loss": 0.2928, + "step": 3171 + }, + { + "epoch": 0.21697790546549012, + "grad_norm": 4.41059898945027, + "learning_rate": 9.111091536533119e-06, + "loss": 0.2345, + "step": 3172 + }, + { + "epoch": 0.21704630959709967, + "grad_norm": 3.9414893860834344, + "learning_rate": 9.110460932337187e-06, + "loss": 0.5932, + "step": 3173 + }, + { + "epoch": 0.21711471372870922, + "grad_norm": 1.6890914353425048, + "learning_rate": 9.109830126379983e-06, + "loss": 0.2288, + "step": 3174 + }, + { + "epoch": 0.21718311786031877, + "grad_norm": 1.784043342880374, + "learning_rate": 9.109199118692472e-06, + "loss": 0.3329, + "step": 3175 + }, + { + "epoch": 0.21725152199192832, + "grad_norm": 1.6157869416697253, + "learning_rate": 9.108567909305623e-06, + "loss": 0.1768, + "step": 3176 + }, + { + "epoch": 0.21731992612353787, + "grad_norm": 2.430838407376704, + "learning_rate": 9.107936498250422e-06, + "loss": 0.8329, + "step": 3177 + }, + { + "epoch": 0.21738833025514742, + "grad_norm": 1.8743635885699625, + "learning_rate": 9.107304885557862e-06, + "loss": 0.485, + "step": 3178 + }, + { + "epoch": 0.21745673438675697, + "grad_norm": 1.9125406404303569, + "learning_rate": 9.106673071258942e-06, + "loss": 0.4174, + "step": 3179 + }, + { + "epoch": 0.21752513851836652, + "grad_norm": 2.065201622746717, + "learning_rate": 9.106041055384678e-06, + "loss": 0.5487, + "step": 3180 + }, + { + "epoch": 0.21759354264997605, + "grad_norm": 1.7903302419432243, + "learning_rate": 9.105408837966091e-06, + "loss": 0.4418, + "step": 3181 + }, + { + "epoch": 0.2176619467815856, + "grad_norm": 2.346899718384281, + "learning_rate": 9.104776419034213e-06, + "loss": 0.3219, + "step": 3182 + }, + { + "epoch": 0.21773035091319515, + "grad_norm": 2.2383631679596907, + "learning_rate": 9.104143798620087e-06, + "loss": 0.588, + "step": 3183 + }, + { + "epoch": 0.2177987550448047, + "grad_norm": 2.023259493829938, + "learning_rate": 9.103510976754763e-06, + "loss": 0.4539, + "step": 3184 + }, + { + "epoch": 0.21786715917641425, + "grad_norm": 1.612841846488815, + "learning_rate": 9.102877953469308e-06, + "loss": 0.2796, + "step": 3185 + }, + { + "epoch": 0.2179355633080238, + "grad_norm": 2.179211595326193, + "learning_rate": 9.102244728794787e-06, + "loss": 0.569, + "step": 3186 + }, + { + "epoch": 0.21800396743963335, + "grad_norm": 1.4940920354384857, + "learning_rate": 9.101611302762287e-06, + "loss": 0.2041, + "step": 3187 + }, + { + "epoch": 0.2180723715712429, + "grad_norm": 1.8046909596463685, + "learning_rate": 9.100977675402898e-06, + "loss": 0.3887, + "step": 3188 + }, + { + "epoch": 0.21814077570285245, + "grad_norm": 1.4416055024706538, + "learning_rate": 9.100343846747722e-06, + "loss": 0.2741, + "step": 3189 + }, + { + "epoch": 0.218209179834462, + "grad_norm": 1.6040642133317542, + "learning_rate": 9.099709816827866e-06, + "loss": 0.3041, + "step": 3190 + }, + { + "epoch": 0.21827758396607155, + "grad_norm": 1.8390441906583965, + "learning_rate": 9.099075585674459e-06, + "loss": 0.4162, + "step": 3191 + }, + { + "epoch": 0.2183459880976811, + "grad_norm": 2.0284564248534642, + "learning_rate": 9.098441153318626e-06, + "loss": 0.5262, + "step": 3192 + }, + { + "epoch": 0.21841439222929065, + "grad_norm": 2.201637063208827, + "learning_rate": 9.09780651979151e-06, + "loss": 0.5932, + "step": 3193 + }, + { + "epoch": 0.2184827963609002, + "grad_norm": 2.8145839123380583, + "learning_rate": 9.097171685124262e-06, + "loss": 0.5768, + "step": 3194 + }, + { + "epoch": 0.21855120049250976, + "grad_norm": 1.7381251961599251, + "learning_rate": 9.096536649348043e-06, + "loss": 0.2384, + "step": 3195 + }, + { + "epoch": 0.2186196046241193, + "grad_norm": 2.1919990594506857, + "learning_rate": 9.095901412494024e-06, + "loss": 0.347, + "step": 3196 + }, + { + "epoch": 0.21868800875572886, + "grad_norm": 2.1330146344434424, + "learning_rate": 9.095265974593383e-06, + "loss": 0.3356, + "step": 3197 + }, + { + "epoch": 0.2187564128873384, + "grad_norm": 2.3608643428947524, + "learning_rate": 9.094630335677312e-06, + "loss": 0.5839, + "step": 3198 + }, + { + "epoch": 0.21882481701894793, + "grad_norm": 2.3185481900051887, + "learning_rate": 9.093994495777014e-06, + "loss": 0.2416, + "step": 3199 + }, + { + "epoch": 0.21889322115055748, + "grad_norm": 2.25802869982595, + "learning_rate": 9.093358454923695e-06, + "loss": 0.5927, + "step": 3200 + }, + { + "epoch": 0.21896162528216703, + "grad_norm": 1.8800571628323728, + "learning_rate": 9.092722213148574e-06, + "loss": 0.2199, + "step": 3201 + }, + { + "epoch": 0.21903002941377658, + "grad_norm": 1.7072713669956296, + "learning_rate": 9.092085770482886e-06, + "loss": 0.2307, + "step": 3202 + }, + { + "epoch": 0.21909843354538613, + "grad_norm": 2.3625212464058416, + "learning_rate": 9.091449126957867e-06, + "loss": 0.4317, + "step": 3203 + }, + { + "epoch": 0.21916683767699568, + "grad_norm": 2.1731802393892443, + "learning_rate": 9.090812282604768e-06, + "loss": 0.3392, + "step": 3204 + }, + { + "epoch": 0.21923524180860524, + "grad_norm": 1.733914460218355, + "learning_rate": 9.090175237454848e-06, + "loss": 0.335, + "step": 3205 + }, + { + "epoch": 0.2193036459402148, + "grad_norm": 1.8065950460097238, + "learning_rate": 9.089537991539376e-06, + "loss": 0.2932, + "step": 3206 + }, + { + "epoch": 0.21937205007182434, + "grad_norm": 2.0575371116985397, + "learning_rate": 9.088900544889631e-06, + "loss": 0.5291, + "step": 3207 + }, + { + "epoch": 0.2194404542034339, + "grad_norm": 1.3290162672244035, + "learning_rate": 9.0882628975369e-06, + "loss": 0.1995, + "step": 3208 + }, + { + "epoch": 0.21950885833504344, + "grad_norm": 2.4923235727550126, + "learning_rate": 9.087625049512489e-06, + "loss": 0.4771, + "step": 3209 + }, + { + "epoch": 0.219577262466653, + "grad_norm": 1.718372190739422, + "learning_rate": 9.086987000847698e-06, + "loss": 0.4767, + "step": 3210 + }, + { + "epoch": 0.21964566659826254, + "grad_norm": 1.8815031722478206, + "learning_rate": 9.08634875157385e-06, + "loss": 0.3989, + "step": 3211 + }, + { + "epoch": 0.2197140707298721, + "grad_norm": 2.248659622110758, + "learning_rate": 9.085710301722273e-06, + "loss": 0.4101, + "step": 3212 + }, + { + "epoch": 0.21978247486148164, + "grad_norm": 1.9068008940083845, + "learning_rate": 9.085071651324304e-06, + "loss": 0.375, + "step": 3213 + }, + { + "epoch": 0.2198508789930912, + "grad_norm": 2.260718033663618, + "learning_rate": 9.084432800411291e-06, + "loss": 0.4684, + "step": 3214 + }, + { + "epoch": 0.21991928312470074, + "grad_norm": 2.0411996712819485, + "learning_rate": 9.083793749014593e-06, + "loss": 0.3654, + "step": 3215 + }, + { + "epoch": 0.2199876872563103, + "grad_norm": 1.8464387255344574, + "learning_rate": 9.083154497165579e-06, + "loss": 0.2675, + "step": 3216 + }, + { + "epoch": 0.22005609138791982, + "grad_norm": 2.3923845609879173, + "learning_rate": 9.082515044895624e-06, + "loss": 0.6587, + "step": 3217 + }, + { + "epoch": 0.22012449551952937, + "grad_norm": 1.919784738914863, + "learning_rate": 9.081875392236115e-06, + "loss": 0.4553, + "step": 3218 + }, + { + "epoch": 0.22019289965113892, + "grad_norm": 1.7777333429945865, + "learning_rate": 9.081235539218451e-06, + "loss": 0.4686, + "step": 3219 + }, + { + "epoch": 0.22026130378274847, + "grad_norm": 1.792328373195642, + "learning_rate": 9.08059548587404e-06, + "loss": 0.4458, + "step": 3220 + }, + { + "epoch": 0.22032970791435802, + "grad_norm": 1.8091881807101762, + "learning_rate": 9.079955232234296e-06, + "loss": 0.3594, + "step": 3221 + }, + { + "epoch": 0.22039811204596757, + "grad_norm": 1.8769190504333366, + "learning_rate": 9.079314778330648e-06, + "loss": 0.3241, + "step": 3222 + }, + { + "epoch": 0.22046651617757712, + "grad_norm": 2.4707840073204106, + "learning_rate": 9.078674124194533e-06, + "loss": 0.7118, + "step": 3223 + }, + { + "epoch": 0.22053492030918667, + "grad_norm": 1.9159562143492268, + "learning_rate": 9.078033269857394e-06, + "loss": 0.4901, + "step": 3224 + }, + { + "epoch": 0.22060332444079622, + "grad_norm": 2.164232381632194, + "learning_rate": 9.07739221535069e-06, + "loss": 0.3482, + "step": 3225 + }, + { + "epoch": 0.22067172857240577, + "grad_norm": 2.069333496380348, + "learning_rate": 9.07675096070589e-06, + "loss": 0.5247, + "step": 3226 + }, + { + "epoch": 0.22074013270401532, + "grad_norm": 2.2843007238709134, + "learning_rate": 9.076109505954462e-06, + "loss": 0.4793, + "step": 3227 + }, + { + "epoch": 0.22080853683562487, + "grad_norm": 1.9571657953059254, + "learning_rate": 9.075467851127899e-06, + "loss": 0.5782, + "step": 3228 + }, + { + "epoch": 0.22087694096723443, + "grad_norm": 1.4118200773471234, + "learning_rate": 9.074825996257693e-06, + "loss": 0.1443, + "step": 3229 + }, + { + "epoch": 0.22094534509884398, + "grad_norm": 1.9028016759239113, + "learning_rate": 9.074183941375349e-06, + "loss": 0.3249, + "step": 3230 + }, + { + "epoch": 0.22101374923045353, + "grad_norm": 2.416143588206016, + "learning_rate": 9.073541686512383e-06, + "loss": 0.6969, + "step": 3231 + }, + { + "epoch": 0.22108215336206308, + "grad_norm": 2.0803376228473907, + "learning_rate": 9.072899231700321e-06, + "loss": 0.5027, + "step": 3232 + }, + { + "epoch": 0.22115055749367263, + "grad_norm": 2.0628656583379286, + "learning_rate": 9.072256576970694e-06, + "loss": 0.3662, + "step": 3233 + }, + { + "epoch": 0.22121896162528218, + "grad_norm": 1.778405247712394, + "learning_rate": 9.071613722355054e-06, + "loss": 0.2602, + "step": 3234 + }, + { + "epoch": 0.22128736575689173, + "grad_norm": 2.094348960579124, + "learning_rate": 9.070970667884948e-06, + "loss": 0.5245, + "step": 3235 + }, + { + "epoch": 0.22135576988850125, + "grad_norm": 2.33112829545223, + "learning_rate": 9.070327413591943e-06, + "loss": 0.4095, + "step": 3236 + }, + { + "epoch": 0.2214241740201108, + "grad_norm": 1.874399929583621, + "learning_rate": 9.069683959507616e-06, + "loss": 0.3246, + "step": 3237 + }, + { + "epoch": 0.22149257815172035, + "grad_norm": 1.828611101421692, + "learning_rate": 9.069040305663545e-06, + "loss": 0.4329, + "step": 3238 + }, + { + "epoch": 0.2215609822833299, + "grad_norm": 2.3308269835392323, + "learning_rate": 9.068396452091328e-06, + "loss": 0.5982, + "step": 3239 + }, + { + "epoch": 0.22162938641493946, + "grad_norm": 2.0448679438696096, + "learning_rate": 9.067752398822567e-06, + "loss": 0.3544, + "step": 3240 + }, + { + "epoch": 0.221697790546549, + "grad_norm": 2.3953062051191436, + "learning_rate": 9.067108145888874e-06, + "loss": 0.1712, + "step": 3241 + }, + { + "epoch": 0.22176619467815856, + "grad_norm": 7.972588994416761, + "learning_rate": 9.066463693321876e-06, + "loss": 0.2694, + "step": 3242 + }, + { + "epoch": 0.2218345988097681, + "grad_norm": 2.8540633613913426, + "learning_rate": 9.065819041153202e-06, + "loss": 0.1963, + "step": 3243 + }, + { + "epoch": 0.22190300294137766, + "grad_norm": 2.3868681116396786, + "learning_rate": 9.065174189414493e-06, + "loss": 0.5598, + "step": 3244 + }, + { + "epoch": 0.2219714070729872, + "grad_norm": 2.4425401785639225, + "learning_rate": 9.064529138137408e-06, + "loss": 0.7171, + "step": 3245 + }, + { + "epoch": 0.22203981120459676, + "grad_norm": 2.775207255760136, + "learning_rate": 9.063883887353604e-06, + "loss": 0.3557, + "step": 3246 + }, + { + "epoch": 0.2221082153362063, + "grad_norm": 2.746281526529575, + "learning_rate": 9.063238437094754e-06, + "loss": 0.8515, + "step": 3247 + }, + { + "epoch": 0.22217661946781586, + "grad_norm": 2.3234053013718357, + "learning_rate": 9.062592787392541e-06, + "loss": 0.4488, + "step": 3248 + }, + { + "epoch": 0.2222450235994254, + "grad_norm": 2.0701299622875124, + "learning_rate": 9.061946938278656e-06, + "loss": 0.5364, + "step": 3249 + }, + { + "epoch": 0.22231342773103496, + "grad_norm": 1.8533923647769388, + "learning_rate": 9.061300889784802e-06, + "loss": 0.5543, + "step": 3250 + }, + { + "epoch": 0.2223818318626445, + "grad_norm": 2.5209063126637634, + "learning_rate": 9.060654641942686e-06, + "loss": 0.6889, + "step": 3251 + }, + { + "epoch": 0.22245023599425406, + "grad_norm": 1.8885801425183881, + "learning_rate": 9.060008194784032e-06, + "loss": 0.4429, + "step": 3252 + }, + { + "epoch": 0.22251864012586361, + "grad_norm": 2.4802077989659113, + "learning_rate": 9.05936154834057e-06, + "loss": 0.7464, + "step": 3253 + }, + { + "epoch": 0.22258704425747314, + "grad_norm": 2.430568222663811, + "learning_rate": 9.05871470264404e-06, + "loss": 0.4412, + "step": 3254 + }, + { + "epoch": 0.2226554483890827, + "grad_norm": 1.9778759704690647, + "learning_rate": 9.058067657726195e-06, + "loss": 0.4952, + "step": 3255 + }, + { + "epoch": 0.22272385252069224, + "grad_norm": 2.165288763655266, + "learning_rate": 9.05742041361879e-06, + "loss": 0.5698, + "step": 3256 + }, + { + "epoch": 0.2227922566523018, + "grad_norm": 1.9764242538535666, + "learning_rate": 9.056772970353602e-06, + "loss": 0.3804, + "step": 3257 + }, + { + "epoch": 0.22286066078391134, + "grad_norm": 2.141221287090019, + "learning_rate": 9.056125327962407e-06, + "loss": 0.3754, + "step": 3258 + }, + { + "epoch": 0.2229290649155209, + "grad_norm": 1.9541271957248965, + "learning_rate": 9.055477486476992e-06, + "loss": 0.5033, + "step": 3259 + }, + { + "epoch": 0.22299746904713044, + "grad_norm": 1.1852053459851444, + "learning_rate": 9.054829445929156e-06, + "loss": 0.1127, + "step": 3260 + }, + { + "epoch": 0.22306587317874, + "grad_norm": 2.0775520997796266, + "learning_rate": 9.054181206350714e-06, + "loss": 0.3306, + "step": 3261 + }, + { + "epoch": 0.22313427731034954, + "grad_norm": 2.6345594451293293, + "learning_rate": 9.053532767773479e-06, + "loss": 0.5851, + "step": 3262 + }, + { + "epoch": 0.2232026814419591, + "grad_norm": 2.4928341806804815, + "learning_rate": 9.052884130229282e-06, + "loss": 0.7575, + "step": 3263 + }, + { + "epoch": 0.22327108557356865, + "grad_norm": 2.0500872718447365, + "learning_rate": 9.05223529374996e-06, + "loss": 0.5438, + "step": 3264 + }, + { + "epoch": 0.2233394897051782, + "grad_norm": 1.7779856350659586, + "learning_rate": 9.051586258367365e-06, + "loss": 0.3358, + "step": 3265 + }, + { + "epoch": 0.22340789383678775, + "grad_norm": 2.007983742176262, + "learning_rate": 9.050937024113347e-06, + "loss": 0.3629, + "step": 3266 + }, + { + "epoch": 0.2234762979683973, + "grad_norm": 1.8992111136856873, + "learning_rate": 9.05028759101978e-06, + "loss": 0.4488, + "step": 3267 + }, + { + "epoch": 0.22354470210000685, + "grad_norm": 1.6345888951526466, + "learning_rate": 9.04963795911854e-06, + "loss": 0.2891, + "step": 3268 + }, + { + "epoch": 0.2236131062316164, + "grad_norm": 2.3794739551623776, + "learning_rate": 9.048988128441513e-06, + "loss": 0.519, + "step": 3269 + }, + { + "epoch": 0.22368151036322595, + "grad_norm": 2.487936645791322, + "learning_rate": 9.048338099020596e-06, + "loss": 0.6965, + "step": 3270 + }, + { + "epoch": 0.2237499144948355, + "grad_norm": 2.033899076375438, + "learning_rate": 9.047687870887697e-06, + "loss": 0.4974, + "step": 3271 + }, + { + "epoch": 0.22381831862644505, + "grad_norm": 2.442814291233218, + "learning_rate": 9.047037444074728e-06, + "loss": 0.8164, + "step": 3272 + }, + { + "epoch": 0.22388672275805457, + "grad_norm": 1.749315502630974, + "learning_rate": 9.046386818613622e-06, + "loss": 0.2673, + "step": 3273 + }, + { + "epoch": 0.22395512688966412, + "grad_norm": 1.8767664575392558, + "learning_rate": 9.04573599453631e-06, + "loss": 0.369, + "step": 3274 + }, + { + "epoch": 0.22402353102127368, + "grad_norm": 2.564285570642786, + "learning_rate": 9.045084971874738e-06, + "loss": 0.2105, + "step": 3275 + }, + { + "epoch": 0.22409193515288323, + "grad_norm": 1.6137765442757859, + "learning_rate": 9.044433750660862e-06, + "loss": 0.3658, + "step": 3276 + }, + { + "epoch": 0.22416033928449278, + "grad_norm": 1.9970288596479577, + "learning_rate": 9.043782330926649e-06, + "loss": 0.6017, + "step": 3277 + }, + { + "epoch": 0.22422874341610233, + "grad_norm": 2.425049712959481, + "learning_rate": 9.04313071270407e-06, + "loss": 0.4471, + "step": 3278 + }, + { + "epoch": 0.22429714754771188, + "grad_norm": 1.9259142181732276, + "learning_rate": 9.042478896025113e-06, + "loss": 0.5007, + "step": 3279 + }, + { + "epoch": 0.22436555167932143, + "grad_norm": 2.8684906038180613, + "learning_rate": 9.04182688092177e-06, + "loss": 0.3606, + "step": 3280 + }, + { + "epoch": 0.22443395581093098, + "grad_norm": 2.064800803394237, + "learning_rate": 9.041174667426044e-06, + "loss": 0.6542, + "step": 3281 + }, + { + "epoch": 0.22450235994254053, + "grad_norm": 2.640121756233104, + "learning_rate": 9.040522255569953e-06, + "loss": 0.5004, + "step": 3282 + }, + { + "epoch": 0.22457076407415008, + "grad_norm": 2.1629935079076814, + "learning_rate": 9.039869645385519e-06, + "loss": 0.7587, + "step": 3283 + }, + { + "epoch": 0.22463916820575963, + "grad_norm": 2.026313255235791, + "learning_rate": 9.039216836904773e-06, + "loss": 0.6402, + "step": 3284 + }, + { + "epoch": 0.22470757233736918, + "grad_norm": 2.688613223015711, + "learning_rate": 9.038563830159758e-06, + "loss": 0.2858, + "step": 3285 + }, + { + "epoch": 0.22477597646897873, + "grad_norm": 3.2712733641939082, + "learning_rate": 9.03791062518253e-06, + "loss": 0.4268, + "step": 3286 + }, + { + "epoch": 0.22484438060058828, + "grad_norm": 2.459763183156365, + "learning_rate": 9.037257222005152e-06, + "loss": 0.5306, + "step": 3287 + }, + { + "epoch": 0.22491278473219783, + "grad_norm": 2.2027816843146786, + "learning_rate": 9.03660362065969e-06, + "loss": 0.6515, + "step": 3288 + }, + { + "epoch": 0.22498118886380739, + "grad_norm": 6.048252944278911, + "learning_rate": 9.035949821178232e-06, + "loss": 0.2498, + "step": 3289 + }, + { + "epoch": 0.22504959299541694, + "grad_norm": 2.047713624804628, + "learning_rate": 9.035295823592867e-06, + "loss": 0.2581, + "step": 3290 + }, + { + "epoch": 0.22511799712702646, + "grad_norm": 1.9035910643297413, + "learning_rate": 9.034641627935695e-06, + "loss": 0.3656, + "step": 3291 + }, + { + "epoch": 0.225186401258636, + "grad_norm": 1.4210341555875012, + "learning_rate": 9.03398723423883e-06, + "loss": 0.2446, + "step": 3292 + }, + { + "epoch": 0.22525480539024556, + "grad_norm": 4.435924122606445, + "learning_rate": 9.033332642534392e-06, + "loss": 0.3584, + "step": 3293 + }, + { + "epoch": 0.2253232095218551, + "grad_norm": 2.1568714849487067, + "learning_rate": 9.03267785285451e-06, + "loss": 0.4738, + "step": 3294 + }, + { + "epoch": 0.22539161365346466, + "grad_norm": 2.5462519345770085, + "learning_rate": 9.032022865231328e-06, + "loss": 0.7708, + "step": 3295 + }, + { + "epoch": 0.2254600177850742, + "grad_norm": 9.316804646224904, + "learning_rate": 9.03136767969699e-06, + "loss": 0.5607, + "step": 3296 + }, + { + "epoch": 0.22552842191668376, + "grad_norm": 2.2453660333419663, + "learning_rate": 9.03071229628366e-06, + "loss": 0.5869, + "step": 3297 + }, + { + "epoch": 0.22559682604829331, + "grad_norm": 1.8251342191722628, + "learning_rate": 9.030056715023508e-06, + "loss": 0.3802, + "step": 3298 + }, + { + "epoch": 0.22566523017990286, + "grad_norm": 1.7196551193829477, + "learning_rate": 9.029400935948712e-06, + "loss": 0.3357, + "step": 3299 + }, + { + "epoch": 0.22573363431151242, + "grad_norm": 2.859569532520539, + "learning_rate": 9.028744959091458e-06, + "loss": 0.317, + "step": 3300 + }, + { + "epoch": 0.22580203844312197, + "grad_norm": 1.4102289492816724, + "learning_rate": 9.028088784483946e-06, + "loss": 0.2475, + "step": 3301 + }, + { + "epoch": 0.22587044257473152, + "grad_norm": 1.7608278817530074, + "learning_rate": 9.027432412158387e-06, + "loss": 0.3985, + "step": 3302 + }, + { + "epoch": 0.22593884670634107, + "grad_norm": 2.165020720215332, + "learning_rate": 9.026775842146998e-06, + "loss": 0.5131, + "step": 3303 + }, + { + "epoch": 0.22600725083795062, + "grad_norm": 2.211020532133989, + "learning_rate": 9.026119074482003e-06, + "loss": 0.5586, + "step": 3304 + }, + { + "epoch": 0.22607565496956017, + "grad_norm": 2.0150058926633427, + "learning_rate": 9.025462109195642e-06, + "loss": 0.4759, + "step": 3305 + }, + { + "epoch": 0.22614405910116972, + "grad_norm": 2.4079440891022146, + "learning_rate": 9.024804946320164e-06, + "loss": 0.7291, + "step": 3306 + }, + { + "epoch": 0.22621246323277927, + "grad_norm": 1.9200491489512321, + "learning_rate": 9.024147585887822e-06, + "loss": 0.3804, + "step": 3307 + }, + { + "epoch": 0.22628086736438882, + "grad_norm": 2.9682021890460613, + "learning_rate": 9.023490027930885e-06, + "loss": 0.5588, + "step": 3308 + }, + { + "epoch": 0.22634927149599837, + "grad_norm": 2.2210649717266566, + "learning_rate": 9.022832272481628e-06, + "loss": 0.6155, + "step": 3309 + }, + { + "epoch": 0.2264176756276079, + "grad_norm": 2.2703978317039777, + "learning_rate": 9.022174319572336e-06, + "loss": 0.6408, + "step": 3310 + }, + { + "epoch": 0.22648607975921745, + "grad_norm": 1.9741674243699576, + "learning_rate": 9.021516169235306e-06, + "loss": 0.2448, + "step": 3311 + }, + { + "epoch": 0.226554483890827, + "grad_norm": 1.8373426376483444, + "learning_rate": 9.020857821502843e-06, + "loss": 0.3171, + "step": 3312 + }, + { + "epoch": 0.22662288802243655, + "grad_norm": 2.15726694368876, + "learning_rate": 9.020199276407262e-06, + "loss": 0.4626, + "step": 3313 + }, + { + "epoch": 0.2266912921540461, + "grad_norm": 1.5647055842499435, + "learning_rate": 9.019540533980886e-06, + "loss": 0.1966, + "step": 3314 + }, + { + "epoch": 0.22675969628565565, + "grad_norm": 1.412773983680754, + "learning_rate": 9.018881594256051e-06, + "loss": 0.3098, + "step": 3315 + }, + { + "epoch": 0.2268281004172652, + "grad_norm": 2.7290442302166142, + "learning_rate": 9.018222457265102e-06, + "loss": 0.6597, + "step": 3316 + }, + { + "epoch": 0.22689650454887475, + "grad_norm": 1.9452778303741405, + "learning_rate": 9.017563123040389e-06, + "loss": 0.3228, + "step": 3317 + }, + { + "epoch": 0.2269649086804843, + "grad_norm": 2.1367798575079004, + "learning_rate": 9.016903591614278e-06, + "loss": 0.5048, + "step": 3318 + }, + { + "epoch": 0.22703331281209385, + "grad_norm": 1.1409247989086029, + "learning_rate": 9.01624386301914e-06, + "loss": 0.1265, + "step": 3319 + }, + { + "epoch": 0.2271017169437034, + "grad_norm": 2.130245588873764, + "learning_rate": 9.015583937287361e-06, + "loss": 0.4483, + "step": 3320 + }, + { + "epoch": 0.22717012107531295, + "grad_norm": 2.1514664885486208, + "learning_rate": 9.01492381445133e-06, + "loss": 0.4697, + "step": 3321 + }, + { + "epoch": 0.2272385252069225, + "grad_norm": 2.1746925253137848, + "learning_rate": 9.014263494543452e-06, + "loss": 0.5771, + "step": 3322 + }, + { + "epoch": 0.22730692933853205, + "grad_norm": 1.8970861275806148, + "learning_rate": 9.013602977596134e-06, + "loss": 0.3632, + "step": 3323 + }, + { + "epoch": 0.2273753334701416, + "grad_norm": 2.1327805778713524, + "learning_rate": 9.012942263641803e-06, + "loss": 0.5719, + "step": 3324 + }, + { + "epoch": 0.22744373760175116, + "grad_norm": 2.04024045907884, + "learning_rate": 9.012281352712886e-06, + "loss": 0.4492, + "step": 3325 + }, + { + "epoch": 0.2275121417333607, + "grad_norm": 2.463377055232673, + "learning_rate": 9.011620244841826e-06, + "loss": 0.6922, + "step": 3326 + }, + { + "epoch": 0.22758054586497026, + "grad_norm": 1.8464603424929322, + "learning_rate": 9.010958940061072e-06, + "loss": 0.4335, + "step": 3327 + }, + { + "epoch": 0.22764894999657978, + "grad_norm": 1.982954436678783, + "learning_rate": 9.010297438403084e-06, + "loss": 0.4192, + "step": 3328 + }, + { + "epoch": 0.22771735412818933, + "grad_norm": 2.1252855045997516, + "learning_rate": 9.009635739900334e-06, + "loss": 0.4631, + "step": 3329 + }, + { + "epoch": 0.22778575825979888, + "grad_norm": 1.8825950417837205, + "learning_rate": 9.008973844585297e-06, + "loss": 0.4194, + "step": 3330 + }, + { + "epoch": 0.22785416239140843, + "grad_norm": 1.7078387884017094, + "learning_rate": 9.008311752490467e-06, + "loss": 0.1694, + "step": 3331 + }, + { + "epoch": 0.22792256652301798, + "grad_norm": 2.089641875026662, + "learning_rate": 9.007649463648338e-06, + "loss": 0.5418, + "step": 3332 + }, + { + "epoch": 0.22799097065462753, + "grad_norm": 1.9464072858139017, + "learning_rate": 9.006986978091423e-06, + "loss": 0.4161, + "step": 3333 + }, + { + "epoch": 0.22805937478623708, + "grad_norm": 2.13606555236751, + "learning_rate": 9.006324295852236e-06, + "loss": 0.3306, + "step": 3334 + }, + { + "epoch": 0.22812777891784664, + "grad_norm": 1.960328148016677, + "learning_rate": 9.005661416963306e-06, + "loss": 0.3742, + "step": 3335 + }, + { + "epoch": 0.22819618304945619, + "grad_norm": 2.0473521746555488, + "learning_rate": 9.004998341457172e-06, + "loss": 0.3926, + "step": 3336 + }, + { + "epoch": 0.22826458718106574, + "grad_norm": 1.896237334672428, + "learning_rate": 9.004335069366378e-06, + "loss": 0.378, + "step": 3337 + }, + { + "epoch": 0.2283329913126753, + "grad_norm": 2.0157877686302523, + "learning_rate": 9.003671600723483e-06, + "loss": 0.3773, + "step": 3338 + }, + { + "epoch": 0.22840139544428484, + "grad_norm": 2.2982147487475197, + "learning_rate": 9.003007935561052e-06, + "loss": 0.5006, + "step": 3339 + }, + { + "epoch": 0.2284697995758944, + "grad_norm": 1.8572355311288813, + "learning_rate": 9.00234407391166e-06, + "loss": 0.2816, + "step": 3340 + }, + { + "epoch": 0.22853820370750394, + "grad_norm": 1.642922736089055, + "learning_rate": 9.001680015807896e-06, + "loss": 0.2061, + "step": 3341 + }, + { + "epoch": 0.2286066078391135, + "grad_norm": 1.3266033684149392, + "learning_rate": 9.00101576128235e-06, + "loss": 0.1835, + "step": 3342 + }, + { + "epoch": 0.22867501197072304, + "grad_norm": 2.0247047572600847, + "learning_rate": 9.000351310367633e-06, + "loss": 0.4853, + "step": 3343 + }, + { + "epoch": 0.2287434161023326, + "grad_norm": 1.9668970622330546, + "learning_rate": 8.999686663096354e-06, + "loss": 0.4362, + "step": 3344 + }, + { + "epoch": 0.22881182023394214, + "grad_norm": 1.8596307334223043, + "learning_rate": 8.99902181950114e-06, + "loss": 0.3818, + "step": 3345 + }, + { + "epoch": 0.22888022436555167, + "grad_norm": 2.114411127817508, + "learning_rate": 8.998356779614623e-06, + "loss": 0.6767, + "step": 3346 + }, + { + "epoch": 0.22894862849716122, + "grad_norm": 1.9233484744953042, + "learning_rate": 8.997691543469448e-06, + "loss": 0.3972, + "step": 3347 + }, + { + "epoch": 0.22901703262877077, + "grad_norm": 1.7329529009696105, + "learning_rate": 8.997026111098268e-06, + "loss": 0.2433, + "step": 3348 + }, + { + "epoch": 0.22908543676038032, + "grad_norm": 1.9594349328750946, + "learning_rate": 8.996360482533746e-06, + "loss": 0.4056, + "step": 3349 + }, + { + "epoch": 0.22915384089198987, + "grad_norm": 2.174683840288519, + "learning_rate": 8.995694657808551e-06, + "loss": 0.5371, + "step": 3350 + }, + { + "epoch": 0.22922224502359942, + "grad_norm": 1.9869230086630392, + "learning_rate": 8.995028636955369e-06, + "loss": 0.3677, + "step": 3351 + }, + { + "epoch": 0.22929064915520897, + "grad_norm": 2.0644788208140974, + "learning_rate": 8.994362420006888e-06, + "loss": 0.6405, + "step": 3352 + }, + { + "epoch": 0.22935905328681852, + "grad_norm": 2.383782839581385, + "learning_rate": 8.993696006995812e-06, + "loss": 0.538, + "step": 3353 + }, + { + "epoch": 0.22942745741842807, + "grad_norm": 2.0139126046381097, + "learning_rate": 8.993029397954852e-06, + "loss": 0.4315, + "step": 3354 + }, + { + "epoch": 0.22949586155003762, + "grad_norm": 1.727038982287421, + "learning_rate": 8.992362592916724e-06, + "loss": 0.4766, + "step": 3355 + }, + { + "epoch": 0.22956426568164717, + "grad_norm": 2.270482377760973, + "learning_rate": 8.991695591914162e-06, + "loss": 0.6876, + "step": 3356 + }, + { + "epoch": 0.22963266981325672, + "grad_norm": 2.160624524040048, + "learning_rate": 8.991028394979906e-06, + "loss": 0.3931, + "step": 3357 + }, + { + "epoch": 0.22970107394486627, + "grad_norm": 1.7022507068334516, + "learning_rate": 8.990361002146701e-06, + "loss": 0.2851, + "step": 3358 + }, + { + "epoch": 0.22976947807647582, + "grad_norm": 1.6666419315489478, + "learning_rate": 8.98969341344731e-06, + "loss": 0.4121, + "step": 3359 + }, + { + "epoch": 0.22983788220808538, + "grad_norm": 2.2633234501394903, + "learning_rate": 8.989025628914503e-06, + "loss": 0.4728, + "step": 3360 + }, + { + "epoch": 0.22990628633969493, + "grad_norm": 2.0439632322033465, + "learning_rate": 8.988357648581052e-06, + "loss": 0.5051, + "step": 3361 + }, + { + "epoch": 0.22997469047130448, + "grad_norm": 2.1117468391382155, + "learning_rate": 8.98768947247975e-06, + "loss": 0.4577, + "step": 3362 + }, + { + "epoch": 0.23004309460291403, + "grad_norm": 2.4423186924085636, + "learning_rate": 8.987021100643392e-06, + "loss": 0.6197, + "step": 3363 + }, + { + "epoch": 0.23011149873452358, + "grad_norm": 1.795248767694322, + "learning_rate": 8.986352533104784e-06, + "loss": 0.4127, + "step": 3364 + }, + { + "epoch": 0.2301799028661331, + "grad_norm": 2.267850992507662, + "learning_rate": 8.985683769896745e-06, + "loss": 0.5527, + "step": 3365 + }, + { + "epoch": 0.23024830699774265, + "grad_norm": 1.5974355139425058, + "learning_rate": 8.9850148110521e-06, + "loss": 0.3084, + "step": 3366 + }, + { + "epoch": 0.2303167111293522, + "grad_norm": 2.5204220342986234, + "learning_rate": 8.984345656603684e-06, + "loss": 0.7087, + "step": 3367 + }, + { + "epoch": 0.23038511526096175, + "grad_norm": 1.5736659572633789, + "learning_rate": 8.983676306584344e-06, + "loss": 0.2137, + "step": 3368 + }, + { + "epoch": 0.2304535193925713, + "grad_norm": 2.4377869901678153, + "learning_rate": 8.983006761026931e-06, + "loss": 0.7134, + "step": 3369 + }, + { + "epoch": 0.23052192352418086, + "grad_norm": 1.6891864786690152, + "learning_rate": 8.982337019964316e-06, + "loss": 0.3526, + "step": 3370 + }, + { + "epoch": 0.2305903276557904, + "grad_norm": 2.6094967648083554, + "learning_rate": 8.981667083429368e-06, + "loss": 0.7118, + "step": 3371 + }, + { + "epoch": 0.23065873178739996, + "grad_norm": 1.4015743780323855, + "learning_rate": 8.980996951454972e-06, + "loss": 0.2604, + "step": 3372 + }, + { + "epoch": 0.2307271359190095, + "grad_norm": 2.591379737200857, + "learning_rate": 8.98032662407402e-06, + "loss": 0.2952, + "step": 3373 + }, + { + "epoch": 0.23079554005061906, + "grad_norm": 1.8721533643978587, + "learning_rate": 8.97965610131942e-06, + "loss": 0.3494, + "step": 3374 + }, + { + "epoch": 0.2308639441822286, + "grad_norm": 2.6754681035041976, + "learning_rate": 8.978985383224079e-06, + "loss": 0.2384, + "step": 3375 + }, + { + "epoch": 0.23093234831383816, + "grad_norm": 2.080896286505453, + "learning_rate": 8.97831446982092e-06, + "loss": 0.5561, + "step": 3376 + }, + { + "epoch": 0.2310007524454477, + "grad_norm": 1.6928843807714589, + "learning_rate": 8.977643361142878e-06, + "loss": 0.2333, + "step": 3377 + }, + { + "epoch": 0.23106915657705726, + "grad_norm": 2.19741318270813, + "learning_rate": 8.976972057222892e-06, + "loss": 0.4187, + "step": 3378 + }, + { + "epoch": 0.2311375607086668, + "grad_norm": 2.068755150699254, + "learning_rate": 8.976300558093911e-06, + "loss": 0.4702, + "step": 3379 + }, + { + "epoch": 0.23120596484027636, + "grad_norm": 2.0735620870868336, + "learning_rate": 8.9756288637889e-06, + "loss": 0.428, + "step": 3380 + }, + { + "epoch": 0.2312743689718859, + "grad_norm": 1.6732120915967175, + "learning_rate": 8.974956974340823e-06, + "loss": 0.3047, + "step": 3381 + }, + { + "epoch": 0.23134277310349546, + "grad_norm": 2.463977395464502, + "learning_rate": 8.974284889782666e-06, + "loss": 0.6124, + "step": 3382 + }, + { + "epoch": 0.231411177235105, + "grad_norm": 2.025724186739052, + "learning_rate": 8.973612610147411e-06, + "loss": 0.3897, + "step": 3383 + }, + { + "epoch": 0.23147958136671454, + "grad_norm": 2.1360594902279666, + "learning_rate": 8.972940135468064e-06, + "loss": 0.3176, + "step": 3384 + }, + { + "epoch": 0.2315479854983241, + "grad_norm": 2.6691371822237273, + "learning_rate": 8.97226746577763e-06, + "loss": 0.5991, + "step": 3385 + }, + { + "epoch": 0.23161638962993364, + "grad_norm": 1.9982622464951447, + "learning_rate": 8.971594601109124e-06, + "loss": 0.3802, + "step": 3386 + }, + { + "epoch": 0.2316847937615432, + "grad_norm": 3.0732841476713033, + "learning_rate": 8.970921541495578e-06, + "loss": 0.4693, + "step": 3387 + }, + { + "epoch": 0.23175319789315274, + "grad_norm": 1.6053108801340668, + "learning_rate": 8.970248286970028e-06, + "loss": 0.2133, + "step": 3388 + }, + { + "epoch": 0.2318216020247623, + "grad_norm": 2.257136514307827, + "learning_rate": 8.96957483756552e-06, + "loss": 0.2612, + "step": 3389 + }, + { + "epoch": 0.23189000615637184, + "grad_norm": 1.828168566910109, + "learning_rate": 8.96890119331511e-06, + "loss": 0.3326, + "step": 3390 + }, + { + "epoch": 0.2319584102879814, + "grad_norm": 1.2761990262013643, + "learning_rate": 8.968227354251863e-06, + "loss": 0.2138, + "step": 3391 + }, + { + "epoch": 0.23202681441959094, + "grad_norm": 2.434888932869393, + "learning_rate": 8.967553320408856e-06, + "loss": 0.2973, + "step": 3392 + }, + { + "epoch": 0.2320952185512005, + "grad_norm": 1.7963381919102603, + "learning_rate": 8.966879091819172e-06, + "loss": 0.1846, + "step": 3393 + }, + { + "epoch": 0.23216362268281004, + "grad_norm": 2.198001484040392, + "learning_rate": 8.966204668515907e-06, + "loss": 0.4422, + "step": 3394 + }, + { + "epoch": 0.2322320268144196, + "grad_norm": 1.3805720746674868, + "learning_rate": 8.965530050532162e-06, + "loss": 0.231, + "step": 3395 + }, + { + "epoch": 0.23230043094602915, + "grad_norm": 1.94542960691516, + "learning_rate": 8.964855237901056e-06, + "loss": 0.58, + "step": 3396 + }, + { + "epoch": 0.2323688350776387, + "grad_norm": 2.494946596767632, + "learning_rate": 8.964180230655708e-06, + "loss": 0.2655, + "step": 3397 + }, + { + "epoch": 0.23243723920924825, + "grad_norm": 2.187214097004231, + "learning_rate": 8.963505028829251e-06, + "loss": 0.558, + "step": 3398 + }, + { + "epoch": 0.2325056433408578, + "grad_norm": 2.2440188900272333, + "learning_rate": 8.962829632454829e-06, + "loss": 0.3503, + "step": 3399 + }, + { + "epoch": 0.23257404747246735, + "grad_norm": 2.332601780965508, + "learning_rate": 8.96215404156559e-06, + "loss": 0.2702, + "step": 3400 + }, + { + "epoch": 0.2326424516040769, + "grad_norm": 2.1880855389708485, + "learning_rate": 8.961478256194699e-06, + "loss": 0.3146, + "step": 3401 + }, + { + "epoch": 0.23271085573568642, + "grad_norm": 2.350418832922536, + "learning_rate": 8.960802276375324e-06, + "loss": 0.5249, + "step": 3402 + }, + { + "epoch": 0.23277925986729597, + "grad_norm": 3.11203494906475, + "learning_rate": 8.960126102140647e-06, + "loss": 0.5532, + "step": 3403 + }, + { + "epoch": 0.23284766399890552, + "grad_norm": 2.60621134914582, + "learning_rate": 8.959449733523858e-06, + "loss": 0.6754, + "step": 3404 + }, + { + "epoch": 0.23291606813051507, + "grad_norm": 2.373899041529407, + "learning_rate": 8.958773170558157e-06, + "loss": 0.8006, + "step": 3405 + }, + { + "epoch": 0.23298447226212463, + "grad_norm": 2.161232750775818, + "learning_rate": 8.95809641327675e-06, + "loss": 0.4757, + "step": 3406 + }, + { + "epoch": 0.23305287639373418, + "grad_norm": 1.7195809436430554, + "learning_rate": 8.957419461712862e-06, + "loss": 0.365, + "step": 3407 + }, + { + "epoch": 0.23312128052534373, + "grad_norm": 2.2681389508589627, + "learning_rate": 8.956742315899713e-06, + "loss": 0.6261, + "step": 3408 + }, + { + "epoch": 0.23318968465695328, + "grad_norm": 1.5552361253561937, + "learning_rate": 8.956064975870543e-06, + "loss": 0.3188, + "step": 3409 + }, + { + "epoch": 0.23325808878856283, + "grad_norm": 1.501496695824175, + "learning_rate": 8.955387441658603e-06, + "loss": 0.367, + "step": 3410 + }, + { + "epoch": 0.23332649292017238, + "grad_norm": 3.017862823784093, + "learning_rate": 8.954709713297147e-06, + "loss": 0.5095, + "step": 3411 + }, + { + "epoch": 0.23339489705178193, + "grad_norm": 1.823334865737959, + "learning_rate": 8.95403179081944e-06, + "loss": 0.3457, + "step": 3412 + }, + { + "epoch": 0.23346330118339148, + "grad_norm": 2.341668266220044, + "learning_rate": 8.953353674258761e-06, + "loss": 0.6586, + "step": 3413 + }, + { + "epoch": 0.23353170531500103, + "grad_norm": 2.2139627228237453, + "learning_rate": 8.952675363648391e-06, + "loss": 0.467, + "step": 3414 + }, + { + "epoch": 0.23360010944661058, + "grad_norm": 1.6632884266861108, + "learning_rate": 8.951996859021627e-06, + "loss": 0.356, + "step": 3415 + }, + { + "epoch": 0.23366851357822013, + "grad_norm": 1.6286796393899936, + "learning_rate": 8.951318160411773e-06, + "loss": 0.3176, + "step": 3416 + }, + { + "epoch": 0.23373691770982968, + "grad_norm": 1.233113762690254, + "learning_rate": 8.950639267852144e-06, + "loss": 0.1299, + "step": 3417 + }, + { + "epoch": 0.23380532184143923, + "grad_norm": 1.2060753407428093, + "learning_rate": 8.949960181376062e-06, + "loss": 0.1308, + "step": 3418 + }, + { + "epoch": 0.23387372597304878, + "grad_norm": 1.6746568175229848, + "learning_rate": 8.949280901016859e-06, + "loss": 0.3181, + "step": 3419 + }, + { + "epoch": 0.2339421301046583, + "grad_norm": 2.0597283893694227, + "learning_rate": 8.94860142680788e-06, + "loss": 0.6307, + "step": 3420 + }, + { + "epoch": 0.23401053423626786, + "grad_norm": 1.931739195188072, + "learning_rate": 8.947921758782474e-06, + "loss": 0.5174, + "step": 3421 + }, + { + "epoch": 0.2340789383678774, + "grad_norm": 2.260295680466207, + "learning_rate": 8.947241896974007e-06, + "loss": 0.6179, + "step": 3422 + }, + { + "epoch": 0.23414734249948696, + "grad_norm": 2.1326145038296134, + "learning_rate": 8.946561841415843e-06, + "loss": 0.6058, + "step": 3423 + }, + { + "epoch": 0.2342157466310965, + "grad_norm": 2.3158316647435626, + "learning_rate": 8.945881592141367e-06, + "loss": 0.4059, + "step": 3424 + }, + { + "epoch": 0.23428415076270606, + "grad_norm": 1.5908517102631663, + "learning_rate": 8.945201149183969e-06, + "loss": 0.3912, + "step": 3425 + }, + { + "epoch": 0.2343525548943156, + "grad_norm": 1.5222171297050058, + "learning_rate": 8.944520512577047e-06, + "loss": 0.3516, + "step": 3426 + }, + { + "epoch": 0.23442095902592516, + "grad_norm": 1.658554522930471, + "learning_rate": 8.943839682354009e-06, + "loss": 0.2345, + "step": 3427 + }, + { + "epoch": 0.2344893631575347, + "grad_norm": 1.6771637875547618, + "learning_rate": 8.943158658548277e-06, + "loss": 0.2239, + "step": 3428 + }, + { + "epoch": 0.23455776728914426, + "grad_norm": 2.097160131508657, + "learning_rate": 8.942477441193274e-06, + "loss": 0.4232, + "step": 3429 + }, + { + "epoch": 0.23462617142075382, + "grad_norm": 2.6755817088327563, + "learning_rate": 8.941796030322443e-06, + "loss": 0.6748, + "step": 3430 + }, + { + "epoch": 0.23469457555236337, + "grad_norm": 2.4119750692065995, + "learning_rate": 8.941114425969227e-06, + "loss": 0.6771, + "step": 3431 + }, + { + "epoch": 0.23476297968397292, + "grad_norm": 2.069204959081689, + "learning_rate": 8.940432628167082e-06, + "loss": 0.4592, + "step": 3432 + }, + { + "epoch": 0.23483138381558247, + "grad_norm": 1.8469926941028285, + "learning_rate": 8.939750636949476e-06, + "loss": 0.2558, + "step": 3433 + }, + { + "epoch": 0.23489978794719202, + "grad_norm": 1.5301811614025054, + "learning_rate": 8.939068452349885e-06, + "loss": 0.3521, + "step": 3434 + }, + { + "epoch": 0.23496819207880157, + "grad_norm": 2.262225580170285, + "learning_rate": 8.938386074401793e-06, + "loss": 0.4404, + "step": 3435 + }, + { + "epoch": 0.23503659621041112, + "grad_norm": 1.8437818831580428, + "learning_rate": 8.937703503138693e-06, + "loss": 0.2454, + "step": 3436 + }, + { + "epoch": 0.23510500034202067, + "grad_norm": 1.9158568049099658, + "learning_rate": 8.937020738594091e-06, + "loss": 0.272, + "step": 3437 + }, + { + "epoch": 0.23517340447363022, + "grad_norm": 2.219182200714685, + "learning_rate": 8.936337780801499e-06, + "loss": 0.7001, + "step": 3438 + }, + { + "epoch": 0.23524180860523974, + "grad_norm": 2.717840557797907, + "learning_rate": 8.935654629794442e-06, + "loss": 0.5437, + "step": 3439 + }, + { + "epoch": 0.2353102127368493, + "grad_norm": 1.9215818197036318, + "learning_rate": 8.93497128560645e-06, + "loss": 0.4161, + "step": 3440 + }, + { + "epoch": 0.23537861686845885, + "grad_norm": 1.7942456246881182, + "learning_rate": 8.934287748271062e-06, + "loss": 0.4807, + "step": 3441 + }, + { + "epoch": 0.2354470210000684, + "grad_norm": 1.6936264971366728, + "learning_rate": 8.933604017821835e-06, + "loss": 0.3914, + "step": 3442 + }, + { + "epoch": 0.23551542513167795, + "grad_norm": 1.4351142023662349, + "learning_rate": 8.93292009429233e-06, + "loss": 0.1955, + "step": 3443 + }, + { + "epoch": 0.2355838292632875, + "grad_norm": 2.1318158626266497, + "learning_rate": 8.932235977716112e-06, + "loss": 0.582, + "step": 3444 + }, + { + "epoch": 0.23565223339489705, + "grad_norm": 1.4893892624172549, + "learning_rate": 8.931551668126765e-06, + "loss": 0.3599, + "step": 3445 + }, + { + "epoch": 0.2357206375265066, + "grad_norm": 1.829489589903938, + "learning_rate": 8.930867165557878e-06, + "loss": 0.3859, + "step": 3446 + }, + { + "epoch": 0.23578904165811615, + "grad_norm": 2.2260143341423797, + "learning_rate": 8.930182470043045e-06, + "loss": 0.3505, + "step": 3447 + }, + { + "epoch": 0.2358574457897257, + "grad_norm": 2.1504201029490324, + "learning_rate": 8.92949758161588e-06, + "loss": 0.8239, + "step": 3448 + }, + { + "epoch": 0.23592584992133525, + "grad_norm": 1.806486178833052, + "learning_rate": 8.92881250031e-06, + "loss": 0.2524, + "step": 3449 + }, + { + "epoch": 0.2359942540529448, + "grad_norm": 2.2913384433706376, + "learning_rate": 8.928127226159027e-06, + "loss": 0.6785, + "step": 3450 + }, + { + "epoch": 0.23606265818455435, + "grad_norm": 2.437190537492992, + "learning_rate": 8.927441759196602e-06, + "loss": 0.7093, + "step": 3451 + }, + { + "epoch": 0.2361310623161639, + "grad_norm": 1.8501020538688582, + "learning_rate": 8.926756099456371e-06, + "loss": 0.3573, + "step": 3452 + }, + { + "epoch": 0.23619946644777345, + "grad_norm": 1.9141066721276214, + "learning_rate": 8.926070246971989e-06, + "loss": 0.5458, + "step": 3453 + }, + { + "epoch": 0.236267870579383, + "grad_norm": 1.5677931299014787, + "learning_rate": 8.92538420177712e-06, + "loss": 0.3073, + "step": 3454 + }, + { + "epoch": 0.23633627471099256, + "grad_norm": 2.0792389372329168, + "learning_rate": 8.924697963905438e-06, + "loss": 0.4628, + "step": 3455 + }, + { + "epoch": 0.2364046788426021, + "grad_norm": 2.3749655419911773, + "learning_rate": 8.924011533390628e-06, + "loss": 0.5311, + "step": 3456 + }, + { + "epoch": 0.23647308297421163, + "grad_norm": 1.781769317080661, + "learning_rate": 8.923324910266385e-06, + "loss": 0.3661, + "step": 3457 + }, + { + "epoch": 0.23654148710582118, + "grad_norm": 2.335064466503436, + "learning_rate": 8.922638094566408e-06, + "loss": 0.3274, + "step": 3458 + }, + { + "epoch": 0.23660989123743073, + "grad_norm": 2.378604456357423, + "learning_rate": 8.92195108632441e-06, + "loss": 0.5092, + "step": 3459 + }, + { + "epoch": 0.23667829536904028, + "grad_norm": 1.8926385017582588, + "learning_rate": 8.921263885574116e-06, + "loss": 0.4745, + "step": 3460 + }, + { + "epoch": 0.23674669950064983, + "grad_norm": 1.8072619519323971, + "learning_rate": 8.920576492349252e-06, + "loss": 0.2016, + "step": 3461 + }, + { + "epoch": 0.23681510363225938, + "grad_norm": 1.7305347976824554, + "learning_rate": 8.919888906683565e-06, + "loss": 0.3337, + "step": 3462 + }, + { + "epoch": 0.23688350776386893, + "grad_norm": 2.063372868009961, + "learning_rate": 8.9192011286108e-06, + "loss": 0.4617, + "step": 3463 + }, + { + "epoch": 0.23695191189547848, + "grad_norm": 2.07306950730929, + "learning_rate": 8.918513158164719e-06, + "loss": 0.4073, + "step": 3464 + }, + { + "epoch": 0.23702031602708803, + "grad_norm": 2.0547263790847117, + "learning_rate": 8.917824995379088e-06, + "loss": 0.5052, + "step": 3465 + }, + { + "epoch": 0.23708872015869759, + "grad_norm": 2.0669693567189666, + "learning_rate": 8.917136640287688e-06, + "loss": 0.5114, + "step": 3466 + }, + { + "epoch": 0.23715712429030714, + "grad_norm": 1.6656906173155586, + "learning_rate": 8.916448092924306e-06, + "loss": 0.404, + "step": 3467 + }, + { + "epoch": 0.2372255284219167, + "grad_norm": 2.0519492102294623, + "learning_rate": 8.91575935332274e-06, + "loss": 0.54, + "step": 3468 + }, + { + "epoch": 0.23729393255352624, + "grad_norm": 1.7812872096408205, + "learning_rate": 8.915070421516797e-06, + "loss": 0.1282, + "step": 3469 + }, + { + "epoch": 0.2373623366851358, + "grad_norm": 2.1465526427640906, + "learning_rate": 8.91438129754029e-06, + "loss": 0.5953, + "step": 3470 + }, + { + "epoch": 0.23743074081674534, + "grad_norm": 1.4524615860199348, + "learning_rate": 8.913691981427046e-06, + "loss": 0.2024, + "step": 3471 + }, + { + "epoch": 0.2374991449483549, + "grad_norm": 2.1604638872224995, + "learning_rate": 8.913002473210902e-06, + "loss": 0.1841, + "step": 3472 + }, + { + "epoch": 0.23756754907996444, + "grad_norm": 1.8156286359412446, + "learning_rate": 8.9123127729257e-06, + "loss": 0.3645, + "step": 3473 + }, + { + "epoch": 0.237635953211574, + "grad_norm": 2.2669253793907593, + "learning_rate": 8.911622880605294e-06, + "loss": 0.5878, + "step": 3474 + }, + { + "epoch": 0.23770435734318354, + "grad_norm": 1.7989701495604407, + "learning_rate": 8.91093279628355e-06, + "loss": 0.4244, + "step": 3475 + }, + { + "epoch": 0.23777276147479307, + "grad_norm": 1.355791976822969, + "learning_rate": 8.910242519994337e-06, + "loss": 0.2167, + "step": 3476 + }, + { + "epoch": 0.23784116560640262, + "grad_norm": 2.176841717108824, + "learning_rate": 8.90955205177154e-06, + "loss": 0.4449, + "step": 3477 + }, + { + "epoch": 0.23790956973801217, + "grad_norm": 1.8556500516956176, + "learning_rate": 8.908861391649048e-06, + "loss": 0.2224, + "step": 3478 + }, + { + "epoch": 0.23797797386962172, + "grad_norm": 1.6924224162104182, + "learning_rate": 8.908170539660766e-06, + "loss": 0.3243, + "step": 3479 + }, + { + "epoch": 0.23804637800123127, + "grad_norm": 7.198276791394299, + "learning_rate": 8.907479495840599e-06, + "loss": 0.6166, + "step": 3480 + }, + { + "epoch": 0.23811478213284082, + "grad_norm": 1.9252567737726682, + "learning_rate": 8.90678826022247e-06, + "loss": 0.395, + "step": 3481 + }, + { + "epoch": 0.23818318626445037, + "grad_norm": 1.200398591645429, + "learning_rate": 8.906096832840307e-06, + "loss": 0.161, + "step": 3482 + }, + { + "epoch": 0.23825159039605992, + "grad_norm": 1.965949361765666, + "learning_rate": 8.905405213728049e-06, + "loss": 0.3921, + "step": 3483 + }, + { + "epoch": 0.23831999452766947, + "grad_norm": 1.4900928645109242, + "learning_rate": 8.904713402919645e-06, + "loss": 0.2268, + "step": 3484 + }, + { + "epoch": 0.23838839865927902, + "grad_norm": 1.9559054677250698, + "learning_rate": 8.904021400449051e-06, + "loss": 0.2168, + "step": 3485 + }, + { + "epoch": 0.23845680279088857, + "grad_norm": 3.453650059201485, + "learning_rate": 8.903329206350234e-06, + "loss": 0.1519, + "step": 3486 + }, + { + "epoch": 0.23852520692249812, + "grad_norm": 1.609538819947294, + "learning_rate": 8.902636820657172e-06, + "loss": 0.4191, + "step": 3487 + }, + { + "epoch": 0.23859361105410767, + "grad_norm": 1.7861413553991983, + "learning_rate": 8.901944243403849e-06, + "loss": 0.3772, + "step": 3488 + }, + { + "epoch": 0.23866201518571722, + "grad_norm": 2.85346747742229, + "learning_rate": 8.901251474624258e-06, + "loss": 0.5848, + "step": 3489 + }, + { + "epoch": 0.23873041931732678, + "grad_norm": 2.0073951851547127, + "learning_rate": 8.900558514352406e-06, + "loss": 0.5408, + "step": 3490 + }, + { + "epoch": 0.23879882344893633, + "grad_norm": 2.168022670590807, + "learning_rate": 8.899865362622308e-06, + "loss": 0.5675, + "step": 3491 + }, + { + "epoch": 0.23886722758054588, + "grad_norm": 2.133138471013745, + "learning_rate": 8.899172019467985e-06, + "loss": 0.5543, + "step": 3492 + }, + { + "epoch": 0.23893563171215543, + "grad_norm": 2.214800798918029, + "learning_rate": 8.898478484923472e-06, + "loss": 0.4982, + "step": 3493 + }, + { + "epoch": 0.23900403584376495, + "grad_norm": 1.4266004325444626, + "learning_rate": 8.897784759022806e-06, + "loss": 0.2206, + "step": 3494 + }, + { + "epoch": 0.2390724399753745, + "grad_norm": 2.0080917513018517, + "learning_rate": 8.897090841800046e-06, + "loss": 0.5116, + "step": 3495 + }, + { + "epoch": 0.23914084410698405, + "grad_norm": 2.2399385185404364, + "learning_rate": 8.896396733289246e-06, + "loss": 0.7554, + "step": 3496 + }, + { + "epoch": 0.2392092482385936, + "grad_norm": 1.836493436362826, + "learning_rate": 8.89570243352448e-06, + "loss": 0.4481, + "step": 3497 + }, + { + "epoch": 0.23927765237020315, + "grad_norm": 2.121832365975706, + "learning_rate": 8.895007942539827e-06, + "loss": 0.659, + "step": 3498 + }, + { + "epoch": 0.2393460565018127, + "grad_norm": 2.168810923130151, + "learning_rate": 8.894313260369372e-06, + "loss": 0.1798, + "step": 3499 + }, + { + "epoch": 0.23941446063342225, + "grad_norm": 2.2206554827113125, + "learning_rate": 8.893618387047221e-06, + "loss": 0.3251, + "step": 3500 + }, + { + "epoch": 0.2394828647650318, + "grad_norm": 2.0358725566708817, + "learning_rate": 8.892923322607474e-06, + "loss": 0.4306, + "step": 3501 + }, + { + "epoch": 0.23955126889664136, + "grad_norm": 2.5399767359471683, + "learning_rate": 8.892228067084253e-06, + "loss": 0.582, + "step": 3502 + }, + { + "epoch": 0.2396196730282509, + "grad_norm": 1.7812705038797931, + "learning_rate": 8.891532620511682e-06, + "loss": 0.4987, + "step": 3503 + }, + { + "epoch": 0.23968807715986046, + "grad_norm": 2.1022286050791794, + "learning_rate": 8.8908369829239e-06, + "loss": 0.4661, + "step": 3504 + }, + { + "epoch": 0.23975648129147, + "grad_norm": 2.2559790905649693, + "learning_rate": 8.89014115435505e-06, + "loss": 0.6198, + "step": 3505 + }, + { + "epoch": 0.23982488542307956, + "grad_norm": 1.329849100210843, + "learning_rate": 8.889445134839285e-06, + "loss": 0.1747, + "step": 3506 + }, + { + "epoch": 0.2398932895546891, + "grad_norm": 2.0023917262014987, + "learning_rate": 8.888748924410772e-06, + "loss": 0.4765, + "step": 3507 + }, + { + "epoch": 0.23996169368629866, + "grad_norm": 3.0550494747869466, + "learning_rate": 8.888052523103683e-06, + "loss": 0.4568, + "step": 3508 + }, + { + "epoch": 0.2400300978179082, + "grad_norm": 2.080641005028506, + "learning_rate": 8.887355930952202e-06, + "loss": 0.3749, + "step": 3509 + }, + { + "epoch": 0.24009850194951776, + "grad_norm": 1.7950566420743268, + "learning_rate": 8.886659147990519e-06, + "loss": 0.4156, + "step": 3510 + }, + { + "epoch": 0.2401669060811273, + "grad_norm": 2.489777155770487, + "learning_rate": 8.885962174252836e-06, + "loss": 0.2623, + "step": 3511 + }, + { + "epoch": 0.24023531021273684, + "grad_norm": 1.3561020836396613, + "learning_rate": 8.885265009773365e-06, + "loss": 0.202, + "step": 3512 + }, + { + "epoch": 0.2403037143443464, + "grad_norm": 1.6589199354547361, + "learning_rate": 8.884567654586326e-06, + "loss": 0.4296, + "step": 3513 + }, + { + "epoch": 0.24037211847595594, + "grad_norm": 1.8550351569861918, + "learning_rate": 8.883870108725948e-06, + "loss": 0.3258, + "step": 3514 + }, + { + "epoch": 0.2404405226075655, + "grad_norm": 1.9880744062052373, + "learning_rate": 8.883172372226469e-06, + "loss": 0.5681, + "step": 3515 + }, + { + "epoch": 0.24050892673917504, + "grad_norm": 1.6554816821152638, + "learning_rate": 8.88247444512214e-06, + "loss": 0.4673, + "step": 3516 + }, + { + "epoch": 0.2405773308707846, + "grad_norm": 1.9123340022750883, + "learning_rate": 8.881776327447216e-06, + "loss": 0.4039, + "step": 3517 + }, + { + "epoch": 0.24064573500239414, + "grad_norm": 1.9309296027394156, + "learning_rate": 8.881078019235964e-06, + "loss": 0.5105, + "step": 3518 + }, + { + "epoch": 0.2407141391340037, + "grad_norm": 2.3592054040850803, + "learning_rate": 8.880379520522664e-06, + "loss": 0.3976, + "step": 3519 + }, + { + "epoch": 0.24078254326561324, + "grad_norm": 3.2328142463896925, + "learning_rate": 8.879680831341597e-06, + "loss": 0.4502, + "step": 3520 + }, + { + "epoch": 0.2408509473972228, + "grad_norm": 1.8010351866706105, + "learning_rate": 8.878981951727059e-06, + "loss": 0.3651, + "step": 3521 + }, + { + "epoch": 0.24091935152883234, + "grad_norm": 1.5119062228603852, + "learning_rate": 8.878282881713358e-06, + "loss": 0.2318, + "step": 3522 + }, + { + "epoch": 0.2409877556604419, + "grad_norm": 1.5358733426700846, + "learning_rate": 8.877583621334803e-06, + "loss": 0.2661, + "step": 3523 + }, + { + "epoch": 0.24105615979205144, + "grad_norm": 2.1984545292903066, + "learning_rate": 8.87688417062572e-06, + "loss": 0.4017, + "step": 3524 + }, + { + "epoch": 0.241124563923661, + "grad_norm": 2.4460702459423853, + "learning_rate": 8.876184529620441e-06, + "loss": 0.3518, + "step": 3525 + }, + { + "epoch": 0.24119296805527055, + "grad_norm": 1.9403287173045176, + "learning_rate": 8.875484698353308e-06, + "loss": 0.4788, + "step": 3526 + }, + { + "epoch": 0.2412613721868801, + "grad_norm": 2.5392021962396703, + "learning_rate": 8.87478467685867e-06, + "loss": 0.4332, + "step": 3527 + }, + { + "epoch": 0.24132977631848965, + "grad_norm": 2.2119583159139817, + "learning_rate": 8.874084465170889e-06, + "loss": 0.4514, + "step": 3528 + }, + { + "epoch": 0.2413981804500992, + "grad_norm": 3.2605707672475877, + "learning_rate": 8.873384063324336e-06, + "loss": 0.306, + "step": 3529 + }, + { + "epoch": 0.24146658458170875, + "grad_norm": 1.8229119196330432, + "learning_rate": 8.872683471353387e-06, + "loss": 0.2694, + "step": 3530 + }, + { + "epoch": 0.24153498871331827, + "grad_norm": 2.6961397394255546, + "learning_rate": 8.871982689292434e-06, + "loss": 0.282, + "step": 3531 + }, + { + "epoch": 0.24160339284492782, + "grad_norm": 2.252243572472373, + "learning_rate": 8.871281717175872e-06, + "loss": 0.4989, + "step": 3532 + }, + { + "epoch": 0.24167179697653737, + "grad_norm": 2.4090495358532826, + "learning_rate": 8.870580555038108e-06, + "loss": 0.457, + "step": 3533 + }, + { + "epoch": 0.24174020110814692, + "grad_norm": 2.2280243207487573, + "learning_rate": 8.86987920291356e-06, + "loss": 0.448, + "step": 3534 + }, + { + "epoch": 0.24180860523975647, + "grad_norm": 1.7555733767110766, + "learning_rate": 8.869177660836654e-06, + "loss": 0.3127, + "step": 3535 + }, + { + "epoch": 0.24187700937136603, + "grad_norm": 2.1088918803761953, + "learning_rate": 8.868475928841825e-06, + "loss": 0.3538, + "step": 3536 + }, + { + "epoch": 0.24194541350297558, + "grad_norm": 2.275315348153571, + "learning_rate": 8.867774006963516e-06, + "loss": 0.3256, + "step": 3537 + }, + { + "epoch": 0.24201381763458513, + "grad_norm": 1.6897004426072812, + "learning_rate": 8.867071895236182e-06, + "loss": 0.3224, + "step": 3538 + }, + { + "epoch": 0.24208222176619468, + "grad_norm": 2.1338883086634133, + "learning_rate": 8.866369593694285e-06, + "loss": 0.4675, + "step": 3539 + }, + { + "epoch": 0.24215062589780423, + "grad_norm": 1.7936946754031582, + "learning_rate": 8.865667102372297e-06, + "loss": 0.4276, + "step": 3540 + }, + { + "epoch": 0.24221903002941378, + "grad_norm": 1.8391339498040395, + "learning_rate": 8.864964421304703e-06, + "loss": 0.2484, + "step": 3541 + }, + { + "epoch": 0.24228743416102333, + "grad_norm": 2.1457707448430114, + "learning_rate": 8.864261550525988e-06, + "loss": 0.42, + "step": 3542 + }, + { + "epoch": 0.24235583829263288, + "grad_norm": 2.178574353035766, + "learning_rate": 8.863558490070659e-06, + "loss": 0.5143, + "step": 3543 + }, + { + "epoch": 0.24242424242424243, + "grad_norm": 1.904052452638138, + "learning_rate": 8.86285523997322e-06, + "loss": 0.3821, + "step": 3544 + }, + { + "epoch": 0.24249264655585198, + "grad_norm": 1.6148464205165685, + "learning_rate": 8.862151800268193e-06, + "loss": 0.3176, + "step": 3545 + }, + { + "epoch": 0.24256105068746153, + "grad_norm": 2.122622148270048, + "learning_rate": 8.861448170990105e-06, + "loss": 0.5881, + "step": 3546 + }, + { + "epoch": 0.24262945481907108, + "grad_norm": 1.9329931771428825, + "learning_rate": 8.860744352173496e-06, + "loss": 0.5848, + "step": 3547 + }, + { + "epoch": 0.24269785895068063, + "grad_norm": 2.98110953026054, + "learning_rate": 8.860040343852909e-06, + "loss": 0.403, + "step": 3548 + }, + { + "epoch": 0.24276626308229016, + "grad_norm": 1.8977634033756967, + "learning_rate": 8.859336146062904e-06, + "loss": 0.4909, + "step": 3549 + }, + { + "epoch": 0.2428346672138997, + "grad_norm": 1.897979838146614, + "learning_rate": 8.858631758838042e-06, + "loss": 0.3486, + "step": 3550 + }, + { + "epoch": 0.24290307134550926, + "grad_norm": 2.465969572176545, + "learning_rate": 8.857927182212902e-06, + "loss": 0.3985, + "step": 3551 + }, + { + "epoch": 0.2429714754771188, + "grad_norm": 2.36527055607157, + "learning_rate": 8.857222416222064e-06, + "loss": 0.4891, + "step": 3552 + }, + { + "epoch": 0.24303987960872836, + "grad_norm": 2.0636484279452265, + "learning_rate": 8.856517460900124e-06, + "loss": 0.4619, + "step": 3553 + }, + { + "epoch": 0.2431082837403379, + "grad_norm": 2.066570928029472, + "learning_rate": 8.855812316281686e-06, + "loss": 0.574, + "step": 3554 + }, + { + "epoch": 0.24317668787194746, + "grad_norm": 2.130048754986459, + "learning_rate": 8.85510698240136e-06, + "loss": 0.4201, + "step": 3555 + }, + { + "epoch": 0.243245092003557, + "grad_norm": 1.6300662642010184, + "learning_rate": 8.854401459293765e-06, + "loss": 0.1855, + "step": 3556 + }, + { + "epoch": 0.24331349613516656, + "grad_norm": 2.197068635375456, + "learning_rate": 8.853695746993534e-06, + "loss": 0.5487, + "step": 3557 + }, + { + "epoch": 0.2433819002667761, + "grad_norm": 2.2177105604536544, + "learning_rate": 8.852989845535306e-06, + "loss": 0.5295, + "step": 3558 + }, + { + "epoch": 0.24345030439838566, + "grad_norm": 2.153151749913958, + "learning_rate": 8.852283754953734e-06, + "loss": 0.4264, + "step": 3559 + }, + { + "epoch": 0.24351870852999521, + "grad_norm": 2.627826752448788, + "learning_rate": 8.85157747528347e-06, + "loss": 0.614, + "step": 3560 + }, + { + "epoch": 0.24358711266160477, + "grad_norm": 2.042064050680563, + "learning_rate": 8.850871006559185e-06, + "loss": 0.2703, + "step": 3561 + }, + { + "epoch": 0.24365551679321432, + "grad_norm": 2.1302803364704355, + "learning_rate": 8.850164348815556e-06, + "loss": 0.6816, + "step": 3562 + }, + { + "epoch": 0.24372392092482387, + "grad_norm": 1.566870741813665, + "learning_rate": 8.849457502087266e-06, + "loss": 0.3601, + "step": 3563 + }, + { + "epoch": 0.24379232505643342, + "grad_norm": 2.070985907671996, + "learning_rate": 8.848750466409015e-06, + "loss": 0.4795, + "step": 3564 + }, + { + "epoch": 0.24386072918804297, + "grad_norm": 1.3008078853068212, + "learning_rate": 8.848043241815504e-06, + "loss": 0.2565, + "step": 3565 + }, + { + "epoch": 0.24392913331965252, + "grad_norm": 2.244847859360083, + "learning_rate": 8.84733582834145e-06, + "loss": 0.2973, + "step": 3566 + }, + { + "epoch": 0.24399753745126207, + "grad_norm": 2.0577640200431495, + "learning_rate": 8.846628226021576e-06, + "loss": 0.4992, + "step": 3567 + }, + { + "epoch": 0.2440659415828716, + "grad_norm": 1.4654695369442992, + "learning_rate": 8.845920434890614e-06, + "loss": 0.1986, + "step": 3568 + }, + { + "epoch": 0.24413434571448114, + "grad_norm": 1.9076511361679034, + "learning_rate": 8.845212454983303e-06, + "loss": 0.3336, + "step": 3569 + }, + { + "epoch": 0.2442027498460907, + "grad_norm": 1.3203770040521974, + "learning_rate": 8.844504286334396e-06, + "loss": 0.2045, + "step": 3570 + }, + { + "epoch": 0.24427115397770024, + "grad_norm": 2.291616596773114, + "learning_rate": 8.843795928978656e-06, + "loss": 0.5738, + "step": 3571 + }, + { + "epoch": 0.2443395581093098, + "grad_norm": 1.8079667745125616, + "learning_rate": 8.843087382950847e-06, + "loss": 0.428, + "step": 3572 + }, + { + "epoch": 0.24440796224091935, + "grad_norm": 2.258755239587226, + "learning_rate": 8.842378648285755e-06, + "loss": 0.3358, + "step": 3573 + }, + { + "epoch": 0.2444763663725289, + "grad_norm": 1.8287948538624164, + "learning_rate": 8.841669725018164e-06, + "loss": 0.324, + "step": 3574 + }, + { + "epoch": 0.24454477050413845, + "grad_norm": 2.0022223519515547, + "learning_rate": 8.84096061318287e-06, + "loss": 0.4338, + "step": 3575 + }, + { + "epoch": 0.244613174635748, + "grad_norm": 1.9930901745002272, + "learning_rate": 8.84025131281468e-06, + "loss": 0.2814, + "step": 3576 + }, + { + "epoch": 0.24468157876735755, + "grad_norm": 1.7194561894701617, + "learning_rate": 8.839541823948412e-06, + "loss": 0.1975, + "step": 3577 + }, + { + "epoch": 0.2447499828989671, + "grad_norm": 1.7087632303207096, + "learning_rate": 8.838832146618893e-06, + "loss": 0.2438, + "step": 3578 + }, + { + "epoch": 0.24481838703057665, + "grad_norm": 1.563705365644008, + "learning_rate": 8.838122280860953e-06, + "loss": 0.2875, + "step": 3579 + }, + { + "epoch": 0.2448867911621862, + "grad_norm": 2.1074728124364968, + "learning_rate": 8.837412226709436e-06, + "loss": 0.4431, + "step": 3580 + }, + { + "epoch": 0.24495519529379575, + "grad_norm": 3.0088904047077967, + "learning_rate": 8.836701984199197e-06, + "loss": 0.4924, + "step": 3581 + }, + { + "epoch": 0.2450235994254053, + "grad_norm": 1.9317998675827164, + "learning_rate": 8.835991553365098e-06, + "loss": 0.4655, + "step": 3582 + }, + { + "epoch": 0.24509200355701485, + "grad_norm": 2.521964147946684, + "learning_rate": 8.83528093424201e-06, + "loss": 0.966, + "step": 3583 + }, + { + "epoch": 0.2451604076886244, + "grad_norm": 2.471765499366185, + "learning_rate": 8.834570126864812e-06, + "loss": 0.2922, + "step": 3584 + }, + { + "epoch": 0.24522881182023396, + "grad_norm": 2.081668840110806, + "learning_rate": 8.833859131268397e-06, + "loss": 0.4676, + "step": 3585 + }, + { + "epoch": 0.24529721595184348, + "grad_norm": 3.158962330718265, + "learning_rate": 8.83314794748766e-06, + "loss": 0.44, + "step": 3586 + }, + { + "epoch": 0.24536562008345303, + "grad_norm": 1.1699487017579209, + "learning_rate": 8.832436575557514e-06, + "loss": 0.2112, + "step": 3587 + }, + { + "epoch": 0.24543402421506258, + "grad_norm": 1.8426439677466124, + "learning_rate": 8.831725015512874e-06, + "loss": 0.3395, + "step": 3588 + }, + { + "epoch": 0.24550242834667213, + "grad_norm": 1.5667330289246892, + "learning_rate": 8.831013267388665e-06, + "loss": 0.1343, + "step": 3589 + }, + { + "epoch": 0.24557083247828168, + "grad_norm": 2.1088482764260665, + "learning_rate": 8.830301331219825e-06, + "loss": 0.4535, + "step": 3590 + }, + { + "epoch": 0.24563923660989123, + "grad_norm": 2.2545126986667325, + "learning_rate": 8.829589207041303e-06, + "loss": 0.7592, + "step": 3591 + }, + { + "epoch": 0.24570764074150078, + "grad_norm": 2.584472027794224, + "learning_rate": 8.828876894888046e-06, + "loss": 0.4832, + "step": 3592 + }, + { + "epoch": 0.24577604487311033, + "grad_norm": 1.8064628843561716, + "learning_rate": 8.828164394795024e-06, + "loss": 0.1889, + "step": 3593 + }, + { + "epoch": 0.24584444900471988, + "grad_norm": 2.299841496199348, + "learning_rate": 8.827451706797206e-06, + "loss": 0.4802, + "step": 3594 + }, + { + "epoch": 0.24591285313632943, + "grad_norm": 2.0616525838164583, + "learning_rate": 8.826738830929576e-06, + "loss": 0.6441, + "step": 3595 + }, + { + "epoch": 0.24598125726793899, + "grad_norm": 3.0284048374340693, + "learning_rate": 8.826025767227126e-06, + "loss": 0.3006, + "step": 3596 + }, + { + "epoch": 0.24604966139954854, + "grad_norm": 1.9074144988627313, + "learning_rate": 8.825312515724854e-06, + "loss": 0.4394, + "step": 3597 + }, + { + "epoch": 0.2461180655311581, + "grad_norm": 2.0539720943339606, + "learning_rate": 8.824599076457772e-06, + "loss": 0.283, + "step": 3598 + }, + { + "epoch": 0.24618646966276764, + "grad_norm": 1.9390710106600655, + "learning_rate": 8.823885449460899e-06, + "loss": 0.5663, + "step": 3599 + }, + { + "epoch": 0.2462548737943772, + "grad_norm": 2.331538402293262, + "learning_rate": 8.823171634769263e-06, + "loss": 0.7592, + "step": 3600 + }, + { + "epoch": 0.24632327792598674, + "grad_norm": 2.773616924660778, + "learning_rate": 8.8224576324179e-06, + "loss": 0.495, + "step": 3601 + }, + { + "epoch": 0.2463916820575963, + "grad_norm": 1.8859808053241347, + "learning_rate": 8.82174344244186e-06, + "loss": 0.3877, + "step": 3602 + }, + { + "epoch": 0.24646008618920584, + "grad_norm": 1.8823191760202191, + "learning_rate": 8.821029064876195e-06, + "loss": 0.3491, + "step": 3603 + }, + { + "epoch": 0.2465284903208154, + "grad_norm": 2.455421296309577, + "learning_rate": 8.820314499755974e-06, + "loss": 0.5469, + "step": 3604 + }, + { + "epoch": 0.24659689445242491, + "grad_norm": 2.1029962130390936, + "learning_rate": 8.819599747116268e-06, + "loss": 0.5027, + "step": 3605 + }, + { + "epoch": 0.24666529858403446, + "grad_norm": 1.9042126015820322, + "learning_rate": 8.81888480699216e-06, + "loss": 0.3612, + "step": 3606 + }, + { + "epoch": 0.24673370271564402, + "grad_norm": 1.9233716316627503, + "learning_rate": 8.818169679418747e-06, + "loss": 0.2815, + "step": 3607 + }, + { + "epoch": 0.24680210684725357, + "grad_norm": 1.7686924683242164, + "learning_rate": 8.817454364431127e-06, + "loss": 0.364, + "step": 3608 + }, + { + "epoch": 0.24687051097886312, + "grad_norm": 2.177605430993771, + "learning_rate": 8.816738862064413e-06, + "loss": 0.389, + "step": 3609 + }, + { + "epoch": 0.24693891511047267, + "grad_norm": 2.5262363784834285, + "learning_rate": 8.816023172353725e-06, + "loss": 0.4238, + "step": 3610 + }, + { + "epoch": 0.24700731924208222, + "grad_norm": 2.172584150339515, + "learning_rate": 8.815307295334193e-06, + "loss": 0.2836, + "step": 3611 + }, + { + "epoch": 0.24707572337369177, + "grad_norm": 2.1906824191482017, + "learning_rate": 8.814591231040953e-06, + "loss": 0.5868, + "step": 3612 + }, + { + "epoch": 0.24714412750530132, + "grad_norm": 1.8116815976095986, + "learning_rate": 8.813874979509155e-06, + "loss": 0.3273, + "step": 3613 + }, + { + "epoch": 0.24721253163691087, + "grad_norm": 1.6358884193760206, + "learning_rate": 8.813158540773957e-06, + "loss": 0.3051, + "step": 3614 + }, + { + "epoch": 0.24728093576852042, + "grad_norm": 1.481421893543876, + "learning_rate": 8.812441914870524e-06, + "loss": 0.198, + "step": 3615 + }, + { + "epoch": 0.24734933990012997, + "grad_norm": 1.52977809910466, + "learning_rate": 8.811725101834034e-06, + "loss": 0.3749, + "step": 3616 + }, + { + "epoch": 0.24741774403173952, + "grad_norm": 1.9626965630855975, + "learning_rate": 8.811008101699666e-06, + "loss": 0.503, + "step": 3617 + }, + { + "epoch": 0.24748614816334907, + "grad_norm": 1.5978698004817273, + "learning_rate": 8.81029091450262e-06, + "loss": 0.2308, + "step": 3618 + }, + { + "epoch": 0.24755455229495862, + "grad_norm": 2.0638228025308925, + "learning_rate": 8.809573540278094e-06, + "loss": 0.476, + "step": 3619 + }, + { + "epoch": 0.24762295642656817, + "grad_norm": 2.42749024665201, + "learning_rate": 8.808855979061302e-06, + "loss": 0.3959, + "step": 3620 + }, + { + "epoch": 0.24769136055817773, + "grad_norm": 1.8209562964096926, + "learning_rate": 8.808138230887468e-06, + "loss": 0.2538, + "step": 3621 + }, + { + "epoch": 0.24775976468978728, + "grad_norm": 1.4896951483533563, + "learning_rate": 8.80742029579182e-06, + "loss": 0.2044, + "step": 3622 + }, + { + "epoch": 0.2478281688213968, + "grad_norm": 2.0361125990781366, + "learning_rate": 8.806702173809596e-06, + "loss": 0.4257, + "step": 3623 + }, + { + "epoch": 0.24789657295300635, + "grad_norm": 1.4364812625629924, + "learning_rate": 8.805983864976048e-06, + "loss": 0.2249, + "step": 3624 + }, + { + "epoch": 0.2479649770846159, + "grad_norm": 1.827740604384312, + "learning_rate": 8.805265369326434e-06, + "loss": 0.196, + "step": 3625 + }, + { + "epoch": 0.24803338121622545, + "grad_norm": 1.5292665641772198, + "learning_rate": 8.804546686896018e-06, + "loss": 0.1674, + "step": 3626 + }, + { + "epoch": 0.248101785347835, + "grad_norm": 2.3815581099856784, + "learning_rate": 8.80382781772008e-06, + "loss": 0.5849, + "step": 3627 + }, + { + "epoch": 0.24817018947944455, + "grad_norm": 2.1781791643334176, + "learning_rate": 8.803108761833905e-06, + "loss": 0.318, + "step": 3628 + }, + { + "epoch": 0.2482385936110541, + "grad_norm": 2.2132555937567613, + "learning_rate": 8.802389519272787e-06, + "loss": 0.6308, + "step": 3629 + }, + { + "epoch": 0.24830699774266365, + "grad_norm": 2.1609627527070225, + "learning_rate": 8.801670090072027e-06, + "loss": 0.4136, + "step": 3630 + }, + { + "epoch": 0.2483754018742732, + "grad_norm": 2.2371014586399025, + "learning_rate": 8.800950474266945e-06, + "loss": 0.6158, + "step": 3631 + }, + { + "epoch": 0.24844380600588276, + "grad_norm": 2.279098031360579, + "learning_rate": 8.800230671892856e-06, + "loss": 0.7064, + "step": 3632 + }, + { + "epoch": 0.2485122101374923, + "grad_norm": 1.8945104774830717, + "learning_rate": 8.799510682985097e-06, + "loss": 0.5501, + "step": 3633 + }, + { + "epoch": 0.24858061426910186, + "grad_norm": 1.6574622207539178, + "learning_rate": 8.798790507579004e-06, + "loss": 0.2101, + "step": 3634 + }, + { + "epoch": 0.2486490184007114, + "grad_norm": 2.1902909764296212, + "learning_rate": 8.798070145709929e-06, + "loss": 0.2838, + "step": 3635 + }, + { + "epoch": 0.24871742253232096, + "grad_norm": 1.7880914709900384, + "learning_rate": 8.797349597413232e-06, + "loss": 0.2373, + "step": 3636 + }, + { + "epoch": 0.2487858266639305, + "grad_norm": 2.38296634615171, + "learning_rate": 8.796628862724276e-06, + "loss": 0.6978, + "step": 3637 + }, + { + "epoch": 0.24885423079554006, + "grad_norm": 1.5133595125858899, + "learning_rate": 8.795907941678446e-06, + "loss": 0.2725, + "step": 3638 + }, + { + "epoch": 0.2489226349271496, + "grad_norm": 1.7269293624047266, + "learning_rate": 8.79518683431112e-06, + "loss": 0.4093, + "step": 3639 + }, + { + "epoch": 0.24899103905875916, + "grad_norm": 1.9576208098415249, + "learning_rate": 8.794465540657701e-06, + "loss": 0.3221, + "step": 3640 + }, + { + "epoch": 0.24905944319036868, + "grad_norm": 1.049447654514485, + "learning_rate": 8.793744060753588e-06, + "loss": 0.22, + "step": 3641 + }, + { + "epoch": 0.24912784732197824, + "grad_norm": 2.3012237511898563, + "learning_rate": 8.793022394634197e-06, + "loss": 0.3883, + "step": 3642 + }, + { + "epoch": 0.24919625145358779, + "grad_norm": 2.11698375580213, + "learning_rate": 8.792300542334951e-06, + "loss": 0.3533, + "step": 3643 + }, + { + "epoch": 0.24926465558519734, + "grad_norm": 1.7337182236243671, + "learning_rate": 8.791578503891281e-06, + "loss": 0.2609, + "step": 3644 + }, + { + "epoch": 0.2493330597168069, + "grad_norm": 1.7511901129530842, + "learning_rate": 8.790856279338629e-06, + "loss": 0.4407, + "step": 3645 + }, + { + "epoch": 0.24940146384841644, + "grad_norm": 2.1596324347639473, + "learning_rate": 8.790133868712445e-06, + "loss": 0.1855, + "step": 3646 + }, + { + "epoch": 0.249469867980026, + "grad_norm": 2.198313484408368, + "learning_rate": 8.789411272048188e-06, + "loss": 0.5711, + "step": 3647 + }, + { + "epoch": 0.24953827211163554, + "grad_norm": 2.2928317287387974, + "learning_rate": 8.788688489381327e-06, + "loss": 0.5492, + "step": 3648 + }, + { + "epoch": 0.2496066762432451, + "grad_norm": 3.2532646948056643, + "learning_rate": 8.78796552074734e-06, + "loss": 0.6677, + "step": 3649 + }, + { + "epoch": 0.24967508037485464, + "grad_norm": 1.8311217679769864, + "learning_rate": 8.787242366181714e-06, + "loss": 0.5966, + "step": 3650 + }, + { + "epoch": 0.2497434845064642, + "grad_norm": 1.8553319298342186, + "learning_rate": 8.786519025719944e-06, + "loss": 0.359, + "step": 3651 + }, + { + "epoch": 0.24981188863807374, + "grad_norm": 1.891693995260158, + "learning_rate": 8.785795499397533e-06, + "loss": 0.2838, + "step": 3652 + }, + { + "epoch": 0.2498802927696833, + "grad_norm": 2.290456666264614, + "learning_rate": 8.78507178725e-06, + "loss": 0.4469, + "step": 3653 + }, + { + "epoch": 0.24994869690129284, + "grad_norm": 2.483378756404268, + "learning_rate": 8.784347889312865e-06, + "loss": 0.4657, + "step": 3654 + }, + { + "epoch": 0.25001710103290237, + "grad_norm": 2.2983699008069536, + "learning_rate": 8.783623805621659e-06, + "loss": 0.5729, + "step": 3655 + }, + { + "epoch": 0.25008550516451195, + "grad_norm": 1.285135735692996, + "learning_rate": 8.782899536211928e-06, + "loss": 0.1799, + "step": 3656 + }, + { + "epoch": 0.25015390929612147, + "grad_norm": 1.8662356759771705, + "learning_rate": 8.78217508111922e-06, + "loss": 0.3413, + "step": 3657 + }, + { + "epoch": 0.25022231342773105, + "grad_norm": 1.7629667543337244, + "learning_rate": 8.781450440379096e-06, + "loss": 0.3712, + "step": 3658 + }, + { + "epoch": 0.25029071755934057, + "grad_norm": 2.2711354690462473, + "learning_rate": 8.780725614027123e-06, + "loss": 0.4553, + "step": 3659 + }, + { + "epoch": 0.25035912169095015, + "grad_norm": 2.0187022280281424, + "learning_rate": 8.780000602098878e-06, + "loss": 0.3502, + "step": 3660 + }, + { + "epoch": 0.25042752582255967, + "grad_norm": 2.039252352736915, + "learning_rate": 8.779275404629954e-06, + "loss": 0.5877, + "step": 3661 + }, + { + "epoch": 0.25049592995416925, + "grad_norm": 2.1675391794027417, + "learning_rate": 8.77855002165594e-06, + "loss": 0.3533, + "step": 3662 + }, + { + "epoch": 0.2505643340857788, + "grad_norm": 1.7882136617531168, + "learning_rate": 8.777824453212447e-06, + "loss": 0.3767, + "step": 3663 + }, + { + "epoch": 0.25063273821738835, + "grad_norm": 2.1116128565423797, + "learning_rate": 8.777098699335084e-06, + "loss": 0.6099, + "step": 3664 + }, + { + "epoch": 0.2507011423489979, + "grad_norm": 2.165364937353115, + "learning_rate": 8.77637276005948e-06, + "loss": 0.352, + "step": 3665 + }, + { + "epoch": 0.25076954648060745, + "grad_norm": 3.2337274133448624, + "learning_rate": 8.775646635421264e-06, + "loss": 0.4948, + "step": 3666 + }, + { + "epoch": 0.250837950612217, + "grad_norm": 1.559403976140635, + "learning_rate": 8.774920325456079e-06, + "loss": 0.276, + "step": 3667 + }, + { + "epoch": 0.25090635474382655, + "grad_norm": 2.167591522664861, + "learning_rate": 8.774193830199576e-06, + "loss": 0.3207, + "step": 3668 + }, + { + "epoch": 0.2509747588754361, + "grad_norm": 1.6440792595234432, + "learning_rate": 8.773467149687413e-06, + "loss": 0.358, + "step": 3669 + }, + { + "epoch": 0.2510431630070456, + "grad_norm": 2.7852819128164215, + "learning_rate": 8.77274028395526e-06, + "loss": 0.8373, + "step": 3670 + }, + { + "epoch": 0.2511115671386552, + "grad_norm": 2.4423113434595027, + "learning_rate": 8.772013233038797e-06, + "loss": 0.474, + "step": 3671 + }, + { + "epoch": 0.2511799712702647, + "grad_norm": 2.3258825421660263, + "learning_rate": 8.771285996973709e-06, + "loss": 0.918, + "step": 3672 + }, + { + "epoch": 0.2512483754018743, + "grad_norm": 2.9545007760664546, + "learning_rate": 8.770558575795693e-06, + "loss": 0.404, + "step": 3673 + }, + { + "epoch": 0.2513167795334838, + "grad_norm": 1.7217458884266403, + "learning_rate": 8.769830969540456e-06, + "loss": 0.3518, + "step": 3674 + }, + { + "epoch": 0.2513851836650934, + "grad_norm": 2.003971775463403, + "learning_rate": 8.769103178243708e-06, + "loss": 0.3623, + "step": 3675 + }, + { + "epoch": 0.2514535877967029, + "grad_norm": 1.8082085615328936, + "learning_rate": 8.768375201941178e-06, + "loss": 0.2018, + "step": 3676 + }, + { + "epoch": 0.2515219919283125, + "grad_norm": 1.8862159485520995, + "learning_rate": 8.767647040668592e-06, + "loss": 0.3639, + "step": 3677 + }, + { + "epoch": 0.251590396059922, + "grad_norm": 2.6315151436656468, + "learning_rate": 8.766918694461698e-06, + "loss": 0.362, + "step": 3678 + }, + { + "epoch": 0.2516588001915316, + "grad_norm": 1.8320863914731722, + "learning_rate": 8.766190163356243e-06, + "loss": 0.2959, + "step": 3679 + }, + { + "epoch": 0.2517272043231411, + "grad_norm": 1.7049511035217206, + "learning_rate": 8.76546144738799e-06, + "loss": 0.3818, + "step": 3680 + }, + { + "epoch": 0.2517956084547507, + "grad_norm": 1.6887435569612037, + "learning_rate": 8.764732546592704e-06, + "loss": 0.3837, + "step": 3681 + }, + { + "epoch": 0.2518640125863602, + "grad_norm": 2.02527917170873, + "learning_rate": 8.764003461006166e-06, + "loss": 0.5322, + "step": 3682 + }, + { + "epoch": 0.2519324167179698, + "grad_norm": 2.003351078873881, + "learning_rate": 8.763274190664162e-06, + "loss": 0.2953, + "step": 3683 + }, + { + "epoch": 0.2520008208495793, + "grad_norm": 2.2859151704835763, + "learning_rate": 8.762544735602487e-06, + "loss": 0.5583, + "step": 3684 + }, + { + "epoch": 0.2520692249811889, + "grad_norm": 2.2749387366042253, + "learning_rate": 8.761815095856947e-06, + "loss": 0.8159, + "step": 3685 + }, + { + "epoch": 0.2521376291127984, + "grad_norm": 1.9396273639391912, + "learning_rate": 8.761085271463359e-06, + "loss": 0.3525, + "step": 3686 + }, + { + "epoch": 0.252206033244408, + "grad_norm": 2.0548205133179693, + "learning_rate": 8.760355262457542e-06, + "loss": 0.6201, + "step": 3687 + }, + { + "epoch": 0.2522744373760175, + "grad_norm": 1.6944557389441395, + "learning_rate": 8.75962506887533e-06, + "loss": 0.3277, + "step": 3688 + }, + { + "epoch": 0.25234284150762704, + "grad_norm": 1.797659329606941, + "learning_rate": 8.758894690752563e-06, + "loss": 0.2138, + "step": 3689 + }, + { + "epoch": 0.2524112456392366, + "grad_norm": 1.6512452390141548, + "learning_rate": 8.758164128125097e-06, + "loss": 0.3783, + "step": 3690 + }, + { + "epoch": 0.25247964977084614, + "grad_norm": 1.7174631416469248, + "learning_rate": 8.757433381028786e-06, + "loss": 0.2687, + "step": 3691 + }, + { + "epoch": 0.2525480539024557, + "grad_norm": 1.8930565433082918, + "learning_rate": 8.7567024494995e-06, + "loss": 0.2716, + "step": 3692 + }, + { + "epoch": 0.25261645803406524, + "grad_norm": 2.1830797221645644, + "learning_rate": 8.755971333573116e-06, + "loss": 0.3664, + "step": 3693 + }, + { + "epoch": 0.2526848621656748, + "grad_norm": 2.4176063574163944, + "learning_rate": 8.755240033285522e-06, + "loss": 0.7091, + "step": 3694 + }, + { + "epoch": 0.25275326629728434, + "grad_norm": 1.527432865947932, + "learning_rate": 8.754508548672613e-06, + "loss": 0.1719, + "step": 3695 + }, + { + "epoch": 0.2528216704288939, + "grad_norm": 1.6138218419632795, + "learning_rate": 8.753776879770295e-06, + "loss": 0.2039, + "step": 3696 + }, + { + "epoch": 0.25289007456050344, + "grad_norm": 2.1521821939306878, + "learning_rate": 8.753045026614479e-06, + "loss": 0.4514, + "step": 3697 + }, + { + "epoch": 0.252958478692113, + "grad_norm": 2.335390428763787, + "learning_rate": 8.752312989241092e-06, + "loss": 0.5962, + "step": 3698 + }, + { + "epoch": 0.25302688282372254, + "grad_norm": 1.9959557218662032, + "learning_rate": 8.751580767686063e-06, + "loss": 0.4497, + "step": 3699 + }, + { + "epoch": 0.2530952869553321, + "grad_norm": 2.0132321118955496, + "learning_rate": 8.750848361985333e-06, + "loss": 0.468, + "step": 3700 + }, + { + "epoch": 0.25316369108694164, + "grad_norm": 2.568965477734852, + "learning_rate": 8.750115772174854e-06, + "loss": 0.8495, + "step": 3701 + }, + { + "epoch": 0.2532320952185512, + "grad_norm": 2.041180019055073, + "learning_rate": 8.749382998290584e-06, + "loss": 0.4763, + "step": 3702 + }, + { + "epoch": 0.25330049935016075, + "grad_norm": 1.70332986798147, + "learning_rate": 8.748650040368489e-06, + "loss": 0.4163, + "step": 3703 + }, + { + "epoch": 0.2533689034817703, + "grad_norm": 1.632620745446328, + "learning_rate": 8.747916898444549e-06, + "loss": 0.2239, + "step": 3704 + }, + { + "epoch": 0.25343730761337985, + "grad_norm": 1.5255900800866282, + "learning_rate": 8.74718357255475e-06, + "loss": 0.3609, + "step": 3705 + }, + { + "epoch": 0.25350571174498937, + "grad_norm": 2.028021443754688, + "learning_rate": 8.746450062735085e-06, + "loss": 0.5544, + "step": 3706 + }, + { + "epoch": 0.25357411587659895, + "grad_norm": 2.2266186757279236, + "learning_rate": 8.745716369021559e-06, + "loss": 0.7075, + "step": 3707 + }, + { + "epoch": 0.25364252000820847, + "grad_norm": 1.986409210980634, + "learning_rate": 8.744982491450187e-06, + "loss": 0.4369, + "step": 3708 + }, + { + "epoch": 0.25371092413981805, + "grad_norm": 1.5007005535792608, + "learning_rate": 8.74424843005699e-06, + "loss": 0.3156, + "step": 3709 + }, + { + "epoch": 0.2537793282714276, + "grad_norm": 1.512980742642881, + "learning_rate": 8.743514184877999e-06, + "loss": 0.2189, + "step": 3710 + }, + { + "epoch": 0.25384773240303715, + "grad_norm": 1.4958015261082451, + "learning_rate": 8.742779755949256e-06, + "loss": 0.2232, + "step": 3711 + }, + { + "epoch": 0.2539161365346467, + "grad_norm": 2.1284142091826848, + "learning_rate": 8.742045143306806e-06, + "loss": 0.6236, + "step": 3712 + }, + { + "epoch": 0.25398454066625625, + "grad_norm": 2.2605434058442757, + "learning_rate": 8.741310346986711e-06, + "loss": 0.4444, + "step": 3713 + }, + { + "epoch": 0.2540529447978658, + "grad_norm": 2.119175758597951, + "learning_rate": 8.74057536702504e-06, + "loss": 0.3769, + "step": 3714 + }, + { + "epoch": 0.25412134892947535, + "grad_norm": 1.8587876682530793, + "learning_rate": 8.739840203457864e-06, + "loss": 0.3487, + "step": 3715 + }, + { + "epoch": 0.2541897530610849, + "grad_norm": 1.9922895068812096, + "learning_rate": 8.739104856321272e-06, + "loss": 0.1879, + "step": 3716 + }, + { + "epoch": 0.25425815719269446, + "grad_norm": 2.435866591110902, + "learning_rate": 8.738369325651359e-06, + "loss": 0.7173, + "step": 3717 + }, + { + "epoch": 0.254326561324304, + "grad_norm": 1.5917684213166399, + "learning_rate": 8.737633611484226e-06, + "loss": 0.2553, + "step": 3718 + }, + { + "epoch": 0.25439496545591356, + "grad_norm": 1.57149044379898, + "learning_rate": 8.736897713855988e-06, + "loss": 0.2554, + "step": 3719 + }, + { + "epoch": 0.2544633695875231, + "grad_norm": 1.8541311401721232, + "learning_rate": 8.736161632802763e-06, + "loss": 0.44, + "step": 3720 + }, + { + "epoch": 0.25453177371913266, + "grad_norm": 1.6829840801965599, + "learning_rate": 8.735425368360684e-06, + "loss": 0.1808, + "step": 3721 + }, + { + "epoch": 0.2546001778507422, + "grad_norm": 1.3613732766937454, + "learning_rate": 8.734688920565888e-06, + "loss": 0.2093, + "step": 3722 + }, + { + "epoch": 0.25466858198235176, + "grad_norm": 1.4646367193786696, + "learning_rate": 8.733952289454528e-06, + "loss": 0.22, + "step": 3723 + }, + { + "epoch": 0.2547369861139613, + "grad_norm": 1.6256174212121057, + "learning_rate": 8.733215475062758e-06, + "loss": 0.2365, + "step": 3724 + }, + { + "epoch": 0.2548053902455708, + "grad_norm": 1.657682975271702, + "learning_rate": 8.732478477426744e-06, + "loss": 0.2447, + "step": 3725 + }, + { + "epoch": 0.2548737943771804, + "grad_norm": 1.8677997104193498, + "learning_rate": 8.731741296582663e-06, + "loss": 0.4387, + "step": 3726 + }, + { + "epoch": 0.2549421985087899, + "grad_norm": 2.1135873343727853, + "learning_rate": 8.731003932566697e-06, + "loss": 0.5533, + "step": 3727 + }, + { + "epoch": 0.2550106026403995, + "grad_norm": 2.759523676736638, + "learning_rate": 8.730266385415043e-06, + "loss": 0.3617, + "step": 3728 + }, + { + "epoch": 0.255079006772009, + "grad_norm": 4.112128709421683, + "learning_rate": 8.729528655163901e-06, + "loss": 0.4859, + "step": 3729 + }, + { + "epoch": 0.2551474109036186, + "grad_norm": 2.419982376758586, + "learning_rate": 8.728790741849485e-06, + "loss": 0.5855, + "step": 3730 + }, + { + "epoch": 0.2552158150352281, + "grad_norm": 1.9982265119447644, + "learning_rate": 8.72805264550801e-06, + "loss": 0.3847, + "step": 3731 + }, + { + "epoch": 0.2552842191668377, + "grad_norm": 2.2566644462290912, + "learning_rate": 8.72731436617571e-06, + "loss": 0.3604, + "step": 3732 + }, + { + "epoch": 0.2553526232984472, + "grad_norm": 2.1020678582644208, + "learning_rate": 8.726575903888821e-06, + "loss": 0.4631, + "step": 3733 + }, + { + "epoch": 0.2554210274300568, + "grad_norm": 1.8410632673866005, + "learning_rate": 8.725837258683591e-06, + "loss": 0.4353, + "step": 3734 + }, + { + "epoch": 0.2554894315616663, + "grad_norm": 2.072987020119765, + "learning_rate": 8.72509843059628e-06, + "loss": 0.4351, + "step": 3735 + }, + { + "epoch": 0.2555578356932759, + "grad_norm": 2.6921516792451534, + "learning_rate": 8.724359419663146e-06, + "loss": 0.4995, + "step": 3736 + }, + { + "epoch": 0.2556262398248854, + "grad_norm": 1.5112624255916651, + "learning_rate": 8.723620225920469e-06, + "loss": 0.1542, + "step": 3737 + }, + { + "epoch": 0.255694643956495, + "grad_norm": 2.4393911117111555, + "learning_rate": 8.722880849404529e-06, + "loss": 0.5003, + "step": 3738 + }, + { + "epoch": 0.2557630480881045, + "grad_norm": 1.8627549233807337, + "learning_rate": 8.722141290151618e-06, + "loss": 0.342, + "step": 3739 + }, + { + "epoch": 0.2558314522197141, + "grad_norm": 2.6753815456820584, + "learning_rate": 8.721401548198042e-06, + "loss": 0.5105, + "step": 3740 + }, + { + "epoch": 0.2558998563513236, + "grad_norm": 2.518497538870404, + "learning_rate": 8.720661623580104e-06, + "loss": 0.3576, + "step": 3741 + }, + { + "epoch": 0.2559682604829332, + "grad_norm": 2.146392688407079, + "learning_rate": 8.719921516334129e-06, + "loss": 0.6106, + "step": 3742 + }, + { + "epoch": 0.2560366646145427, + "grad_norm": 2.485822040267069, + "learning_rate": 8.71918122649644e-06, + "loss": 0.7004, + "step": 3743 + }, + { + "epoch": 0.25610506874615224, + "grad_norm": 1.9206860826848366, + "learning_rate": 8.71844075410338e-06, + "loss": 0.508, + "step": 3744 + }, + { + "epoch": 0.2561734728777618, + "grad_norm": 2.329333394986348, + "learning_rate": 8.71770009919129e-06, + "loss": 0.6457, + "step": 3745 + }, + { + "epoch": 0.25624187700937134, + "grad_norm": 1.8954936070723056, + "learning_rate": 8.716959261796526e-06, + "loss": 0.3805, + "step": 3746 + }, + { + "epoch": 0.2563102811409809, + "grad_norm": 2.273370230554536, + "learning_rate": 8.716218241955455e-06, + "loss": 0.4966, + "step": 3747 + }, + { + "epoch": 0.25637868527259045, + "grad_norm": 2.283750176463671, + "learning_rate": 8.715477039704443e-06, + "loss": 0.4653, + "step": 3748 + }, + { + "epoch": 0.2564470894042, + "grad_norm": 1.39575788367551, + "learning_rate": 8.71473565507988e-06, + "loss": 0.2971, + "step": 3749 + }, + { + "epoch": 0.25651549353580955, + "grad_norm": 2.2552370278087186, + "learning_rate": 8.713994088118149e-06, + "loss": 0.469, + "step": 3750 + }, + { + "epoch": 0.2565838976674191, + "grad_norm": 2.596040214425104, + "learning_rate": 8.713252338855656e-06, + "loss": 0.7142, + "step": 3751 + }, + { + "epoch": 0.25665230179902865, + "grad_norm": 2.1290524073834263, + "learning_rate": 8.712510407328807e-06, + "loss": 0.5432, + "step": 3752 + }, + { + "epoch": 0.2567207059306382, + "grad_norm": 2.321450304490317, + "learning_rate": 8.71176829357402e-06, + "loss": 0.7617, + "step": 3753 + }, + { + "epoch": 0.25678911006224775, + "grad_norm": 1.9196335584789157, + "learning_rate": 8.71102599762772e-06, + "loss": 0.3639, + "step": 3754 + }, + { + "epoch": 0.25685751419385733, + "grad_norm": 2.1197299471843696, + "learning_rate": 8.710283519526344e-06, + "loss": 0.5749, + "step": 3755 + }, + { + "epoch": 0.25692591832546685, + "grad_norm": 1.6805559884964918, + "learning_rate": 8.709540859306336e-06, + "loss": 0.3122, + "step": 3756 + }, + { + "epoch": 0.25699432245707643, + "grad_norm": 1.9838112458339603, + "learning_rate": 8.708798017004149e-06, + "loss": 0.2606, + "step": 3757 + }, + { + "epoch": 0.25706272658868595, + "grad_norm": 1.7567792101910173, + "learning_rate": 8.708054992656247e-06, + "loss": 0.3571, + "step": 3758 + }, + { + "epoch": 0.25713113072029553, + "grad_norm": 1.822708517634056, + "learning_rate": 8.7073117862991e-06, + "loss": 0.329, + "step": 3759 + }, + { + "epoch": 0.25719953485190505, + "grad_norm": 1.4015993836396776, + "learning_rate": 8.706568397969187e-06, + "loss": 0.3071, + "step": 3760 + }, + { + "epoch": 0.2572679389835146, + "grad_norm": 2.2762195043390685, + "learning_rate": 8.705824827703e-06, + "loss": 0.7629, + "step": 3761 + }, + { + "epoch": 0.25733634311512416, + "grad_norm": 2.6367764954071706, + "learning_rate": 8.705081075537033e-06, + "loss": 0.3541, + "step": 3762 + }, + { + "epoch": 0.2574047472467337, + "grad_norm": 2.0008557917581555, + "learning_rate": 8.704337141507796e-06, + "loss": 0.4508, + "step": 3763 + }, + { + "epoch": 0.25747315137834326, + "grad_norm": 1.753745511342216, + "learning_rate": 8.703593025651805e-06, + "loss": 0.3974, + "step": 3764 + }, + { + "epoch": 0.2575415555099528, + "grad_norm": 2.048584784502026, + "learning_rate": 8.702848728005582e-06, + "loss": 0.5299, + "step": 3765 + }, + { + "epoch": 0.25760995964156236, + "grad_norm": 2.261239094842328, + "learning_rate": 8.702104248605663e-06, + "loss": 0.5388, + "step": 3766 + }, + { + "epoch": 0.2576783637731719, + "grad_norm": 1.8413010998929729, + "learning_rate": 8.701359587488592e-06, + "loss": 0.3819, + "step": 3767 + }, + { + "epoch": 0.25774676790478146, + "grad_norm": 2.3412005991288587, + "learning_rate": 8.700614744690916e-06, + "loss": 0.2172, + "step": 3768 + }, + { + "epoch": 0.257815172036391, + "grad_norm": 1.8809207339790488, + "learning_rate": 8.699869720249201e-06, + "loss": 0.3209, + "step": 3769 + }, + { + "epoch": 0.25788357616800056, + "grad_norm": 2.485911883612426, + "learning_rate": 8.69912451420001e-06, + "loss": 0.5342, + "step": 3770 + }, + { + "epoch": 0.2579519802996101, + "grad_norm": 1.7415387694916884, + "learning_rate": 8.698379126579928e-06, + "loss": 0.2969, + "step": 3771 + }, + { + "epoch": 0.25802038443121966, + "grad_norm": 1.8338953469862282, + "learning_rate": 8.69763355742554e-06, + "loss": 0.1724, + "step": 3772 + }, + { + "epoch": 0.2580887885628292, + "grad_norm": 2.5803881681993546, + "learning_rate": 8.69688780677344e-06, + "loss": 0.6057, + "step": 3773 + }, + { + "epoch": 0.25815719269443876, + "grad_norm": 2.0541530102500887, + "learning_rate": 8.696141874660233e-06, + "loss": 0.4126, + "step": 3774 + }, + { + "epoch": 0.2582255968260483, + "grad_norm": 2.337944068393449, + "learning_rate": 8.695395761122535e-06, + "loss": 0.587, + "step": 3775 + }, + { + "epoch": 0.25829400095765787, + "grad_norm": 1.968884138048676, + "learning_rate": 8.69464946619697e-06, + "loss": 0.5089, + "step": 3776 + }, + { + "epoch": 0.2583624050892674, + "grad_norm": 2.6022034391096223, + "learning_rate": 8.693902989920165e-06, + "loss": 0.7562, + "step": 3777 + }, + { + "epoch": 0.25843080922087697, + "grad_norm": 2.392131991150399, + "learning_rate": 8.693156332328765e-06, + "loss": 0.8327, + "step": 3778 + }, + { + "epoch": 0.2584992133524865, + "grad_norm": 1.6956156841898158, + "learning_rate": 8.69240949345942e-06, + "loss": 0.3107, + "step": 3779 + }, + { + "epoch": 0.258567617484096, + "grad_norm": 2.060832838129626, + "learning_rate": 8.691662473348783e-06, + "loss": 0.3524, + "step": 3780 + }, + { + "epoch": 0.2586360216157056, + "grad_norm": 1.8884068868284047, + "learning_rate": 8.690915272033528e-06, + "loss": 0.37, + "step": 3781 + }, + { + "epoch": 0.2587044257473151, + "grad_norm": 1.4164256797842358, + "learning_rate": 8.690167889550327e-06, + "loss": 0.2325, + "step": 3782 + }, + { + "epoch": 0.2587728298789247, + "grad_norm": 1.8767353689721784, + "learning_rate": 8.689420325935866e-06, + "loss": 0.2096, + "step": 3783 + }, + { + "epoch": 0.2588412340105342, + "grad_norm": 1.393956212753046, + "learning_rate": 8.68867258122684e-06, + "loss": 0.1875, + "step": 3784 + }, + { + "epoch": 0.2589096381421438, + "grad_norm": 2.1067997647921053, + "learning_rate": 8.68792465545995e-06, + "loss": 0.3897, + "step": 3785 + }, + { + "epoch": 0.2589780422737533, + "grad_norm": 1.8632475032189542, + "learning_rate": 8.68717654867191e-06, + "loss": 0.3883, + "step": 3786 + }, + { + "epoch": 0.2590464464053629, + "grad_norm": 1.7184449445331187, + "learning_rate": 8.68642826089944e-06, + "loss": 0.3526, + "step": 3787 + }, + { + "epoch": 0.2591148505369724, + "grad_norm": 2.0258202497183246, + "learning_rate": 8.68567979217927e-06, + "loss": 0.5084, + "step": 3788 + }, + { + "epoch": 0.259183254668582, + "grad_norm": 1.9287518267582706, + "learning_rate": 8.684931142548137e-06, + "loss": 0.2246, + "step": 3789 + }, + { + "epoch": 0.2592516588001915, + "grad_norm": 2.3037049405422967, + "learning_rate": 8.684182312042789e-06, + "loss": 0.2189, + "step": 3790 + }, + { + "epoch": 0.2593200629318011, + "grad_norm": 1.878456220443114, + "learning_rate": 8.683433300699982e-06, + "loss": 0.3433, + "step": 3791 + }, + { + "epoch": 0.2593884670634106, + "grad_norm": 1.3875051068736077, + "learning_rate": 8.682684108556483e-06, + "loss": 0.1318, + "step": 3792 + }, + { + "epoch": 0.2594568711950202, + "grad_norm": 2.1891093158548887, + "learning_rate": 8.681934735649064e-06, + "loss": 0.5984, + "step": 3793 + }, + { + "epoch": 0.2595252753266297, + "grad_norm": 1.2248973112467503, + "learning_rate": 8.681185182014507e-06, + "loss": 0.1175, + "step": 3794 + }, + { + "epoch": 0.2595936794582393, + "grad_norm": 2.1675285442303123, + "learning_rate": 8.680435447689606e-06, + "loss": 0.4836, + "step": 3795 + }, + { + "epoch": 0.2596620835898488, + "grad_norm": 2.153402871000302, + "learning_rate": 8.67968553271116e-06, + "loss": 0.5, + "step": 3796 + }, + { + "epoch": 0.2597304877214584, + "grad_norm": 2.175436016478446, + "learning_rate": 8.67893543711598e-06, + "loss": 0.5922, + "step": 3797 + }, + { + "epoch": 0.2597988918530679, + "grad_norm": 1.499194415651855, + "learning_rate": 8.678185160940883e-06, + "loss": 0.3355, + "step": 3798 + }, + { + "epoch": 0.25986729598467745, + "grad_norm": 2.0722922522068266, + "learning_rate": 8.677434704222697e-06, + "loss": 0.4605, + "step": 3799 + }, + { + "epoch": 0.259935700116287, + "grad_norm": 2.323172614333435, + "learning_rate": 8.676684066998255e-06, + "loss": 0.6241, + "step": 3800 + }, + { + "epoch": 0.26000410424789655, + "grad_norm": 9.519955468694734, + "learning_rate": 8.675933249304407e-06, + "loss": 0.502, + "step": 3801 + }, + { + "epoch": 0.26007250837950613, + "grad_norm": 2.1312801389716625, + "learning_rate": 8.675182251178005e-06, + "loss": 0.3534, + "step": 3802 + }, + { + "epoch": 0.26014091251111565, + "grad_norm": 1.9934119257801661, + "learning_rate": 8.674431072655909e-06, + "loss": 0.5786, + "step": 3803 + }, + { + "epoch": 0.26020931664272523, + "grad_norm": 2.7630081080588, + "learning_rate": 8.673679713774992e-06, + "loss": 0.3604, + "step": 3804 + }, + { + "epoch": 0.26027772077433475, + "grad_norm": 2.0349233644170783, + "learning_rate": 8.672928174572136e-06, + "loss": 0.5232, + "step": 3805 + }, + { + "epoch": 0.26034612490594433, + "grad_norm": 2.036252939902397, + "learning_rate": 8.672176455084229e-06, + "loss": 0.6172, + "step": 3806 + }, + { + "epoch": 0.26041452903755385, + "grad_norm": 2.3644884194777975, + "learning_rate": 8.671424555348167e-06, + "loss": 0.3678, + "step": 3807 + }, + { + "epoch": 0.26048293316916343, + "grad_norm": 2.4554214965759433, + "learning_rate": 8.67067247540086e-06, + "loss": 0.5772, + "step": 3808 + }, + { + "epoch": 0.26055133730077296, + "grad_norm": 2.5063302184226055, + "learning_rate": 8.669920215279222e-06, + "loss": 0.3721, + "step": 3809 + }, + { + "epoch": 0.26061974143238253, + "grad_norm": 1.826949467062984, + "learning_rate": 8.669167775020178e-06, + "loss": 0.3768, + "step": 3810 + }, + { + "epoch": 0.26068814556399206, + "grad_norm": 1.6256497349630066, + "learning_rate": 8.668415154660662e-06, + "loss": 0.4092, + "step": 3811 + }, + { + "epoch": 0.26075654969560164, + "grad_norm": 1.7576536827397595, + "learning_rate": 8.667662354237614e-06, + "loss": 0.2178, + "step": 3812 + }, + { + "epoch": 0.26082495382721116, + "grad_norm": 2.060150705011086, + "learning_rate": 8.666909373787989e-06, + "loss": 0.4869, + "step": 3813 + }, + { + "epoch": 0.26089335795882074, + "grad_norm": 1.984831412084155, + "learning_rate": 8.666156213348743e-06, + "loss": 0.3708, + "step": 3814 + }, + { + "epoch": 0.26096176209043026, + "grad_norm": 1.9266886865480586, + "learning_rate": 8.665402872956847e-06, + "loss": 0.3288, + "step": 3815 + }, + { + "epoch": 0.26103016622203984, + "grad_norm": 2.1908461591065347, + "learning_rate": 8.664649352649276e-06, + "loss": 0.2639, + "step": 3816 + }, + { + "epoch": 0.26109857035364936, + "grad_norm": 1.8620266639324567, + "learning_rate": 8.663895652463021e-06, + "loss": 0.4257, + "step": 3817 + }, + { + "epoch": 0.2611669744852589, + "grad_norm": 2.1656954832290705, + "learning_rate": 8.663141772435075e-06, + "loss": 0.591, + "step": 3818 + }, + { + "epoch": 0.26123537861686846, + "grad_norm": 1.3493180298628091, + "learning_rate": 8.662387712602438e-06, + "loss": 0.2619, + "step": 3819 + }, + { + "epoch": 0.261303782748478, + "grad_norm": 2.665904485041762, + "learning_rate": 8.66163347300213e-06, + "loss": 0.3421, + "step": 3820 + }, + { + "epoch": 0.26137218688008756, + "grad_norm": 1.9290154711615592, + "learning_rate": 8.660879053671168e-06, + "loss": 0.4318, + "step": 3821 + }, + { + "epoch": 0.2614405910116971, + "grad_norm": 2.7018155360599807, + "learning_rate": 8.660124454646584e-06, + "loss": 0.2322, + "step": 3822 + }, + { + "epoch": 0.26150899514330667, + "grad_norm": 2.349457302229959, + "learning_rate": 8.659369675965418e-06, + "loss": 0.5342, + "step": 3823 + }, + { + "epoch": 0.2615773992749162, + "grad_norm": 1.8366194696176317, + "learning_rate": 8.658614717664718e-06, + "loss": 0.3883, + "step": 3824 + }, + { + "epoch": 0.26164580340652577, + "grad_norm": 2.0240216976763095, + "learning_rate": 8.657859579781538e-06, + "loss": 0.5571, + "step": 3825 + }, + { + "epoch": 0.2617142075381353, + "grad_norm": 1.8285551869187844, + "learning_rate": 8.657104262352946e-06, + "loss": 0.3653, + "step": 3826 + }, + { + "epoch": 0.26178261166974487, + "grad_norm": 2.005179612255417, + "learning_rate": 8.656348765416019e-06, + "loss": 0.3802, + "step": 3827 + }, + { + "epoch": 0.2618510158013544, + "grad_norm": 1.9189289936798664, + "learning_rate": 8.655593089007836e-06, + "loss": 0.2619, + "step": 3828 + }, + { + "epoch": 0.26191941993296397, + "grad_norm": 2.127379109414131, + "learning_rate": 8.654837233165494e-06, + "loss": 0.5341, + "step": 3829 + }, + { + "epoch": 0.2619878240645735, + "grad_norm": 1.6009877755392659, + "learning_rate": 8.65408119792609e-06, + "loss": 0.2322, + "step": 3830 + }, + { + "epoch": 0.26205622819618307, + "grad_norm": 1.4619977541037399, + "learning_rate": 8.653324983326737e-06, + "loss": 0.2392, + "step": 3831 + }, + { + "epoch": 0.2621246323277926, + "grad_norm": 1.2345202185520918, + "learning_rate": 8.652568589404549e-06, + "loss": 0.2034, + "step": 3832 + }, + { + "epoch": 0.2621930364594022, + "grad_norm": 2.6308600206947808, + "learning_rate": 8.65181201619666e-06, + "loss": 0.557, + "step": 3833 + }, + { + "epoch": 0.2622614405910117, + "grad_norm": 1.688455492178315, + "learning_rate": 8.651055263740201e-06, + "loss": 0.1623, + "step": 3834 + }, + { + "epoch": 0.2623298447226212, + "grad_norm": 1.9866932483675976, + "learning_rate": 8.650298332072318e-06, + "loss": 0.4874, + "step": 3835 + }, + { + "epoch": 0.2623982488542308, + "grad_norm": 2.258626794325928, + "learning_rate": 8.649541221230169e-06, + "loss": 0.3789, + "step": 3836 + }, + { + "epoch": 0.2624666529858403, + "grad_norm": 1.9610611925970278, + "learning_rate": 8.64878393125091e-06, + "loss": 0.3832, + "step": 3837 + }, + { + "epoch": 0.2625350571174499, + "grad_norm": 2.2543048340930087, + "learning_rate": 8.648026462171717e-06, + "loss": 0.6259, + "step": 3838 + }, + { + "epoch": 0.2626034612490594, + "grad_norm": 1.5975898015350474, + "learning_rate": 8.64726881402977e-06, + "loss": 0.1636, + "step": 3839 + }, + { + "epoch": 0.262671865380669, + "grad_norm": 1.5962105624734764, + "learning_rate": 8.646510986862256e-06, + "loss": 0.2852, + "step": 3840 + }, + { + "epoch": 0.2627402695122785, + "grad_norm": 1.9310832516430922, + "learning_rate": 8.645752980706374e-06, + "loss": 0.4397, + "step": 3841 + }, + { + "epoch": 0.2628086736438881, + "grad_norm": 2.2147478422629576, + "learning_rate": 8.644994795599331e-06, + "loss": 0.493, + "step": 3842 + }, + { + "epoch": 0.2628770777754976, + "grad_norm": 2.1284704409749673, + "learning_rate": 8.644236431578343e-06, + "loss": 0.6002, + "step": 3843 + }, + { + "epoch": 0.2629454819071072, + "grad_norm": 2.060186208102829, + "learning_rate": 8.643477888680632e-06, + "loss": 0.4839, + "step": 3844 + }, + { + "epoch": 0.2630138860387167, + "grad_norm": 6.195336225158751, + "learning_rate": 8.642719166943433e-06, + "loss": 0.5023, + "step": 3845 + }, + { + "epoch": 0.2630822901703263, + "grad_norm": 2.117289903141984, + "learning_rate": 8.641960266403983e-06, + "loss": 0.6401, + "step": 3846 + }, + { + "epoch": 0.26315069430193583, + "grad_norm": 1.869967944213416, + "learning_rate": 8.64120118709954e-06, + "loss": 0.3164, + "step": 3847 + }, + { + "epoch": 0.2632190984335454, + "grad_norm": 1.607202109311805, + "learning_rate": 8.640441929067358e-06, + "loss": 0.293, + "step": 3848 + }, + { + "epoch": 0.26328750256515493, + "grad_norm": 1.9545599739561677, + "learning_rate": 8.639682492344708e-06, + "loss": 0.4276, + "step": 3849 + }, + { + "epoch": 0.2633559066967645, + "grad_norm": 2.437758982178156, + "learning_rate": 8.638922876968865e-06, + "loss": 0.7525, + "step": 3850 + }, + { + "epoch": 0.26342431082837403, + "grad_norm": 1.5856593033655244, + "learning_rate": 8.638163082977115e-06, + "loss": 0.2553, + "step": 3851 + }, + { + "epoch": 0.2634927149599836, + "grad_norm": 2.433192965599134, + "learning_rate": 8.637403110406751e-06, + "loss": 0.537, + "step": 3852 + }, + { + "epoch": 0.26356111909159313, + "grad_norm": 2.5949626294056296, + "learning_rate": 8.636642959295078e-06, + "loss": 0.9505, + "step": 3853 + }, + { + "epoch": 0.26362952322320266, + "grad_norm": 1.90582216652219, + "learning_rate": 8.635882629679408e-06, + "loss": 0.5669, + "step": 3854 + }, + { + "epoch": 0.26369792735481223, + "grad_norm": 2.1725265534175233, + "learning_rate": 8.635122121597061e-06, + "loss": 0.5309, + "step": 3855 + }, + { + "epoch": 0.26376633148642176, + "grad_norm": 2.035341905996142, + "learning_rate": 8.634361435085368e-06, + "loss": 0.5246, + "step": 3856 + }, + { + "epoch": 0.26383473561803134, + "grad_norm": 1.6533394568518407, + "learning_rate": 8.633600570181663e-06, + "loss": 0.1804, + "step": 3857 + }, + { + "epoch": 0.26390313974964086, + "grad_norm": 3.295835132132199, + "learning_rate": 8.632839526923297e-06, + "loss": 0.1985, + "step": 3858 + }, + { + "epoch": 0.26397154388125044, + "grad_norm": 1.2468129327809534, + "learning_rate": 8.632078305347623e-06, + "loss": 0.1854, + "step": 3859 + }, + { + "epoch": 0.26403994801285996, + "grad_norm": 1.9565702797735496, + "learning_rate": 8.631316905492009e-06, + "loss": 0.3297, + "step": 3860 + }, + { + "epoch": 0.26410835214446954, + "grad_norm": 1.955008883779249, + "learning_rate": 8.630555327393823e-06, + "loss": 0.2879, + "step": 3861 + }, + { + "epoch": 0.26417675627607906, + "grad_norm": 2.0488992575051435, + "learning_rate": 8.629793571090451e-06, + "loss": 0.4984, + "step": 3862 + }, + { + "epoch": 0.26424516040768864, + "grad_norm": 1.7760875832074836, + "learning_rate": 8.629031636619283e-06, + "loss": 0.3188, + "step": 3863 + }, + { + "epoch": 0.26431356453929816, + "grad_norm": 2.067728755155042, + "learning_rate": 8.628269524017717e-06, + "loss": 0.5573, + "step": 3864 + }, + { + "epoch": 0.26438196867090774, + "grad_norm": 1.944857944480552, + "learning_rate": 8.627507233323163e-06, + "loss": 0.4424, + "step": 3865 + }, + { + "epoch": 0.26445037280251726, + "grad_norm": 1.23743311675388, + "learning_rate": 8.626744764573037e-06, + "loss": 0.169, + "step": 3866 + }, + { + "epoch": 0.26451877693412684, + "grad_norm": 1.5815296508190633, + "learning_rate": 8.625982117804764e-06, + "loss": 0.2638, + "step": 3867 + }, + { + "epoch": 0.26458718106573637, + "grad_norm": 2.821175662435869, + "learning_rate": 8.625219293055778e-06, + "loss": 0.4743, + "step": 3868 + }, + { + "epoch": 0.26465558519734594, + "grad_norm": 1.9781892964910344, + "learning_rate": 8.624456290363523e-06, + "loss": 0.3754, + "step": 3869 + }, + { + "epoch": 0.26472398932895547, + "grad_norm": 2.5484299330420686, + "learning_rate": 8.623693109765452e-06, + "loss": 0.6238, + "step": 3870 + }, + { + "epoch": 0.26479239346056505, + "grad_norm": 2.287668584906689, + "learning_rate": 8.622929751299025e-06, + "loss": 0.3879, + "step": 3871 + }, + { + "epoch": 0.26486079759217457, + "grad_norm": 2.4621015248089524, + "learning_rate": 8.622166215001711e-06, + "loss": 0.4079, + "step": 3872 + }, + { + "epoch": 0.2649292017237841, + "grad_norm": 1.3934450888357568, + "learning_rate": 8.621402500910986e-06, + "loss": 0.1583, + "step": 3873 + }, + { + "epoch": 0.26499760585539367, + "grad_norm": 1.6018696885796704, + "learning_rate": 8.62063860906434e-06, + "loss": 0.2791, + "step": 3874 + }, + { + "epoch": 0.2650660099870032, + "grad_norm": 3.406317893935247, + "learning_rate": 8.619874539499266e-06, + "loss": 0.6191, + "step": 3875 + }, + { + "epoch": 0.26513441411861277, + "grad_norm": 1.9182451247159906, + "learning_rate": 8.619110292253273e-06, + "loss": 0.2444, + "step": 3876 + }, + { + "epoch": 0.2652028182502223, + "grad_norm": 1.3062672986933945, + "learning_rate": 8.618345867363866e-06, + "loss": 0.1437, + "step": 3877 + }, + { + "epoch": 0.2652712223818319, + "grad_norm": 2.3875945961271943, + "learning_rate": 8.617581264868573e-06, + "loss": 0.5152, + "step": 3878 + }, + { + "epoch": 0.2653396265134414, + "grad_norm": 2.530157076987974, + "learning_rate": 8.616816484804922e-06, + "loss": 0.8101, + "step": 3879 + }, + { + "epoch": 0.265408030645051, + "grad_norm": 1.905944854514485, + "learning_rate": 8.616051527210453e-06, + "loss": 0.3887, + "step": 3880 + }, + { + "epoch": 0.2654764347766605, + "grad_norm": 2.2708170099867475, + "learning_rate": 8.615286392122712e-06, + "loss": 0.4395, + "step": 3881 + }, + { + "epoch": 0.2655448389082701, + "grad_norm": 1.658408782391197, + "learning_rate": 8.614521079579258e-06, + "loss": 0.1705, + "step": 3882 + }, + { + "epoch": 0.2656132430398796, + "grad_norm": 1.942266745573328, + "learning_rate": 8.613755589617657e-06, + "loss": 0.3971, + "step": 3883 + }, + { + "epoch": 0.2656816471714892, + "grad_norm": 1.8224431449226826, + "learning_rate": 8.612989922275477e-06, + "loss": 0.4277, + "step": 3884 + }, + { + "epoch": 0.2657500513030987, + "grad_norm": 2.6065149966115304, + "learning_rate": 8.612224077590308e-06, + "loss": 0.8541, + "step": 3885 + }, + { + "epoch": 0.2658184554347083, + "grad_norm": 2.110347061117121, + "learning_rate": 8.611458055599735e-06, + "loss": 0.4799, + "step": 3886 + }, + { + "epoch": 0.2658868595663178, + "grad_norm": 1.4312031117717292, + "learning_rate": 8.610691856341364e-06, + "loss": 0.3196, + "step": 3887 + }, + { + "epoch": 0.2659552636979274, + "grad_norm": 2.0278286436710773, + "learning_rate": 8.609925479852798e-06, + "loss": 0.6279, + "step": 3888 + }, + { + "epoch": 0.2660236678295369, + "grad_norm": 2.2387219672522, + "learning_rate": 8.60915892617166e-06, + "loss": 0.5679, + "step": 3889 + }, + { + "epoch": 0.2660920719611464, + "grad_norm": 1.941374227770128, + "learning_rate": 8.608392195335573e-06, + "loss": 0.441, + "step": 3890 + }, + { + "epoch": 0.266160476092756, + "grad_norm": 2.2857101421233494, + "learning_rate": 8.60762528738217e-06, + "loss": 0.5554, + "step": 3891 + }, + { + "epoch": 0.2662288802243655, + "grad_norm": 2.195736613321091, + "learning_rate": 8.606858202349099e-06, + "loss": 0.6298, + "step": 3892 + }, + { + "epoch": 0.2662972843559751, + "grad_norm": 1.5587885885425532, + "learning_rate": 8.60609094027401e-06, + "loss": 0.355, + "step": 3893 + }, + { + "epoch": 0.26636568848758463, + "grad_norm": 1.7134342157498346, + "learning_rate": 8.605323501194562e-06, + "loss": 0.1745, + "step": 3894 + }, + { + "epoch": 0.2664340926191942, + "grad_norm": 2.077042444678629, + "learning_rate": 8.60455588514843e-06, + "loss": 0.4935, + "step": 3895 + }, + { + "epoch": 0.26650249675080373, + "grad_norm": 3.6354414256800336, + "learning_rate": 8.603788092173286e-06, + "loss": 0.5822, + "step": 3896 + }, + { + "epoch": 0.2665709008824133, + "grad_norm": 1.6882665535544443, + "learning_rate": 8.603020122306821e-06, + "loss": 0.2442, + "step": 3897 + }, + { + "epoch": 0.26663930501402283, + "grad_norm": 1.673763095834394, + "learning_rate": 8.602251975586731e-06, + "loss": 0.2906, + "step": 3898 + }, + { + "epoch": 0.2667077091456324, + "grad_norm": 1.669072562065785, + "learning_rate": 8.601483652050717e-06, + "loss": 0.238, + "step": 3899 + }, + { + "epoch": 0.26677611327724193, + "grad_norm": 1.6922579376410996, + "learning_rate": 8.600715151736494e-06, + "loss": 0.4097, + "step": 3900 + }, + { + "epoch": 0.2668445174088515, + "grad_norm": 1.7096846709718694, + "learning_rate": 8.599946474681785e-06, + "loss": 0.3562, + "step": 3901 + }, + { + "epoch": 0.26691292154046103, + "grad_norm": 1.8082657407416736, + "learning_rate": 8.599177620924317e-06, + "loss": 0.3458, + "step": 3902 + }, + { + "epoch": 0.2669813256720706, + "grad_norm": 2.473066088081072, + "learning_rate": 8.598408590501832e-06, + "loss": 0.3083, + "step": 3903 + }, + { + "epoch": 0.26704972980368014, + "grad_norm": 2.54277741610395, + "learning_rate": 8.597639383452078e-06, + "loss": 0.595, + "step": 3904 + }, + { + "epoch": 0.2671181339352897, + "grad_norm": 2.319413155324377, + "learning_rate": 8.59686999981281e-06, + "loss": 0.537, + "step": 3905 + }, + { + "epoch": 0.26718653806689924, + "grad_norm": 1.9968734905374521, + "learning_rate": 8.596100439621793e-06, + "loss": 0.5454, + "step": 3906 + }, + { + "epoch": 0.2672549421985088, + "grad_norm": 1.4992530572334326, + "learning_rate": 8.595330702916799e-06, + "loss": 0.2688, + "step": 3907 + }, + { + "epoch": 0.26732334633011834, + "grad_norm": 2.0812412573942933, + "learning_rate": 8.594560789735615e-06, + "loss": 0.4504, + "step": 3908 + }, + { + "epoch": 0.26739175046172786, + "grad_norm": 2.558563057518825, + "learning_rate": 8.59379070011603e-06, + "loss": 0.8897, + "step": 3909 + }, + { + "epoch": 0.26746015459333744, + "grad_norm": 1.004069482023372, + "learning_rate": 8.593020434095841e-06, + "loss": 0.1314, + "step": 3910 + }, + { + "epoch": 0.26752855872494696, + "grad_norm": 2.0079041359915397, + "learning_rate": 8.592249991712859e-06, + "loss": 0.5178, + "step": 3911 + }, + { + "epoch": 0.26759696285655654, + "grad_norm": 2.1369007214428817, + "learning_rate": 8.5914793730049e-06, + "loss": 0.5574, + "step": 3912 + }, + { + "epoch": 0.26766536698816606, + "grad_norm": 2.103515532023393, + "learning_rate": 8.590708578009792e-06, + "loss": 0.2078, + "step": 3913 + }, + { + "epoch": 0.26773377111977564, + "grad_norm": 2.46978534041782, + "learning_rate": 8.589937606765367e-06, + "loss": 0.4073, + "step": 3914 + }, + { + "epoch": 0.26780217525138517, + "grad_norm": 1.9110494175387152, + "learning_rate": 8.58916645930947e-06, + "loss": 0.416, + "step": 3915 + }, + { + "epoch": 0.26787057938299474, + "grad_norm": 1.658881804526956, + "learning_rate": 8.588395135679949e-06, + "loss": 0.3177, + "step": 3916 + }, + { + "epoch": 0.26793898351460427, + "grad_norm": 1.739872514840667, + "learning_rate": 8.587623635914668e-06, + "loss": 0.3393, + "step": 3917 + }, + { + "epoch": 0.26800738764621385, + "grad_norm": 1.7476860014864672, + "learning_rate": 8.586851960051495e-06, + "loss": 0.3033, + "step": 3918 + }, + { + "epoch": 0.26807579177782337, + "grad_norm": 1.9247172397236543, + "learning_rate": 8.586080108128304e-06, + "loss": 0.5074, + "step": 3919 + }, + { + "epoch": 0.26814419590943295, + "grad_norm": 1.7074336225864346, + "learning_rate": 8.585308080182987e-06, + "loss": 0.2669, + "step": 3920 + }, + { + "epoch": 0.26821260004104247, + "grad_norm": 2.0397790840853576, + "learning_rate": 8.584535876253435e-06, + "loss": 0.2908, + "step": 3921 + }, + { + "epoch": 0.26828100417265205, + "grad_norm": 2.0386903741146707, + "learning_rate": 8.583763496377553e-06, + "loss": 0.3386, + "step": 3922 + }, + { + "epoch": 0.26834940830426157, + "grad_norm": 1.6815852910418312, + "learning_rate": 8.582990940593254e-06, + "loss": 0.3571, + "step": 3923 + }, + { + "epoch": 0.26841781243587115, + "grad_norm": 1.5893370760602008, + "learning_rate": 8.582218208938455e-06, + "loss": 0.3061, + "step": 3924 + }, + { + "epoch": 0.2684862165674807, + "grad_norm": 1.648904815531504, + "learning_rate": 8.58144530145109e-06, + "loss": 0.2769, + "step": 3925 + }, + { + "epoch": 0.26855462069909025, + "grad_norm": 2.008462978632934, + "learning_rate": 8.580672218169095e-06, + "loss": 0.5699, + "step": 3926 + }, + { + "epoch": 0.2686230248306998, + "grad_norm": 2.138884512028475, + "learning_rate": 8.579898959130414e-06, + "loss": 0.4186, + "step": 3927 + }, + { + "epoch": 0.2686914289623093, + "grad_norm": 1.6897165390787237, + "learning_rate": 8.57912552437301e-06, + "loss": 0.2451, + "step": 3928 + }, + { + "epoch": 0.2687598330939189, + "grad_norm": 2.2800145134965843, + "learning_rate": 8.578351913934837e-06, + "loss": 0.5711, + "step": 3929 + }, + { + "epoch": 0.2688282372255284, + "grad_norm": 1.740951826892215, + "learning_rate": 8.577578127853874e-06, + "loss": 0.3552, + "step": 3930 + }, + { + "epoch": 0.268896641357138, + "grad_norm": 2.1306702611205193, + "learning_rate": 8.5768041661681e-06, + "loss": 0.4907, + "step": 3931 + }, + { + "epoch": 0.2689650454887475, + "grad_norm": 2.0091472560445176, + "learning_rate": 8.576030028915507e-06, + "loss": 0.3814, + "step": 3932 + }, + { + "epoch": 0.2690334496203571, + "grad_norm": 2.179000595856857, + "learning_rate": 8.575255716134091e-06, + "loss": 0.7577, + "step": 3933 + }, + { + "epoch": 0.2691018537519666, + "grad_norm": 2.0447238709667714, + "learning_rate": 8.57448122786186e-06, + "loss": 0.4723, + "step": 3934 + }, + { + "epoch": 0.2691702578835762, + "grad_norm": 1.6486858614722504, + "learning_rate": 8.573706564136828e-06, + "loss": 0.284, + "step": 3935 + }, + { + "epoch": 0.2692386620151857, + "grad_norm": 1.9723785091559676, + "learning_rate": 8.572931724997021e-06, + "loss": 0.2637, + "step": 3936 + }, + { + "epoch": 0.2693070661467953, + "grad_norm": 2.1279486856229606, + "learning_rate": 8.572156710480472e-06, + "loss": 0.582, + "step": 3937 + }, + { + "epoch": 0.2693754702784048, + "grad_norm": 1.338422391143996, + "learning_rate": 8.571381520625222e-06, + "loss": 0.3125, + "step": 3938 + }, + { + "epoch": 0.2694438744100144, + "grad_norm": 2.282206795198353, + "learning_rate": 8.57060615546932e-06, + "loss": 0.5787, + "step": 3939 + }, + { + "epoch": 0.2695122785416239, + "grad_norm": 1.351371150721508, + "learning_rate": 8.569830615050827e-06, + "loss": 0.3315, + "step": 3940 + }, + { + "epoch": 0.2695806826732335, + "grad_norm": 1.7527139918836292, + "learning_rate": 8.569054899407806e-06, + "loss": 0.3104, + "step": 3941 + }, + { + "epoch": 0.269649086804843, + "grad_norm": 2.60834043517583, + "learning_rate": 8.56827900857834e-06, + "loss": 0.2023, + "step": 3942 + }, + { + "epoch": 0.2697174909364526, + "grad_norm": 1.983742350255282, + "learning_rate": 8.567502942600508e-06, + "loss": 0.4199, + "step": 3943 + }, + { + "epoch": 0.2697858950680621, + "grad_norm": 1.959871678774161, + "learning_rate": 8.566726701512402e-06, + "loss": 0.4878, + "step": 3944 + }, + { + "epoch": 0.2698542991996717, + "grad_norm": 2.3387877193958877, + "learning_rate": 8.565950285352127e-06, + "loss": 0.566, + "step": 3945 + }, + { + "epoch": 0.2699227033312812, + "grad_norm": 1.9256012041218211, + "learning_rate": 8.565173694157791e-06, + "loss": 0.3312, + "step": 3946 + }, + { + "epoch": 0.26999110746289073, + "grad_norm": 1.8402221112459054, + "learning_rate": 8.564396927967518e-06, + "loss": 0.3693, + "step": 3947 + }, + { + "epoch": 0.2700595115945003, + "grad_norm": 2.6713605686665622, + "learning_rate": 8.563619986819429e-06, + "loss": 0.3114, + "step": 3948 + }, + { + "epoch": 0.27012791572610984, + "grad_norm": 2.066191690082871, + "learning_rate": 8.562842870751662e-06, + "loss": 0.317, + "step": 3949 + }, + { + "epoch": 0.2701963198577194, + "grad_norm": 1.778274870726061, + "learning_rate": 8.562065579802362e-06, + "loss": 0.287, + "step": 3950 + }, + { + "epoch": 0.27026472398932894, + "grad_norm": 2.053023694498807, + "learning_rate": 8.561288114009682e-06, + "loss": 0.3211, + "step": 3951 + }, + { + "epoch": 0.2703331281209385, + "grad_norm": 1.973252623650423, + "learning_rate": 8.560510473411786e-06, + "loss": 0.442, + "step": 3952 + }, + { + "epoch": 0.27040153225254804, + "grad_norm": 4.041339947908623, + "learning_rate": 8.559732658046842e-06, + "loss": 0.4209, + "step": 3953 + }, + { + "epoch": 0.2704699363841576, + "grad_norm": 2.7671850890499665, + "learning_rate": 8.558954667953028e-06, + "loss": 0.3396, + "step": 3954 + }, + { + "epoch": 0.27053834051576714, + "grad_norm": 2.2073887695789445, + "learning_rate": 8.558176503168533e-06, + "loss": 0.3087, + "step": 3955 + }, + { + "epoch": 0.2706067446473767, + "grad_norm": 1.5711279549993908, + "learning_rate": 8.557398163731553e-06, + "loss": 0.3441, + "step": 3956 + }, + { + "epoch": 0.27067514877898624, + "grad_norm": 1.86764242703443, + "learning_rate": 8.556619649680294e-06, + "loss": 0.1573, + "step": 3957 + }, + { + "epoch": 0.2707435529105958, + "grad_norm": 3.298457303839305, + "learning_rate": 8.555840961052968e-06, + "loss": 0.452, + "step": 3958 + }, + { + "epoch": 0.27081195704220534, + "grad_norm": 1.6690283186054578, + "learning_rate": 8.555062097887796e-06, + "loss": 0.176, + "step": 3959 + }, + { + "epoch": 0.2708803611738149, + "grad_norm": 1.7855543489869328, + "learning_rate": 8.554283060223008e-06, + "loss": 0.3476, + "step": 3960 + }, + { + "epoch": 0.27094876530542444, + "grad_norm": 1.6605008653984812, + "learning_rate": 8.553503848096845e-06, + "loss": 0.3402, + "step": 3961 + }, + { + "epoch": 0.271017169437034, + "grad_norm": 2.264013822465326, + "learning_rate": 8.552724461547553e-06, + "loss": 0.5354, + "step": 3962 + }, + { + "epoch": 0.27108557356864355, + "grad_norm": 4.844696346251014, + "learning_rate": 8.551944900613389e-06, + "loss": 0.6512, + "step": 3963 + }, + { + "epoch": 0.27115397770025307, + "grad_norm": 2.0858768604336793, + "learning_rate": 8.551165165332616e-06, + "loss": 0.2883, + "step": 3964 + }, + { + "epoch": 0.27122238183186265, + "grad_norm": 2.3863953591380898, + "learning_rate": 8.55038525574351e-06, + "loss": 0.2383, + "step": 3965 + }, + { + "epoch": 0.27129078596347217, + "grad_norm": 2.035820129682618, + "learning_rate": 8.54960517188435e-06, + "loss": 0.5152, + "step": 3966 + }, + { + "epoch": 0.27135919009508175, + "grad_norm": 2.012885272127952, + "learning_rate": 8.548824913793427e-06, + "loss": 0.2606, + "step": 3967 + }, + { + "epoch": 0.27142759422669127, + "grad_norm": 2.604573434850468, + "learning_rate": 8.54804448150904e-06, + "loss": 0.777, + "step": 3968 + }, + { + "epoch": 0.27149599835830085, + "grad_norm": 1.9864854103314018, + "learning_rate": 8.547263875069498e-06, + "loss": 0.4099, + "step": 3969 + }, + { + "epoch": 0.2715644024899104, + "grad_norm": 2.227033685720208, + "learning_rate": 8.546483094513114e-06, + "loss": 0.2138, + "step": 3970 + }, + { + "epoch": 0.27163280662151995, + "grad_norm": 1.7141394765672784, + "learning_rate": 8.545702139878214e-06, + "loss": 0.2615, + "step": 3971 + }, + { + "epoch": 0.2717012107531295, + "grad_norm": 2.357802252118369, + "learning_rate": 8.544921011203131e-06, + "loss": 0.6132, + "step": 3972 + }, + { + "epoch": 0.27176961488473905, + "grad_norm": 1.9984053870895104, + "learning_rate": 8.544139708526208e-06, + "loss": 0.2884, + "step": 3973 + }, + { + "epoch": 0.2718380190163486, + "grad_norm": 1.8475937419423125, + "learning_rate": 8.543358231885791e-06, + "loss": 0.3825, + "step": 3974 + }, + { + "epoch": 0.27190642314795815, + "grad_norm": 1.2600380115118532, + "learning_rate": 8.542576581320243e-06, + "loss": 0.0832, + "step": 3975 + }, + { + "epoch": 0.2719748272795677, + "grad_norm": 1.7710769420309165, + "learning_rate": 8.541794756867929e-06, + "loss": 0.2836, + "step": 3976 + }, + { + "epoch": 0.27204323141117726, + "grad_norm": 2.249115505147794, + "learning_rate": 8.541012758567227e-06, + "loss": 0.2999, + "step": 3977 + }, + { + "epoch": 0.2721116355427868, + "grad_norm": 2.4016601334173466, + "learning_rate": 8.540230586456517e-06, + "loss": 0.4719, + "step": 3978 + }, + { + "epoch": 0.27218003967439636, + "grad_norm": 1.9094039395038769, + "learning_rate": 8.539448240574196e-06, + "loss": 0.462, + "step": 3979 + }, + { + "epoch": 0.2722484438060059, + "grad_norm": 2.035230111742235, + "learning_rate": 8.538665720958662e-06, + "loss": 0.5887, + "step": 3980 + }, + { + "epoch": 0.27231684793761546, + "grad_norm": 3.9640322902492136, + "learning_rate": 8.537883027648328e-06, + "loss": 0.7707, + "step": 3981 + }, + { + "epoch": 0.272385252069225, + "grad_norm": 1.899097210264163, + "learning_rate": 8.537100160681608e-06, + "loss": 0.3501, + "step": 3982 + }, + { + "epoch": 0.2724536562008345, + "grad_norm": 1.8938322842539188, + "learning_rate": 8.536317120096934e-06, + "loss": 0.3595, + "step": 3983 + }, + { + "epoch": 0.2725220603324441, + "grad_norm": 2.1053016681231114, + "learning_rate": 8.535533905932739e-06, + "loss": 0.4734, + "step": 3984 + }, + { + "epoch": 0.2725904644640536, + "grad_norm": 1.8570397522458475, + "learning_rate": 8.534750518227466e-06, + "loss": 0.3609, + "step": 3985 + }, + { + "epoch": 0.2726588685956632, + "grad_norm": 2.1360829608670713, + "learning_rate": 8.533966957019568e-06, + "loss": 0.6145, + "step": 3986 + }, + { + "epoch": 0.2727272727272727, + "grad_norm": 2.279267808586659, + "learning_rate": 8.533183222347507e-06, + "loss": 0.7605, + "step": 3987 + }, + { + "epoch": 0.2727956768588823, + "grad_norm": 1.274278670094343, + "learning_rate": 8.532399314249752e-06, + "loss": 0.187, + "step": 3988 + }, + { + "epoch": 0.2728640809904918, + "grad_norm": 2.3171625559127236, + "learning_rate": 8.531615232764782e-06, + "loss": 0.4577, + "step": 3989 + }, + { + "epoch": 0.2729324851221014, + "grad_norm": 1.6265859338353053, + "learning_rate": 8.530830977931081e-06, + "loss": 0.2598, + "step": 3990 + }, + { + "epoch": 0.2730008892537109, + "grad_norm": 1.3942460869158277, + "learning_rate": 8.530046549787145e-06, + "loss": 0.1703, + "step": 3991 + }, + { + "epoch": 0.2730692933853205, + "grad_norm": 2.273142058062274, + "learning_rate": 8.52926194837148e-06, + "loss": 0.5864, + "step": 3992 + }, + { + "epoch": 0.27313769751693, + "grad_norm": 1.6585896089909125, + "learning_rate": 8.528477173722593e-06, + "loss": 0.4135, + "step": 3993 + }, + { + "epoch": 0.2732061016485396, + "grad_norm": 1.6458331343996757, + "learning_rate": 8.527692225879009e-06, + "loss": 0.358, + "step": 3994 + }, + { + "epoch": 0.2732745057801491, + "grad_norm": 2.559274984624956, + "learning_rate": 8.526907104879256e-06, + "loss": 0.6316, + "step": 3995 + }, + { + "epoch": 0.2733429099117587, + "grad_norm": 2.019083781446827, + "learning_rate": 8.526121810761872e-06, + "loss": 0.5041, + "step": 3996 + }, + { + "epoch": 0.2734113140433682, + "grad_norm": 1.8611948940312513, + "learning_rate": 8.525336343565401e-06, + "loss": 0.2788, + "step": 3997 + }, + { + "epoch": 0.2734797181749778, + "grad_norm": 1.7746393273840537, + "learning_rate": 8.524550703328396e-06, + "loss": 0.4737, + "step": 3998 + }, + { + "epoch": 0.2735481223065873, + "grad_norm": 2.223573907567021, + "learning_rate": 8.523764890089425e-06, + "loss": 0.4509, + "step": 3999 + }, + { + "epoch": 0.2736165264381969, + "grad_norm": 1.9988170418460935, + "learning_rate": 8.522978903887058e-06, + "loss": 0.2827, + "step": 4000 + }, + { + "epoch": 0.2736849305698064, + "grad_norm": 2.1774531058781443, + "learning_rate": 8.522192744759873e-06, + "loss": 0.5359, + "step": 4001 + }, + { + "epoch": 0.27375333470141594, + "grad_norm": 1.9537563257304151, + "learning_rate": 8.521406412746459e-06, + "loss": 0.298, + "step": 4002 + }, + { + "epoch": 0.2738217388330255, + "grad_norm": 1.6607440827794158, + "learning_rate": 8.520619907885415e-06, + "loss": 0.1874, + "step": 4003 + }, + { + "epoch": 0.27389014296463504, + "grad_norm": 2.1996836975418135, + "learning_rate": 8.519833230215344e-06, + "loss": 0.5481, + "step": 4004 + }, + { + "epoch": 0.2739585470962446, + "grad_norm": 2.0714405225658012, + "learning_rate": 8.51904637977486e-06, + "loss": 0.5405, + "step": 4005 + }, + { + "epoch": 0.27402695122785414, + "grad_norm": 1.3854013229419146, + "learning_rate": 8.51825935660259e-06, + "loss": 0.1506, + "step": 4006 + }, + { + "epoch": 0.2740953553594637, + "grad_norm": 2.0045308335084013, + "learning_rate": 8.517472160737158e-06, + "loss": 0.3127, + "step": 4007 + }, + { + "epoch": 0.27416375949107324, + "grad_norm": 2.5733821389289253, + "learning_rate": 8.516684792217207e-06, + "loss": 0.5858, + "step": 4008 + }, + { + "epoch": 0.2742321636226828, + "grad_norm": 2.038168700310326, + "learning_rate": 8.515897251081385e-06, + "loss": 0.5607, + "step": 4009 + }, + { + "epoch": 0.27430056775429235, + "grad_norm": 2.5289273967925285, + "learning_rate": 8.515109537368346e-06, + "loss": 0.593, + "step": 4010 + }, + { + "epoch": 0.2743689718859019, + "grad_norm": 1.5915884794249189, + "learning_rate": 8.514321651116759e-06, + "loss": 0.1916, + "step": 4011 + }, + { + "epoch": 0.27443737601751145, + "grad_norm": 1.8759264643908664, + "learning_rate": 8.513533592365295e-06, + "loss": 0.5041, + "step": 4012 + }, + { + "epoch": 0.274505780149121, + "grad_norm": 2.348675276596448, + "learning_rate": 8.512745361152635e-06, + "loss": 0.316, + "step": 4013 + }, + { + "epoch": 0.27457418428073055, + "grad_norm": 1.812211704345769, + "learning_rate": 8.51195695751747e-06, + "loss": 0.2817, + "step": 4014 + }, + { + "epoch": 0.2746425884123401, + "grad_norm": 1.8946870090912453, + "learning_rate": 8.511168381498498e-06, + "loss": 0.2751, + "step": 4015 + }, + { + "epoch": 0.27471099254394965, + "grad_norm": 2.5336099972881803, + "learning_rate": 8.510379633134428e-06, + "loss": 0.7069, + "step": 4016 + }, + { + "epoch": 0.27477939667555923, + "grad_norm": 1.918468072383233, + "learning_rate": 8.50959071246397e-06, + "loss": 0.2063, + "step": 4017 + }, + { + "epoch": 0.27484780080716875, + "grad_norm": 1.5397481939974433, + "learning_rate": 8.508801619525856e-06, + "loss": 0.3477, + "step": 4018 + }, + { + "epoch": 0.2749162049387783, + "grad_norm": 1.5957555493786164, + "learning_rate": 8.508012354358815e-06, + "loss": 0.2833, + "step": 4019 + }, + { + "epoch": 0.27498460907038785, + "grad_norm": 1.8752827709278095, + "learning_rate": 8.507222917001584e-06, + "loss": 0.3164, + "step": 4020 + }, + { + "epoch": 0.2750530132019974, + "grad_norm": 2.3953023294430755, + "learning_rate": 8.50643330749292e-06, + "loss": 0.6674, + "step": 4021 + }, + { + "epoch": 0.27512141733360695, + "grad_norm": 1.4448143203566457, + "learning_rate": 8.505643525871575e-06, + "loss": 0.2512, + "step": 4022 + }, + { + "epoch": 0.2751898214652165, + "grad_norm": 2.3958137836211537, + "learning_rate": 8.504853572176318e-06, + "loss": 0.3274, + "step": 4023 + }, + { + "epoch": 0.27525822559682606, + "grad_norm": 1.6158249546843375, + "learning_rate": 8.504063446445922e-06, + "loss": 0.4531, + "step": 4024 + }, + { + "epoch": 0.2753266297284356, + "grad_norm": 2.481225638666359, + "learning_rate": 8.503273148719172e-06, + "loss": 0.7795, + "step": 4025 + }, + { + "epoch": 0.27539503386004516, + "grad_norm": 2.039065602179612, + "learning_rate": 8.502482679034858e-06, + "loss": 0.5923, + "step": 4026 + }, + { + "epoch": 0.2754634379916547, + "grad_norm": 2.106041138647946, + "learning_rate": 8.501692037431782e-06, + "loss": 0.5683, + "step": 4027 + }, + { + "epoch": 0.27553184212326426, + "grad_norm": 2.023126143725346, + "learning_rate": 8.500901223948748e-06, + "loss": 0.2927, + "step": 4028 + }, + { + "epoch": 0.2756002462548738, + "grad_norm": 2.1137845749737174, + "learning_rate": 8.50011023862458e-06, + "loss": 0.2528, + "step": 4029 + }, + { + "epoch": 0.27566865038648336, + "grad_norm": 2.475150340504079, + "learning_rate": 8.499319081498099e-06, + "loss": 0.2594, + "step": 4030 + }, + { + "epoch": 0.2757370545180929, + "grad_norm": 2.088033437528286, + "learning_rate": 8.498527752608137e-06, + "loss": 0.4571, + "step": 4031 + }, + { + "epoch": 0.27580545864970246, + "grad_norm": 1.7296423693371519, + "learning_rate": 8.49773625199354e-06, + "loss": 0.4935, + "step": 4032 + }, + { + "epoch": 0.275873862781312, + "grad_norm": 2.2130822000294392, + "learning_rate": 8.496944579693157e-06, + "loss": 0.6412, + "step": 4033 + }, + { + "epoch": 0.27594226691292156, + "grad_norm": 2.244757983197649, + "learning_rate": 8.49615273574585e-06, + "loss": 0.6112, + "step": 4034 + }, + { + "epoch": 0.2760106710445311, + "grad_norm": 2.467555964510652, + "learning_rate": 8.495360720190482e-06, + "loss": 0.735, + "step": 4035 + }, + { + "epoch": 0.27607907517614066, + "grad_norm": 1.7573084596021098, + "learning_rate": 8.49456853306593e-06, + "loss": 0.4287, + "step": 4036 + }, + { + "epoch": 0.2761474793077502, + "grad_norm": 1.6119061074984462, + "learning_rate": 8.493776174411081e-06, + "loss": 0.3576, + "step": 4037 + }, + { + "epoch": 0.2762158834393597, + "grad_norm": 1.6860877105361514, + "learning_rate": 8.492983644264825e-06, + "loss": 0.3221, + "step": 4038 + }, + { + "epoch": 0.2762842875709693, + "grad_norm": 2.1011941438066826, + "learning_rate": 8.492190942666065e-06, + "loss": 0.5524, + "step": 4039 + }, + { + "epoch": 0.2763526917025788, + "grad_norm": 2.0289818791867678, + "learning_rate": 8.491398069653709e-06, + "loss": 0.4679, + "step": 4040 + }, + { + "epoch": 0.2764210958341884, + "grad_norm": 2.3171900023165852, + "learning_rate": 8.490605025266678e-06, + "loss": 0.624, + "step": 4041 + }, + { + "epoch": 0.2764894999657979, + "grad_norm": 1.7906469217260057, + "learning_rate": 8.489811809543894e-06, + "loss": 0.4645, + "step": 4042 + }, + { + "epoch": 0.2765579040974075, + "grad_norm": 1.8803601359977, + "learning_rate": 8.489018422524295e-06, + "loss": 0.2604, + "step": 4043 + }, + { + "epoch": 0.276626308229017, + "grad_norm": 1.9193337122651082, + "learning_rate": 8.488224864246825e-06, + "loss": 0.2345, + "step": 4044 + }, + { + "epoch": 0.2766947123606266, + "grad_norm": 1.8609389091313406, + "learning_rate": 8.487431134750432e-06, + "loss": 0.4577, + "step": 4045 + }, + { + "epoch": 0.2767631164922361, + "grad_norm": 1.802470265360157, + "learning_rate": 8.48663723407408e-06, + "loss": 0.4449, + "step": 4046 + }, + { + "epoch": 0.2768315206238457, + "grad_norm": 1.8467293525363682, + "learning_rate": 8.485843162256734e-06, + "loss": 0.3386, + "step": 4047 + }, + { + "epoch": 0.2768999247554552, + "grad_norm": 1.8138061618325485, + "learning_rate": 8.485048919337373e-06, + "loss": 0.4671, + "step": 4048 + }, + { + "epoch": 0.2769683288870648, + "grad_norm": 2.2821064543224634, + "learning_rate": 8.484254505354983e-06, + "loss": 0.5631, + "step": 4049 + }, + { + "epoch": 0.2770367330186743, + "grad_norm": 2.4445079067818827, + "learning_rate": 8.483459920348557e-06, + "loss": 0.3773, + "step": 4050 + }, + { + "epoch": 0.2771051371502839, + "grad_norm": 2.8209538985274714, + "learning_rate": 8.482665164357096e-06, + "loss": 0.6124, + "step": 4051 + }, + { + "epoch": 0.2771735412818934, + "grad_norm": 1.633024744904836, + "learning_rate": 8.481870237419609e-06, + "loss": 0.182, + "step": 4052 + }, + { + "epoch": 0.277241945413503, + "grad_norm": 2.6411564760489563, + "learning_rate": 8.481075139575118e-06, + "loss": 0.3542, + "step": 4053 + }, + { + "epoch": 0.2773103495451125, + "grad_norm": 2.0628671319364567, + "learning_rate": 8.480279870862652e-06, + "loss": 0.331, + "step": 4054 + }, + { + "epoch": 0.2773787536767221, + "grad_norm": 2.313439706659353, + "learning_rate": 8.479484431321242e-06, + "loss": 0.6242, + "step": 4055 + }, + { + "epoch": 0.2774471578083316, + "grad_norm": 2.0737717823990818, + "learning_rate": 8.478688820989933e-06, + "loss": 0.3573, + "step": 4056 + }, + { + "epoch": 0.27751556193994115, + "grad_norm": 2.1572919898690635, + "learning_rate": 8.47789303990778e-06, + "loss": 0.5475, + "step": 4057 + }, + { + "epoch": 0.2775839660715507, + "grad_norm": 2.910592721043769, + "learning_rate": 8.47709708811384e-06, + "loss": 0.9712, + "step": 4058 + }, + { + "epoch": 0.27765237020316025, + "grad_norm": 1.924033863665544, + "learning_rate": 8.476300965647186e-06, + "loss": 0.3551, + "step": 4059 + }, + { + "epoch": 0.2777207743347698, + "grad_norm": 2.325023543843661, + "learning_rate": 8.475504672546892e-06, + "loss": 0.7078, + "step": 4060 + }, + { + "epoch": 0.27778917846637935, + "grad_norm": 1.849005780932053, + "learning_rate": 8.474708208852047e-06, + "loss": 0.3059, + "step": 4061 + }, + { + "epoch": 0.27785758259798893, + "grad_norm": 2.2722384774663724, + "learning_rate": 8.473911574601743e-06, + "loss": 0.5917, + "step": 4062 + }, + { + "epoch": 0.27792598672959845, + "grad_norm": 1.2776441745763565, + "learning_rate": 8.473114769835085e-06, + "loss": 0.2215, + "step": 4063 + }, + { + "epoch": 0.27799439086120803, + "grad_norm": 1.9818215273882767, + "learning_rate": 8.472317794591181e-06, + "loss": 0.6721, + "step": 4064 + }, + { + "epoch": 0.27806279499281755, + "grad_norm": 1.8604669690317843, + "learning_rate": 8.471520648909152e-06, + "loss": 0.2155, + "step": 4065 + }, + { + "epoch": 0.27813119912442713, + "grad_norm": 2.218933359816863, + "learning_rate": 8.470723332828127e-06, + "loss": 0.6166, + "step": 4066 + }, + { + "epoch": 0.27819960325603665, + "grad_norm": 2.6182327574525766, + "learning_rate": 8.46992584638724e-06, + "loss": 0.4962, + "step": 4067 + }, + { + "epoch": 0.27826800738764623, + "grad_norm": 1.6222565903975816, + "learning_rate": 8.469128189625637e-06, + "loss": 0.3151, + "step": 4068 + }, + { + "epoch": 0.27833641151925576, + "grad_norm": 1.9011502428707177, + "learning_rate": 8.46833036258247e-06, + "loss": 0.3057, + "step": 4069 + }, + { + "epoch": 0.27840481565086533, + "grad_norm": 1.8006908617483794, + "learning_rate": 8.4675323652969e-06, + "loss": 0.3945, + "step": 4070 + }, + { + "epoch": 0.27847321978247486, + "grad_norm": 1.7293878628301012, + "learning_rate": 8.466734197808097e-06, + "loss": 0.3161, + "step": 4071 + }, + { + "epoch": 0.27854162391408444, + "grad_norm": 1.6813944548905777, + "learning_rate": 8.465935860155238e-06, + "loss": 0.384, + "step": 4072 + }, + { + "epoch": 0.27861002804569396, + "grad_norm": 1.8363182809286305, + "learning_rate": 8.465137352377513e-06, + "loss": 0.2569, + "step": 4073 + }, + { + "epoch": 0.27867843217730354, + "grad_norm": 1.7872823829497215, + "learning_rate": 8.464338674514112e-06, + "loss": 0.3335, + "step": 4074 + }, + { + "epoch": 0.27874683630891306, + "grad_norm": 2.234181262333466, + "learning_rate": 8.463539826604238e-06, + "loss": 0.5054, + "step": 4075 + }, + { + "epoch": 0.2788152404405226, + "grad_norm": 2.3745319225182056, + "learning_rate": 8.462740808687107e-06, + "loss": 0.5139, + "step": 4076 + }, + { + "epoch": 0.27888364457213216, + "grad_norm": 2.926044131438993, + "learning_rate": 8.461941620801934e-06, + "loss": 0.4182, + "step": 4077 + }, + { + "epoch": 0.2789520487037417, + "grad_norm": 2.668233185337626, + "learning_rate": 8.461142262987949e-06, + "loss": 0.4781, + "step": 4078 + }, + { + "epoch": 0.27902045283535126, + "grad_norm": 2.4918538351525945, + "learning_rate": 8.460342735284388e-06, + "loss": 0.7855, + "step": 4079 + }, + { + "epoch": 0.2790888569669608, + "grad_norm": 1.7591610419300763, + "learning_rate": 8.459543037730497e-06, + "loss": 0.3107, + "step": 4080 + }, + { + "epoch": 0.27915726109857036, + "grad_norm": 2.1872175579292077, + "learning_rate": 8.458743170365526e-06, + "loss": 0.312, + "step": 4081 + }, + { + "epoch": 0.2792256652301799, + "grad_norm": 2.2964410436792018, + "learning_rate": 8.457943133228738e-06, + "loss": 0.4929, + "step": 4082 + }, + { + "epoch": 0.27929406936178947, + "grad_norm": 2.237784671788875, + "learning_rate": 8.457142926359405e-06, + "loss": 0.5937, + "step": 4083 + }, + { + "epoch": 0.279362473493399, + "grad_norm": 2.285686131526415, + "learning_rate": 8.456342549796801e-06, + "loss": 0.5045, + "step": 4084 + }, + { + "epoch": 0.27943087762500857, + "grad_norm": 2.3145147870698373, + "learning_rate": 8.455542003580216e-06, + "loss": 0.4255, + "step": 4085 + }, + { + "epoch": 0.2794992817566181, + "grad_norm": 1.7042169089205512, + "learning_rate": 8.45474128774894e-06, + "loss": 0.3672, + "step": 4086 + }, + { + "epoch": 0.27956768588822767, + "grad_norm": 2.3585325736053155, + "learning_rate": 8.45394040234228e-06, + "loss": 0.4304, + "step": 4087 + }, + { + "epoch": 0.2796360900198372, + "grad_norm": 1.6113487797193595, + "learning_rate": 8.453139347399546e-06, + "loss": 0.4435, + "step": 4088 + }, + { + "epoch": 0.27970449415144677, + "grad_norm": 2.350000344185094, + "learning_rate": 8.452338122960057e-06, + "loss": 0.6332, + "step": 4089 + }, + { + "epoch": 0.2797728982830563, + "grad_norm": 1.9452345958874901, + "learning_rate": 8.451536729063144e-06, + "loss": 0.4469, + "step": 4090 + }, + { + "epoch": 0.27984130241466587, + "grad_norm": 2.069563503453263, + "learning_rate": 8.45073516574814e-06, + "loss": 0.5193, + "step": 4091 + }, + { + "epoch": 0.2799097065462754, + "grad_norm": 2.2902623081992513, + "learning_rate": 8.44993343305439e-06, + "loss": 0.6213, + "step": 4092 + }, + { + "epoch": 0.2799781106778849, + "grad_norm": 1.9731162639614233, + "learning_rate": 8.449131531021248e-06, + "loss": 0.5108, + "step": 4093 + }, + { + "epoch": 0.2800465148094945, + "grad_norm": 1.6106505639790065, + "learning_rate": 8.448329459688077e-06, + "loss": 0.2499, + "step": 4094 + }, + { + "epoch": 0.280114918941104, + "grad_norm": 2.2416420981031036, + "learning_rate": 8.44752721909424e-06, + "loss": 0.6474, + "step": 4095 + }, + { + "epoch": 0.2801833230727136, + "grad_norm": 2.012645300660736, + "learning_rate": 8.44672480927912e-06, + "loss": 0.4403, + "step": 4096 + }, + { + "epoch": 0.2802517272043231, + "grad_norm": 2.188178502169387, + "learning_rate": 8.445922230282105e-06, + "loss": 0.3434, + "step": 4097 + }, + { + "epoch": 0.2803201313359327, + "grad_norm": 2.416703104013036, + "learning_rate": 8.445119482142585e-06, + "loss": 0.3338, + "step": 4098 + }, + { + "epoch": 0.2803885354675422, + "grad_norm": 1.7904099924513388, + "learning_rate": 8.444316564899966e-06, + "loss": 0.2934, + "step": 4099 + }, + { + "epoch": 0.2804569395991518, + "grad_norm": 2.1816367351664643, + "learning_rate": 8.443513478593656e-06, + "loss": 0.4908, + "step": 4100 + }, + { + "epoch": 0.2805253437307613, + "grad_norm": 2.140584263100705, + "learning_rate": 8.442710223263076e-06, + "loss": 0.5406, + "step": 4101 + }, + { + "epoch": 0.2805937478623709, + "grad_norm": 1.9241069456106052, + "learning_rate": 8.441906798947654e-06, + "loss": 0.5469, + "step": 4102 + }, + { + "epoch": 0.2806621519939804, + "grad_norm": 1.6151488580172009, + "learning_rate": 8.441103205686824e-06, + "loss": 0.357, + "step": 4103 + }, + { + "epoch": 0.28073055612559, + "grad_norm": 2.2508203017803705, + "learning_rate": 8.440299443520034e-06, + "loss": 0.7182, + "step": 4104 + }, + { + "epoch": 0.2807989602571995, + "grad_norm": 2.16893372238477, + "learning_rate": 8.439495512486733e-06, + "loss": 0.4302, + "step": 4105 + }, + { + "epoch": 0.2808673643888091, + "grad_norm": 2.2147816509711347, + "learning_rate": 8.438691412626384e-06, + "loss": 0.6532, + "step": 4106 + }, + { + "epoch": 0.2809357685204186, + "grad_norm": 2.1570034507499263, + "learning_rate": 8.437887143978456e-06, + "loss": 0.3895, + "step": 4107 + }, + { + "epoch": 0.2810041726520282, + "grad_norm": 1.5978381209629355, + "learning_rate": 8.437082706582423e-06, + "loss": 0.298, + "step": 4108 + }, + { + "epoch": 0.28107257678363773, + "grad_norm": 1.8389026235284895, + "learning_rate": 8.436278100477775e-06, + "loss": 0.281, + "step": 4109 + }, + { + "epoch": 0.2811409809152473, + "grad_norm": 1.9880357602000212, + "learning_rate": 8.435473325704004e-06, + "loss": 0.4908, + "step": 4110 + }, + { + "epoch": 0.28120938504685683, + "grad_norm": 1.9768503454445665, + "learning_rate": 8.434668382300611e-06, + "loss": 0.5159, + "step": 4111 + }, + { + "epoch": 0.28127778917846635, + "grad_norm": 2.2344071979470277, + "learning_rate": 8.43386327030711e-06, + "loss": 0.4593, + "step": 4112 + }, + { + "epoch": 0.28134619331007593, + "grad_norm": 2.1511174012368635, + "learning_rate": 8.433057989763016e-06, + "loss": 0.5133, + "step": 4113 + }, + { + "epoch": 0.28141459744168545, + "grad_norm": 2.3899637878639393, + "learning_rate": 8.432252540707859e-06, + "loss": 0.6514, + "step": 4114 + }, + { + "epoch": 0.28148300157329503, + "grad_norm": 2.3224471031741265, + "learning_rate": 8.431446923181173e-06, + "loss": 0.2898, + "step": 4115 + }, + { + "epoch": 0.28155140570490456, + "grad_norm": 2.3305245904629768, + "learning_rate": 8.430641137222501e-06, + "loss": 0.5183, + "step": 4116 + }, + { + "epoch": 0.28161980983651413, + "grad_norm": 1.5602725734430862, + "learning_rate": 8.429835182871397e-06, + "loss": 0.325, + "step": 4117 + }, + { + "epoch": 0.28168821396812366, + "grad_norm": 2.4019491848593035, + "learning_rate": 8.42902906016742e-06, + "loss": 0.713, + "step": 4118 + }, + { + "epoch": 0.28175661809973324, + "grad_norm": 2.401432062849629, + "learning_rate": 8.428222769150137e-06, + "loss": 0.3908, + "step": 4119 + }, + { + "epoch": 0.28182502223134276, + "grad_norm": 1.7240273777791484, + "learning_rate": 8.427416309859125e-06, + "loss": 0.1954, + "step": 4120 + }, + { + "epoch": 0.28189342636295234, + "grad_norm": 1.8023962114931131, + "learning_rate": 8.42660968233397e-06, + "loss": 0.3664, + "step": 4121 + }, + { + "epoch": 0.28196183049456186, + "grad_norm": 1.7133643648971502, + "learning_rate": 8.425802886614265e-06, + "loss": 0.3924, + "step": 4122 + }, + { + "epoch": 0.28203023462617144, + "grad_norm": 1.4058545033295495, + "learning_rate": 8.424995922739611e-06, + "loss": 0.2838, + "step": 4123 + }, + { + "epoch": 0.28209863875778096, + "grad_norm": 2.178022357049151, + "learning_rate": 8.424188790749619e-06, + "loss": 0.6621, + "step": 4124 + }, + { + "epoch": 0.28216704288939054, + "grad_norm": 1.6335670060349836, + "learning_rate": 8.423381490683906e-06, + "loss": 0.3421, + "step": 4125 + }, + { + "epoch": 0.28223544702100006, + "grad_norm": 4.401542638071007, + "learning_rate": 8.422574022582097e-06, + "loss": 0.5527, + "step": 4126 + }, + { + "epoch": 0.28230385115260964, + "grad_norm": 1.3843434501350882, + "learning_rate": 8.421766386483827e-06, + "loss": 0.2349, + "step": 4127 + }, + { + "epoch": 0.28237225528421916, + "grad_norm": 2.110302300961654, + "learning_rate": 8.420958582428743e-06, + "loss": 0.399, + "step": 4128 + }, + { + "epoch": 0.28244065941582874, + "grad_norm": 2.3966754691020236, + "learning_rate": 8.420150610456489e-06, + "loss": 0.6423, + "step": 4129 + }, + { + "epoch": 0.28250906354743827, + "grad_norm": 2.673963271655248, + "learning_rate": 8.419342470606726e-06, + "loss": 0.8446, + "step": 4130 + }, + { + "epoch": 0.2825774676790478, + "grad_norm": 2.5870171451155293, + "learning_rate": 8.418534162919126e-06, + "loss": 0.7076, + "step": 4131 + }, + { + "epoch": 0.28264587181065737, + "grad_norm": 1.9487789858191062, + "learning_rate": 8.417725687433359e-06, + "loss": 0.5491, + "step": 4132 + }, + { + "epoch": 0.2827142759422669, + "grad_norm": 1.9327793756852223, + "learning_rate": 8.416917044189113e-06, + "loss": 0.4239, + "step": 4133 + }, + { + "epoch": 0.28278268007387647, + "grad_norm": 2.01439070428283, + "learning_rate": 8.416108233226078e-06, + "loss": 0.5589, + "step": 4134 + }, + { + "epoch": 0.282851084205486, + "grad_norm": 2.0003070632832083, + "learning_rate": 8.415299254583953e-06, + "loss": 0.5984, + "step": 4135 + }, + { + "epoch": 0.28291948833709557, + "grad_norm": 1.7936762118364995, + "learning_rate": 8.414490108302449e-06, + "loss": 0.5169, + "step": 4136 + }, + { + "epoch": 0.2829878924687051, + "grad_norm": 1.5726188310320701, + "learning_rate": 8.413680794421282e-06, + "loss": 0.3568, + "step": 4137 + }, + { + "epoch": 0.28305629660031467, + "grad_norm": 2.293906703995961, + "learning_rate": 8.412871312980178e-06, + "loss": 0.5575, + "step": 4138 + }, + { + "epoch": 0.2831247007319242, + "grad_norm": 1.9476006352655, + "learning_rate": 8.412061664018869e-06, + "loss": 0.3621, + "step": 4139 + }, + { + "epoch": 0.2831931048635338, + "grad_norm": 1.56233609531945, + "learning_rate": 8.411251847577096e-06, + "loss": 0.3105, + "step": 4140 + }, + { + "epoch": 0.2832615089951433, + "grad_norm": 1.8140447298636775, + "learning_rate": 8.41044186369461e-06, + "loss": 0.5005, + "step": 4141 + }, + { + "epoch": 0.2833299131267529, + "grad_norm": 2.0569415257260273, + "learning_rate": 8.409631712411168e-06, + "loss": 0.6165, + "step": 4142 + }, + { + "epoch": 0.2833983172583624, + "grad_norm": 1.8217105244876917, + "learning_rate": 8.40882139376654e-06, + "loss": 0.4191, + "step": 4143 + }, + { + "epoch": 0.283466721389972, + "grad_norm": 2.163582245870588, + "learning_rate": 8.408010907800492e-06, + "loss": 0.2474, + "step": 4144 + }, + { + "epoch": 0.2835351255215815, + "grad_norm": 2.608310333786334, + "learning_rate": 8.407200254552814e-06, + "loss": 0.8638, + "step": 4145 + }, + { + "epoch": 0.2836035296531911, + "grad_norm": 2.0929376980057555, + "learning_rate": 8.406389434063294e-06, + "loss": 0.535, + "step": 4146 + }, + { + "epoch": 0.2836719337848006, + "grad_norm": 1.9193059296075488, + "learning_rate": 8.405578446371731e-06, + "loss": 0.5292, + "step": 4147 + }, + { + "epoch": 0.2837403379164101, + "grad_norm": 1.993799783868383, + "learning_rate": 8.404767291517935e-06, + "loss": 0.5152, + "step": 4148 + }, + { + "epoch": 0.2838087420480197, + "grad_norm": 1.4610081441394114, + "learning_rate": 8.403955969541716e-06, + "loss": 0.2931, + "step": 4149 + }, + { + "epoch": 0.2838771461796292, + "grad_norm": 2.2347643619687814, + "learning_rate": 8.403144480482902e-06, + "loss": 0.4409, + "step": 4150 + }, + { + "epoch": 0.2839455503112388, + "grad_norm": 1.13888112348974, + "learning_rate": 8.402332824381321e-06, + "loss": 0.2184, + "step": 4151 + }, + { + "epoch": 0.2840139544428483, + "grad_norm": 1.890824602955798, + "learning_rate": 8.401521001276817e-06, + "loss": 0.3012, + "step": 4152 + }, + { + "epoch": 0.2840823585744579, + "grad_norm": 2.1931297202397864, + "learning_rate": 8.400709011209236e-06, + "loss": 0.4819, + "step": 4153 + }, + { + "epoch": 0.28415076270606743, + "grad_norm": 1.5808734189053728, + "learning_rate": 8.399896854218436e-06, + "loss": 0.3238, + "step": 4154 + }, + { + "epoch": 0.284219166837677, + "grad_norm": 2.4528268402065336, + "learning_rate": 8.39908453034428e-06, + "loss": 0.7785, + "step": 4155 + }, + { + "epoch": 0.28428757096928653, + "grad_norm": 1.5501666809067915, + "learning_rate": 8.39827203962664e-06, + "loss": 0.2812, + "step": 4156 + }, + { + "epoch": 0.2843559751008961, + "grad_norm": 2.6187010473715175, + "learning_rate": 8.397459382105399e-06, + "loss": 0.4141, + "step": 4157 + }, + { + "epoch": 0.28442437923250563, + "grad_norm": 2.52097506295475, + "learning_rate": 8.396646557820447e-06, + "loss": 0.8656, + "step": 4158 + }, + { + "epoch": 0.2844927833641152, + "grad_norm": 1.3060719716665607, + "learning_rate": 8.395833566811677e-06, + "loss": 0.1829, + "step": 4159 + }, + { + "epoch": 0.28456118749572473, + "grad_norm": 1.6104947472388662, + "learning_rate": 8.395020409118999e-06, + "loss": 0.3187, + "step": 4160 + }, + { + "epoch": 0.2846295916273343, + "grad_norm": 2.283573371882417, + "learning_rate": 8.394207084782324e-06, + "loss": 0.6696, + "step": 4161 + }, + { + "epoch": 0.28469799575894383, + "grad_norm": 1.703242052937933, + "learning_rate": 8.393393593841576e-06, + "loss": 0.4131, + "step": 4162 + }, + { + "epoch": 0.2847663998905534, + "grad_norm": 2.4995730631581763, + "learning_rate": 8.392579936336682e-06, + "loss": 0.2998, + "step": 4163 + }, + { + "epoch": 0.28483480402216294, + "grad_norm": 2.495831083228785, + "learning_rate": 8.391766112307584e-06, + "loss": 0.661, + "step": 4164 + }, + { + "epoch": 0.2849032081537725, + "grad_norm": 1.9411777116208224, + "learning_rate": 8.390952121794224e-06, + "loss": 0.4025, + "step": 4165 + }, + { + "epoch": 0.28497161228538204, + "grad_norm": 2.3187660657578033, + "learning_rate": 8.390137964836562e-06, + "loss": 0.6449, + "step": 4166 + }, + { + "epoch": 0.28504001641699156, + "grad_norm": 2.3289903050366654, + "learning_rate": 8.389323641474557e-06, + "loss": 0.4605, + "step": 4167 + }, + { + "epoch": 0.28510842054860114, + "grad_norm": 1.9465664648192331, + "learning_rate": 8.388509151748181e-06, + "loss": 0.3894, + "step": 4168 + }, + { + "epoch": 0.28517682468021066, + "grad_norm": 2.3445658980252837, + "learning_rate": 8.38769449569741e-06, + "loss": 0.6695, + "step": 4169 + }, + { + "epoch": 0.28524522881182024, + "grad_norm": 1.7159728154522285, + "learning_rate": 8.386879673362238e-06, + "loss": 0.2343, + "step": 4170 + }, + { + "epoch": 0.28531363294342976, + "grad_norm": 1.2907388451037143, + "learning_rate": 8.386064684782655e-06, + "loss": 0.1368, + "step": 4171 + }, + { + "epoch": 0.28538203707503934, + "grad_norm": 3.0110186246703243, + "learning_rate": 8.385249529998667e-06, + "loss": 0.4849, + "step": 4172 + }, + { + "epoch": 0.28545044120664886, + "grad_norm": 2.167139234878621, + "learning_rate": 8.384434209050285e-06, + "loss": 0.3429, + "step": 4173 + }, + { + "epoch": 0.28551884533825844, + "grad_norm": 2.713926547328887, + "learning_rate": 8.383618721977526e-06, + "loss": 0.471, + "step": 4174 + }, + { + "epoch": 0.28558724946986797, + "grad_norm": 2.1791225165993735, + "learning_rate": 8.382803068820423e-06, + "loss": 0.4166, + "step": 4175 + }, + { + "epoch": 0.28565565360147754, + "grad_norm": 1.742308479864121, + "learning_rate": 8.38198724961901e-06, + "loss": 0.3643, + "step": 4176 + }, + { + "epoch": 0.28572405773308707, + "grad_norm": 1.7399518932209126, + "learning_rate": 8.38117126441333e-06, + "loss": 0.28, + "step": 4177 + }, + { + "epoch": 0.28579246186469665, + "grad_norm": 1.8589346789061705, + "learning_rate": 8.380355113243439e-06, + "loss": 0.4034, + "step": 4178 + }, + { + "epoch": 0.28586086599630617, + "grad_norm": 2.098496856474868, + "learning_rate": 8.379538796149395e-06, + "loss": 0.5879, + "step": 4179 + }, + { + "epoch": 0.28592927012791575, + "grad_norm": 2.811145286662251, + "learning_rate": 8.378722313171266e-06, + "loss": 0.3499, + "step": 4180 + }, + { + "epoch": 0.28599767425952527, + "grad_norm": 1.955549764383777, + "learning_rate": 8.37790566434913e-06, + "loss": 0.4565, + "step": 4181 + }, + { + "epoch": 0.28606607839113485, + "grad_norm": 2.2005116505779254, + "learning_rate": 8.377088849723073e-06, + "loss": 0.2858, + "step": 4182 + }, + { + "epoch": 0.28613448252274437, + "grad_norm": 1.425094695249263, + "learning_rate": 8.376271869333187e-06, + "loss": 0.2561, + "step": 4183 + }, + { + "epoch": 0.28620288665435395, + "grad_norm": 1.4577263613455604, + "learning_rate": 8.375454723219575e-06, + "loss": 0.3004, + "step": 4184 + }, + { + "epoch": 0.2862712907859635, + "grad_norm": 1.8696585947002373, + "learning_rate": 8.374637411422345e-06, + "loss": 0.3453, + "step": 4185 + }, + { + "epoch": 0.286339694917573, + "grad_norm": 1.6087999751739868, + "learning_rate": 8.373819933981614e-06, + "loss": 0.304, + "step": 4186 + }, + { + "epoch": 0.2864080990491826, + "grad_norm": 1.3461765015904332, + "learning_rate": 8.37300229093751e-06, + "loss": 0.211, + "step": 4187 + }, + { + "epoch": 0.2864765031807921, + "grad_norm": 2.3113247423790333, + "learning_rate": 8.372184482330163e-06, + "loss": 0.7829, + "step": 4188 + }, + { + "epoch": 0.2865449073124017, + "grad_norm": 2.4299086503825347, + "learning_rate": 8.37136650819972e-06, + "loss": 0.3432, + "step": 4189 + }, + { + "epoch": 0.2866133114440112, + "grad_norm": 2.127279496527193, + "learning_rate": 8.370548368586326e-06, + "loss": 0.5279, + "step": 4190 + }, + { + "epoch": 0.2866817155756208, + "grad_norm": 2.202348058073195, + "learning_rate": 8.369730063530144e-06, + "loss": 0.7508, + "step": 4191 + }, + { + "epoch": 0.2867501197072303, + "grad_norm": 1.7069811562666828, + "learning_rate": 8.368911593071338e-06, + "loss": 0.3523, + "step": 4192 + }, + { + "epoch": 0.2868185238388399, + "grad_norm": 1.7720509456895166, + "learning_rate": 8.368092957250082e-06, + "loss": 0.1784, + "step": 4193 + }, + { + "epoch": 0.2868869279704494, + "grad_norm": 2.195175696077844, + "learning_rate": 8.36727415610656e-06, + "loss": 0.5405, + "step": 4194 + }, + { + "epoch": 0.286955332102059, + "grad_norm": 2.2183708552006216, + "learning_rate": 8.366455189680962e-06, + "loss": 0.5921, + "step": 4195 + }, + { + "epoch": 0.2870237362336685, + "grad_norm": 2.2344152306889615, + "learning_rate": 8.365636058013487e-06, + "loss": 0.7171, + "step": 4196 + }, + { + "epoch": 0.2870921403652781, + "grad_norm": 1.5306171890486286, + "learning_rate": 8.36481676114434e-06, + "loss": 0.2509, + "step": 4197 + }, + { + "epoch": 0.2871605444968876, + "grad_norm": 2.3119551880319107, + "learning_rate": 8.363997299113739e-06, + "loss": 0.7046, + "step": 4198 + }, + { + "epoch": 0.2872289486284972, + "grad_norm": 1.4238559107953848, + "learning_rate": 8.363177671961908e-06, + "loss": 0.2438, + "step": 4199 + }, + { + "epoch": 0.2872973527601067, + "grad_norm": 1.5349998297365075, + "learning_rate": 8.362357879729073e-06, + "loss": 0.1741, + "step": 4200 + }, + { + "epoch": 0.2873657568917163, + "grad_norm": 2.3054318406018024, + "learning_rate": 8.361537922455479e-06, + "loss": 0.5545, + "step": 4201 + }, + { + "epoch": 0.2874341610233258, + "grad_norm": 1.9841801405022892, + "learning_rate": 8.36071780018137e-06, + "loss": 0.3537, + "step": 4202 + }, + { + "epoch": 0.2875025651549354, + "grad_norm": 1.8679162761955297, + "learning_rate": 8.359897512947003e-06, + "loss": 0.4998, + "step": 4203 + }, + { + "epoch": 0.2875709692865449, + "grad_norm": 1.635554903610637, + "learning_rate": 8.359077060792642e-06, + "loss": 0.1648, + "step": 4204 + }, + { + "epoch": 0.28763937341815443, + "grad_norm": 2.097758878474063, + "learning_rate": 8.358256443758556e-06, + "loss": 0.576, + "step": 4205 + }, + { + "epoch": 0.287707777549764, + "grad_norm": 1.7896781345240282, + "learning_rate": 8.357435661885028e-06, + "loss": 0.3028, + "step": 4206 + }, + { + "epoch": 0.28777618168137353, + "grad_norm": 1.8777409844057, + "learning_rate": 8.356614715212344e-06, + "loss": 0.294, + "step": 4207 + }, + { + "epoch": 0.2878445858129831, + "grad_norm": 2.2050487425124583, + "learning_rate": 8.355793603780802e-06, + "loss": 0.4668, + "step": 4208 + }, + { + "epoch": 0.28791298994459263, + "grad_norm": 1.7501763059156414, + "learning_rate": 8.354972327630705e-06, + "loss": 0.3902, + "step": 4209 + }, + { + "epoch": 0.2879813940762022, + "grad_norm": 2.1790158867565412, + "learning_rate": 8.354150886802364e-06, + "loss": 0.3205, + "step": 4210 + }, + { + "epoch": 0.28804979820781174, + "grad_norm": 1.623771202882821, + "learning_rate": 8.353329281336103e-06, + "loss": 0.3174, + "step": 4211 + }, + { + "epoch": 0.2881182023394213, + "grad_norm": 1.985694187695154, + "learning_rate": 8.352507511272246e-06, + "loss": 0.4936, + "step": 4212 + }, + { + "epoch": 0.28818660647103084, + "grad_norm": 2.50738049288914, + "learning_rate": 8.351685576651129e-06, + "loss": 0.8515, + "step": 4213 + }, + { + "epoch": 0.2882550106026404, + "grad_norm": 2.169975839762738, + "learning_rate": 8.350863477513103e-06, + "loss": 0.3748, + "step": 4214 + }, + { + "epoch": 0.28832341473424994, + "grad_norm": 1.7289149539692215, + "learning_rate": 8.350041213898511e-06, + "loss": 0.1452, + "step": 4215 + }, + { + "epoch": 0.2883918188658595, + "grad_norm": 1.2406379523670137, + "learning_rate": 8.349218785847724e-06, + "loss": 0.2184, + "step": 4216 + }, + { + "epoch": 0.28846022299746904, + "grad_norm": 1.9994329453299855, + "learning_rate": 8.348396193401105e-06, + "loss": 0.6004, + "step": 4217 + }, + { + "epoch": 0.2885286271290786, + "grad_norm": 2.3136502446701668, + "learning_rate": 8.34757343659903e-06, + "loss": 0.4803, + "step": 4218 + }, + { + "epoch": 0.28859703126068814, + "grad_norm": 2.5708032388847464, + "learning_rate": 8.346750515481888e-06, + "loss": 0.5428, + "step": 4219 + }, + { + "epoch": 0.2886654353922977, + "grad_norm": 1.7346408958064088, + "learning_rate": 8.345927430090066e-06, + "loss": 0.3521, + "step": 4220 + }, + { + "epoch": 0.28873383952390724, + "grad_norm": 1.4533900152572374, + "learning_rate": 8.34510418046397e-06, + "loss": 0.2265, + "step": 4221 + }, + { + "epoch": 0.28880224365551677, + "grad_norm": 2.216531843567439, + "learning_rate": 8.344280766644009e-06, + "loss": 0.3612, + "step": 4222 + }, + { + "epoch": 0.28887064778712634, + "grad_norm": 2.4983267549038377, + "learning_rate": 8.343457188670597e-06, + "loss": 0.3604, + "step": 4223 + }, + { + "epoch": 0.28893905191873587, + "grad_norm": 2.400578394499258, + "learning_rate": 8.342633446584161e-06, + "loss": 0.3681, + "step": 4224 + }, + { + "epoch": 0.28900745605034545, + "grad_norm": 1.8954525899728862, + "learning_rate": 8.341809540425132e-06, + "loss": 0.3372, + "step": 4225 + }, + { + "epoch": 0.28907586018195497, + "grad_norm": 2.065084680190315, + "learning_rate": 8.340985470233955e-06, + "loss": 0.5333, + "step": 4226 + }, + { + "epoch": 0.28914426431356455, + "grad_norm": 2.3203304848792765, + "learning_rate": 8.34016123605108e-06, + "loss": 0.6356, + "step": 4227 + }, + { + "epoch": 0.28921266844517407, + "grad_norm": 2.0099852491070007, + "learning_rate": 8.339336837916959e-06, + "loss": 0.4522, + "step": 4228 + }, + { + "epoch": 0.28928107257678365, + "grad_norm": 1.931751394615445, + "learning_rate": 8.338512275872062e-06, + "loss": 0.447, + "step": 4229 + }, + { + "epoch": 0.28934947670839317, + "grad_norm": 1.606986011077933, + "learning_rate": 8.33768754995686e-06, + "loss": 0.2214, + "step": 4230 + }, + { + "epoch": 0.28941788084000275, + "grad_norm": 2.2244879846049024, + "learning_rate": 8.336862660211836e-06, + "loss": 0.6711, + "step": 4231 + }, + { + "epoch": 0.2894862849716123, + "grad_norm": 2.7594907155296315, + "learning_rate": 8.33603760667748e-06, + "loss": 0.332, + "step": 4232 + }, + { + "epoch": 0.28955468910322185, + "grad_norm": 2.237573301989978, + "learning_rate": 8.335212389394287e-06, + "loss": 0.3426, + "step": 4233 + }, + { + "epoch": 0.2896230932348314, + "grad_norm": 2.142369513149125, + "learning_rate": 8.334387008402766e-06, + "loss": 0.6322, + "step": 4234 + }, + { + "epoch": 0.28969149736644095, + "grad_norm": 2.1197547514948027, + "learning_rate": 8.333561463743426e-06, + "loss": 0.518, + "step": 4235 + }, + { + "epoch": 0.2897599014980505, + "grad_norm": 1.6334381980850716, + "learning_rate": 8.332735755456795e-06, + "loss": 0.1891, + "step": 4236 + }, + { + "epoch": 0.28982830562966005, + "grad_norm": 2.003266230468049, + "learning_rate": 8.3319098835834e-06, + "loss": 0.3946, + "step": 4237 + }, + { + "epoch": 0.2898967097612696, + "grad_norm": 1.7915015149146043, + "learning_rate": 8.331083848163776e-06, + "loss": 0.3487, + "step": 4238 + }, + { + "epoch": 0.28996511389287916, + "grad_norm": 2.0491802760245363, + "learning_rate": 8.330257649238472e-06, + "loss": 0.348, + "step": 4239 + }, + { + "epoch": 0.2900335180244887, + "grad_norm": 2.0572860552715277, + "learning_rate": 8.32943128684804e-06, + "loss": 0.3045, + "step": 4240 + }, + { + "epoch": 0.2901019221560982, + "grad_norm": 2.353623212832393, + "learning_rate": 8.328604761033043e-06, + "loss": 0.5121, + "step": 4241 + }, + { + "epoch": 0.2901703262877078, + "grad_norm": 2.2892623739243487, + "learning_rate": 8.327778071834052e-06, + "loss": 0.3792, + "step": 4242 + }, + { + "epoch": 0.2902387304193173, + "grad_norm": 2.2083688176801015, + "learning_rate": 8.326951219291644e-06, + "loss": 0.3551, + "step": 4243 + }, + { + "epoch": 0.2903071345509269, + "grad_norm": 2.162373372460861, + "learning_rate": 8.326124203446401e-06, + "loss": 0.5873, + "step": 4244 + }, + { + "epoch": 0.2903755386825364, + "grad_norm": 2.0229677443009417, + "learning_rate": 8.325297024338922e-06, + "loss": 0.4499, + "step": 4245 + }, + { + "epoch": 0.290443942814146, + "grad_norm": 1.6488267384776105, + "learning_rate": 8.32446968200981e-06, + "loss": 0.3085, + "step": 4246 + }, + { + "epoch": 0.2905123469457555, + "grad_norm": 1.9920925192537045, + "learning_rate": 8.323642176499671e-06, + "loss": 0.5982, + "step": 4247 + }, + { + "epoch": 0.2905807510773651, + "grad_norm": 1.896251416655321, + "learning_rate": 8.322814507849123e-06, + "loss": 0.4709, + "step": 4248 + }, + { + "epoch": 0.2906491552089746, + "grad_norm": 1.6000682372949973, + "learning_rate": 8.321986676098792e-06, + "loss": 0.1881, + "step": 4249 + }, + { + "epoch": 0.2907175593405842, + "grad_norm": 1.8194034500413716, + "learning_rate": 8.321158681289316e-06, + "loss": 0.3991, + "step": 4250 + }, + { + "epoch": 0.2907859634721937, + "grad_norm": 2.257117088431183, + "learning_rate": 8.320330523461334e-06, + "loss": 0.3365, + "step": 4251 + }, + { + "epoch": 0.2908543676038033, + "grad_norm": 1.7177890150134834, + "learning_rate": 8.319502202655497e-06, + "loss": 0.2664, + "step": 4252 + }, + { + "epoch": 0.2909227717354128, + "grad_norm": 2.3317622440143864, + "learning_rate": 8.31867371891246e-06, + "loss": 0.7409, + "step": 4253 + }, + { + "epoch": 0.2909911758670224, + "grad_norm": 1.9327535171575474, + "learning_rate": 8.317845072272891e-06, + "loss": 0.4642, + "step": 4254 + }, + { + "epoch": 0.2910595799986319, + "grad_norm": 1.699230746522846, + "learning_rate": 8.317016262777465e-06, + "loss": 0.2492, + "step": 4255 + }, + { + "epoch": 0.2911279841302415, + "grad_norm": 2.024505995904773, + "learning_rate": 8.316187290466862e-06, + "loss": 0.3576, + "step": 4256 + }, + { + "epoch": 0.291196388261851, + "grad_norm": 1.938863397623191, + "learning_rate": 8.315358155381776e-06, + "loss": 0.4811, + "step": 4257 + }, + { + "epoch": 0.2912647923934606, + "grad_norm": 1.8005909105441809, + "learning_rate": 8.3145288575629e-06, + "loss": 0.4435, + "step": 4258 + }, + { + "epoch": 0.2913331965250701, + "grad_norm": 3.5573638398578398, + "learning_rate": 8.313699397050941e-06, + "loss": 1.0087, + "step": 4259 + }, + { + "epoch": 0.29140160065667964, + "grad_norm": 2.358038902915069, + "learning_rate": 8.312869773886616e-06, + "loss": 0.5006, + "step": 4260 + }, + { + "epoch": 0.2914700047882892, + "grad_norm": 2.121016983975454, + "learning_rate": 8.312039988110643e-06, + "loss": 0.5361, + "step": 4261 + }, + { + "epoch": 0.29153840891989874, + "grad_norm": 1.292001112938094, + "learning_rate": 8.311210039763755e-06, + "loss": 0.1555, + "step": 4262 + }, + { + "epoch": 0.2916068130515083, + "grad_norm": 1.795591534542707, + "learning_rate": 8.310379928886688e-06, + "loss": 0.4918, + "step": 4263 + }, + { + "epoch": 0.29167521718311784, + "grad_norm": 1.9349695301182788, + "learning_rate": 8.309549655520188e-06, + "loss": 0.4361, + "step": 4264 + }, + { + "epoch": 0.2917436213147274, + "grad_norm": 2.2711629057867166, + "learning_rate": 8.308719219705011e-06, + "loss": 0.268, + "step": 4265 + }, + { + "epoch": 0.29181202544633694, + "grad_norm": 2.3433193510503436, + "learning_rate": 8.307888621481915e-06, + "loss": 0.7995, + "step": 4266 + }, + { + "epoch": 0.2918804295779465, + "grad_norm": 2.8401753416971145, + "learning_rate": 8.307057860891673e-06, + "loss": 0.2654, + "step": 4267 + }, + { + "epoch": 0.29194883370955604, + "grad_norm": 3.3190314660265985, + "learning_rate": 8.306226937975061e-06, + "loss": 0.4586, + "step": 4268 + }, + { + "epoch": 0.2920172378411656, + "grad_norm": 1.975219345587815, + "learning_rate": 8.305395852772866e-06, + "loss": 0.4478, + "step": 4269 + }, + { + "epoch": 0.29208564197277515, + "grad_norm": 1.9921560172790986, + "learning_rate": 8.30456460532588e-06, + "loss": 0.1036, + "step": 4270 + }, + { + "epoch": 0.2921540461043847, + "grad_norm": 1.8542943349740098, + "learning_rate": 8.303733195674906e-06, + "loss": 0.428, + "step": 4271 + }, + { + "epoch": 0.29222245023599425, + "grad_norm": 1.818507195681101, + "learning_rate": 8.302901623860753e-06, + "loss": 0.4165, + "step": 4272 + }, + { + "epoch": 0.2922908543676038, + "grad_norm": 1.8578758246599578, + "learning_rate": 8.302069889924241e-06, + "loss": 0.4651, + "step": 4273 + }, + { + "epoch": 0.29235925849921335, + "grad_norm": 2.387328299494811, + "learning_rate": 8.30123799390619e-06, + "loss": 0.7392, + "step": 4274 + }, + { + "epoch": 0.2924276626308229, + "grad_norm": 1.7988156246782305, + "learning_rate": 8.300405935847438e-06, + "loss": 0.3316, + "step": 4275 + }, + { + "epoch": 0.29249606676243245, + "grad_norm": 2.4072256401632837, + "learning_rate": 8.299573715788825e-06, + "loss": 0.5606, + "step": 4276 + }, + { + "epoch": 0.292564470894042, + "grad_norm": 2.3288208670386115, + "learning_rate": 8.298741333771201e-06, + "loss": 0.6703, + "step": 4277 + }, + { + "epoch": 0.29263287502565155, + "grad_norm": 1.827084799656847, + "learning_rate": 8.297908789835422e-06, + "loss": 0.1663, + "step": 4278 + }, + { + "epoch": 0.2927012791572611, + "grad_norm": 2.0890772474477415, + "learning_rate": 8.297076084022355e-06, + "loss": 0.3662, + "step": 4279 + }, + { + "epoch": 0.29276968328887065, + "grad_norm": 1.9040239813480455, + "learning_rate": 8.296243216372872e-06, + "loss": 0.3356, + "step": 4280 + }, + { + "epoch": 0.2928380874204802, + "grad_norm": 2.183724203290367, + "learning_rate": 8.295410186927855e-06, + "loss": 0.6626, + "step": 4281 + }, + { + "epoch": 0.29290649155208975, + "grad_norm": 4.493730309964523, + "learning_rate": 8.294576995728193e-06, + "loss": 0.4732, + "step": 4282 + }, + { + "epoch": 0.2929748956836993, + "grad_norm": 1.5075513831043796, + "learning_rate": 8.293743642814782e-06, + "loss": 0.1904, + "step": 4283 + }, + { + "epoch": 0.29304329981530886, + "grad_norm": 2.2641799132846825, + "learning_rate": 8.292910128228527e-06, + "loss": 0.7955, + "step": 4284 + }, + { + "epoch": 0.2931117039469184, + "grad_norm": 2.252314774257065, + "learning_rate": 8.292076452010343e-06, + "loss": 0.5315, + "step": 4285 + }, + { + "epoch": 0.29318010807852796, + "grad_norm": 1.5503608412415573, + "learning_rate": 8.29124261420115e-06, + "loss": 0.1853, + "step": 4286 + }, + { + "epoch": 0.2932485122101375, + "grad_norm": 1.4973477120890917, + "learning_rate": 8.290408614841873e-06, + "loss": 0.3419, + "step": 4287 + }, + { + "epoch": 0.29331691634174706, + "grad_norm": 1.7144125046419079, + "learning_rate": 8.289574453973456e-06, + "loss": 0.1803, + "step": 4288 + }, + { + "epoch": 0.2933853204733566, + "grad_norm": 2.0756160677539897, + "learning_rate": 8.288740131636837e-06, + "loss": 0.3281, + "step": 4289 + }, + { + "epoch": 0.29345372460496616, + "grad_norm": 1.9429088012686906, + "learning_rate": 8.287905647872971e-06, + "loss": 0.4387, + "step": 4290 + }, + { + "epoch": 0.2935221287365757, + "grad_norm": 1.8759406869248965, + "learning_rate": 8.28707100272282e-06, + "loss": 0.399, + "step": 4291 + }, + { + "epoch": 0.29359053286818526, + "grad_norm": 2.0632655247497453, + "learning_rate": 8.286236196227351e-06, + "loss": 0.5884, + "step": 4292 + }, + { + "epoch": 0.2936589369997948, + "grad_norm": 2.5903492890123747, + "learning_rate": 8.285401228427542e-06, + "loss": 0.3341, + "step": 4293 + }, + { + "epoch": 0.29372734113140436, + "grad_norm": 2.0620843710578467, + "learning_rate": 8.28456609936437e-06, + "loss": 0.2397, + "step": 4294 + }, + { + "epoch": 0.2937957452630139, + "grad_norm": 2.059405429221652, + "learning_rate": 8.283730809078838e-06, + "loss": 0.3444, + "step": 4295 + }, + { + "epoch": 0.2938641493946234, + "grad_norm": 2.0624455946913707, + "learning_rate": 8.28289535761194e-06, + "loss": 0.2707, + "step": 4296 + }, + { + "epoch": 0.293932553526233, + "grad_norm": 2.257741728996762, + "learning_rate": 8.282059745004684e-06, + "loss": 0.456, + "step": 4297 + }, + { + "epoch": 0.2940009576578425, + "grad_norm": 2.1774083256940657, + "learning_rate": 8.281223971298087e-06, + "loss": 0.501, + "step": 4298 + }, + { + "epoch": 0.2940693617894521, + "grad_norm": 2.2894287141918404, + "learning_rate": 8.280388036533171e-06, + "loss": 0.5861, + "step": 4299 + }, + { + "epoch": 0.2941377659210616, + "grad_norm": 2.377115007465866, + "learning_rate": 8.279551940750974e-06, + "loss": 0.3272, + "step": 4300 + }, + { + "epoch": 0.2942061700526712, + "grad_norm": 2.076296711969806, + "learning_rate": 8.278715683992527e-06, + "loss": 0.5942, + "step": 4301 + }, + { + "epoch": 0.2942745741842807, + "grad_norm": 2.1518249786550507, + "learning_rate": 8.277879266298883e-06, + "loss": 0.5353, + "step": 4302 + }, + { + "epoch": 0.2943429783158903, + "grad_norm": 1.7580745883105517, + "learning_rate": 8.277042687711094e-06, + "loss": 0.2309, + "step": 4303 + }, + { + "epoch": 0.2944113824474998, + "grad_norm": 2.7770968728523604, + "learning_rate": 8.276205948270226e-06, + "loss": 0.4932, + "step": 4304 + }, + { + "epoch": 0.2944797865791094, + "grad_norm": 2.437518706616732, + "learning_rate": 8.275369048017351e-06, + "loss": 0.7513, + "step": 4305 + }, + { + "epoch": 0.2945481907107189, + "grad_norm": 1.6968506370467478, + "learning_rate": 8.274531986993546e-06, + "loss": 0.3614, + "step": 4306 + }, + { + "epoch": 0.2946165948423285, + "grad_norm": 2.1908608368135956, + "learning_rate": 8.273694765239899e-06, + "loss": 0.4226, + "step": 4307 + }, + { + "epoch": 0.294684998973938, + "grad_norm": 1.7696711672272214, + "learning_rate": 8.272857382797502e-06, + "loss": 0.3582, + "step": 4308 + }, + { + "epoch": 0.2947534031055476, + "grad_norm": 1.5428710182238696, + "learning_rate": 8.27201983970746e-06, + "loss": 0.297, + "step": 4309 + }, + { + "epoch": 0.2948218072371571, + "grad_norm": 2.310998345296887, + "learning_rate": 8.271182136010886e-06, + "loss": 0.5221, + "step": 4310 + }, + { + "epoch": 0.2948902113687667, + "grad_norm": 1.748163651982507, + "learning_rate": 8.270344271748898e-06, + "loss": 0.3283, + "step": 4311 + }, + { + "epoch": 0.2949586155003762, + "grad_norm": 2.056920328710087, + "learning_rate": 8.269506246962618e-06, + "loss": 0.5783, + "step": 4312 + }, + { + "epoch": 0.2950270196319858, + "grad_norm": 2.2029816770265627, + "learning_rate": 8.268668061693184e-06, + "loss": 0.6617, + "step": 4313 + }, + { + "epoch": 0.2950954237635953, + "grad_norm": 2.5912122555786374, + "learning_rate": 8.267829715981736e-06, + "loss": 0.2526, + "step": 4314 + }, + { + "epoch": 0.29516382789520484, + "grad_norm": 1.1924929525009174, + "learning_rate": 8.266991209869429e-06, + "loss": 0.1645, + "step": 4315 + }, + { + "epoch": 0.2952322320268144, + "grad_norm": 1.857009938285869, + "learning_rate": 8.266152543397414e-06, + "loss": 0.3681, + "step": 4316 + }, + { + "epoch": 0.29530063615842395, + "grad_norm": 1.7782536326041058, + "learning_rate": 8.265313716606859e-06, + "loss": 0.2788, + "step": 4317 + }, + { + "epoch": 0.2953690402900335, + "grad_norm": 2.0827124413496874, + "learning_rate": 8.264474729538943e-06, + "loss": 0.5187, + "step": 4318 + }, + { + "epoch": 0.29543744442164305, + "grad_norm": 2.696528268363171, + "learning_rate": 8.26363558223484e-06, + "loss": 0.8324, + "step": 4319 + }, + { + "epoch": 0.2955058485532526, + "grad_norm": 1.2267639911886532, + "learning_rate": 8.262796274735742e-06, + "loss": 0.1857, + "step": 4320 + }, + { + "epoch": 0.29557425268486215, + "grad_norm": 2.373088249191226, + "learning_rate": 8.26195680708285e-06, + "loss": 0.3583, + "step": 4321 + }, + { + "epoch": 0.2956426568164717, + "grad_norm": 1.954302059260493, + "learning_rate": 8.261117179317364e-06, + "loss": 0.3484, + "step": 4322 + }, + { + "epoch": 0.29571106094808125, + "grad_norm": 2.6199948299699107, + "learning_rate": 8.2602773914805e-06, + "loss": 0.6237, + "step": 4323 + }, + { + "epoch": 0.29577946507969083, + "grad_norm": 1.9661181167939457, + "learning_rate": 8.259437443613478e-06, + "loss": 0.4168, + "step": 4324 + }, + { + "epoch": 0.29584786921130035, + "grad_norm": 2.058879025434105, + "learning_rate": 8.258597335757525e-06, + "loss": 0.5, + "step": 4325 + }, + { + "epoch": 0.29591627334290993, + "grad_norm": 1.586371594951006, + "learning_rate": 8.25775706795388e-06, + "loss": 0.2173, + "step": 4326 + }, + { + "epoch": 0.29598467747451945, + "grad_norm": 3.5586338145641885, + "learning_rate": 8.256916640243788e-06, + "loss": 0.461, + "step": 4327 + }, + { + "epoch": 0.29605308160612903, + "grad_norm": 2.7076843486734354, + "learning_rate": 8.256076052668498e-06, + "loss": 0.4748, + "step": 4328 + }, + { + "epoch": 0.29612148573773855, + "grad_norm": 1.6675261595960311, + "learning_rate": 8.255235305269274e-06, + "loss": 0.3555, + "step": 4329 + }, + { + "epoch": 0.29618988986934813, + "grad_norm": 2.191543166835338, + "learning_rate": 8.25439439808738e-06, + "loss": 0.4884, + "step": 4330 + }, + { + "epoch": 0.29625829400095766, + "grad_norm": 2.078035255874923, + "learning_rate": 8.253553331164095e-06, + "loss": 0.4847, + "step": 4331 + }, + { + "epoch": 0.29632669813256723, + "grad_norm": 1.8370789733227029, + "learning_rate": 8.252712104540699e-06, + "loss": 0.3902, + "step": 4332 + }, + { + "epoch": 0.29639510226417676, + "grad_norm": 1.7048896169728807, + "learning_rate": 8.251870718258489e-06, + "loss": 0.4388, + "step": 4333 + }, + { + "epoch": 0.2964635063957863, + "grad_norm": 2.1105030222588472, + "learning_rate": 8.251029172358757e-06, + "loss": 0.3154, + "step": 4334 + }, + { + "epoch": 0.29653191052739586, + "grad_norm": 2.3625160933254232, + "learning_rate": 8.250187466882816e-06, + "loss": 0.72, + "step": 4335 + }, + { + "epoch": 0.2966003146590054, + "grad_norm": 4.998768464054578, + "learning_rate": 8.24934560187198e-06, + "loss": 0.6977, + "step": 4336 + }, + { + "epoch": 0.29666871879061496, + "grad_norm": 2.2860629345212704, + "learning_rate": 8.24850357736757e-06, + "loss": 0.3318, + "step": 4337 + }, + { + "epoch": 0.2967371229222245, + "grad_norm": 1.9619562990418649, + "learning_rate": 8.24766139341092e-06, + "loss": 0.2558, + "step": 4338 + }, + { + "epoch": 0.29680552705383406, + "grad_norm": 1.9329022760596308, + "learning_rate": 8.246819050043363e-06, + "loss": 0.4529, + "step": 4339 + }, + { + "epoch": 0.2968739311854436, + "grad_norm": 1.6989296809769683, + "learning_rate": 8.245976547306247e-06, + "loss": 0.2839, + "step": 4340 + }, + { + "epoch": 0.29694233531705316, + "grad_norm": 74.13475291530439, + "learning_rate": 8.24513388524093e-06, + "loss": 0.7148, + "step": 4341 + }, + { + "epoch": 0.2970107394486627, + "grad_norm": 2.368658798293908, + "learning_rate": 8.244291063888772e-06, + "loss": 0.432, + "step": 4342 + }, + { + "epoch": 0.29707914358027226, + "grad_norm": 1.6126354023913037, + "learning_rate": 8.243448083291141e-06, + "loss": 0.207, + "step": 4343 + }, + { + "epoch": 0.2971475477118818, + "grad_norm": 2.493028953109837, + "learning_rate": 8.242604943489414e-06, + "loss": 0.8736, + "step": 4344 + }, + { + "epoch": 0.29721595184349137, + "grad_norm": 2.0584072195014707, + "learning_rate": 8.24176164452498e-06, + "loss": 0.5452, + "step": 4345 + }, + { + "epoch": 0.2972843559751009, + "grad_norm": 1.853881096458275, + "learning_rate": 8.240918186439231e-06, + "loss": 0.4272, + "step": 4346 + }, + { + "epoch": 0.29735276010671047, + "grad_norm": 1.7381484913224434, + "learning_rate": 8.240074569273566e-06, + "loss": 0.1981, + "step": 4347 + }, + { + "epoch": 0.29742116423832, + "grad_norm": 2.082854900658267, + "learning_rate": 8.239230793069394e-06, + "loss": 0.5135, + "step": 4348 + }, + { + "epoch": 0.29748956836992957, + "grad_norm": 2.028148572223436, + "learning_rate": 8.238386857868134e-06, + "loss": 0.32, + "step": 4349 + }, + { + "epoch": 0.2975579725015391, + "grad_norm": 2.2091076088823383, + "learning_rate": 8.237542763711208e-06, + "loss": 0.3765, + "step": 4350 + }, + { + "epoch": 0.2976263766331486, + "grad_norm": 1.7951859745828895, + "learning_rate": 8.236698510640051e-06, + "loss": 0.3417, + "step": 4351 + }, + { + "epoch": 0.2976947807647582, + "grad_norm": 1.5669117929869507, + "learning_rate": 8.235854098696099e-06, + "loss": 0.1974, + "step": 4352 + }, + { + "epoch": 0.2977631848963677, + "grad_norm": 1.7397109977529026, + "learning_rate": 8.235009527920804e-06, + "loss": 0.36, + "step": 4353 + }, + { + "epoch": 0.2978315890279773, + "grad_norm": 2.4185650356798054, + "learning_rate": 8.234164798355617e-06, + "loss": 0.4686, + "step": 4354 + }, + { + "epoch": 0.2978999931595868, + "grad_norm": 1.3590308060571978, + "learning_rate": 8.233319910042005e-06, + "loss": 0.1674, + "step": 4355 + }, + { + "epoch": 0.2979683972911964, + "grad_norm": 1.955752406043604, + "learning_rate": 8.23247486302144e-06, + "loss": 0.2577, + "step": 4356 + }, + { + "epoch": 0.2980368014228059, + "grad_norm": 2.525508624135376, + "learning_rate": 8.2316296573354e-06, + "loss": 0.5884, + "step": 4357 + }, + { + "epoch": 0.2981052055544155, + "grad_norm": 2.0623221031687002, + "learning_rate": 8.230784293025368e-06, + "loss": 0.5085, + "step": 4358 + }, + { + "epoch": 0.298173609686025, + "grad_norm": 1.7520429870966217, + "learning_rate": 8.229938770132843e-06, + "loss": 0.2808, + "step": 4359 + }, + { + "epoch": 0.2982420138176346, + "grad_norm": 2.2018801303666238, + "learning_rate": 8.229093088699327e-06, + "loss": 0.6929, + "step": 4360 + }, + { + "epoch": 0.2983104179492441, + "grad_norm": 2.027752642107411, + "learning_rate": 8.228247248766328e-06, + "loss": 0.5933, + "step": 4361 + }, + { + "epoch": 0.2983788220808537, + "grad_norm": 1.4501424966050405, + "learning_rate": 8.227401250375365e-06, + "loss": 0.289, + "step": 4362 + }, + { + "epoch": 0.2984472262124632, + "grad_norm": 2.369933823454317, + "learning_rate": 8.226555093567962e-06, + "loss": 0.3363, + "step": 4363 + }, + { + "epoch": 0.2985156303440728, + "grad_norm": 1.58936549123484, + "learning_rate": 8.225708778385657e-06, + "loss": 0.2733, + "step": 4364 + }, + { + "epoch": 0.2985840344756823, + "grad_norm": 1.8733070240606982, + "learning_rate": 8.224862304869986e-06, + "loss": 0.3434, + "step": 4365 + }, + { + "epoch": 0.2986524386072919, + "grad_norm": 1.3526371076832686, + "learning_rate": 8.224015673062503e-06, + "loss": 0.133, + "step": 4366 + }, + { + "epoch": 0.2987208427389014, + "grad_norm": 2.02209859905799, + "learning_rate": 8.22316888300476e-06, + "loss": 0.3318, + "step": 4367 + }, + { + "epoch": 0.298789246870511, + "grad_norm": 1.8284435891229491, + "learning_rate": 8.222321934738325e-06, + "loss": 0.4746, + "step": 4368 + }, + { + "epoch": 0.29885765100212053, + "grad_norm": 2.2140107009471004, + "learning_rate": 8.221474828304767e-06, + "loss": 0.5164, + "step": 4369 + }, + { + "epoch": 0.29892605513373005, + "grad_norm": 1.5986526385495314, + "learning_rate": 8.22062756374567e-06, + "loss": 0.265, + "step": 4370 + }, + { + "epoch": 0.29899445926533963, + "grad_norm": 2.1921438923833176, + "learning_rate": 8.219780141102618e-06, + "loss": 0.5913, + "step": 4371 + }, + { + "epoch": 0.29906286339694915, + "grad_norm": 2.151430663264606, + "learning_rate": 8.218932560417211e-06, + "loss": 0.5523, + "step": 4372 + }, + { + "epoch": 0.29913126752855873, + "grad_norm": 1.2049904718779152, + "learning_rate": 8.218084821731048e-06, + "loss": 0.1531, + "step": 4373 + }, + { + "epoch": 0.29919967166016825, + "grad_norm": 2.182361132048017, + "learning_rate": 8.217236925085742e-06, + "loss": 0.3253, + "step": 4374 + }, + { + "epoch": 0.29926807579177783, + "grad_norm": 2.0076195293865244, + "learning_rate": 8.216388870522913e-06, + "loss": 0.5847, + "step": 4375 + }, + { + "epoch": 0.29933647992338736, + "grad_norm": 2.0571656421900766, + "learning_rate": 8.215540658084187e-06, + "loss": 0.3833, + "step": 4376 + }, + { + "epoch": 0.29940488405499693, + "grad_norm": 1.9589040212990523, + "learning_rate": 8.214692287811197e-06, + "loss": 0.4325, + "step": 4377 + }, + { + "epoch": 0.29947328818660646, + "grad_norm": 2.0713425862736212, + "learning_rate": 8.213843759745584e-06, + "loss": 0.4695, + "step": 4378 + }, + { + "epoch": 0.29954169231821604, + "grad_norm": 1.8431779107247328, + "learning_rate": 8.212995073929002e-06, + "loss": 0.3731, + "step": 4379 + }, + { + "epoch": 0.29961009644982556, + "grad_norm": 2.622526854367581, + "learning_rate": 8.212146230403107e-06, + "loss": 0.4554, + "step": 4380 + }, + { + "epoch": 0.29967850058143514, + "grad_norm": 1.9546500222221364, + "learning_rate": 8.211297229209562e-06, + "loss": 0.4238, + "step": 4381 + }, + { + "epoch": 0.29974690471304466, + "grad_norm": 1.6152324515057956, + "learning_rate": 8.210448070390042e-06, + "loss": 0.284, + "step": 4382 + }, + { + "epoch": 0.29981530884465424, + "grad_norm": 2.312837550297307, + "learning_rate": 8.209598753986227e-06, + "loss": 0.6271, + "step": 4383 + }, + { + "epoch": 0.29988371297626376, + "grad_norm": 1.8739257019701363, + "learning_rate": 8.20874928003981e-06, + "loss": 0.101, + "step": 4384 + }, + { + "epoch": 0.29995211710787334, + "grad_norm": 2.0203124680633424, + "learning_rate": 8.20789964859248e-06, + "loss": 0.3607, + "step": 4385 + }, + { + "epoch": 0.30002052123948286, + "grad_norm": 1.7459106648016218, + "learning_rate": 8.207049859685944e-06, + "loss": 0.4022, + "step": 4386 + }, + { + "epoch": 0.30008892537109244, + "grad_norm": 1.660896544610951, + "learning_rate": 8.206199913361917e-06, + "loss": 0.3422, + "step": 4387 + }, + { + "epoch": 0.30015732950270196, + "grad_norm": 1.930870473475956, + "learning_rate": 8.205349809662113e-06, + "loss": 0.4775, + "step": 4388 + }, + { + "epoch": 0.3002257336343115, + "grad_norm": 1.903616289639257, + "learning_rate": 8.204499548628263e-06, + "loss": 0.6349, + "step": 4389 + }, + { + "epoch": 0.30029413776592107, + "grad_norm": 1.6905821054414703, + "learning_rate": 8.2036491303021e-06, + "loss": 0.4977, + "step": 4390 + }, + { + "epoch": 0.3003625418975306, + "grad_norm": 2.30496268165131, + "learning_rate": 8.202798554725369e-06, + "loss": 0.3744, + "step": 4391 + }, + { + "epoch": 0.30043094602914017, + "grad_norm": 1.7662211525741984, + "learning_rate": 8.201947821939817e-06, + "loss": 0.2902, + "step": 4392 + }, + { + "epoch": 0.3004993501607497, + "grad_norm": 2.572153290284858, + "learning_rate": 8.201096931987205e-06, + "loss": 0.7839, + "step": 4393 + }, + { + "epoch": 0.30056775429235927, + "grad_norm": 2.059921406556222, + "learning_rate": 8.200245884909297e-06, + "loss": 0.2444, + "step": 4394 + }, + { + "epoch": 0.3006361584239688, + "grad_norm": 1.728593167219811, + "learning_rate": 8.199394680747867e-06, + "loss": 0.449, + "step": 4395 + }, + { + "epoch": 0.30070456255557837, + "grad_norm": 2.2971678047648787, + "learning_rate": 8.198543319544696e-06, + "loss": 0.6019, + "step": 4396 + }, + { + "epoch": 0.3007729666871879, + "grad_norm": 2.1019930274463907, + "learning_rate": 8.197691801341576e-06, + "loss": 0.5252, + "step": 4397 + }, + { + "epoch": 0.30084137081879747, + "grad_norm": 1.7935050461487811, + "learning_rate": 8.196840126180298e-06, + "loss": 0.3881, + "step": 4398 + }, + { + "epoch": 0.300909774950407, + "grad_norm": 1.7720947601976345, + "learning_rate": 8.19598829410267e-06, + "loss": 0.1926, + "step": 4399 + }, + { + "epoch": 0.3009781790820166, + "grad_norm": 2.2026009578926513, + "learning_rate": 8.195136305150505e-06, + "loss": 0.5675, + "step": 4400 + }, + { + "epoch": 0.3010465832136261, + "grad_norm": 2.04019199526712, + "learning_rate": 8.194284159365617e-06, + "loss": 0.3717, + "step": 4401 + }, + { + "epoch": 0.3011149873452357, + "grad_norm": 1.8897905833479838, + "learning_rate": 8.19343185678984e-06, + "loss": 0.3539, + "step": 4402 + }, + { + "epoch": 0.3011833914768452, + "grad_norm": 2.3352192217751977, + "learning_rate": 8.192579397465006e-06, + "loss": 0.4605, + "step": 4403 + }, + { + "epoch": 0.3012517956084548, + "grad_norm": 1.613931847977869, + "learning_rate": 8.19172678143296e-06, + "loss": 0.343, + "step": 4404 + }, + { + "epoch": 0.3013201997400643, + "grad_norm": 2.312964605363656, + "learning_rate": 8.19087400873555e-06, + "loss": 0.6483, + "step": 4405 + }, + { + "epoch": 0.3013886038716738, + "grad_norm": 2.09756129243479, + "learning_rate": 8.190021079414635e-06, + "loss": 0.4645, + "step": 4406 + }, + { + "epoch": 0.3014570080032834, + "grad_norm": 1.1232074340819378, + "learning_rate": 8.189167993512079e-06, + "loss": 0.0889, + "step": 4407 + }, + { + "epoch": 0.3015254121348929, + "grad_norm": 2.42645334832209, + "learning_rate": 8.18831475106976e-06, + "loss": 0.6407, + "step": 4408 + }, + { + "epoch": 0.3015938162665025, + "grad_norm": 1.6176158964182306, + "learning_rate": 8.187461352129556e-06, + "loss": 0.2283, + "step": 4409 + }, + { + "epoch": 0.301662220398112, + "grad_norm": 1.8495000627677023, + "learning_rate": 8.186607796733357e-06, + "loss": 0.4907, + "step": 4410 + }, + { + "epoch": 0.3017306245297216, + "grad_norm": 1.5034396178373945, + "learning_rate": 8.185754084923059e-06, + "loss": 0.2273, + "step": 4411 + }, + { + "epoch": 0.3017990286613311, + "grad_norm": 2.147273879081319, + "learning_rate": 8.184900216740566e-06, + "loss": 0.3235, + "step": 4412 + }, + { + "epoch": 0.3018674327929407, + "grad_norm": 1.9322014853588152, + "learning_rate": 8.18404619222779e-06, + "loss": 0.4948, + "step": 4413 + }, + { + "epoch": 0.3019358369245502, + "grad_norm": 2.3909459459621565, + "learning_rate": 8.183192011426655e-06, + "loss": 0.5859, + "step": 4414 + }, + { + "epoch": 0.3020042410561598, + "grad_norm": 1.7599068943777556, + "learning_rate": 8.182337674379082e-06, + "loss": 0.3916, + "step": 4415 + }, + { + "epoch": 0.30207264518776933, + "grad_norm": 1.8775764248121665, + "learning_rate": 8.181483181127008e-06, + "loss": 0.5185, + "step": 4416 + }, + { + "epoch": 0.3021410493193789, + "grad_norm": 1.9599412994085184, + "learning_rate": 8.180628531712377e-06, + "loss": 0.4642, + "step": 4417 + }, + { + "epoch": 0.30220945345098843, + "grad_norm": 1.6756500068258415, + "learning_rate": 8.179773726177138e-06, + "loss": 0.3175, + "step": 4418 + }, + { + "epoch": 0.302277857582598, + "grad_norm": 1.6490687567277906, + "learning_rate": 8.178918764563251e-06, + "loss": 0.3052, + "step": 4419 + }, + { + "epoch": 0.30234626171420753, + "grad_norm": 1.5708322234322685, + "learning_rate": 8.17806364691268e-06, + "loss": 0.2695, + "step": 4420 + }, + { + "epoch": 0.3024146658458171, + "grad_norm": 2.5013296465701242, + "learning_rate": 8.177208373267398e-06, + "loss": 0.4475, + "step": 4421 + }, + { + "epoch": 0.30248306997742663, + "grad_norm": 1.9553590204997542, + "learning_rate": 8.176352943669385e-06, + "loss": 0.554, + "step": 4422 + }, + { + "epoch": 0.3025514741090362, + "grad_norm": 1.5355494046730196, + "learning_rate": 8.175497358160633e-06, + "loss": 0.2233, + "step": 4423 + }, + { + "epoch": 0.30261987824064573, + "grad_norm": 1.9059135226036026, + "learning_rate": 8.174641616783136e-06, + "loss": 0.2438, + "step": 4424 + }, + { + "epoch": 0.30268828237225526, + "grad_norm": 1.7345058374628326, + "learning_rate": 8.173785719578899e-06, + "loss": 0.475, + "step": 4425 + }, + { + "epoch": 0.30275668650386484, + "grad_norm": 2.28246028357941, + "learning_rate": 8.172929666589932e-06, + "loss": 0.4751, + "step": 4426 + }, + { + "epoch": 0.30282509063547436, + "grad_norm": 1.5405058410401316, + "learning_rate": 8.172073457858254e-06, + "loss": 0.344, + "step": 4427 + }, + { + "epoch": 0.30289349476708394, + "grad_norm": 1.7661192801333832, + "learning_rate": 8.171217093425893e-06, + "loss": 0.3552, + "step": 4428 + }, + { + "epoch": 0.30296189889869346, + "grad_norm": 1.9913726523952235, + "learning_rate": 8.170360573334885e-06, + "loss": 0.449, + "step": 4429 + }, + { + "epoch": 0.30303030303030304, + "grad_norm": 1.9387308064500015, + "learning_rate": 8.16950389762727e-06, + "loss": 0.4594, + "step": 4430 + }, + { + "epoch": 0.30309870716191256, + "grad_norm": 1.6597993265071551, + "learning_rate": 8.1686470663451e-06, + "loss": 0.2702, + "step": 4431 + }, + { + "epoch": 0.30316711129352214, + "grad_norm": 1.6982587185055649, + "learning_rate": 8.167790079530427e-06, + "loss": 0.2416, + "step": 4432 + }, + { + "epoch": 0.30323551542513166, + "grad_norm": 2.4890489935882654, + "learning_rate": 8.166932937225321e-06, + "loss": 0.5799, + "step": 4433 + }, + { + "epoch": 0.30330391955674124, + "grad_norm": 1.4262014292463008, + "learning_rate": 8.166075639471855e-06, + "loss": 0.2726, + "step": 4434 + }, + { + "epoch": 0.30337232368835076, + "grad_norm": 2.2539731847122932, + "learning_rate": 8.165218186312107e-06, + "loss": 0.2212, + "step": 4435 + }, + { + "epoch": 0.30344072781996034, + "grad_norm": 1.6079711485461092, + "learning_rate": 8.164360577788165e-06, + "loss": 0.1582, + "step": 4436 + }, + { + "epoch": 0.30350913195156987, + "grad_norm": 2.353259342533104, + "learning_rate": 8.163502813942126e-06, + "loss": 0.3764, + "step": 4437 + }, + { + "epoch": 0.30357753608317944, + "grad_norm": 1.958309688746648, + "learning_rate": 8.162644894816093e-06, + "loss": 0.3797, + "step": 4438 + }, + { + "epoch": 0.30364594021478897, + "grad_norm": 2.042251752851007, + "learning_rate": 8.161786820452176e-06, + "loss": 0.3717, + "step": 4439 + }, + { + "epoch": 0.30371434434639855, + "grad_norm": 1.9927656681096648, + "learning_rate": 8.160928590892493e-06, + "loss": 0.4331, + "step": 4440 + }, + { + "epoch": 0.30378274847800807, + "grad_norm": 2.4495939183405255, + "learning_rate": 8.160070206179171e-06, + "loss": 0.4531, + "step": 4441 + }, + { + "epoch": 0.30385115260961765, + "grad_norm": 1.8473006101777323, + "learning_rate": 8.159211666354343e-06, + "loss": 0.3024, + "step": 4442 + }, + { + "epoch": 0.30391955674122717, + "grad_norm": 2.4605720248634997, + "learning_rate": 8.158352971460151e-06, + "loss": 0.4442, + "step": 4443 + }, + { + "epoch": 0.3039879608728367, + "grad_norm": 2.3168462010639757, + "learning_rate": 8.157494121538745e-06, + "loss": 0.2245, + "step": 4444 + }, + { + "epoch": 0.30405636500444627, + "grad_norm": 2.024181538471333, + "learning_rate": 8.15663511663228e-06, + "loss": 0.2154, + "step": 4445 + }, + { + "epoch": 0.3041247691360558, + "grad_norm": 1.6695756519859057, + "learning_rate": 8.155775956782919e-06, + "loss": 0.3972, + "step": 4446 + }, + { + "epoch": 0.3041931732676654, + "grad_norm": 1.8811901273102944, + "learning_rate": 8.154916642032835e-06, + "loss": 0.5617, + "step": 4447 + }, + { + "epoch": 0.3042615773992749, + "grad_norm": 5.189400910619835, + "learning_rate": 8.154057172424208e-06, + "loss": 0.5288, + "step": 4448 + }, + { + "epoch": 0.3043299815308845, + "grad_norm": 1.8595624716999797, + "learning_rate": 8.153197547999223e-06, + "loss": 0.4717, + "step": 4449 + }, + { + "epoch": 0.304398385662494, + "grad_norm": 1.9064392128827181, + "learning_rate": 8.152337768800079e-06, + "loss": 0.3011, + "step": 4450 + }, + { + "epoch": 0.3044667897941036, + "grad_norm": 1.9612497559833557, + "learning_rate": 8.15147783486897e-06, + "loss": 0.2908, + "step": 4451 + }, + { + "epoch": 0.3045351939257131, + "grad_norm": 1.8459401074409312, + "learning_rate": 8.150617746248113e-06, + "loss": 0.2606, + "step": 4452 + }, + { + "epoch": 0.3046035980573227, + "grad_norm": 2.038597338882273, + "learning_rate": 8.149757502979723e-06, + "loss": 0.3377, + "step": 4453 + }, + { + "epoch": 0.3046720021889322, + "grad_norm": 1.7836501615127542, + "learning_rate": 8.148897105106026e-06, + "loss": 0.4503, + "step": 4454 + }, + { + "epoch": 0.3047404063205418, + "grad_norm": 2.207384131923524, + "learning_rate": 8.148036552669251e-06, + "loss": 0.2828, + "step": 4455 + }, + { + "epoch": 0.3048088104521513, + "grad_norm": 2.1627158647082783, + "learning_rate": 8.14717584571164e-06, + "loss": 0.5051, + "step": 4456 + }, + { + "epoch": 0.3048772145837609, + "grad_norm": 2.2759242336667507, + "learning_rate": 8.146314984275443e-06, + "loss": 0.3701, + "step": 4457 + }, + { + "epoch": 0.3049456187153704, + "grad_norm": 2.594295639647555, + "learning_rate": 8.145453968402912e-06, + "loss": 0.4882, + "step": 4458 + }, + { + "epoch": 0.30501402284698, + "grad_norm": 1.6390768375987872, + "learning_rate": 8.14459279813631e-06, + "loss": 0.3151, + "step": 4459 + }, + { + "epoch": 0.3050824269785895, + "grad_norm": 1.7340368766080847, + "learning_rate": 8.14373147351791e-06, + "loss": 0.2929, + "step": 4460 + }, + { + "epoch": 0.3051508311101991, + "grad_norm": 2.482117106700088, + "learning_rate": 8.142869994589988e-06, + "loss": 0.5146, + "step": 4461 + }, + { + "epoch": 0.3052192352418086, + "grad_norm": 1.5139027401004181, + "learning_rate": 8.14200836139483e-06, + "loss": 0.2934, + "step": 4462 + }, + { + "epoch": 0.30528763937341813, + "grad_norm": 1.0501478131240127, + "learning_rate": 8.14114657397473e-06, + "loss": 0.1227, + "step": 4463 + }, + { + "epoch": 0.3053560435050277, + "grad_norm": 2.323153442381116, + "learning_rate": 8.140284632371985e-06, + "loss": 0.518, + "step": 4464 + }, + { + "epoch": 0.30542444763663723, + "grad_norm": 2.508968324593584, + "learning_rate": 8.139422536628907e-06, + "loss": 0.6906, + "step": 4465 + }, + { + "epoch": 0.3054928517682468, + "grad_norm": 2.0020137145014973, + "learning_rate": 8.138560286787811e-06, + "loss": 0.275, + "step": 4466 + }, + { + "epoch": 0.30556125589985633, + "grad_norm": 1.573974377428135, + "learning_rate": 8.13769788289102e-06, + "loss": 0.3212, + "step": 4467 + }, + { + "epoch": 0.3056296600314659, + "grad_norm": 1.7554243098271483, + "learning_rate": 8.136835324980865e-06, + "loss": 0.2467, + "step": 4468 + }, + { + "epoch": 0.30569806416307543, + "grad_norm": 3.6028264366182774, + "learning_rate": 8.135972613099687e-06, + "loss": 0.3687, + "step": 4469 + }, + { + "epoch": 0.305766468294685, + "grad_norm": 2.359667816686245, + "learning_rate": 8.135109747289828e-06, + "loss": 0.2905, + "step": 4470 + }, + { + "epoch": 0.30583487242629454, + "grad_norm": 2.1294855572008524, + "learning_rate": 8.134246727593643e-06, + "loss": 0.5357, + "step": 4471 + }, + { + "epoch": 0.3059032765579041, + "grad_norm": 2.0303742354831384, + "learning_rate": 8.133383554053493e-06, + "loss": 0.4908, + "step": 4472 + }, + { + "epoch": 0.30597168068951364, + "grad_norm": 2.4251683208705144, + "learning_rate": 8.132520226711748e-06, + "loss": 0.3328, + "step": 4473 + }, + { + "epoch": 0.3060400848211232, + "grad_norm": 1.8437874192142258, + "learning_rate": 8.131656745610782e-06, + "loss": 0.3199, + "step": 4474 + }, + { + "epoch": 0.30610848895273274, + "grad_norm": 1.8258980326424283, + "learning_rate": 8.130793110792983e-06, + "loss": 0.2281, + "step": 4475 + }, + { + "epoch": 0.3061768930843423, + "grad_norm": 1.4406785989128679, + "learning_rate": 8.129929322300738e-06, + "loss": 0.2911, + "step": 4476 + }, + { + "epoch": 0.30624529721595184, + "grad_norm": 2.285736753627926, + "learning_rate": 8.129065380176448e-06, + "loss": 0.7736, + "step": 4477 + }, + { + "epoch": 0.3063137013475614, + "grad_norm": 1.4014326366067464, + "learning_rate": 8.12820128446252e-06, + "loss": 0.1455, + "step": 4478 + }, + { + "epoch": 0.30638210547917094, + "grad_norm": 4.121127825410811, + "learning_rate": 8.127337035201365e-06, + "loss": 0.372, + "step": 4479 + }, + { + "epoch": 0.30645050961078046, + "grad_norm": 1.890097524817507, + "learning_rate": 8.126472632435408e-06, + "loss": 0.213, + "step": 4480 + }, + { + "epoch": 0.30651891374239004, + "grad_norm": 2.0097407533165046, + "learning_rate": 8.125608076207077e-06, + "loss": 0.4082, + "step": 4481 + }, + { + "epoch": 0.30658731787399957, + "grad_norm": 2.2149935744000784, + "learning_rate": 8.124743366558805e-06, + "loss": 0.3953, + "step": 4482 + }, + { + "epoch": 0.30665572200560914, + "grad_norm": 1.7534587518617353, + "learning_rate": 8.123878503533042e-06, + "loss": 0.277, + "step": 4483 + }, + { + "epoch": 0.30672412613721867, + "grad_norm": 2.386125631353597, + "learning_rate": 8.123013487172239e-06, + "loss": 0.3303, + "step": 4484 + }, + { + "epoch": 0.30679253026882825, + "grad_norm": 1.6247145603815099, + "learning_rate": 8.12214831751885e-06, + "loss": 0.3605, + "step": 4485 + }, + { + "epoch": 0.30686093440043777, + "grad_norm": 2.0785796449720775, + "learning_rate": 8.121282994615345e-06, + "loss": 0.4675, + "step": 4486 + }, + { + "epoch": 0.30692933853204735, + "grad_norm": 1.7001068118493798, + "learning_rate": 8.120417518504198e-06, + "loss": 0.275, + "step": 4487 + }, + { + "epoch": 0.30699774266365687, + "grad_norm": 1.5930261885019337, + "learning_rate": 8.119551889227892e-06, + "loss": 0.285, + "step": 4488 + }, + { + "epoch": 0.30706614679526645, + "grad_norm": 1.8442870485353697, + "learning_rate": 8.118686106828914e-06, + "loss": 0.1491, + "step": 4489 + }, + { + "epoch": 0.30713455092687597, + "grad_norm": 2.0290074036433987, + "learning_rate": 8.117820171349764e-06, + "loss": 0.5202, + "step": 4490 + }, + { + "epoch": 0.30720295505848555, + "grad_norm": 1.0318797836360345, + "learning_rate": 8.116954082832944e-06, + "loss": 0.0816, + "step": 4491 + }, + { + "epoch": 0.3072713591900951, + "grad_norm": 1.8490385929113953, + "learning_rate": 8.116087841320963e-06, + "loss": 0.4121, + "step": 4492 + }, + { + "epoch": 0.30733976332170465, + "grad_norm": 2.5400800631982143, + "learning_rate": 8.115221446856345e-06, + "loss": 0.6038, + "step": 4493 + }, + { + "epoch": 0.3074081674533142, + "grad_norm": 4.121074210025692, + "learning_rate": 8.114354899481616e-06, + "loss": 0.7128, + "step": 4494 + }, + { + "epoch": 0.30747657158492375, + "grad_norm": 1.6590236300788335, + "learning_rate": 8.113488199239308e-06, + "loss": 0.2079, + "step": 4495 + }, + { + "epoch": 0.3075449757165333, + "grad_norm": 2.116498891465176, + "learning_rate": 8.112621346171967e-06, + "loss": 0.5699, + "step": 4496 + }, + { + "epoch": 0.30761337984814285, + "grad_norm": 2.2104613192365914, + "learning_rate": 8.111754340322136e-06, + "loss": 0.5457, + "step": 4497 + }, + { + "epoch": 0.3076817839797524, + "grad_norm": 1.9413475472922002, + "learning_rate": 8.110887181732379e-06, + "loss": 0.3384, + "step": 4498 + }, + { + "epoch": 0.3077501881113619, + "grad_norm": 1.5796436896130528, + "learning_rate": 8.110019870445254e-06, + "loss": 0.119, + "step": 4499 + }, + { + "epoch": 0.3078185922429715, + "grad_norm": 2.0681095710886708, + "learning_rate": 8.109152406503336e-06, + "loss": 0.4871, + "step": 4500 + }, + { + "epoch": 0.307886996374581, + "grad_norm": 1.6207315453947233, + "learning_rate": 8.108284789949206e-06, + "loss": 0.3344, + "step": 4501 + }, + { + "epoch": 0.3079554005061906, + "grad_norm": 2.378249668157004, + "learning_rate": 8.107417020825446e-06, + "loss": 0.7089, + "step": 4502 + }, + { + "epoch": 0.3080238046378001, + "grad_norm": 1.4805854308515949, + "learning_rate": 8.106549099174655e-06, + "loss": 0.2845, + "step": 4503 + }, + { + "epoch": 0.3080922087694097, + "grad_norm": 2.0623316696140344, + "learning_rate": 8.105681025039431e-06, + "loss": 0.3278, + "step": 4504 + }, + { + "epoch": 0.3081606129010192, + "grad_norm": 2.05788910476271, + "learning_rate": 8.104812798462388e-06, + "loss": 0.2659, + "step": 4505 + }, + { + "epoch": 0.3082290170326288, + "grad_norm": 1.9026840187634606, + "learning_rate": 8.103944419486138e-06, + "loss": 0.4966, + "step": 4506 + }, + { + "epoch": 0.3082974211642383, + "grad_norm": 1.915109052453869, + "learning_rate": 8.103075888153307e-06, + "loss": 0.3593, + "step": 4507 + }, + { + "epoch": 0.3083658252958479, + "grad_norm": 2.2883592898995992, + "learning_rate": 8.102207204506528e-06, + "loss": 0.5713, + "step": 4508 + }, + { + "epoch": 0.3084342294274574, + "grad_norm": 2.5248510812525127, + "learning_rate": 8.101338368588436e-06, + "loss": 0.5181, + "step": 4509 + }, + { + "epoch": 0.308502633559067, + "grad_norm": 2.1174646315835437, + "learning_rate": 8.100469380441684e-06, + "loss": 0.442, + "step": 4510 + }, + { + "epoch": 0.3085710376906765, + "grad_norm": 5.043574254733256, + "learning_rate": 8.099600240108922e-06, + "loss": 0.3628, + "step": 4511 + }, + { + "epoch": 0.3086394418222861, + "grad_norm": 1.8519623401141423, + "learning_rate": 8.09873094763281e-06, + "loss": 0.2375, + "step": 4512 + }, + { + "epoch": 0.3087078459538956, + "grad_norm": 1.921444092491352, + "learning_rate": 8.097861503056023e-06, + "loss": 0.4832, + "step": 4513 + }, + { + "epoch": 0.3087762500855052, + "grad_norm": 2.2266568950252594, + "learning_rate": 8.096991906421232e-06, + "loss": 0.3719, + "step": 4514 + }, + { + "epoch": 0.3088446542171147, + "grad_norm": 1.9877252607028297, + "learning_rate": 8.096122157771124e-06, + "loss": 0.4754, + "step": 4515 + }, + { + "epoch": 0.3089130583487243, + "grad_norm": 1.8788968440908913, + "learning_rate": 8.095252257148388e-06, + "loss": 0.3752, + "step": 4516 + }, + { + "epoch": 0.3089814624803338, + "grad_norm": 1.958674823800914, + "learning_rate": 8.094382204595725e-06, + "loss": 0.3408, + "step": 4517 + }, + { + "epoch": 0.30904986661194334, + "grad_norm": 1.609402739646877, + "learning_rate": 8.093512000155841e-06, + "loss": 0.3228, + "step": 4518 + }, + { + "epoch": 0.3091182707435529, + "grad_norm": 2.0768448231577117, + "learning_rate": 8.092641643871451e-06, + "loss": 0.3566, + "step": 4519 + }, + { + "epoch": 0.30918667487516244, + "grad_norm": 1.6830248502143619, + "learning_rate": 8.091771135785273e-06, + "loss": 0.3222, + "step": 4520 + }, + { + "epoch": 0.309255079006772, + "grad_norm": 2.561617420143436, + "learning_rate": 8.090900475940038e-06, + "loss": 0.6836, + "step": 4521 + }, + { + "epoch": 0.30932348313838154, + "grad_norm": 2.5812507754086864, + "learning_rate": 8.090029664378483e-06, + "loss": 0.6325, + "step": 4522 + }, + { + "epoch": 0.3093918872699911, + "grad_norm": 1.7546442429058802, + "learning_rate": 8.089158701143349e-06, + "loss": 0.4089, + "step": 4523 + }, + { + "epoch": 0.30946029140160064, + "grad_norm": 2.6815605568375864, + "learning_rate": 8.08828758627739e-06, + "loss": 0.4804, + "step": 4524 + }, + { + "epoch": 0.3095286955332102, + "grad_norm": 1.6473734074321562, + "learning_rate": 8.087416319823361e-06, + "loss": 0.3537, + "step": 4525 + }, + { + "epoch": 0.30959709966481974, + "grad_norm": 1.5075827742285135, + "learning_rate": 8.086544901824032e-06, + "loss": 0.2326, + "step": 4526 + }, + { + "epoch": 0.3096655037964293, + "grad_norm": 1.9215366717017346, + "learning_rate": 8.085673332322173e-06, + "loss": 0.4389, + "step": 4527 + }, + { + "epoch": 0.30973390792803884, + "grad_norm": 1.7695413677608418, + "learning_rate": 8.084801611360567e-06, + "loss": 0.2684, + "step": 4528 + }, + { + "epoch": 0.3098023120596484, + "grad_norm": 2.2181695796800196, + "learning_rate": 8.083929738982e-06, + "loss": 0.6196, + "step": 4529 + }, + { + "epoch": 0.30987071619125794, + "grad_norm": 2.91251534715792, + "learning_rate": 8.083057715229272e-06, + "loss": 0.7181, + "step": 4530 + }, + { + "epoch": 0.3099391203228675, + "grad_norm": 2.156787539250602, + "learning_rate": 8.082185540145181e-06, + "loss": 0.3952, + "step": 4531 + }, + { + "epoch": 0.31000752445447705, + "grad_norm": 1.791209743883282, + "learning_rate": 8.081313213772541e-06, + "loss": 0.4125, + "step": 4532 + }, + { + "epoch": 0.3100759285860866, + "grad_norm": 3.191996698376913, + "learning_rate": 8.08044073615417e-06, + "loss": 0.6563, + "step": 4533 + }, + { + "epoch": 0.31014433271769615, + "grad_norm": 1.8837397931321718, + "learning_rate": 8.079568107332892e-06, + "loss": 0.4225, + "step": 4534 + }, + { + "epoch": 0.3102127368493057, + "grad_norm": 1.8918126372035742, + "learning_rate": 8.07869532735154e-06, + "loss": 0.3639, + "step": 4535 + }, + { + "epoch": 0.31028114098091525, + "grad_norm": 2.7023434694394233, + "learning_rate": 8.077822396252954e-06, + "loss": 0.5519, + "step": 4536 + }, + { + "epoch": 0.31034954511252477, + "grad_norm": 2.240402358391167, + "learning_rate": 8.076949314079982e-06, + "loss": 0.5382, + "step": 4537 + }, + { + "epoch": 0.31041794924413435, + "grad_norm": 1.7593006933961852, + "learning_rate": 8.07607608087548e-06, + "loss": 0.3053, + "step": 4538 + }, + { + "epoch": 0.3104863533757439, + "grad_norm": 1.475996178807989, + "learning_rate": 8.07520269668231e-06, + "loss": 0.2837, + "step": 4539 + }, + { + "epoch": 0.31055475750735345, + "grad_norm": 2.069282768686207, + "learning_rate": 8.07432916154334e-06, + "loss": 0.5269, + "step": 4540 + }, + { + "epoch": 0.310623161638963, + "grad_norm": 2.074596754928873, + "learning_rate": 8.07345547550145e-06, + "loss": 0.3029, + "step": 4541 + }, + { + "epoch": 0.31069156577057255, + "grad_norm": 2.0575469248995524, + "learning_rate": 8.072581638599524e-06, + "loss": 0.6423, + "step": 4542 + }, + { + "epoch": 0.3107599699021821, + "grad_norm": 1.7134971545234479, + "learning_rate": 8.071707650880452e-06, + "loss": 0.2057, + "step": 4543 + }, + { + "epoch": 0.31082837403379165, + "grad_norm": 2.054010214338191, + "learning_rate": 8.070833512387138e-06, + "loss": 0.522, + "step": 4544 + }, + { + "epoch": 0.3108967781654012, + "grad_norm": 2.606856472705898, + "learning_rate": 8.069959223162485e-06, + "loss": 0.4849, + "step": 4545 + }, + { + "epoch": 0.31096518229701076, + "grad_norm": 2.0908705212240712, + "learning_rate": 8.069084783249408e-06, + "loss": 0.4616, + "step": 4546 + }, + { + "epoch": 0.3110335864286203, + "grad_norm": 1.8329155481578185, + "learning_rate": 8.068210192690831e-06, + "loss": 0.2995, + "step": 4547 + }, + { + "epoch": 0.31110199056022986, + "grad_norm": 1.943009349796085, + "learning_rate": 8.06733545152968e-06, + "loss": 0.2828, + "step": 4548 + }, + { + "epoch": 0.3111703946918394, + "grad_norm": 2.2425330025446297, + "learning_rate": 8.066460559808893e-06, + "loss": 0.5798, + "step": 4549 + }, + { + "epoch": 0.31123879882344896, + "grad_norm": 1.7998424615068027, + "learning_rate": 8.065585517571415e-06, + "loss": 0.2031, + "step": 4550 + }, + { + "epoch": 0.3113072029550585, + "grad_norm": 2.026268212532498, + "learning_rate": 8.064710324860195e-06, + "loss": 0.5321, + "step": 4551 + }, + { + "epoch": 0.31137560708666806, + "grad_norm": 1.433689553388421, + "learning_rate": 8.063834981718195e-06, + "loss": 0.318, + "step": 4552 + }, + { + "epoch": 0.3114440112182776, + "grad_norm": 2.2092156664360525, + "learning_rate": 8.062959488188378e-06, + "loss": 0.6107, + "step": 4553 + }, + { + "epoch": 0.3115124153498871, + "grad_norm": 1.7787819974776453, + "learning_rate": 8.062083844313718e-06, + "loss": 0.2198, + "step": 4554 + }, + { + "epoch": 0.3115808194814967, + "grad_norm": 2.0727280542134143, + "learning_rate": 8.061208050137197e-06, + "loss": 0.5609, + "step": 4555 + }, + { + "epoch": 0.3116492236131062, + "grad_norm": 1.9436713538557662, + "learning_rate": 8.0603321057018e-06, + "loss": 0.5549, + "step": 4556 + }, + { + "epoch": 0.3117176277447158, + "grad_norm": 1.5471926035572463, + "learning_rate": 8.059456011050528e-06, + "loss": 0.3057, + "step": 4557 + }, + { + "epoch": 0.3117860318763253, + "grad_norm": 2.584391832123895, + "learning_rate": 8.05857976622638e-06, + "loss": 0.5781, + "step": 4558 + }, + { + "epoch": 0.3118544360079349, + "grad_norm": 2.7738256854406647, + "learning_rate": 8.057703371272368e-06, + "loss": 0.6961, + "step": 4559 + }, + { + "epoch": 0.3119228401395444, + "grad_norm": 1.225602849714456, + "learning_rate": 8.056826826231507e-06, + "loss": 0.1984, + "step": 4560 + }, + { + "epoch": 0.311991244271154, + "grad_norm": 1.6791150204246867, + "learning_rate": 8.055950131146826e-06, + "loss": 0.3083, + "step": 4561 + }, + { + "epoch": 0.3120596484027635, + "grad_norm": 2.4129310536591073, + "learning_rate": 8.055073286061354e-06, + "loss": 0.5884, + "step": 4562 + }, + { + "epoch": 0.3121280525343731, + "grad_norm": 1.962391303022601, + "learning_rate": 8.054196291018134e-06, + "loss": 0.5272, + "step": 4563 + }, + { + "epoch": 0.3121964566659826, + "grad_norm": 1.820750777237462, + "learning_rate": 8.053319146060212e-06, + "loss": 0.3918, + "step": 4564 + }, + { + "epoch": 0.3122648607975922, + "grad_norm": 2.081906255867056, + "learning_rate": 8.052441851230639e-06, + "loss": 0.6461, + "step": 4565 + }, + { + "epoch": 0.3123332649292017, + "grad_norm": 2.1477211156647202, + "learning_rate": 8.051564406572482e-06, + "loss": 0.4994, + "step": 4566 + }, + { + "epoch": 0.3124016690608113, + "grad_norm": 1.4963697970625458, + "learning_rate": 8.050686812128807e-06, + "loss": 0.29, + "step": 4567 + }, + { + "epoch": 0.3124700731924208, + "grad_norm": 1.5436388854142804, + "learning_rate": 8.049809067942693e-06, + "loss": 0.1881, + "step": 4568 + }, + { + "epoch": 0.3125384773240304, + "grad_norm": 2.180766327472033, + "learning_rate": 8.048931174057222e-06, + "loss": 0.5535, + "step": 4569 + }, + { + "epoch": 0.3126068814556399, + "grad_norm": 2.159409081649674, + "learning_rate": 8.048053130515487e-06, + "loss": 0.4971, + "step": 4570 + }, + { + "epoch": 0.3126752855872495, + "grad_norm": 2.2175626935988073, + "learning_rate": 8.047174937360582e-06, + "loss": 0.4667, + "step": 4571 + }, + { + "epoch": 0.312743689718859, + "grad_norm": 1.6619992606314014, + "learning_rate": 8.046296594635618e-06, + "loss": 0.3341, + "step": 4572 + }, + { + "epoch": 0.31281209385046854, + "grad_norm": 2.276841090167361, + "learning_rate": 8.045418102383708e-06, + "loss": 0.7905, + "step": 4573 + }, + { + "epoch": 0.3128804979820781, + "grad_norm": 2.34996904520182, + "learning_rate": 8.044539460647971e-06, + "loss": 0.5689, + "step": 4574 + }, + { + "epoch": 0.31294890211368764, + "grad_norm": 1.4224779218366574, + "learning_rate": 8.043660669471536e-06, + "loss": 0.2838, + "step": 4575 + }, + { + "epoch": 0.3130173062452972, + "grad_norm": 1.4200548637060242, + "learning_rate": 8.042781728897535e-06, + "loss": 0.1755, + "step": 4576 + }, + { + "epoch": 0.31308571037690675, + "grad_norm": 3.18491297811082, + "learning_rate": 8.041902638969116e-06, + "loss": 0.6327, + "step": 4577 + }, + { + "epoch": 0.3131541145085163, + "grad_norm": 1.842922530134382, + "learning_rate": 8.041023399729425e-06, + "loss": 0.1914, + "step": 4578 + }, + { + "epoch": 0.31322251864012585, + "grad_norm": 1.4435583759478028, + "learning_rate": 8.040144011221621e-06, + "loss": 0.2308, + "step": 4579 + }, + { + "epoch": 0.3132909227717354, + "grad_norm": 2.101753123405507, + "learning_rate": 8.039264473488869e-06, + "loss": 0.4887, + "step": 4580 + }, + { + "epoch": 0.31335932690334495, + "grad_norm": 2.399361572188305, + "learning_rate": 8.038384786574338e-06, + "loss": 0.6392, + "step": 4581 + }, + { + "epoch": 0.3134277310349545, + "grad_norm": 2.0339892697041875, + "learning_rate": 8.037504950521212e-06, + "loss": 0.2171, + "step": 4582 + }, + { + "epoch": 0.31349613516656405, + "grad_norm": 2.0447607113526973, + "learning_rate": 8.036624965372672e-06, + "loss": 0.3521, + "step": 4583 + }, + { + "epoch": 0.31356453929817363, + "grad_norm": 1.8418537871385454, + "learning_rate": 8.035744831171919e-06, + "loss": 0.3481, + "step": 4584 + }, + { + "epoch": 0.31363294342978315, + "grad_norm": 1.97208835245597, + "learning_rate": 8.034864547962147e-06, + "loss": 0.26, + "step": 4585 + }, + { + "epoch": 0.31370134756139273, + "grad_norm": 2.2469744506441485, + "learning_rate": 8.033984115786569e-06, + "loss": 0.5187, + "step": 4586 + }, + { + "epoch": 0.31376975169300225, + "grad_norm": 2.9427103442069864, + "learning_rate": 8.033103534688399e-06, + "loss": 0.5159, + "step": 4587 + }, + { + "epoch": 0.31383815582461183, + "grad_norm": 2.3452154028894037, + "learning_rate": 8.03222280471086e-06, + "loss": 0.5748, + "step": 4588 + }, + { + "epoch": 0.31390655995622135, + "grad_norm": 1.2776656605533854, + "learning_rate": 8.031341925897186e-06, + "loss": 0.0881, + "step": 4589 + }, + { + "epoch": 0.31397496408783093, + "grad_norm": 2.1764370941708577, + "learning_rate": 8.03046089829061e-06, + "loss": 0.6363, + "step": 4590 + }, + { + "epoch": 0.31404336821944046, + "grad_norm": 2.336962089252301, + "learning_rate": 8.02957972193438e-06, + "loss": 0.7016, + "step": 4591 + }, + { + "epoch": 0.31411177235105, + "grad_norm": 2.048877308185183, + "learning_rate": 8.028698396871747e-06, + "loss": 0.3529, + "step": 4592 + }, + { + "epoch": 0.31418017648265956, + "grad_norm": 1.9128949399575785, + "learning_rate": 8.027816923145973e-06, + "loss": 0.2368, + "step": 4593 + }, + { + "epoch": 0.3142485806142691, + "grad_norm": 1.6434299197242472, + "learning_rate": 8.02693530080032e-06, + "loss": 0.2131, + "step": 4594 + }, + { + "epoch": 0.31431698474587866, + "grad_norm": 1.975768710179958, + "learning_rate": 8.026053529878068e-06, + "loss": 0.371, + "step": 4595 + }, + { + "epoch": 0.3143853888774882, + "grad_norm": 2.1536551148208076, + "learning_rate": 8.025171610422496e-06, + "loss": 0.5171, + "step": 4596 + }, + { + "epoch": 0.31445379300909776, + "grad_norm": 3.2409203753406546, + "learning_rate": 8.024289542476891e-06, + "loss": 0.2016, + "step": 4597 + }, + { + "epoch": 0.3145221971407073, + "grad_norm": 1.8625618367522763, + "learning_rate": 8.023407326084553e-06, + "loss": 0.4492, + "step": 4598 + }, + { + "epoch": 0.31459060127231686, + "grad_norm": 1.5119573616753372, + "learning_rate": 8.022524961288783e-06, + "loss": 0.1791, + "step": 4599 + }, + { + "epoch": 0.3146590054039264, + "grad_norm": 2.0042853600976596, + "learning_rate": 8.021642448132894e-06, + "loss": 0.3751, + "step": 4600 + }, + { + "epoch": 0.31472740953553596, + "grad_norm": 1.4804097241836465, + "learning_rate": 8.020759786660199e-06, + "loss": 0.2496, + "step": 4601 + }, + { + "epoch": 0.3147958136671455, + "grad_norm": 2.073101806200369, + "learning_rate": 8.019876976914029e-06, + "loss": 0.4509, + "step": 4602 + }, + { + "epoch": 0.31486421779875506, + "grad_norm": 2.5570665475359555, + "learning_rate": 8.018994018937713e-06, + "loss": 0.4008, + "step": 4603 + }, + { + "epoch": 0.3149326219303646, + "grad_norm": 1.6518740155393543, + "learning_rate": 8.018110912774592e-06, + "loss": 0.3352, + "step": 4604 + }, + { + "epoch": 0.31500102606197417, + "grad_norm": 2.3837136862658816, + "learning_rate": 8.017227658468015e-06, + "loss": 0.5796, + "step": 4605 + }, + { + "epoch": 0.3150694301935837, + "grad_norm": 1.5894806046453045, + "learning_rate": 8.01634425606133e-06, + "loss": 0.2769, + "step": 4606 + }, + { + "epoch": 0.31513783432519327, + "grad_norm": 1.566937109302434, + "learning_rate": 8.015460705597907e-06, + "loss": 0.216, + "step": 4607 + }, + { + "epoch": 0.3152062384568028, + "grad_norm": 1.674209658072488, + "learning_rate": 8.01457700712111e-06, + "loss": 0.4572, + "step": 4608 + }, + { + "epoch": 0.3152746425884123, + "grad_norm": 2.0002647202358594, + "learning_rate": 8.013693160674317e-06, + "loss": 0.4645, + "step": 4609 + }, + { + "epoch": 0.3153430467200219, + "grad_norm": 2.1271767827079664, + "learning_rate": 8.01280916630091e-06, + "loss": 0.6164, + "step": 4610 + }, + { + "epoch": 0.3154114508516314, + "grad_norm": 2.306004917101692, + "learning_rate": 8.011925024044279e-06, + "loss": 0.5042, + "step": 4611 + }, + { + "epoch": 0.315479854983241, + "grad_norm": 2.148666612372968, + "learning_rate": 8.011040733947826e-06, + "loss": 0.4217, + "step": 4612 + }, + { + "epoch": 0.3155482591148505, + "grad_norm": 2.1999092644165064, + "learning_rate": 8.010156296054952e-06, + "loss": 0.4071, + "step": 4613 + }, + { + "epoch": 0.3156166632464601, + "grad_norm": 2.0597885652881898, + "learning_rate": 8.009271710409071e-06, + "loss": 0.3424, + "step": 4614 + }, + { + "epoch": 0.3156850673780696, + "grad_norm": 1.9338456731504097, + "learning_rate": 8.008386977053604e-06, + "loss": 0.2706, + "step": 4615 + }, + { + "epoch": 0.3157534715096792, + "grad_norm": 2.077902208303061, + "learning_rate": 8.007502096031975e-06, + "loss": 0.4559, + "step": 4616 + }, + { + "epoch": 0.3158218756412887, + "grad_norm": 1.7837749872014663, + "learning_rate": 8.006617067387622e-06, + "loss": 0.4452, + "step": 4617 + }, + { + "epoch": 0.3158902797728983, + "grad_norm": 2.0983437418496473, + "learning_rate": 8.005731891163984e-06, + "loss": 0.4312, + "step": 4618 + }, + { + "epoch": 0.3159586839045078, + "grad_norm": 2.4050232249504613, + "learning_rate": 8.004846567404509e-06, + "loss": 0.4991, + "step": 4619 + }, + { + "epoch": 0.3160270880361174, + "grad_norm": 1.960130190989669, + "learning_rate": 8.003961096152654e-06, + "loss": 0.499, + "step": 4620 + }, + { + "epoch": 0.3160954921677269, + "grad_norm": 1.9366827825643371, + "learning_rate": 8.003075477451886e-06, + "loss": 0.4664, + "step": 4621 + }, + { + "epoch": 0.3161638962993365, + "grad_norm": 1.6542225700544015, + "learning_rate": 8.002189711345668e-06, + "loss": 0.2341, + "step": 4622 + }, + { + "epoch": 0.316232300430946, + "grad_norm": 2.2950004758969396, + "learning_rate": 8.001303797877485e-06, + "loss": 0.5045, + "step": 4623 + }, + { + "epoch": 0.3163007045625556, + "grad_norm": 2.338174163782774, + "learning_rate": 8.000417737090815e-06, + "loss": 0.7243, + "step": 4624 + }, + { + "epoch": 0.3163691086941651, + "grad_norm": 2.1480925404435594, + "learning_rate": 7.999531529029154e-06, + "loss": 0.5434, + "step": 4625 + }, + { + "epoch": 0.3164375128257747, + "grad_norm": 1.6002282912027828, + "learning_rate": 7.998645173736003e-06, + "loss": 0.2241, + "step": 4626 + }, + { + "epoch": 0.3165059169573842, + "grad_norm": 1.5394898707145492, + "learning_rate": 7.997758671254864e-06, + "loss": 0.3735, + "step": 4627 + }, + { + "epoch": 0.31657432108899375, + "grad_norm": 2.252810458106826, + "learning_rate": 7.996872021629255e-06, + "loss": 0.3945, + "step": 4628 + }, + { + "epoch": 0.3166427252206033, + "grad_norm": 1.2726832061767364, + "learning_rate": 7.995985224902697e-06, + "loss": 0.1989, + "step": 4629 + }, + { + "epoch": 0.31671112935221285, + "grad_norm": 1.8105305009449235, + "learning_rate": 7.995098281118716e-06, + "loss": 0.3034, + "step": 4630 + }, + { + "epoch": 0.31677953348382243, + "grad_norm": 1.5355553059818352, + "learning_rate": 7.994211190320847e-06, + "loss": 0.1302, + "step": 4631 + }, + { + "epoch": 0.31684793761543195, + "grad_norm": 1.6169609196737627, + "learning_rate": 7.993323952552634e-06, + "loss": 0.3969, + "step": 4632 + }, + { + "epoch": 0.31691634174704153, + "grad_norm": 2.21252873450548, + "learning_rate": 7.992436567857627e-06, + "loss": 0.606, + "step": 4633 + }, + { + "epoch": 0.31698474587865105, + "grad_norm": 1.883962691449425, + "learning_rate": 7.991549036279382e-06, + "loss": 0.2931, + "step": 4634 + }, + { + "epoch": 0.31705315001026063, + "grad_norm": 1.9459316463330345, + "learning_rate": 7.990661357861464e-06, + "loss": 0.4667, + "step": 4635 + }, + { + "epoch": 0.31712155414187015, + "grad_norm": 1.9829812298971825, + "learning_rate": 7.989773532647447e-06, + "loss": 0.4981, + "step": 4636 + }, + { + "epoch": 0.31718995827347973, + "grad_norm": 1.9346333993622316, + "learning_rate": 7.988885560680906e-06, + "loss": 0.577, + "step": 4637 + }, + { + "epoch": 0.31725836240508926, + "grad_norm": 2.2552780261000898, + "learning_rate": 7.987997442005428e-06, + "loss": 0.517, + "step": 4638 + }, + { + "epoch": 0.31732676653669883, + "grad_norm": 1.6341435469128776, + "learning_rate": 7.98710917666461e-06, + "loss": 0.4441, + "step": 4639 + }, + { + "epoch": 0.31739517066830836, + "grad_norm": 2.3361123220544786, + "learning_rate": 7.986220764702046e-06, + "loss": 0.3124, + "step": 4640 + }, + { + "epoch": 0.31746357479991794, + "grad_norm": 1.9079073751917162, + "learning_rate": 7.985332206161347e-06, + "loss": 0.3723, + "step": 4641 + }, + { + "epoch": 0.31753197893152746, + "grad_norm": 1.849434752715667, + "learning_rate": 7.984443501086127e-06, + "loss": 0.2555, + "step": 4642 + }, + { + "epoch": 0.31760038306313704, + "grad_norm": 1.8460405677047653, + "learning_rate": 7.983554649520009e-06, + "loss": 0.1623, + "step": 4643 + }, + { + "epoch": 0.31766878719474656, + "grad_norm": 2.861593235480272, + "learning_rate": 7.98266565150662e-06, + "loss": 0.3709, + "step": 4644 + }, + { + "epoch": 0.31773719132635614, + "grad_norm": 2.178947186282102, + "learning_rate": 7.981776507089599e-06, + "loss": 0.3737, + "step": 4645 + }, + { + "epoch": 0.31780559545796566, + "grad_norm": 1.8489449867698116, + "learning_rate": 7.980887216312587e-06, + "loss": 0.4952, + "step": 4646 + }, + { + "epoch": 0.3178739995895752, + "grad_norm": 2.397948153026968, + "learning_rate": 7.979997779219237e-06, + "loss": 0.59, + "step": 4647 + }, + { + "epoch": 0.31794240372118476, + "grad_norm": 1.935262664961438, + "learning_rate": 7.979108195853208e-06, + "loss": 0.3892, + "step": 4648 + }, + { + "epoch": 0.3180108078527943, + "grad_norm": 2.239844274105015, + "learning_rate": 7.978218466258158e-06, + "loss": 0.3762, + "step": 4649 + }, + { + "epoch": 0.31807921198440386, + "grad_norm": 2.099673091897471, + "learning_rate": 7.977328590477766e-06, + "loss": 0.4111, + "step": 4650 + }, + { + "epoch": 0.3181476161160134, + "grad_norm": 2.3477866887116154, + "learning_rate": 7.976438568555712e-06, + "loss": 0.7316, + "step": 4651 + }, + { + "epoch": 0.31821602024762297, + "grad_norm": 1.98464073444397, + "learning_rate": 7.975548400535677e-06, + "loss": 0.3227, + "step": 4652 + }, + { + "epoch": 0.3182844243792325, + "grad_norm": 2.298115135993361, + "learning_rate": 7.974658086461358e-06, + "loss": 0.5964, + "step": 4653 + }, + { + "epoch": 0.31835282851084207, + "grad_norm": 1.8831994205225568, + "learning_rate": 7.973767626376456e-06, + "loss": 0.3711, + "step": 4654 + }, + { + "epoch": 0.3184212326424516, + "grad_norm": 1.821733165932264, + "learning_rate": 7.97287702032468e-06, + "loss": 0.3453, + "step": 4655 + }, + { + "epoch": 0.31848963677406117, + "grad_norm": 2.821375448423535, + "learning_rate": 7.97198626834974e-06, + "loss": 0.5577, + "step": 4656 + }, + { + "epoch": 0.3185580409056707, + "grad_norm": 2.1825142718806534, + "learning_rate": 7.971095370495366e-06, + "loss": 0.5849, + "step": 4657 + }, + { + "epoch": 0.31862644503728027, + "grad_norm": 2.006300245380762, + "learning_rate": 7.970204326805284e-06, + "loss": 0.4618, + "step": 4658 + }, + { + "epoch": 0.3186948491688898, + "grad_norm": 1.81110315922865, + "learning_rate": 7.969313137323228e-06, + "loss": 0.3222, + "step": 4659 + }, + { + "epoch": 0.31876325330049937, + "grad_norm": 1.4162799062416336, + "learning_rate": 7.968421802092947e-06, + "loss": 0.1538, + "step": 4660 + }, + { + "epoch": 0.3188316574321089, + "grad_norm": 1.7337389157593421, + "learning_rate": 7.967530321158188e-06, + "loss": 0.354, + "step": 4661 + }, + { + "epoch": 0.3189000615637185, + "grad_norm": 1.6678214577671429, + "learning_rate": 7.966638694562712e-06, + "loss": 0.3882, + "step": 4662 + }, + { + "epoch": 0.318968465695328, + "grad_norm": 1.7030024156288703, + "learning_rate": 7.965746922350283e-06, + "loss": 0.4361, + "step": 4663 + }, + { + "epoch": 0.3190368698269376, + "grad_norm": 1.9737800720634586, + "learning_rate": 7.964855004564675e-06, + "loss": 0.4353, + "step": 4664 + }, + { + "epoch": 0.3191052739585471, + "grad_norm": 1.9422495447684403, + "learning_rate": 7.963962941249663e-06, + "loss": 0.5343, + "step": 4665 + }, + { + "epoch": 0.3191736780901566, + "grad_norm": 1.84578663066899, + "learning_rate": 7.963070732449039e-06, + "loss": 0.2398, + "step": 4666 + }, + { + "epoch": 0.3192420822217662, + "grad_norm": 2.061106331976055, + "learning_rate": 7.962178378206595e-06, + "loss": 0.3069, + "step": 4667 + }, + { + "epoch": 0.3193104863533757, + "grad_norm": 2.14545140216135, + "learning_rate": 7.961285878566132e-06, + "loss": 0.2753, + "step": 4668 + }, + { + "epoch": 0.3193788904849853, + "grad_norm": 2.407004225794938, + "learning_rate": 7.960393233571458e-06, + "loss": 0.7628, + "step": 4669 + }, + { + "epoch": 0.3194472946165948, + "grad_norm": 1.9749045976702233, + "learning_rate": 7.959500443266387e-06, + "loss": 0.4568, + "step": 4670 + }, + { + "epoch": 0.3195156987482044, + "grad_norm": 1.7425667473995172, + "learning_rate": 7.958607507694745e-06, + "loss": 0.3321, + "step": 4671 + }, + { + "epoch": 0.3195841028798139, + "grad_norm": 2.520338103749058, + "learning_rate": 7.957714426900358e-06, + "loss": 0.479, + "step": 4672 + }, + { + "epoch": 0.3196525070114235, + "grad_norm": 2.368828478547339, + "learning_rate": 7.956821200927064e-06, + "loss": 0.7207, + "step": 4673 + }, + { + "epoch": 0.319720911143033, + "grad_norm": 1.9389743694224175, + "learning_rate": 7.95592782981871e-06, + "loss": 0.4765, + "step": 4674 + }, + { + "epoch": 0.3197893152746426, + "grad_norm": 1.9159455087258235, + "learning_rate": 7.955034313619142e-06, + "loss": 0.4362, + "step": 4675 + }, + { + "epoch": 0.31985771940625213, + "grad_norm": 1.4080836474225695, + "learning_rate": 7.95414065237222e-06, + "loss": 0.1895, + "step": 4676 + }, + { + "epoch": 0.3199261235378617, + "grad_norm": 3.114165488041188, + "learning_rate": 7.953246846121809e-06, + "loss": 0.3217, + "step": 4677 + }, + { + "epoch": 0.31999452766947123, + "grad_norm": 2.123937060474269, + "learning_rate": 7.952352894911783e-06, + "loss": 0.3677, + "step": 4678 + }, + { + "epoch": 0.3200629318010808, + "grad_norm": 1.9524923978576953, + "learning_rate": 7.95145879878602e-06, + "loss": 0.3921, + "step": 4679 + }, + { + "epoch": 0.32013133593269033, + "grad_norm": 2.1481995780292085, + "learning_rate": 7.950564557788405e-06, + "loss": 0.3676, + "step": 4680 + }, + { + "epoch": 0.3201997400642999, + "grad_norm": 1.835323283753102, + "learning_rate": 7.949670171962836e-06, + "loss": 0.3956, + "step": 4681 + }, + { + "epoch": 0.32026814419590943, + "grad_norm": 2.26849186482273, + "learning_rate": 7.94877564135321e-06, + "loss": 0.62, + "step": 4682 + }, + { + "epoch": 0.32033654832751896, + "grad_norm": 2.182581846219878, + "learning_rate": 7.947880966003435e-06, + "loss": 0.382, + "step": 4683 + }, + { + "epoch": 0.32040495245912853, + "grad_norm": 1.5652411044466175, + "learning_rate": 7.946986145957428e-06, + "loss": 0.3355, + "step": 4684 + }, + { + "epoch": 0.32047335659073806, + "grad_norm": 1.8565359881077546, + "learning_rate": 7.94609118125911e-06, + "loss": 0.3406, + "step": 4685 + }, + { + "epoch": 0.32054176072234764, + "grad_norm": 2.413889144179052, + "learning_rate": 7.94519607195241e-06, + "loss": 0.4716, + "step": 4686 + }, + { + "epoch": 0.32061016485395716, + "grad_norm": 1.8393275028671892, + "learning_rate": 7.944300818081266e-06, + "loss": 0.179, + "step": 4687 + }, + { + "epoch": 0.32067856898556674, + "grad_norm": 1.4350491575675233, + "learning_rate": 7.943405419689619e-06, + "loss": 0.3386, + "step": 4688 + }, + { + "epoch": 0.32074697311717626, + "grad_norm": 1.7482102472266168, + "learning_rate": 7.94250987682142e-06, + "loss": 0.3456, + "step": 4689 + }, + { + "epoch": 0.32081537724878584, + "grad_norm": 1.7301159707927884, + "learning_rate": 7.941614189520625e-06, + "loss": 0.3531, + "step": 4690 + }, + { + "epoch": 0.32088378138039536, + "grad_norm": 1.4702694267344525, + "learning_rate": 7.940718357831206e-06, + "loss": 0.3368, + "step": 4691 + }, + { + "epoch": 0.32095218551200494, + "grad_norm": 2.2478776299661116, + "learning_rate": 7.939822381797124e-06, + "loss": 0.3345, + "step": 4692 + }, + { + "epoch": 0.32102058964361446, + "grad_norm": 2.1130874936307618, + "learning_rate": 7.938926261462366e-06, + "loss": 0.1641, + "step": 4693 + }, + { + "epoch": 0.32108899377522404, + "grad_norm": 1.8200398105810318, + "learning_rate": 7.938029996870916e-06, + "loss": 0.4561, + "step": 4694 + }, + { + "epoch": 0.32115739790683356, + "grad_norm": 2.2148473897698855, + "learning_rate": 7.937133588066763e-06, + "loss": 0.4922, + "step": 4695 + }, + { + "epoch": 0.32122580203844314, + "grad_norm": 1.956046894303641, + "learning_rate": 7.936237035093913e-06, + "loss": 0.5497, + "step": 4696 + }, + { + "epoch": 0.32129420617005267, + "grad_norm": 1.334445477127574, + "learning_rate": 7.93534033799637e-06, + "loss": 0.1746, + "step": 4697 + }, + { + "epoch": 0.32136261030166224, + "grad_norm": 1.9361454493313146, + "learning_rate": 7.934443496818147e-06, + "loss": 0.1957, + "step": 4698 + }, + { + "epoch": 0.32143101443327177, + "grad_norm": 1.8753712683904153, + "learning_rate": 7.933546511603269e-06, + "loss": 0.4519, + "step": 4699 + }, + { + "epoch": 0.32149941856488135, + "grad_norm": 1.9956057524955066, + "learning_rate": 7.93264938239576e-06, + "loss": 0.3593, + "step": 4700 + }, + { + "epoch": 0.32156782269649087, + "grad_norm": 2.6951371343203365, + "learning_rate": 7.93175210923966e-06, + "loss": 0.6585, + "step": 4701 + }, + { + "epoch": 0.3216362268281004, + "grad_norm": 1.663590272423463, + "learning_rate": 7.930854692179009e-06, + "loss": 0.4229, + "step": 4702 + }, + { + "epoch": 0.32170463095970997, + "grad_norm": 2.374376905419935, + "learning_rate": 7.929957131257855e-06, + "loss": 0.5821, + "step": 4703 + }, + { + "epoch": 0.3217730350913195, + "grad_norm": 1.9394132757943063, + "learning_rate": 7.929059426520259e-06, + "loss": 0.4716, + "step": 4704 + }, + { + "epoch": 0.32184143922292907, + "grad_norm": 1.9524934930302216, + "learning_rate": 7.928161578010279e-06, + "loss": 0.266, + "step": 4705 + }, + { + "epoch": 0.3219098433545386, + "grad_norm": 2.1212654255051566, + "learning_rate": 7.927263585771989e-06, + "loss": 0.5178, + "step": 4706 + }, + { + "epoch": 0.3219782474861482, + "grad_norm": 3.95763425007715, + "learning_rate": 7.92636544984947e-06, + "loss": 0.3659, + "step": 4707 + }, + { + "epoch": 0.3220466516177577, + "grad_norm": 2.005844922185458, + "learning_rate": 7.925467170286801e-06, + "loss": 0.4486, + "step": 4708 + }, + { + "epoch": 0.3221150557493673, + "grad_norm": 2.0728324129403872, + "learning_rate": 7.924568747128077e-06, + "loss": 0.457, + "step": 4709 + }, + { + "epoch": 0.3221834598809768, + "grad_norm": 1.5640292980621926, + "learning_rate": 7.923670180417394e-06, + "loss": 0.2752, + "step": 4710 + }, + { + "epoch": 0.3222518640125864, + "grad_norm": 2.2388129476582628, + "learning_rate": 7.922771470198862e-06, + "loss": 0.4465, + "step": 4711 + }, + { + "epoch": 0.3223202681441959, + "grad_norm": 1.7985701608987477, + "learning_rate": 7.921872616516593e-06, + "loss": 0.1991, + "step": 4712 + }, + { + "epoch": 0.3223886722758055, + "grad_norm": 1.907750750580406, + "learning_rate": 7.920973619414705e-06, + "loss": 0.3558, + "step": 4713 + }, + { + "epoch": 0.322457076407415, + "grad_norm": 2.0946880488200765, + "learning_rate": 7.920074478937327e-06, + "loss": 0.5731, + "step": 4714 + }, + { + "epoch": 0.3225254805390246, + "grad_norm": 1.6546683451738855, + "learning_rate": 7.919175195128593e-06, + "loss": 0.2967, + "step": 4715 + }, + { + "epoch": 0.3225938846706341, + "grad_norm": 3.0437966372541725, + "learning_rate": 7.918275768032646e-06, + "loss": 0.304, + "step": 4716 + }, + { + "epoch": 0.3226622888022437, + "grad_norm": 1.9950125391883973, + "learning_rate": 7.91737619769363e-06, + "loss": 0.4079, + "step": 4717 + }, + { + "epoch": 0.3227306929338532, + "grad_norm": 2.333438991527252, + "learning_rate": 7.916476484155702e-06, + "loss": 0.6038, + "step": 4718 + }, + { + "epoch": 0.3227990970654628, + "grad_norm": 1.7869215757158052, + "learning_rate": 7.915576627463024e-06, + "loss": 0.328, + "step": 4719 + }, + { + "epoch": 0.3228675011970723, + "grad_norm": 3.516932858807556, + "learning_rate": 7.914676627659766e-06, + "loss": 0.3012, + "step": 4720 + }, + { + "epoch": 0.3229359053286818, + "grad_norm": 1.6683880191917209, + "learning_rate": 7.913776484790105e-06, + "loss": 0.3137, + "step": 4721 + }, + { + "epoch": 0.3230043094602914, + "grad_norm": 2.009252758964969, + "learning_rate": 7.912876198898225e-06, + "loss": 0.3073, + "step": 4722 + }, + { + "epoch": 0.32307271359190093, + "grad_norm": 2.733263083946951, + "learning_rate": 7.911975770028314e-06, + "loss": 0.4796, + "step": 4723 + }, + { + "epoch": 0.3231411177235105, + "grad_norm": 2.235531321006042, + "learning_rate": 7.911075198224571e-06, + "loss": 0.5775, + "step": 4724 + }, + { + "epoch": 0.32320952185512003, + "grad_norm": 2.0047052437642967, + "learning_rate": 7.910174483531199e-06, + "loss": 0.3449, + "step": 4725 + }, + { + "epoch": 0.3232779259867296, + "grad_norm": 1.8800346491542566, + "learning_rate": 7.90927362599241e-06, + "loss": 0.358, + "step": 4726 + }, + { + "epoch": 0.32334633011833913, + "grad_norm": 1.854420052361352, + "learning_rate": 7.908372625652423e-06, + "loss": 0.2923, + "step": 4727 + }, + { + "epoch": 0.3234147342499487, + "grad_norm": 2.0925696661391697, + "learning_rate": 7.907471482555464e-06, + "loss": 0.663, + "step": 4728 + }, + { + "epoch": 0.32348313838155823, + "grad_norm": 1.6816148213213387, + "learning_rate": 7.906570196745763e-06, + "loss": 0.2264, + "step": 4729 + }, + { + "epoch": 0.3235515425131678, + "grad_norm": 1.6007402146694334, + "learning_rate": 7.905668768267562e-06, + "loss": 0.2837, + "step": 4730 + }, + { + "epoch": 0.32361994664477733, + "grad_norm": 1.687799674375439, + "learning_rate": 7.904767197165107e-06, + "loss": 0.3387, + "step": 4731 + }, + { + "epoch": 0.3236883507763869, + "grad_norm": 2.267857917918533, + "learning_rate": 7.90386548348265e-06, + "loss": 0.5334, + "step": 4732 + }, + { + "epoch": 0.32375675490799644, + "grad_norm": 1.7155262972540322, + "learning_rate": 7.902963627264453e-06, + "loss": 0.2899, + "step": 4733 + }, + { + "epoch": 0.323825159039606, + "grad_norm": 2.2038308561749607, + "learning_rate": 7.902061628554783e-06, + "loss": 0.1532, + "step": 4734 + }, + { + "epoch": 0.32389356317121554, + "grad_norm": 1.5406598924299097, + "learning_rate": 7.901159487397914e-06, + "loss": 0.2171, + "step": 4735 + }, + { + "epoch": 0.3239619673028251, + "grad_norm": 2.069765486517255, + "learning_rate": 7.900257203838127e-06, + "loss": 0.5203, + "step": 4736 + }, + { + "epoch": 0.32403037143443464, + "grad_norm": 2.140451131846052, + "learning_rate": 7.899354777919712e-06, + "loss": 0.4667, + "step": 4737 + }, + { + "epoch": 0.32409877556604416, + "grad_norm": 2.0639244420375182, + "learning_rate": 7.898452209686964e-06, + "loss": 0.5945, + "step": 4738 + }, + { + "epoch": 0.32416717969765374, + "grad_norm": 1.8634671372432259, + "learning_rate": 7.897549499184184e-06, + "loss": 0.4363, + "step": 4739 + }, + { + "epoch": 0.32423558382926326, + "grad_norm": 1.9661689748517939, + "learning_rate": 7.896646646455683e-06, + "loss": 0.5698, + "step": 4740 + }, + { + "epoch": 0.32430398796087284, + "grad_norm": 1.683244204246756, + "learning_rate": 7.895743651545779e-06, + "loss": 0.342, + "step": 4741 + }, + { + "epoch": 0.32437239209248236, + "grad_norm": 1.5413150907339788, + "learning_rate": 7.894840514498792e-06, + "loss": 0.1845, + "step": 4742 + }, + { + "epoch": 0.32444079622409194, + "grad_norm": 1.460985590913308, + "learning_rate": 7.893937235359053e-06, + "loss": 0.1108, + "step": 4743 + }, + { + "epoch": 0.32450920035570147, + "grad_norm": 2.2281192824944416, + "learning_rate": 7.893033814170899e-06, + "loss": 0.5791, + "step": 4744 + }, + { + "epoch": 0.32457760448731104, + "grad_norm": 1.4233756059240128, + "learning_rate": 7.892130250978675e-06, + "loss": 0.3124, + "step": 4745 + }, + { + "epoch": 0.32464600861892057, + "grad_norm": 1.7185482600330673, + "learning_rate": 7.891226545826733e-06, + "loss": 0.3585, + "step": 4746 + }, + { + "epoch": 0.32471441275053015, + "grad_norm": 2.238040658160043, + "learning_rate": 7.890322698759432e-06, + "loss": 0.5998, + "step": 4747 + }, + { + "epoch": 0.32478281688213967, + "grad_norm": 1.727780279437636, + "learning_rate": 7.889418709821133e-06, + "loss": 0.3375, + "step": 4748 + }, + { + "epoch": 0.32485122101374925, + "grad_norm": 1.2949809525536096, + "learning_rate": 7.888514579056213e-06, + "loss": 0.2146, + "step": 4749 + }, + { + "epoch": 0.32491962514535877, + "grad_norm": 1.6439177576189297, + "learning_rate": 7.887610306509048e-06, + "loss": 0.4479, + "step": 4750 + }, + { + "epoch": 0.32498802927696835, + "grad_norm": 1.88257145426885, + "learning_rate": 7.886705892224024e-06, + "loss": 0.4742, + "step": 4751 + }, + { + "epoch": 0.32505643340857787, + "grad_norm": 2.9598995191041113, + "learning_rate": 7.885801336245538e-06, + "loss": 0.508, + "step": 4752 + }, + { + "epoch": 0.32512483754018745, + "grad_norm": 1.9477774120642406, + "learning_rate": 7.884896638617985e-06, + "loss": 0.4285, + "step": 4753 + }, + { + "epoch": 0.325193241671797, + "grad_norm": 2.3454720083911256, + "learning_rate": 7.883991799385773e-06, + "loss": 0.4776, + "step": 4754 + }, + { + "epoch": 0.32526164580340655, + "grad_norm": 1.9481318621840427, + "learning_rate": 7.883086818593317e-06, + "loss": 0.4545, + "step": 4755 + }, + { + "epoch": 0.3253300499350161, + "grad_norm": 3.0556084593373773, + "learning_rate": 7.88218169628504e-06, + "loss": 0.4911, + "step": 4756 + }, + { + "epoch": 0.3253984540666256, + "grad_norm": 1.810651543646207, + "learning_rate": 7.881276432505366e-06, + "loss": 0.2619, + "step": 4757 + }, + { + "epoch": 0.3254668581982352, + "grad_norm": 1.690175595461533, + "learning_rate": 7.880371027298732e-06, + "loss": 0.402, + "step": 4758 + }, + { + "epoch": 0.3255352623298447, + "grad_norm": 2.236215273745812, + "learning_rate": 7.879465480709577e-06, + "loss": 0.5421, + "step": 4759 + }, + { + "epoch": 0.3256036664614543, + "grad_norm": 2.3241316642794936, + "learning_rate": 7.878559792782353e-06, + "loss": 0.6203, + "step": 4760 + }, + { + "epoch": 0.3256720705930638, + "grad_norm": 2.195600148018806, + "learning_rate": 7.877653963561513e-06, + "loss": 0.4907, + "step": 4761 + }, + { + "epoch": 0.3257404747246734, + "grad_norm": 1.3408098824714252, + "learning_rate": 7.876747993091521e-06, + "loss": 0.1936, + "step": 4762 + }, + { + "epoch": 0.3258088788562829, + "grad_norm": 2.1865105366635067, + "learning_rate": 7.875841881416846e-06, + "loss": 0.4323, + "step": 4763 + }, + { + "epoch": 0.3258772829878925, + "grad_norm": 1.780200649160814, + "learning_rate": 7.874935628581963e-06, + "loss": 0.3851, + "step": 4764 + }, + { + "epoch": 0.325945687119502, + "grad_norm": 2.503425057960861, + "learning_rate": 7.874029234631357e-06, + "loss": 0.4301, + "step": 4765 + }, + { + "epoch": 0.3260140912511116, + "grad_norm": 2.559819226494715, + "learning_rate": 7.873122699609518e-06, + "loss": 0.3924, + "step": 4766 + }, + { + "epoch": 0.3260824953827211, + "grad_norm": 1.7247406336083129, + "learning_rate": 7.872216023560944e-06, + "loss": 0.3117, + "step": 4767 + }, + { + "epoch": 0.3261508995143307, + "grad_norm": 1.7887891535602443, + "learning_rate": 7.871309206530136e-06, + "loss": 0.3155, + "step": 4768 + }, + { + "epoch": 0.3262193036459402, + "grad_norm": 2.037067295857787, + "learning_rate": 7.870402248561607e-06, + "loss": 0.4577, + "step": 4769 + }, + { + "epoch": 0.3262877077775498, + "grad_norm": 2.250406890741492, + "learning_rate": 7.869495149699874e-06, + "loss": 0.4129, + "step": 4770 + }, + { + "epoch": 0.3263561119091593, + "grad_norm": 1.5473329656056223, + "learning_rate": 7.868587909989465e-06, + "loss": 0.3421, + "step": 4771 + }, + { + "epoch": 0.3264245160407689, + "grad_norm": 1.561476229013098, + "learning_rate": 7.867680529474906e-06, + "loss": 0.2819, + "step": 4772 + }, + { + "epoch": 0.3264929201723784, + "grad_norm": 1.967854254414246, + "learning_rate": 7.866773008200741e-06, + "loss": 0.4414, + "step": 4773 + }, + { + "epoch": 0.326561324303988, + "grad_norm": 2.038100549182141, + "learning_rate": 7.865865346211512e-06, + "loss": 0.3689, + "step": 4774 + }, + { + "epoch": 0.3266297284355975, + "grad_norm": 2.1960320871517744, + "learning_rate": 7.864957543551772e-06, + "loss": 0.3817, + "step": 4775 + }, + { + "epoch": 0.32669813256720703, + "grad_norm": 1.803362000424687, + "learning_rate": 7.864049600266082e-06, + "loss": 0.2961, + "step": 4776 + }, + { + "epoch": 0.3267665366988166, + "grad_norm": 1.9158967639401607, + "learning_rate": 7.863141516399008e-06, + "loss": 0.2314, + "step": 4777 + }, + { + "epoch": 0.32683494083042613, + "grad_norm": 1.7164254508248942, + "learning_rate": 7.862233291995121e-06, + "loss": 0.1907, + "step": 4778 + }, + { + "epoch": 0.3269033449620357, + "grad_norm": 1.581676372709162, + "learning_rate": 7.861324927099004e-06, + "loss": 0.1681, + "step": 4779 + }, + { + "epoch": 0.32697174909364524, + "grad_norm": 1.9703653233198624, + "learning_rate": 7.86041642175524e-06, + "loss": 0.5447, + "step": 4780 + }, + { + "epoch": 0.3270401532252548, + "grad_norm": 1.3786167186616654, + "learning_rate": 7.859507776008427e-06, + "loss": 0.2159, + "step": 4781 + }, + { + "epoch": 0.32710855735686434, + "grad_norm": 1.9224396085616886, + "learning_rate": 7.858598989903164e-06, + "loss": 0.4429, + "step": 4782 + }, + { + "epoch": 0.3271769614884739, + "grad_norm": 2.185576656026152, + "learning_rate": 7.857690063484058e-06, + "loss": 0.5988, + "step": 4783 + }, + { + "epoch": 0.32724536562008344, + "grad_norm": 1.7410349946873505, + "learning_rate": 7.856780996795724e-06, + "loss": 0.2947, + "step": 4784 + }, + { + "epoch": 0.327313769751693, + "grad_norm": 1.7605675321373877, + "learning_rate": 7.855871789882783e-06, + "loss": 0.5546, + "step": 4785 + }, + { + "epoch": 0.32738217388330254, + "grad_norm": 1.8661594872840648, + "learning_rate": 7.854962442789866e-06, + "loss": 0.2179, + "step": 4786 + }, + { + "epoch": 0.3274505780149121, + "grad_norm": 1.795933406254988, + "learning_rate": 7.854052955561604e-06, + "loss": 0.3358, + "step": 4787 + }, + { + "epoch": 0.32751898214652164, + "grad_norm": 1.859765664901685, + "learning_rate": 7.85314332824264e-06, + "loss": 0.2062, + "step": 4788 + }, + { + "epoch": 0.3275873862781312, + "grad_norm": 1.6918278844788408, + "learning_rate": 7.852233560877624e-06, + "loss": 0.5043, + "step": 4789 + }, + { + "epoch": 0.32765579040974074, + "grad_norm": 2.0022665904192847, + "learning_rate": 7.851323653511213e-06, + "loss": 0.3111, + "step": 4790 + }, + { + "epoch": 0.3277241945413503, + "grad_norm": 2.182030099171227, + "learning_rate": 7.850413606188067e-06, + "loss": 0.4442, + "step": 4791 + }, + { + "epoch": 0.32779259867295985, + "grad_norm": 2.2634422195624477, + "learning_rate": 7.849503418952858e-06, + "loss": 0.4388, + "step": 4792 + }, + { + "epoch": 0.3278610028045694, + "grad_norm": 2.1713657536833813, + "learning_rate": 7.848593091850261e-06, + "loss": 0.4752, + "step": 4793 + }, + { + "epoch": 0.32792940693617895, + "grad_norm": 2.634630498114518, + "learning_rate": 7.847682624924957e-06, + "loss": 0.2995, + "step": 4794 + }, + { + "epoch": 0.32799781106778847, + "grad_norm": 1.8909649306895433, + "learning_rate": 7.84677201822164e-06, + "loss": 0.1965, + "step": 4795 + }, + { + "epoch": 0.32806621519939805, + "grad_norm": 1.6059416100099981, + "learning_rate": 7.845861271785006e-06, + "loss": 0.3124, + "step": 4796 + }, + { + "epoch": 0.32813461933100757, + "grad_norm": 2.0175778258804438, + "learning_rate": 7.844950385659758e-06, + "loss": 0.3325, + "step": 4797 + }, + { + "epoch": 0.32820302346261715, + "grad_norm": 2.5720770613971196, + "learning_rate": 7.844039359890608e-06, + "loss": 0.3774, + "step": 4798 + }, + { + "epoch": 0.3282714275942267, + "grad_norm": 2.3076971413946876, + "learning_rate": 7.84312819452227e-06, + "loss": 0.5903, + "step": 4799 + }, + { + "epoch": 0.32833983172583625, + "grad_norm": 1.9168399083756633, + "learning_rate": 7.842216889599473e-06, + "loss": 0.5645, + "step": 4800 + }, + { + "epoch": 0.3284082358574458, + "grad_norm": 2.2267658876918914, + "learning_rate": 7.841305445166943e-06, + "loss": 0.6197, + "step": 4801 + }, + { + "epoch": 0.32847663998905535, + "grad_norm": 1.377038804222586, + "learning_rate": 7.840393861269424e-06, + "loss": 0.3169, + "step": 4802 + }, + { + "epoch": 0.3285450441206649, + "grad_norm": 2.569601140241768, + "learning_rate": 7.839482137951658e-06, + "loss": 0.4306, + "step": 4803 + }, + { + "epoch": 0.32861344825227445, + "grad_norm": 1.54026130277356, + "learning_rate": 7.838570275258396e-06, + "loss": 0.1388, + "step": 4804 + }, + { + "epoch": 0.328681852383884, + "grad_norm": 1.9508194552626859, + "learning_rate": 7.837658273234399e-06, + "loss": 0.3776, + "step": 4805 + }, + { + "epoch": 0.32875025651549356, + "grad_norm": 2.1012282513418246, + "learning_rate": 7.83674613192443e-06, + "loss": 0.3475, + "step": 4806 + }, + { + "epoch": 0.3288186606471031, + "grad_norm": 1.27655583616873, + "learning_rate": 7.835833851373262e-06, + "loss": 0.1799, + "step": 4807 + }, + { + "epoch": 0.32888706477871266, + "grad_norm": 1.551733373277867, + "learning_rate": 7.834921431625675e-06, + "loss": 0.3368, + "step": 4808 + }, + { + "epoch": 0.3289554689103222, + "grad_norm": 1.936057669086576, + "learning_rate": 7.834008872726453e-06, + "loss": 0.4493, + "step": 4809 + }, + { + "epoch": 0.32902387304193176, + "grad_norm": 1.8920093031938299, + "learning_rate": 7.833096174720393e-06, + "loss": 0.3604, + "step": 4810 + }, + { + "epoch": 0.3290922771735413, + "grad_norm": 1.7977288166437817, + "learning_rate": 7.832183337652291e-06, + "loss": 0.1937, + "step": 4811 + }, + { + "epoch": 0.3291606813051508, + "grad_norm": 2.275261029658206, + "learning_rate": 7.831270361566952e-06, + "loss": 0.2117, + "step": 4812 + }, + { + "epoch": 0.3292290854367604, + "grad_norm": 2.2317296621174303, + "learning_rate": 7.830357246509194e-06, + "loss": 0.3966, + "step": 4813 + }, + { + "epoch": 0.3292974895683699, + "grad_norm": 2.408677041354863, + "learning_rate": 7.829443992523833e-06, + "loss": 0.5506, + "step": 4814 + }, + { + "epoch": 0.3293658936999795, + "grad_norm": 2.5364475635140646, + "learning_rate": 7.8285305996557e-06, + "loss": 0.4688, + "step": 4815 + }, + { + "epoch": 0.329434297831589, + "grad_norm": 1.886735754153629, + "learning_rate": 7.827617067949624e-06, + "loss": 0.287, + "step": 4816 + }, + { + "epoch": 0.3295027019631986, + "grad_norm": 1.9301313360893506, + "learning_rate": 7.82670339745045e-06, + "loss": 0.3733, + "step": 4817 + }, + { + "epoch": 0.3295711060948081, + "grad_norm": 1.846472701844209, + "learning_rate": 7.825789588203022e-06, + "loss": 0.2866, + "step": 4818 + }, + { + "epoch": 0.3296395102264177, + "grad_norm": 3.0743773636379412, + "learning_rate": 7.824875640252195e-06, + "loss": 0.3274, + "step": 4819 + }, + { + "epoch": 0.3297079143580272, + "grad_norm": 2.2301958054058972, + "learning_rate": 7.823961553642832e-06, + "loss": 0.2948, + "step": 4820 + }, + { + "epoch": 0.3297763184896368, + "grad_norm": 1.6459907796452011, + "learning_rate": 7.823047328419796e-06, + "loss": 0.1765, + "step": 4821 + }, + { + "epoch": 0.3298447226212463, + "grad_norm": 2.790016048711675, + "learning_rate": 7.822132964627968e-06, + "loss": 0.8474, + "step": 4822 + }, + { + "epoch": 0.3299131267528559, + "grad_norm": 1.7793542787774368, + "learning_rate": 7.821218462312226e-06, + "loss": 0.3323, + "step": 4823 + }, + { + "epoch": 0.3299815308844654, + "grad_norm": 2.4611849948015223, + "learning_rate": 7.820303821517457e-06, + "loss": 0.4117, + "step": 4824 + }, + { + "epoch": 0.330049935016075, + "grad_norm": 1.6155304412544615, + "learning_rate": 7.819389042288557e-06, + "loss": 0.3759, + "step": 4825 + }, + { + "epoch": 0.3301183391476845, + "grad_norm": 2.454266209815294, + "learning_rate": 7.81847412467043e-06, + "loss": 0.2919, + "step": 4826 + }, + { + "epoch": 0.3301867432792941, + "grad_norm": 1.6799819954264985, + "learning_rate": 7.817559068707981e-06, + "loss": 0.287, + "step": 4827 + }, + { + "epoch": 0.3302551474109036, + "grad_norm": 2.18240712496947, + "learning_rate": 7.816643874446127e-06, + "loss": 0.4596, + "step": 4828 + }, + { + "epoch": 0.3303235515425132, + "grad_norm": 2.73497704553822, + "learning_rate": 7.81572854192979e-06, + "loss": 0.4095, + "step": 4829 + }, + { + "epoch": 0.3303919556741227, + "grad_norm": 1.9445354181155343, + "learning_rate": 7.814813071203901e-06, + "loss": 0.5223, + "step": 4830 + }, + { + "epoch": 0.33046035980573224, + "grad_norm": 2.410001085130756, + "learning_rate": 7.813897462313393e-06, + "loss": 0.693, + "step": 4831 + }, + { + "epoch": 0.3305287639373418, + "grad_norm": 2.611036385650294, + "learning_rate": 7.81298171530321e-06, + "loss": 0.4666, + "step": 4832 + }, + { + "epoch": 0.33059716806895134, + "grad_norm": 1.5403790144084468, + "learning_rate": 7.812065830218298e-06, + "loss": 0.356, + "step": 4833 + }, + { + "epoch": 0.3306655722005609, + "grad_norm": 1.6920838705580379, + "learning_rate": 7.811149807103617e-06, + "loss": 0.4196, + "step": 4834 + }, + { + "epoch": 0.33073397633217044, + "grad_norm": 1.5159018868269905, + "learning_rate": 7.81023364600413e-06, + "loss": 0.2093, + "step": 4835 + }, + { + "epoch": 0.33080238046378, + "grad_norm": 2.1790037142243577, + "learning_rate": 7.809317346964804e-06, + "loss": 0.5187, + "step": 4836 + }, + { + "epoch": 0.33087078459538954, + "grad_norm": 1.5435346989940564, + "learning_rate": 7.808400910030617e-06, + "loss": 0.2725, + "step": 4837 + }, + { + "epoch": 0.3309391887269991, + "grad_norm": 1.731316756947936, + "learning_rate": 7.80748433524655e-06, + "loss": 0.3338, + "step": 4838 + }, + { + "epoch": 0.33100759285860865, + "grad_norm": 1.9732830222733004, + "learning_rate": 7.806567622657598e-06, + "loss": 0.4087, + "step": 4839 + }, + { + "epoch": 0.3310759969902182, + "grad_norm": 2.4866926191265133, + "learning_rate": 7.805650772308753e-06, + "loss": 0.7641, + "step": 4840 + }, + { + "epoch": 0.33114440112182775, + "grad_norm": 2.3287458296086823, + "learning_rate": 7.80473378424502e-06, + "loss": 0.5614, + "step": 4841 + }, + { + "epoch": 0.3312128052534373, + "grad_norm": 1.7659549278286324, + "learning_rate": 7.803816658511408e-06, + "loss": 0.4991, + "step": 4842 + }, + { + "epoch": 0.33128120938504685, + "grad_norm": 2.183318835645934, + "learning_rate": 7.802899395152933e-06, + "loss": 0.2476, + "step": 4843 + }, + { + "epoch": 0.3313496135166564, + "grad_norm": 2.032932662930904, + "learning_rate": 7.801981994214623e-06, + "loss": 0.6202, + "step": 4844 + }, + { + "epoch": 0.33141801764826595, + "grad_norm": 1.6564550542752985, + "learning_rate": 7.801064455741505e-06, + "loss": 0.2397, + "step": 4845 + }, + { + "epoch": 0.33148642177987553, + "grad_norm": 1.9580499392422586, + "learning_rate": 7.800146779778619e-06, + "loss": 0.484, + "step": 4846 + }, + { + "epoch": 0.33155482591148505, + "grad_norm": 2.3260925240999333, + "learning_rate": 7.799228966371006e-06, + "loss": 0.6276, + "step": 4847 + }, + { + "epoch": 0.33162323004309463, + "grad_norm": 2.0186346266170507, + "learning_rate": 7.798311015563718e-06, + "loss": 0.5776, + "step": 4848 + }, + { + "epoch": 0.33169163417470415, + "grad_norm": 1.7079165016050772, + "learning_rate": 7.797392927401814e-06, + "loss": 0.1583, + "step": 4849 + }, + { + "epoch": 0.3317600383063137, + "grad_norm": 2.2115927827105857, + "learning_rate": 7.796474701930355e-06, + "loss": 0.2352, + "step": 4850 + }, + { + "epoch": 0.33182844243792325, + "grad_norm": 1.8444425486427387, + "learning_rate": 7.795556339194415e-06, + "loss": 0.3763, + "step": 4851 + }, + { + "epoch": 0.3318968465695328, + "grad_norm": 1.9466870824949525, + "learning_rate": 7.794637839239068e-06, + "loss": 0.2171, + "step": 4852 + }, + { + "epoch": 0.33196525070114236, + "grad_norm": 2.3640802089030406, + "learning_rate": 7.793719202109402e-06, + "loss": 0.2291, + "step": 4853 + }, + { + "epoch": 0.3320336548327519, + "grad_norm": 2.638150381740212, + "learning_rate": 7.792800427850507e-06, + "loss": 0.3557, + "step": 4854 + }, + { + "epoch": 0.33210205896436146, + "grad_norm": 2.0117789620582704, + "learning_rate": 7.79188151650748e-06, + "loss": 0.476, + "step": 4855 + }, + { + "epoch": 0.332170463095971, + "grad_norm": 3.417129553088564, + "learning_rate": 7.790962468125428e-06, + "loss": 0.4473, + "step": 4856 + }, + { + "epoch": 0.33223886722758056, + "grad_norm": 1.8613869496643949, + "learning_rate": 7.79004328274946e-06, + "loss": 0.3812, + "step": 4857 + }, + { + "epoch": 0.3323072713591901, + "grad_norm": 1.8843868724961408, + "learning_rate": 7.789123960424693e-06, + "loss": 0.3735, + "step": 4858 + }, + { + "epoch": 0.33237567549079966, + "grad_norm": 2.0230604063392863, + "learning_rate": 7.788204501196255e-06, + "loss": 0.4352, + "step": 4859 + }, + { + "epoch": 0.3324440796224092, + "grad_norm": 1.888573282029442, + "learning_rate": 7.787284905109275e-06, + "loss": 0.4187, + "step": 4860 + }, + { + "epoch": 0.33251248375401876, + "grad_norm": 2.358968064917728, + "learning_rate": 7.786365172208894e-06, + "loss": 0.5485, + "step": 4861 + }, + { + "epoch": 0.3325808878856283, + "grad_norm": 1.9174920944588043, + "learning_rate": 7.785445302540253e-06, + "loss": 0.3407, + "step": 4862 + }, + { + "epoch": 0.33264929201723786, + "grad_norm": 1.6585371087744025, + "learning_rate": 7.784525296148507e-06, + "loss": 0.2727, + "step": 4863 + }, + { + "epoch": 0.3327176961488474, + "grad_norm": 2.007097176211136, + "learning_rate": 7.783605153078813e-06, + "loss": 0.4379, + "step": 4864 + }, + { + "epoch": 0.33278610028045696, + "grad_norm": 1.7594421800759366, + "learning_rate": 7.782684873376336e-06, + "loss": 0.4312, + "step": 4865 + }, + { + "epoch": 0.3328545044120665, + "grad_norm": 1.5440682658412295, + "learning_rate": 7.781764457086248e-06, + "loss": 0.3367, + "step": 4866 + }, + { + "epoch": 0.332922908543676, + "grad_norm": 2.2634349926159527, + "learning_rate": 7.780843904253727e-06, + "loss": 0.2461, + "step": 4867 + }, + { + "epoch": 0.3329913126752856, + "grad_norm": 2.107578543990231, + "learning_rate": 7.77992321492396e-06, + "loss": 0.3931, + "step": 4868 + }, + { + "epoch": 0.3330597168068951, + "grad_norm": 2.0981529542835133, + "learning_rate": 7.779002389142137e-06, + "loss": 0.5083, + "step": 4869 + }, + { + "epoch": 0.3331281209385047, + "grad_norm": 2.0254603404275375, + "learning_rate": 7.778081426953456e-06, + "loss": 0.3235, + "step": 4870 + }, + { + "epoch": 0.3331965250701142, + "grad_norm": 1.892033095954702, + "learning_rate": 7.777160328403123e-06, + "loss": 0.4627, + "step": 4871 + }, + { + "epoch": 0.3332649292017238, + "grad_norm": 1.8731687583687082, + "learning_rate": 7.776239093536353e-06, + "loss": 0.291, + "step": 4872 + }, + { + "epoch": 0.3333333333333333, + "grad_norm": 2.201432714433848, + "learning_rate": 7.77531772239836e-06, + "loss": 0.4699, + "step": 4873 + }, + { + "epoch": 0.3334017374649429, + "grad_norm": 2.0510595586941456, + "learning_rate": 7.774396215034371e-06, + "loss": 0.538, + "step": 4874 + }, + { + "epoch": 0.3334701415965524, + "grad_norm": 2.4674577469575407, + "learning_rate": 7.77347457148962e-06, + "loss": 0.4575, + "step": 4875 + }, + { + "epoch": 0.333538545728162, + "grad_norm": 2.2551656771113735, + "learning_rate": 7.772552791809343e-06, + "loss": 0.5914, + "step": 4876 + }, + { + "epoch": 0.3336069498597715, + "grad_norm": 2.235275920556355, + "learning_rate": 7.771630876038787e-06, + "loss": 0.5088, + "step": 4877 + }, + { + "epoch": 0.3336753539913811, + "grad_norm": 1.5479808957913435, + "learning_rate": 7.7707088242232e-06, + "loss": 0.1771, + "step": 4878 + }, + { + "epoch": 0.3337437581229906, + "grad_norm": 2.7962493622957747, + "learning_rate": 7.769786636407849e-06, + "loss": 0.5375, + "step": 4879 + }, + { + "epoch": 0.3338121622546002, + "grad_norm": 1.983410443148018, + "learning_rate": 7.768864312637993e-06, + "loss": 0.2595, + "step": 4880 + }, + { + "epoch": 0.3338805663862097, + "grad_norm": 1.7592437088731778, + "learning_rate": 7.767941852958906e-06, + "loss": 0.3582, + "step": 4881 + }, + { + "epoch": 0.3339489705178193, + "grad_norm": 1.9963308070410397, + "learning_rate": 7.767019257415866e-06, + "loss": 0.4396, + "step": 4882 + }, + { + "epoch": 0.3340173746494288, + "grad_norm": 2.4684806809618802, + "learning_rate": 7.766096526054158e-06, + "loss": 0.5751, + "step": 4883 + }, + { + "epoch": 0.3340857787810384, + "grad_norm": 1.8219862332257086, + "learning_rate": 7.765173658919079e-06, + "loss": 0.2222, + "step": 4884 + }, + { + "epoch": 0.3341541829126479, + "grad_norm": 1.91730976632011, + "learning_rate": 7.764250656055921e-06, + "loss": 0.3482, + "step": 4885 + }, + { + "epoch": 0.33422258704425745, + "grad_norm": 2.2815808944577185, + "learning_rate": 7.763327517509993e-06, + "loss": 0.5396, + "step": 4886 + }, + { + "epoch": 0.334290991175867, + "grad_norm": 1.7682706751232922, + "learning_rate": 7.762404243326607e-06, + "loss": 0.3935, + "step": 4887 + }, + { + "epoch": 0.33435939530747655, + "grad_norm": 2.2441819754061965, + "learning_rate": 7.76148083355108e-06, + "loss": 0.505, + "step": 4888 + }, + { + "epoch": 0.3344277994390861, + "grad_norm": 2.03303071609336, + "learning_rate": 7.76055728822874e-06, + "loss": 0.4564, + "step": 4889 + }, + { + "epoch": 0.33449620357069565, + "grad_norm": 2.355312215686855, + "learning_rate": 7.759633607404915e-06, + "loss": 0.4878, + "step": 4890 + }, + { + "epoch": 0.33456460770230523, + "grad_norm": 1.541512801948523, + "learning_rate": 7.75870979112495e-06, + "loss": 0.3232, + "step": 4891 + }, + { + "epoch": 0.33463301183391475, + "grad_norm": 1.9697882926773502, + "learning_rate": 7.757785839434185e-06, + "loss": 0.4142, + "step": 4892 + }, + { + "epoch": 0.33470141596552433, + "grad_norm": 2.223020104793544, + "learning_rate": 7.756861752377973e-06, + "loss": 0.2901, + "step": 4893 + }, + { + "epoch": 0.33476982009713385, + "grad_norm": 1.934064803881004, + "learning_rate": 7.755937530001676e-06, + "loss": 0.2083, + "step": 4894 + }, + { + "epoch": 0.33483822422874343, + "grad_norm": 2.046238185013047, + "learning_rate": 7.755013172350654e-06, + "loss": 0.2223, + "step": 4895 + }, + { + "epoch": 0.33490662836035295, + "grad_norm": 2.0301257910535777, + "learning_rate": 7.754088679470282e-06, + "loss": 0.5576, + "step": 4896 + }, + { + "epoch": 0.33497503249196253, + "grad_norm": 2.1325949044368766, + "learning_rate": 7.753164051405942e-06, + "loss": 0.4547, + "step": 4897 + }, + { + "epoch": 0.33504343662357206, + "grad_norm": 2.034489886823818, + "learning_rate": 7.75223928820301e-06, + "loss": 0.4394, + "step": 4898 + }, + { + "epoch": 0.33511184075518163, + "grad_norm": 1.5995423345991278, + "learning_rate": 7.751314389906887e-06, + "loss": 0.2326, + "step": 4899 + }, + { + "epoch": 0.33518024488679116, + "grad_norm": 2.022119879227926, + "learning_rate": 7.750389356562965e-06, + "loss": 0.3161, + "step": 4900 + }, + { + "epoch": 0.33524864901840074, + "grad_norm": 2.069855635994682, + "learning_rate": 7.749464188216656e-06, + "loss": 0.602, + "step": 4901 + }, + { + "epoch": 0.33531705315001026, + "grad_norm": 2.255122288096865, + "learning_rate": 7.748538884913363e-06, + "loss": 0.6735, + "step": 4902 + }, + { + "epoch": 0.33538545728161984, + "grad_norm": 1.732641384664032, + "learning_rate": 7.747613446698513e-06, + "loss": 0.3401, + "step": 4903 + }, + { + "epoch": 0.33545386141322936, + "grad_norm": 2.6456011998227784, + "learning_rate": 7.746687873617525e-06, + "loss": 0.7393, + "step": 4904 + }, + { + "epoch": 0.3355222655448389, + "grad_norm": 1.7494524116056411, + "learning_rate": 7.74576216571583e-06, + "loss": 0.2592, + "step": 4905 + }, + { + "epoch": 0.33559066967644846, + "grad_norm": 2.313148298117874, + "learning_rate": 7.744836323038873e-06, + "loss": 0.5523, + "step": 4906 + }, + { + "epoch": 0.335659073808058, + "grad_norm": 2.2853216192939096, + "learning_rate": 7.743910345632093e-06, + "loss": 0.3274, + "step": 4907 + }, + { + "epoch": 0.33572747793966756, + "grad_norm": 1.414083583421704, + "learning_rate": 7.742984233540942e-06, + "loss": 0.2265, + "step": 4908 + }, + { + "epoch": 0.3357958820712771, + "grad_norm": 2.102036036352177, + "learning_rate": 7.74205798681088e-06, + "loss": 0.3384, + "step": 4909 + }, + { + "epoch": 0.33586428620288666, + "grad_norm": 1.8803773284071763, + "learning_rate": 7.741131605487371e-06, + "loss": 0.1614, + "step": 4910 + }, + { + "epoch": 0.3359326903344962, + "grad_norm": 1.5837787372395082, + "learning_rate": 7.740205089615887e-06, + "loss": 0.231, + "step": 4911 + }, + { + "epoch": 0.33600109446610577, + "grad_norm": 2.15616721187828, + "learning_rate": 7.739278439241903e-06, + "loss": 0.5869, + "step": 4912 + }, + { + "epoch": 0.3360694985977153, + "grad_norm": 1.5607574318456503, + "learning_rate": 7.738351654410905e-06, + "loss": 0.1677, + "step": 4913 + }, + { + "epoch": 0.33613790272932487, + "grad_norm": 2.121429192027799, + "learning_rate": 7.737424735168386e-06, + "loss": 0.6756, + "step": 4914 + }, + { + "epoch": 0.3362063068609344, + "grad_norm": 1.6704626267484806, + "learning_rate": 7.736497681559843e-06, + "loss": 0.4227, + "step": 4915 + }, + { + "epoch": 0.33627471099254397, + "grad_norm": 1.5313830682778464, + "learning_rate": 7.735570493630777e-06, + "loss": 0.2331, + "step": 4916 + }, + { + "epoch": 0.3363431151241535, + "grad_norm": 2.1474050850259903, + "learning_rate": 7.734643171426702e-06, + "loss": 0.6473, + "step": 4917 + }, + { + "epoch": 0.33641151925576307, + "grad_norm": 2.507625929407718, + "learning_rate": 7.733715714993134e-06, + "loss": 0.4378, + "step": 4918 + }, + { + "epoch": 0.3364799233873726, + "grad_norm": 2.199502855450805, + "learning_rate": 7.7327881243756e-06, + "loss": 0.6914, + "step": 4919 + }, + { + "epoch": 0.33654832751898217, + "grad_norm": 1.9630128926775872, + "learning_rate": 7.731860399619626e-06, + "loss": 0.4313, + "step": 4920 + }, + { + "epoch": 0.3366167316505917, + "grad_norm": 2.5537377438132087, + "learning_rate": 7.730932540770753e-06, + "loss": 0.7376, + "step": 4921 + }, + { + "epoch": 0.3366851357822013, + "grad_norm": 1.6283715191440302, + "learning_rate": 7.730004547874524e-06, + "loss": 0.3047, + "step": 4922 + }, + { + "epoch": 0.3367535399138108, + "grad_norm": 1.9318468776397348, + "learning_rate": 7.729076420976487e-06, + "loss": 0.4038, + "step": 4923 + }, + { + "epoch": 0.3368219440454203, + "grad_norm": 2.0694831104519085, + "learning_rate": 7.728148160122201e-06, + "loss": 0.4686, + "step": 4924 + }, + { + "epoch": 0.3368903481770299, + "grad_norm": 2.3208543164133935, + "learning_rate": 7.72721976535723e-06, + "loss": 0.5557, + "step": 4925 + }, + { + "epoch": 0.3369587523086394, + "grad_norm": 2.317649096185567, + "learning_rate": 7.726291236727143e-06, + "loss": 0.4542, + "step": 4926 + }, + { + "epoch": 0.337027156440249, + "grad_norm": 1.589160795931915, + "learning_rate": 7.725362574277517e-06, + "loss": 0.2953, + "step": 4927 + }, + { + "epoch": 0.3370955605718585, + "grad_norm": 1.9009173409206575, + "learning_rate": 7.724433778053934e-06, + "loss": 0.4312, + "step": 4928 + }, + { + "epoch": 0.3371639647034681, + "grad_norm": 2.3071397569189287, + "learning_rate": 7.723504848101987e-06, + "loss": 0.5837, + "step": 4929 + }, + { + "epoch": 0.3372323688350776, + "grad_norm": 2.0553223532054377, + "learning_rate": 7.722575784467268e-06, + "loss": 0.5965, + "step": 4930 + }, + { + "epoch": 0.3373007729666872, + "grad_norm": 1.9327264092776697, + "learning_rate": 7.721646587195384e-06, + "loss": 0.4858, + "step": 4931 + }, + { + "epoch": 0.3373691770982967, + "grad_norm": 1.92907009209159, + "learning_rate": 7.720717256331943e-06, + "loss": 0.44, + "step": 4932 + }, + { + "epoch": 0.3374375812299063, + "grad_norm": 1.9075765918463528, + "learning_rate": 7.719787791922559e-06, + "loss": 0.4599, + "step": 4933 + }, + { + "epoch": 0.3375059853615158, + "grad_norm": 1.8374645456799177, + "learning_rate": 7.71885819401286e-06, + "loss": 0.3386, + "step": 4934 + }, + { + "epoch": 0.3375743894931254, + "grad_norm": 2.4568422302412967, + "learning_rate": 7.717928462648468e-06, + "loss": 0.7101, + "step": 4935 + }, + { + "epoch": 0.3376427936247349, + "grad_norm": 2.0774736136437473, + "learning_rate": 7.716998597875025e-06, + "loss": 0.6396, + "step": 4936 + }, + { + "epoch": 0.3377111977563445, + "grad_norm": 2.6024680351599248, + "learning_rate": 7.71606859973817e-06, + "loss": 0.6476, + "step": 4937 + }, + { + "epoch": 0.33777960188795403, + "grad_norm": 1.8417827973783476, + "learning_rate": 7.71513846828355e-06, + "loss": 0.3501, + "step": 4938 + }, + { + "epoch": 0.3378480060195636, + "grad_norm": 1.9845237150560555, + "learning_rate": 7.714208203556825e-06, + "loss": 0.4051, + "step": 4939 + }, + { + "epoch": 0.33791641015117313, + "grad_norm": 2.2071687571132013, + "learning_rate": 7.713277805603654e-06, + "loss": 0.4401, + "step": 4940 + }, + { + "epoch": 0.33798481428278265, + "grad_norm": 1.7087716411609857, + "learning_rate": 7.712347274469706e-06, + "loss": 0.3717, + "step": 4941 + }, + { + "epoch": 0.33805321841439223, + "grad_norm": 2.1364030273498007, + "learning_rate": 7.711416610200656e-06, + "loss": 0.5749, + "step": 4942 + }, + { + "epoch": 0.33812162254600175, + "grad_norm": 2.2203412998379926, + "learning_rate": 7.710485812842185e-06, + "loss": 0.7238, + "step": 4943 + }, + { + "epoch": 0.33819002667761133, + "grad_norm": 2.26428507940886, + "learning_rate": 7.709554882439982e-06, + "loss": 0.5672, + "step": 4944 + }, + { + "epoch": 0.33825843080922086, + "grad_norm": 2.3531070212073786, + "learning_rate": 7.70862381903974e-06, + "loss": 0.459, + "step": 4945 + }, + { + "epoch": 0.33832683494083043, + "grad_norm": 2.3102025076111876, + "learning_rate": 7.707692622687164e-06, + "loss": 0.343, + "step": 4946 + }, + { + "epoch": 0.33839523907243996, + "grad_norm": 1.6394793733972792, + "learning_rate": 7.706761293427956e-06, + "loss": 0.3274, + "step": 4947 + }, + { + "epoch": 0.33846364320404954, + "grad_norm": 1.6928609402651615, + "learning_rate": 7.705829831307834e-06, + "loss": 0.2119, + "step": 4948 + }, + { + "epoch": 0.33853204733565906, + "grad_norm": 1.7566824141424788, + "learning_rate": 7.704898236372517e-06, + "loss": 0.2965, + "step": 4949 + }, + { + "epoch": 0.33860045146726864, + "grad_norm": 2.3909519757326327, + "learning_rate": 7.703966508667733e-06, + "loss": 0.4784, + "step": 4950 + }, + { + "epoch": 0.33866885559887816, + "grad_norm": 2.1889781181414576, + "learning_rate": 7.703034648239216e-06, + "loss": 0.5561, + "step": 4951 + }, + { + "epoch": 0.33873725973048774, + "grad_norm": 2.0379633955770733, + "learning_rate": 7.702102655132704e-06, + "loss": 0.3001, + "step": 4952 + }, + { + "epoch": 0.33880566386209726, + "grad_norm": 2.521626947865022, + "learning_rate": 7.701170529393947e-06, + "loss": 0.3249, + "step": 4953 + }, + { + "epoch": 0.33887406799370684, + "grad_norm": 2.238067933015238, + "learning_rate": 7.700238271068697e-06, + "loss": 0.4023, + "step": 4954 + }, + { + "epoch": 0.33894247212531636, + "grad_norm": 1.6960437797168046, + "learning_rate": 7.699305880202711e-06, + "loss": 0.2738, + "step": 4955 + }, + { + "epoch": 0.33901087625692594, + "grad_norm": 2.349229561540129, + "learning_rate": 7.69837335684176e-06, + "loss": 0.4709, + "step": 4956 + }, + { + "epoch": 0.33907928038853546, + "grad_norm": 2.106594085745168, + "learning_rate": 7.697440701031612e-06, + "loss": 0.3818, + "step": 4957 + }, + { + "epoch": 0.33914768452014504, + "grad_norm": 2.0731759080970162, + "learning_rate": 7.696507912818052e-06, + "loss": 0.4134, + "step": 4958 + }, + { + "epoch": 0.33921608865175457, + "grad_norm": 1.7363518798027044, + "learning_rate": 7.69557499224686e-06, + "loss": 0.3158, + "step": 4959 + }, + { + "epoch": 0.3392844927833641, + "grad_norm": 2.0053943737639517, + "learning_rate": 7.694641939363831e-06, + "loss": 0.361, + "step": 4960 + }, + { + "epoch": 0.33935289691497367, + "grad_norm": 1.8900346149448621, + "learning_rate": 7.693708754214764e-06, + "loss": 0.3245, + "step": 4961 + }, + { + "epoch": 0.3394213010465832, + "grad_norm": 1.9721648595662176, + "learning_rate": 7.692775436845462e-06, + "loss": 0.3849, + "step": 4962 + }, + { + "epoch": 0.33948970517819277, + "grad_norm": 1.9162659632222565, + "learning_rate": 7.691841987301737e-06, + "loss": 0.427, + "step": 4963 + }, + { + "epoch": 0.3395581093098023, + "grad_norm": 1.9205285216981742, + "learning_rate": 7.690908405629413e-06, + "loss": 0.2156, + "step": 4964 + }, + { + "epoch": 0.33962651344141187, + "grad_norm": 2.2577800286409886, + "learning_rate": 7.689974691874307e-06, + "loss": 0.515, + "step": 4965 + }, + { + "epoch": 0.3396949175730214, + "grad_norm": 1.7810724362328676, + "learning_rate": 7.689040846082254e-06, + "loss": 0.3982, + "step": 4966 + }, + { + "epoch": 0.33976332170463097, + "grad_norm": 1.931284456038271, + "learning_rate": 7.688106868299092e-06, + "loss": 0.3065, + "step": 4967 + }, + { + "epoch": 0.3398317258362405, + "grad_norm": 1.7958973588468332, + "learning_rate": 7.687172758570663e-06, + "loss": 0.363, + "step": 4968 + }, + { + "epoch": 0.3399001299678501, + "grad_norm": 2.016764902514952, + "learning_rate": 7.686238516942819e-06, + "loss": 0.4607, + "step": 4969 + }, + { + "epoch": 0.3399685340994596, + "grad_norm": 2.6251052568540705, + "learning_rate": 7.685304143461418e-06, + "loss": 0.4174, + "step": 4970 + }, + { + "epoch": 0.3400369382310692, + "grad_norm": 2.3595666618056845, + "learning_rate": 7.684369638172321e-06, + "loss": 0.609, + "step": 4971 + }, + { + "epoch": 0.3401053423626787, + "grad_norm": 1.908771629335856, + "learning_rate": 7.6834350011214e-06, + "loss": 0.2829, + "step": 4972 + }, + { + "epoch": 0.3401737464942883, + "grad_norm": 1.598769562499415, + "learning_rate": 7.68250023235453e-06, + "loss": 0.2853, + "step": 4973 + }, + { + "epoch": 0.3402421506258978, + "grad_norm": 1.981961554842008, + "learning_rate": 7.681565331917596e-06, + "loss": 0.4527, + "step": 4974 + }, + { + "epoch": 0.3403105547575074, + "grad_norm": 2.155316313095448, + "learning_rate": 7.680630299856486e-06, + "loss": 0.5949, + "step": 4975 + }, + { + "epoch": 0.3403789588891169, + "grad_norm": 2.6527028241034536, + "learning_rate": 7.679695136217098e-06, + "loss": 0.6276, + "step": 4976 + }, + { + "epoch": 0.3404473630207265, + "grad_norm": 2.0209211956337993, + "learning_rate": 7.678759841045329e-06, + "loss": 0.3852, + "step": 4977 + }, + { + "epoch": 0.340515767152336, + "grad_norm": 2.3358294101572814, + "learning_rate": 7.677824414387093e-06, + "loss": 0.6362, + "step": 4978 + }, + { + "epoch": 0.3405841712839455, + "grad_norm": 1.828908300093097, + "learning_rate": 7.676888856288307e-06, + "loss": 0.371, + "step": 4979 + }, + { + "epoch": 0.3406525754155551, + "grad_norm": 2.4998943306525403, + "learning_rate": 7.675953166794886e-06, + "loss": 0.3675, + "step": 4980 + }, + { + "epoch": 0.3407209795471646, + "grad_norm": 1.6775889905436367, + "learning_rate": 7.67501734595276e-06, + "loss": 0.2225, + "step": 4981 + }, + { + "epoch": 0.3407893836787742, + "grad_norm": 2.2882652652035, + "learning_rate": 7.674081393807868e-06, + "loss": 0.5632, + "step": 4982 + }, + { + "epoch": 0.34085778781038373, + "grad_norm": 2.050373283860867, + "learning_rate": 7.673145310406148e-06, + "loss": 0.499, + "step": 4983 + }, + { + "epoch": 0.3409261919419933, + "grad_norm": 2.309433105135243, + "learning_rate": 7.672209095793545e-06, + "loss": 0.5887, + "step": 4984 + }, + { + "epoch": 0.34099459607360283, + "grad_norm": 2.800574086187672, + "learning_rate": 7.671272750016018e-06, + "loss": 0.2783, + "step": 4985 + }, + { + "epoch": 0.3410630002052124, + "grad_norm": 1.646520924190793, + "learning_rate": 7.670336273119525e-06, + "loss": 0.249, + "step": 4986 + }, + { + "epoch": 0.34113140433682193, + "grad_norm": 1.6059198256323755, + "learning_rate": 7.669399665150032e-06, + "loss": 0.3278, + "step": 4987 + }, + { + "epoch": 0.3411998084684315, + "grad_norm": 1.7503915948914062, + "learning_rate": 7.668462926153514e-06, + "loss": 0.2616, + "step": 4988 + }, + { + "epoch": 0.34126821260004103, + "grad_norm": 2.463771054648708, + "learning_rate": 7.667526056175949e-06, + "loss": 0.3977, + "step": 4989 + }, + { + "epoch": 0.3413366167316506, + "grad_norm": 2.2837669616167475, + "learning_rate": 7.666589055263323e-06, + "loss": 0.2579, + "step": 4990 + }, + { + "epoch": 0.34140502086326013, + "grad_norm": 1.5865279165108683, + "learning_rate": 7.665651923461632e-06, + "loss": 0.3535, + "step": 4991 + }, + { + "epoch": 0.3414734249948697, + "grad_norm": 2.1507228972268067, + "learning_rate": 7.66471466081687e-06, + "loss": 0.487, + "step": 4992 + }, + { + "epoch": 0.34154182912647923, + "grad_norm": 2.529064789364098, + "learning_rate": 7.663777267375045e-06, + "loss": 0.8036, + "step": 4993 + }, + { + "epoch": 0.3416102332580888, + "grad_norm": 1.6746086828907985, + "learning_rate": 7.662839743182168e-06, + "loss": 0.4808, + "step": 4994 + }, + { + "epoch": 0.34167863738969834, + "grad_norm": 1.7831652784561705, + "learning_rate": 7.66190208828426e-06, + "loss": 0.3385, + "step": 4995 + }, + { + "epoch": 0.34174704152130786, + "grad_norm": 1.9690821685317237, + "learning_rate": 7.66096430272734e-06, + "loss": 0.5352, + "step": 4996 + }, + { + "epoch": 0.34181544565291744, + "grad_norm": 1.57261172448093, + "learning_rate": 7.660026386557444e-06, + "loss": 0.3407, + "step": 4997 + }, + { + "epoch": 0.34188384978452696, + "grad_norm": 2.1724961853871805, + "learning_rate": 7.659088339820609e-06, + "loss": 0.4874, + "step": 4998 + }, + { + "epoch": 0.34195225391613654, + "grad_norm": 1.988808462697067, + "learning_rate": 7.658150162562875e-06, + "loss": 0.5044, + "step": 4999 + }, + { + "epoch": 0.34202065804774606, + "grad_norm": 1.6954942157588198, + "learning_rate": 7.657211854830297e-06, + "loss": 0.4161, + "step": 5000 + }, + { + "epoch": 0.34208906217935564, + "grad_norm": 2.2653993165310244, + "learning_rate": 7.65627341666893e-06, + "loss": 0.5284, + "step": 5001 + }, + { + "epoch": 0.34215746631096516, + "grad_norm": 2.3775313778139404, + "learning_rate": 7.655334848124836e-06, + "loss": 0.7241, + "step": 5002 + }, + { + "epoch": 0.34222587044257474, + "grad_norm": 1.8173104035528451, + "learning_rate": 7.654396149244083e-06, + "loss": 0.3734, + "step": 5003 + }, + { + "epoch": 0.34229427457418427, + "grad_norm": 2.3567151050758213, + "learning_rate": 7.653457320072752e-06, + "loss": 0.7378, + "step": 5004 + }, + { + "epoch": 0.34236267870579384, + "grad_norm": 1.339954733075088, + "learning_rate": 7.652518360656921e-06, + "loss": 0.1562, + "step": 5005 + }, + { + "epoch": 0.34243108283740337, + "grad_norm": 1.8664145166494257, + "learning_rate": 7.651579271042682e-06, + "loss": 0.4622, + "step": 5006 + }, + { + "epoch": 0.34249948696901295, + "grad_norm": 2.851377559569576, + "learning_rate": 7.650640051276124e-06, + "loss": 0.7488, + "step": 5007 + }, + { + "epoch": 0.34256789110062247, + "grad_norm": 2.0978164185479296, + "learning_rate": 7.649700701403356e-06, + "loss": 0.6525, + "step": 5008 + }, + { + "epoch": 0.34263629523223205, + "grad_norm": 2.254299678215752, + "learning_rate": 7.648761221470481e-06, + "loss": 0.6544, + "step": 5009 + }, + { + "epoch": 0.34270469936384157, + "grad_norm": 1.642559123978274, + "learning_rate": 7.647821611523614e-06, + "loss": 0.4376, + "step": 5010 + }, + { + "epoch": 0.34277310349545115, + "grad_norm": 1.9471988723929696, + "learning_rate": 7.646881871608877e-06, + "loss": 0.5264, + "step": 5011 + }, + { + "epoch": 0.34284150762706067, + "grad_norm": 2.151377299973751, + "learning_rate": 7.645942001772397e-06, + "loss": 0.5463, + "step": 5012 + }, + { + "epoch": 0.34290991175867025, + "grad_norm": 2.5071349134734664, + "learning_rate": 7.645002002060304e-06, + "loss": 0.6444, + "step": 5013 + }, + { + "epoch": 0.3429783158902798, + "grad_norm": 1.8207170667198422, + "learning_rate": 7.644061872518742e-06, + "loss": 0.4599, + "step": 5014 + }, + { + "epoch": 0.3430467200218893, + "grad_norm": 2.672580112865895, + "learning_rate": 7.643121613193856e-06, + "loss": 0.1837, + "step": 5015 + }, + { + "epoch": 0.3431151241534989, + "grad_norm": 1.662561939992997, + "learning_rate": 7.642181224131795e-06, + "loss": 0.3331, + "step": 5016 + }, + { + "epoch": 0.3431835282851084, + "grad_norm": 2.104113901927703, + "learning_rate": 7.641240705378722e-06, + "loss": 0.4562, + "step": 5017 + }, + { + "epoch": 0.343251932416718, + "grad_norm": 1.96660002563368, + "learning_rate": 7.640300056980801e-06, + "loss": 0.2952, + "step": 5018 + }, + { + "epoch": 0.3433203365483275, + "grad_norm": 2.2781076376976794, + "learning_rate": 7.639359278984202e-06, + "loss": 0.6892, + "step": 5019 + }, + { + "epoch": 0.3433887406799371, + "grad_norm": 2.4134118027228473, + "learning_rate": 7.638418371435105e-06, + "loss": 0.5197, + "step": 5020 + }, + { + "epoch": 0.3434571448115466, + "grad_norm": 2.8191562934937986, + "learning_rate": 7.637477334379695e-06, + "loss": 0.693, + "step": 5021 + }, + { + "epoch": 0.3435255489431562, + "grad_norm": 1.7634593037094408, + "learning_rate": 7.63653616786416e-06, + "loss": 0.2759, + "step": 5022 + }, + { + "epoch": 0.3435939530747657, + "grad_norm": 3.15547960151531, + "learning_rate": 7.635594871934697e-06, + "loss": 0.271, + "step": 5023 + }, + { + "epoch": 0.3436623572063753, + "grad_norm": 1.8538069300163555, + "learning_rate": 7.634653446637512e-06, + "loss": 0.2401, + "step": 5024 + }, + { + "epoch": 0.3437307613379848, + "grad_norm": 2.1144812767569148, + "learning_rate": 7.633711892018814e-06, + "loss": 0.5998, + "step": 5025 + }, + { + "epoch": 0.3437991654695944, + "grad_norm": 2.2594819282354353, + "learning_rate": 7.632770208124815e-06, + "loss": 0.5682, + "step": 5026 + }, + { + "epoch": 0.3438675696012039, + "grad_norm": 2.2365791796229804, + "learning_rate": 7.631828395001743e-06, + "loss": 0.286, + "step": 5027 + }, + { + "epoch": 0.3439359737328135, + "grad_norm": 2.621323480217402, + "learning_rate": 7.630886452695825e-06, + "loss": 0.7521, + "step": 5028 + }, + { + "epoch": 0.344004377864423, + "grad_norm": 1.9261688941030466, + "learning_rate": 7.629944381253295e-06, + "loss": 0.3059, + "step": 5029 + }, + { + "epoch": 0.3440727819960326, + "grad_norm": 2.8612613396776108, + "learning_rate": 7.629002180720395e-06, + "loss": 0.2895, + "step": 5030 + }, + { + "epoch": 0.3441411861276421, + "grad_norm": 1.7903804505540961, + "learning_rate": 7.628059851143372e-06, + "loss": 0.4609, + "step": 5031 + }, + { + "epoch": 0.3442095902592517, + "grad_norm": 2.040844022116806, + "learning_rate": 7.627117392568482e-06, + "loss": 0.5673, + "step": 5032 + }, + { + "epoch": 0.3442779943908612, + "grad_norm": 2.562904930685311, + "learning_rate": 7.626174805041981e-06, + "loss": 0.1546, + "step": 5033 + }, + { + "epoch": 0.34434639852247073, + "grad_norm": 1.9218834357347894, + "learning_rate": 7.625232088610141e-06, + "loss": 0.3925, + "step": 5034 + }, + { + "epoch": 0.3444148026540803, + "grad_norm": 2.8348680297681432, + "learning_rate": 7.624289243319235e-06, + "loss": 0.4488, + "step": 5035 + }, + { + "epoch": 0.34448320678568983, + "grad_norm": 1.7746144267167123, + "learning_rate": 7.623346269215537e-06, + "loss": 0.3828, + "step": 5036 + }, + { + "epoch": 0.3445516109172994, + "grad_norm": 1.86503484167991, + "learning_rate": 7.622403166345338e-06, + "loss": 0.3952, + "step": 5037 + }, + { + "epoch": 0.34462001504890893, + "grad_norm": 2.1812511940053825, + "learning_rate": 7.621459934754928e-06, + "loss": 0.5601, + "step": 5038 + }, + { + "epoch": 0.3446884191805185, + "grad_norm": 1.8789512981796004, + "learning_rate": 7.620516574490604e-06, + "loss": 0.3498, + "step": 5039 + }, + { + "epoch": 0.34475682331212804, + "grad_norm": 1.9258392433580376, + "learning_rate": 7.619573085598675e-06, + "loss": 0.5505, + "step": 5040 + }, + { + "epoch": 0.3448252274437376, + "grad_norm": 1.6281606536666278, + "learning_rate": 7.618629468125447e-06, + "loss": 0.3618, + "step": 5041 + }, + { + "epoch": 0.34489363157534714, + "grad_norm": 2.3029487451841004, + "learning_rate": 7.61768572211724e-06, + "loss": 0.574, + "step": 5042 + }, + { + "epoch": 0.3449620357069567, + "grad_norm": 1.8356356961163052, + "learning_rate": 7.616741847620377e-06, + "loss": 0.3414, + "step": 5043 + }, + { + "epoch": 0.34503043983856624, + "grad_norm": 2.0312281937529084, + "learning_rate": 7.6157978446811875e-06, + "loss": 0.43, + "step": 5044 + }, + { + "epoch": 0.3450988439701758, + "grad_norm": 2.2951934394054248, + "learning_rate": 7.6148537133460086e-06, + "loss": 0.3115, + "step": 5045 + }, + { + "epoch": 0.34516724810178534, + "grad_norm": 1.7093114964895333, + "learning_rate": 7.613909453661183e-06, + "loss": 0.2949, + "step": 5046 + }, + { + "epoch": 0.3452356522333949, + "grad_norm": 2.2534651729644595, + "learning_rate": 7.612965065673058e-06, + "loss": 0.4656, + "step": 5047 + }, + { + "epoch": 0.34530405636500444, + "grad_norm": 1.5579906793125495, + "learning_rate": 7.6120205494279905e-06, + "loss": 0.2093, + "step": 5048 + }, + { + "epoch": 0.345372460496614, + "grad_norm": 1.5976111491956222, + "learning_rate": 7.61107590497234e-06, + "loss": 0.1501, + "step": 5049 + }, + { + "epoch": 0.34544086462822354, + "grad_norm": 4.045804592833235, + "learning_rate": 7.610131132352478e-06, + "loss": 0.2029, + "step": 5050 + }, + { + "epoch": 0.3455092687598331, + "grad_norm": 1.8420311423834037, + "learning_rate": 7.609186231614774e-06, + "loss": 0.4557, + "step": 5051 + }, + { + "epoch": 0.34557767289144264, + "grad_norm": 2.002750394283518, + "learning_rate": 7.608241202805608e-06, + "loss": 0.2504, + "step": 5052 + }, + { + "epoch": 0.34564607702305217, + "grad_norm": 1.804175677939575, + "learning_rate": 7.607296045971373e-06, + "loss": 0.3009, + "step": 5053 + }, + { + "epoch": 0.34571448115466175, + "grad_norm": 2.1726112627174863, + "learning_rate": 7.6063507611584554e-06, + "loss": 0.4926, + "step": 5054 + }, + { + "epoch": 0.34578288528627127, + "grad_norm": 1.7864052015768774, + "learning_rate": 7.605405348413256e-06, + "loss": 0.2251, + "step": 5055 + }, + { + "epoch": 0.34585128941788085, + "grad_norm": 2.305842132995031, + "learning_rate": 7.604459807782181e-06, + "loss": 0.4259, + "step": 5056 + }, + { + "epoch": 0.34591969354949037, + "grad_norm": 1.7383409318429472, + "learning_rate": 7.603514139311643e-06, + "loss": 0.1961, + "step": 5057 + }, + { + "epoch": 0.34598809768109995, + "grad_norm": 2.5119877221830964, + "learning_rate": 7.602568343048058e-06, + "loss": 0.4487, + "step": 5058 + }, + { + "epoch": 0.34605650181270947, + "grad_norm": 2.1752120768259067, + "learning_rate": 7.601622419037851e-06, + "loss": 0.6118, + "step": 5059 + }, + { + "epoch": 0.34612490594431905, + "grad_norm": 2.274870948222786, + "learning_rate": 7.600676367327452e-06, + "loss": 0.2816, + "step": 5060 + }, + { + "epoch": 0.3461933100759286, + "grad_norm": 1.8174422396332368, + "learning_rate": 7.599730187963299e-06, + "loss": 0.3894, + "step": 5061 + }, + { + "epoch": 0.34626171420753815, + "grad_norm": 3.214199275035253, + "learning_rate": 7.598783880991833e-06, + "loss": 0.5053, + "step": 5062 + }, + { + "epoch": 0.3463301183391477, + "grad_norm": 1.7702044791575147, + "learning_rate": 7.597837446459506e-06, + "loss": 0.5407, + "step": 5063 + }, + { + "epoch": 0.34639852247075725, + "grad_norm": 2.5275783636281504, + "learning_rate": 7.596890884412773e-06, + "loss": 0.5092, + "step": 5064 + }, + { + "epoch": 0.3464669266023668, + "grad_norm": 2.1380495184872523, + "learning_rate": 7.595944194898095e-06, + "loss": 0.2265, + "step": 5065 + }, + { + "epoch": 0.34653533073397635, + "grad_norm": 2.2630218612674415, + "learning_rate": 7.594997377961939e-06, + "loss": 0.4191, + "step": 5066 + }, + { + "epoch": 0.3466037348655859, + "grad_norm": 2.1437081040940065, + "learning_rate": 7.594050433650781e-06, + "loss": 0.4447, + "step": 5067 + }, + { + "epoch": 0.34667213899719546, + "grad_norm": 1.7275061553011815, + "learning_rate": 7.593103362011102e-06, + "loss": 0.1821, + "step": 5068 + }, + { + "epoch": 0.346740543128805, + "grad_norm": 2.616644743252492, + "learning_rate": 7.592156163089387e-06, + "loss": 0.4621, + "step": 5069 + }, + { + "epoch": 0.3468089472604145, + "grad_norm": 2.1082967015396177, + "learning_rate": 7.5912088369321315e-06, + "loss": 0.2916, + "step": 5070 + }, + { + "epoch": 0.3468773513920241, + "grad_norm": 1.8895050191545106, + "learning_rate": 7.590261383585832e-06, + "loss": 0.4939, + "step": 5071 + }, + { + "epoch": 0.3469457555236336, + "grad_norm": 5.019918036494659, + "learning_rate": 7.589313803096996e-06, + "loss": 0.5849, + "step": 5072 + }, + { + "epoch": 0.3470141596552432, + "grad_norm": 1.746076647783538, + "learning_rate": 7.588366095512134e-06, + "loss": 0.2961, + "step": 5073 + }, + { + "epoch": 0.3470825637868527, + "grad_norm": 1.9468838270547535, + "learning_rate": 7.587418260877767e-06, + "loss": 0.1447, + "step": 5074 + }, + { + "epoch": 0.3471509679184623, + "grad_norm": 1.391619907726701, + "learning_rate": 7.586470299240416e-06, + "loss": 0.2528, + "step": 5075 + }, + { + "epoch": 0.3472193720500718, + "grad_norm": 1.4260097216094858, + "learning_rate": 7.585522210646612e-06, + "loss": 0.3257, + "step": 5076 + }, + { + "epoch": 0.3472877761816814, + "grad_norm": 1.9363845836972808, + "learning_rate": 7.5845739951428944e-06, + "loss": 0.4095, + "step": 5077 + }, + { + "epoch": 0.3473561803132909, + "grad_norm": 1.829278500497451, + "learning_rate": 7.583625652775803e-06, + "loss": 0.3803, + "step": 5078 + }, + { + "epoch": 0.3474245844449005, + "grad_norm": 2.1617640191050764, + "learning_rate": 7.582677183591889e-06, + "loss": 0.4112, + "step": 5079 + }, + { + "epoch": 0.34749298857651, + "grad_norm": 2.1401813910606866, + "learning_rate": 7.581728587637707e-06, + "loss": 0.5066, + "step": 5080 + }, + { + "epoch": 0.3475613927081196, + "grad_norm": 2.451280258639731, + "learning_rate": 7.580779864959819e-06, + "loss": 0.7859, + "step": 5081 + }, + { + "epoch": 0.3476297968397291, + "grad_norm": 2.0301339971778574, + "learning_rate": 7.579831015604792e-06, + "loss": 0.3852, + "step": 5082 + }, + { + "epoch": 0.3476982009713387, + "grad_norm": 2.7901258979185117, + "learning_rate": 7.578882039619202e-06, + "loss": 0.3572, + "step": 5083 + }, + { + "epoch": 0.3477666051029482, + "grad_norm": 2.3408056011223057, + "learning_rate": 7.577932937049628e-06, + "loss": 0.4596, + "step": 5084 + }, + { + "epoch": 0.3478350092345578, + "grad_norm": 1.85513343592023, + "learning_rate": 7.576983707942655e-06, + "loss": 0.3893, + "step": 5085 + }, + { + "epoch": 0.3479034133661673, + "grad_norm": 2.076936008798247, + "learning_rate": 7.576034352344879e-06, + "loss": 0.4972, + "step": 5086 + }, + { + "epoch": 0.3479718174977769, + "grad_norm": 3.208664305240717, + "learning_rate": 7.575084870302898e-06, + "loss": 0.3119, + "step": 5087 + }, + { + "epoch": 0.3480402216293864, + "grad_norm": 2.0751405093866557, + "learning_rate": 7.574135261863315e-06, + "loss": 0.3967, + "step": 5088 + }, + { + "epoch": 0.34810862576099594, + "grad_norm": 2.253180891235272, + "learning_rate": 7.573185527072745e-06, + "loss": 0.3718, + "step": 5089 + }, + { + "epoch": 0.3481770298926055, + "grad_norm": 1.5055026140150727, + "learning_rate": 7.572235665977803e-06, + "loss": 0.2726, + "step": 5090 + }, + { + "epoch": 0.34824543402421504, + "grad_norm": 3.1076164256633136, + "learning_rate": 7.571285678625112e-06, + "loss": 0.5411, + "step": 5091 + }, + { + "epoch": 0.3483138381558246, + "grad_norm": 1.6586115303829445, + "learning_rate": 7.570335565061304e-06, + "loss": 0.3095, + "step": 5092 + }, + { + "epoch": 0.34838224228743414, + "grad_norm": 1.7331289420373646, + "learning_rate": 7.569385325333014e-06, + "loss": 0.4161, + "step": 5093 + }, + { + "epoch": 0.3484506464190437, + "grad_norm": 2.485001443060412, + "learning_rate": 7.568434959486885e-06, + "loss": 0.4685, + "step": 5094 + }, + { + "epoch": 0.34851905055065324, + "grad_norm": 1.6279356006132768, + "learning_rate": 7.567484467569566e-06, + "loss": 0.431, + "step": 5095 + }, + { + "epoch": 0.3485874546822628, + "grad_norm": 2.0451666276522604, + "learning_rate": 7.5665338496277105e-06, + "loss": 0.4366, + "step": 5096 + }, + { + "epoch": 0.34865585881387234, + "grad_norm": 2.643393256194523, + "learning_rate": 7.565583105707981e-06, + "loss": 0.452, + "step": 5097 + }, + { + "epoch": 0.3487242629454819, + "grad_norm": 2.1147872085819928, + "learning_rate": 7.564632235857044e-06, + "loss": 0.4425, + "step": 5098 + }, + { + "epoch": 0.34879266707709145, + "grad_norm": 2.4444143211851923, + "learning_rate": 7.563681240121569e-06, + "loss": 0.8471, + "step": 5099 + }, + { + "epoch": 0.348861071208701, + "grad_norm": 2.308835576877038, + "learning_rate": 7.562730118548243e-06, + "loss": 0.5621, + "step": 5100 + }, + { + "epoch": 0.34892947534031055, + "grad_norm": 2.359371832662313, + "learning_rate": 7.561778871183745e-06, + "loss": 0.6838, + "step": 5101 + }, + { + "epoch": 0.3489978794719201, + "grad_norm": 2.063485415900501, + "learning_rate": 7.5608274980747686e-06, + "loss": 0.5562, + "step": 5102 + }, + { + "epoch": 0.34906628360352965, + "grad_norm": 3.049574629247441, + "learning_rate": 7.559875999268014e-06, + "loss": 0.4005, + "step": 5103 + }, + { + "epoch": 0.3491346877351392, + "grad_norm": 2.005594951134692, + "learning_rate": 7.558924374810184e-06, + "loss": 0.4886, + "step": 5104 + }, + { + "epoch": 0.34920309186674875, + "grad_norm": 2.3574126590273656, + "learning_rate": 7.557972624747989e-06, + "loss": 0.5979, + "step": 5105 + }, + { + "epoch": 0.34927149599835833, + "grad_norm": 3.054801655853415, + "learning_rate": 7.557020749128143e-06, + "loss": 0.2475, + "step": 5106 + }, + { + "epoch": 0.34933990012996785, + "grad_norm": 2.006898642346959, + "learning_rate": 7.5560687479973734e-06, + "loss": 0.5359, + "step": 5107 + }, + { + "epoch": 0.3494083042615774, + "grad_norm": 1.9005421718222935, + "learning_rate": 7.555116621402405e-06, + "loss": 0.4205, + "step": 5108 + }, + { + "epoch": 0.34947670839318695, + "grad_norm": 2.109600584895267, + "learning_rate": 7.554164369389976e-06, + "loss": 0.5858, + "step": 5109 + }, + { + "epoch": 0.3495451125247965, + "grad_norm": 1.774629058157745, + "learning_rate": 7.553211992006825e-06, + "loss": 0.4204, + "step": 5110 + }, + { + "epoch": 0.34961351665640605, + "grad_norm": 2.4446449282612623, + "learning_rate": 7.5522594892996985e-06, + "loss": 0.315, + "step": 5111 + }, + { + "epoch": 0.3496819207880156, + "grad_norm": 1.7350022947154686, + "learning_rate": 7.551306861315353e-06, + "loss": 0.3548, + "step": 5112 + }, + { + "epoch": 0.34975032491962516, + "grad_norm": 2.355493168996883, + "learning_rate": 7.550354108100546e-06, + "loss": 0.5037, + "step": 5113 + }, + { + "epoch": 0.3498187290512347, + "grad_norm": 1.9164325715413946, + "learning_rate": 7.549401229702045e-06, + "loss": 0.2462, + "step": 5114 + }, + { + "epoch": 0.34988713318284426, + "grad_norm": 1.8153042803824675, + "learning_rate": 7.5484482261666205e-06, + "loss": 0.2744, + "step": 5115 + }, + { + "epoch": 0.3499555373144538, + "grad_norm": 2.210276508262085, + "learning_rate": 7.547495097541049e-06, + "loss": 0.4664, + "step": 5116 + }, + { + "epoch": 0.35002394144606336, + "grad_norm": 1.7358645924280696, + "learning_rate": 7.546541843872119e-06, + "loss": 0.1949, + "step": 5117 + }, + { + "epoch": 0.3500923455776729, + "grad_norm": 1.9598098453355697, + "learning_rate": 7.545588465206614e-06, + "loss": 0.4737, + "step": 5118 + }, + { + "epoch": 0.35016074970928246, + "grad_norm": 2.420407167286246, + "learning_rate": 7.54463496159134e-06, + "loss": 0.5429, + "step": 5119 + }, + { + "epoch": 0.350229153840892, + "grad_norm": 2.3050969293574326, + "learning_rate": 7.5436813330730916e-06, + "loss": 0.486, + "step": 5120 + }, + { + "epoch": 0.35029755797250156, + "grad_norm": 3.469093013677682, + "learning_rate": 7.5427275796986785e-06, + "loss": 0.4475, + "step": 5121 + }, + { + "epoch": 0.3503659621041111, + "grad_norm": 1.824529078224943, + "learning_rate": 7.541773701514919e-06, + "loss": 0.299, + "step": 5122 + }, + { + "epoch": 0.35043436623572066, + "grad_norm": 2.5862683317955355, + "learning_rate": 7.540819698568632e-06, + "loss": 0.5611, + "step": 5123 + }, + { + "epoch": 0.3505027703673302, + "grad_norm": 1.7892804450336042, + "learning_rate": 7.5398655709066444e-06, + "loss": 0.4716, + "step": 5124 + }, + { + "epoch": 0.3505711744989397, + "grad_norm": 1.9819487547406176, + "learning_rate": 7.53891131857579e-06, + "loss": 0.3042, + "step": 5125 + }, + { + "epoch": 0.3506395786305493, + "grad_norm": 1.629384197125745, + "learning_rate": 7.537956941622907e-06, + "loss": 0.1983, + "step": 5126 + }, + { + "epoch": 0.3507079827621588, + "grad_norm": 1.5657015854079077, + "learning_rate": 7.537002440094843e-06, + "loss": 0.3707, + "step": 5127 + }, + { + "epoch": 0.3507763868937684, + "grad_norm": 1.3967676475629487, + "learning_rate": 7.536047814038446e-06, + "loss": 0.3253, + "step": 5128 + }, + { + "epoch": 0.3508447910253779, + "grad_norm": 1.6194473155602647, + "learning_rate": 7.535093063500578e-06, + "loss": 0.2964, + "step": 5129 + }, + { + "epoch": 0.3509131951569875, + "grad_norm": 2.2026910286805763, + "learning_rate": 7.534138188528099e-06, + "loss": 0.5866, + "step": 5130 + }, + { + "epoch": 0.350981599288597, + "grad_norm": 2.3772289520436702, + "learning_rate": 7.533183189167882e-06, + "loss": 0.7167, + "step": 5131 + }, + { + "epoch": 0.3510500034202066, + "grad_norm": 1.7297893507966957, + "learning_rate": 7.5322280654668e-06, + "loss": 0.3635, + "step": 5132 + }, + { + "epoch": 0.3511184075518161, + "grad_norm": 1.9032432888959434, + "learning_rate": 7.531272817471737e-06, + "loss": 0.3677, + "step": 5133 + }, + { + "epoch": 0.3511868116834257, + "grad_norm": 2.152859869967538, + "learning_rate": 7.5303174452295826e-06, + "loss": 0.7398, + "step": 5134 + }, + { + "epoch": 0.3512552158150352, + "grad_norm": 1.7305018634943172, + "learning_rate": 7.529361948787227e-06, + "loss": 0.3867, + "step": 5135 + }, + { + "epoch": 0.3513236199466448, + "grad_norm": 1.9368083626903194, + "learning_rate": 7.528406328191573e-06, + "loss": 0.5184, + "step": 5136 + }, + { + "epoch": 0.3513920240782543, + "grad_norm": 1.7792088123558907, + "learning_rate": 7.527450583489529e-06, + "loss": 0.1437, + "step": 5137 + }, + { + "epoch": 0.3514604282098639, + "grad_norm": 1.8838407787226605, + "learning_rate": 7.526494714728002e-06, + "loss": 0.2559, + "step": 5138 + }, + { + "epoch": 0.3515288323414734, + "grad_norm": 1.9921124388073805, + "learning_rate": 7.525538721953915e-06, + "loss": 0.3525, + "step": 5139 + }, + { + "epoch": 0.351597236473083, + "grad_norm": 1.7540706353582523, + "learning_rate": 7.5245826052141945e-06, + "loss": 0.359, + "step": 5140 + }, + { + "epoch": 0.3516656406046925, + "grad_norm": 2.513274368120186, + "learning_rate": 7.523626364555766e-06, + "loss": 0.6178, + "step": 5141 + }, + { + "epoch": 0.3517340447363021, + "grad_norm": 1.5197987235242938, + "learning_rate": 7.522670000025569e-06, + "loss": 0.2162, + "step": 5142 + }, + { + "epoch": 0.3518024488679116, + "grad_norm": 2.6156694251222126, + "learning_rate": 7.521713511670548e-06, + "loss": 0.7778, + "step": 5143 + }, + { + "epoch": 0.35187085299952114, + "grad_norm": 1.7499093134407198, + "learning_rate": 7.520756899537651e-06, + "loss": 0.3246, + "step": 5144 + }, + { + "epoch": 0.3519392571311307, + "grad_norm": 2.5779521074988456, + "learning_rate": 7.51980016367383e-06, + "loss": 0.6471, + "step": 5145 + }, + { + "epoch": 0.35200766126274025, + "grad_norm": 1.3044377676151233, + "learning_rate": 7.5188433041260515e-06, + "loss": 0.2691, + "step": 5146 + }, + { + "epoch": 0.3520760653943498, + "grad_norm": 1.9080908282531601, + "learning_rate": 7.517886320941279e-06, + "loss": 0.3979, + "step": 5147 + }, + { + "epoch": 0.35214446952595935, + "grad_norm": 1.3812367356643023, + "learning_rate": 7.516929214166487e-06, + "loss": 0.3304, + "step": 5148 + }, + { + "epoch": 0.3522128736575689, + "grad_norm": 2.2443824866814537, + "learning_rate": 7.515971983848657e-06, + "loss": 0.267, + "step": 5149 + }, + { + "epoch": 0.35228127778917845, + "grad_norm": 2.165773441819945, + "learning_rate": 7.515014630034771e-06, + "loss": 0.4644, + "step": 5150 + }, + { + "epoch": 0.352349681920788, + "grad_norm": 1.349114599222628, + "learning_rate": 7.514057152771822e-06, + "loss": 0.2096, + "step": 5151 + }, + { + "epoch": 0.35241808605239755, + "grad_norm": 1.5752982988646334, + "learning_rate": 7.513099552106808e-06, + "loss": 0.2807, + "step": 5152 + }, + { + "epoch": 0.35248649018400713, + "grad_norm": 2.1660490317215015, + "learning_rate": 7.512141828086732e-06, + "loss": 0.6052, + "step": 5153 + }, + { + "epoch": 0.35255489431561665, + "grad_norm": 1.680644583270573, + "learning_rate": 7.511183980758604e-06, + "loss": 0.3051, + "step": 5154 + }, + { + "epoch": 0.35262329844722623, + "grad_norm": 1.8209937871322144, + "learning_rate": 7.5102260101694394e-06, + "loss": 0.4879, + "step": 5155 + }, + { + "epoch": 0.35269170257883575, + "grad_norm": 2.0246688159870088, + "learning_rate": 7.509267916366262e-06, + "loss": 0.4061, + "step": 5156 + }, + { + "epoch": 0.35276010671044533, + "grad_norm": 2.623301819592253, + "learning_rate": 7.508309699396098e-06, + "loss": 0.8071, + "step": 5157 + }, + { + "epoch": 0.35282851084205485, + "grad_norm": 1.84545038501095, + "learning_rate": 7.50735135930598e-06, + "loss": 0.3312, + "step": 5158 + }, + { + "epoch": 0.35289691497366443, + "grad_norm": 1.6798700721345419, + "learning_rate": 7.506392896142951e-06, + "loss": 0.3647, + "step": 5159 + }, + { + "epoch": 0.35296531910527396, + "grad_norm": 1.8179665088844665, + "learning_rate": 7.505434309954054e-06, + "loss": 0.2306, + "step": 5160 + }, + { + "epoch": 0.35303372323688353, + "grad_norm": 2.3665422120302018, + "learning_rate": 7.5044756007863425e-06, + "loss": 0.6385, + "step": 5161 + }, + { + "epoch": 0.35310212736849306, + "grad_norm": 2.290883534453703, + "learning_rate": 7.503516768686876e-06, + "loss": 0.5036, + "step": 5162 + }, + { + "epoch": 0.3531705315001026, + "grad_norm": 2.536322503235693, + "learning_rate": 7.502557813702716e-06, + "loss": 0.5911, + "step": 5163 + }, + { + "epoch": 0.35323893563171216, + "grad_norm": 1.6215877561574406, + "learning_rate": 7.501598735880934e-06, + "loss": 0.3302, + "step": 5164 + }, + { + "epoch": 0.3533073397633217, + "grad_norm": 2.010311701279441, + "learning_rate": 7.500639535268606e-06, + "loss": 0.3717, + "step": 5165 + }, + { + "epoch": 0.35337574389493126, + "grad_norm": 2.2091934887624674, + "learning_rate": 7.499680211912814e-06, + "loss": 0.3538, + "step": 5166 + }, + { + "epoch": 0.3534441480265408, + "grad_norm": 1.714283880022748, + "learning_rate": 7.498720765860646e-06, + "loss": 0.2903, + "step": 5167 + }, + { + "epoch": 0.35351255215815036, + "grad_norm": 1.479015075752244, + "learning_rate": 7.497761197159197e-06, + "loss": 0.1082, + "step": 5168 + }, + { + "epoch": 0.3535809562897599, + "grad_norm": 2.165612061917408, + "learning_rate": 7.4968015058555674e-06, + "loss": 0.4426, + "step": 5169 + }, + { + "epoch": 0.35364936042136946, + "grad_norm": 2.209031540479283, + "learning_rate": 7.495841691996862e-06, + "loss": 0.3228, + "step": 5170 + }, + { + "epoch": 0.353717764552979, + "grad_norm": 2.105307415388008, + "learning_rate": 7.494881755630194e-06, + "loss": 0.6662, + "step": 5171 + }, + { + "epoch": 0.35378616868458856, + "grad_norm": 2.11224464616877, + "learning_rate": 7.493921696802684e-06, + "loss": 0.2448, + "step": 5172 + }, + { + "epoch": 0.3538545728161981, + "grad_norm": 2.0942076745068054, + "learning_rate": 7.492961515561452e-06, + "loss": 0.5907, + "step": 5173 + }, + { + "epoch": 0.35392297694780767, + "grad_norm": 2.01134467489487, + "learning_rate": 7.492001211953631e-06, + "loss": 0.3403, + "step": 5174 + }, + { + "epoch": 0.3539913810794172, + "grad_norm": 2.480588180189923, + "learning_rate": 7.491040786026356e-06, + "loss": 0.6653, + "step": 5175 + }, + { + "epoch": 0.35405978521102677, + "grad_norm": 2.859709225074868, + "learning_rate": 7.490080237826771e-06, + "loss": 0.3995, + "step": 5176 + }, + { + "epoch": 0.3541281893426363, + "grad_norm": 3.5845146348223214, + "learning_rate": 7.489119567402023e-06, + "loss": 0.4734, + "step": 5177 + }, + { + "epoch": 0.35419659347424587, + "grad_norm": 2.1260490072205567, + "learning_rate": 7.488158774799268e-06, + "loss": 0.135, + "step": 5178 + }, + { + "epoch": 0.3542649976058554, + "grad_norm": 2.325548221317216, + "learning_rate": 7.487197860065664e-06, + "loss": 0.2967, + "step": 5179 + }, + { + "epoch": 0.35433340173746497, + "grad_norm": 2.0913718795374603, + "learning_rate": 7.486236823248379e-06, + "loss": 0.5903, + "step": 5180 + }, + { + "epoch": 0.3544018058690745, + "grad_norm": 2.1859178713993628, + "learning_rate": 7.4852756643945835e-06, + "loss": 0.5788, + "step": 5181 + }, + { + "epoch": 0.354470210000684, + "grad_norm": 2.5318299500396733, + "learning_rate": 7.484314383551459e-06, + "loss": 0.7276, + "step": 5182 + }, + { + "epoch": 0.3545386141322936, + "grad_norm": 1.2373685090210458, + "learning_rate": 7.4833529807661865e-06, + "loss": 0.1781, + "step": 5183 + }, + { + "epoch": 0.3546070182639031, + "grad_norm": 1.552917316144243, + "learning_rate": 7.482391456085958e-06, + "loss": 0.341, + "step": 5184 + }, + { + "epoch": 0.3546754223955127, + "grad_norm": 1.2447451832370782, + "learning_rate": 7.481429809557969e-06, + "loss": 0.1433, + "step": 5185 + }, + { + "epoch": 0.3547438265271222, + "grad_norm": 1.783563654742188, + "learning_rate": 7.4804680412294225e-06, + "loss": 0.354, + "step": 5186 + }, + { + "epoch": 0.3548122306587318, + "grad_norm": 2.3057111051058987, + "learning_rate": 7.479506151147529e-06, + "loss": 0.5223, + "step": 5187 + }, + { + "epoch": 0.3548806347903413, + "grad_norm": 2.0891261214481127, + "learning_rate": 7.478544139359496e-06, + "loss": 0.4233, + "step": 5188 + }, + { + "epoch": 0.3549490389219509, + "grad_norm": 2.584012672229008, + "learning_rate": 7.477582005912552e-06, + "loss": 0.6134, + "step": 5189 + }, + { + "epoch": 0.3550174430535604, + "grad_norm": 1.952542683578084, + "learning_rate": 7.4766197508539155e-06, + "loss": 0.1475, + "step": 5190 + }, + { + "epoch": 0.35508584718517, + "grad_norm": 1.8629392473974944, + "learning_rate": 7.4756573742308224e-06, + "loss": 0.4219, + "step": 5191 + }, + { + "epoch": 0.3551542513167795, + "grad_norm": 2.8869874284600696, + "learning_rate": 7.474694876090513e-06, + "loss": 0.676, + "step": 5192 + }, + { + "epoch": 0.3552226554483891, + "grad_norm": 1.1710645766184093, + "learning_rate": 7.473732256480228e-06, + "loss": 0.1867, + "step": 5193 + }, + { + "epoch": 0.3552910595799986, + "grad_norm": 2.1817033258373857, + "learning_rate": 7.472769515447218e-06, + "loss": 0.3512, + "step": 5194 + }, + { + "epoch": 0.3553594637116082, + "grad_norm": 1.8581012482188453, + "learning_rate": 7.471806653038738e-06, + "loss": 0.4641, + "step": 5195 + }, + { + "epoch": 0.3554278678432177, + "grad_norm": 2.457274935817534, + "learning_rate": 7.470843669302053e-06, + "loss": 0.6205, + "step": 5196 + }, + { + "epoch": 0.3554962719748273, + "grad_norm": 1.8022865615389967, + "learning_rate": 7.46988056428443e-06, + "loss": 0.3783, + "step": 5197 + }, + { + "epoch": 0.3555646761064368, + "grad_norm": 2.4801266051285737, + "learning_rate": 7.468917338033139e-06, + "loss": 0.7068, + "step": 5198 + }, + { + "epoch": 0.35563308023804635, + "grad_norm": 1.6910105500086718, + "learning_rate": 7.4679539905954655e-06, + "loss": 0.3566, + "step": 5199 + }, + { + "epoch": 0.35570148436965593, + "grad_norm": 2.516296356369205, + "learning_rate": 7.4669905220186914e-06, + "loss": 0.427, + "step": 5200 + }, + { + "epoch": 0.35576988850126545, + "grad_norm": 2.593753722774178, + "learning_rate": 7.466026932350109e-06, + "loss": 0.4353, + "step": 5201 + }, + { + "epoch": 0.35583829263287503, + "grad_norm": 1.772123728220052, + "learning_rate": 7.465063221637016e-06, + "loss": 0.3362, + "step": 5202 + }, + { + "epoch": 0.35590669676448455, + "grad_norm": 2.0498461671424315, + "learning_rate": 7.464099389926719e-06, + "loss": 0.3339, + "step": 5203 + }, + { + "epoch": 0.35597510089609413, + "grad_norm": 1.3861133032795554, + "learning_rate": 7.463135437266522e-06, + "loss": 0.1657, + "step": 5204 + }, + { + "epoch": 0.35604350502770366, + "grad_norm": 2.2683964976719753, + "learning_rate": 7.462171363703745e-06, + "loss": 0.4916, + "step": 5205 + }, + { + "epoch": 0.35611190915931323, + "grad_norm": 1.7770401160816367, + "learning_rate": 7.461207169285708e-06, + "loss": 0.4055, + "step": 5206 + }, + { + "epoch": 0.35618031329092276, + "grad_norm": 1.9663906681805938, + "learning_rate": 7.460242854059737e-06, + "loss": 0.4324, + "step": 5207 + }, + { + "epoch": 0.35624871742253233, + "grad_norm": 1.9503926505397835, + "learning_rate": 7.459278418073167e-06, + "loss": 0.3181, + "step": 5208 + }, + { + "epoch": 0.35631712155414186, + "grad_norm": 1.652734994223255, + "learning_rate": 7.458313861373336e-06, + "loss": 0.3037, + "step": 5209 + }, + { + "epoch": 0.35638552568575144, + "grad_norm": 2.0989076337694046, + "learning_rate": 7.4573491840075905e-06, + "loss": 0.6398, + "step": 5210 + }, + { + "epoch": 0.35645392981736096, + "grad_norm": 4.750379697782253, + "learning_rate": 7.45638438602328e-06, + "loss": 0.486, + "step": 5211 + }, + { + "epoch": 0.35652233394897054, + "grad_norm": 2.484452036496931, + "learning_rate": 7.4554194674677615e-06, + "loss": 0.6208, + "step": 5212 + }, + { + "epoch": 0.35659073808058006, + "grad_norm": 1.3763468225931292, + "learning_rate": 7.4544544283884e-06, + "loss": 0.2803, + "step": 5213 + }, + { + "epoch": 0.35665914221218964, + "grad_norm": 2.075444166633361, + "learning_rate": 7.453489268832561e-06, + "loss": 0.4624, + "step": 5214 + }, + { + "epoch": 0.35672754634379916, + "grad_norm": 1.6951558352742482, + "learning_rate": 7.452523988847622e-06, + "loss": 0.3936, + "step": 5215 + }, + { + "epoch": 0.35679595047540874, + "grad_norm": 2.3700154946369976, + "learning_rate": 7.451558588480962e-06, + "loss": 0.2128, + "step": 5216 + }, + { + "epoch": 0.35686435460701826, + "grad_norm": 2.012062572985811, + "learning_rate": 7.450593067779969e-06, + "loss": 0.5494, + "step": 5217 + }, + { + "epoch": 0.3569327587386278, + "grad_norm": 2.2394649444752575, + "learning_rate": 7.449627426792033e-06, + "loss": 0.5017, + "step": 5218 + }, + { + "epoch": 0.35700116287023737, + "grad_norm": 2.077028114123062, + "learning_rate": 7.4486616655645565e-06, + "loss": 0.5862, + "step": 5219 + }, + { + "epoch": 0.3570695670018469, + "grad_norm": 2.0009805950163257, + "learning_rate": 7.447695784144937e-06, + "loss": 0.4561, + "step": 5220 + }, + { + "epoch": 0.35713797113345647, + "grad_norm": 1.707556925127725, + "learning_rate": 7.44672978258059e-06, + "loss": 0.4574, + "step": 5221 + }, + { + "epoch": 0.357206375265066, + "grad_norm": 1.9538075284487368, + "learning_rate": 7.445763660918932e-06, + "loss": 0.5072, + "step": 5222 + }, + { + "epoch": 0.35727477939667557, + "grad_norm": 1.8139634102858042, + "learning_rate": 7.444797419207381e-06, + "loss": 0.4539, + "step": 5223 + }, + { + "epoch": 0.3573431835282851, + "grad_norm": 1.789570632137513, + "learning_rate": 7.443831057493368e-06, + "loss": 0.4696, + "step": 5224 + }, + { + "epoch": 0.35741158765989467, + "grad_norm": 1.7906159648700606, + "learning_rate": 7.442864575824324e-06, + "loss": 0.2436, + "step": 5225 + }, + { + "epoch": 0.3574799917915042, + "grad_norm": 1.5713981391406868, + "learning_rate": 7.441897974247692e-06, + "loss": 0.2178, + "step": 5226 + }, + { + "epoch": 0.35754839592311377, + "grad_norm": 2.2403179808283675, + "learning_rate": 7.440931252810914e-06, + "loss": 0.6332, + "step": 5227 + }, + { + "epoch": 0.3576168000547233, + "grad_norm": 2.0124836958052974, + "learning_rate": 7.439964411561445e-06, + "loss": 0.6644, + "step": 5228 + }, + { + "epoch": 0.3576852041863329, + "grad_norm": 1.3590193973950602, + "learning_rate": 7.4389974505467396e-06, + "loss": 0.1775, + "step": 5229 + }, + { + "epoch": 0.3577536083179424, + "grad_norm": 1.8665188146652532, + "learning_rate": 7.4380303698142605e-06, + "loss": 0.3281, + "step": 5230 + }, + { + "epoch": 0.357822012449552, + "grad_norm": 1.9960083915241642, + "learning_rate": 7.4370631694114785e-06, + "loss": 0.5037, + "step": 5231 + }, + { + "epoch": 0.3578904165811615, + "grad_norm": 1.5020666439180343, + "learning_rate": 7.436095849385867e-06, + "loss": 0.2707, + "step": 5232 + }, + { + "epoch": 0.3579588207127711, + "grad_norm": 2.3065826964311382, + "learning_rate": 7.435128409784908e-06, + "loss": 0.7527, + "step": 5233 + }, + { + "epoch": 0.3580272248443806, + "grad_norm": 1.394630343124796, + "learning_rate": 7.434160850656088e-06, + "loss": 0.3085, + "step": 5234 + }, + { + "epoch": 0.3580956289759902, + "grad_norm": 1.9384967370508837, + "learning_rate": 7.433193172046897e-06, + "loss": 0.5509, + "step": 5235 + }, + { + "epoch": 0.3581640331075997, + "grad_norm": 1.4913857987661723, + "learning_rate": 7.432225374004838e-06, + "loss": 0.2937, + "step": 5236 + }, + { + "epoch": 0.3582324372392092, + "grad_norm": 2.0169800151547417, + "learning_rate": 7.431257456577411e-06, + "loss": 0.4593, + "step": 5237 + }, + { + "epoch": 0.3583008413708188, + "grad_norm": 2.28436547967516, + "learning_rate": 7.430289419812128e-06, + "loss": 0.5818, + "step": 5238 + }, + { + "epoch": 0.3583692455024283, + "grad_norm": 1.6149950945745735, + "learning_rate": 7.4293212637565045e-06, + "loss": 0.2529, + "step": 5239 + }, + { + "epoch": 0.3584376496340379, + "grad_norm": 1.875273843477128, + "learning_rate": 7.428352988458063e-06, + "loss": 0.3642, + "step": 5240 + }, + { + "epoch": 0.3585060537656474, + "grad_norm": 1.8453075651123834, + "learning_rate": 7.427384593964329e-06, + "loss": 0.4383, + "step": 5241 + }, + { + "epoch": 0.358574457897257, + "grad_norm": 2.270325493824603, + "learning_rate": 7.426416080322838e-06, + "loss": 0.1768, + "step": 5242 + }, + { + "epoch": 0.3586428620288665, + "grad_norm": 2.02369504563573, + "learning_rate": 7.425447447581129e-06, + "loss": 0.4255, + "step": 5243 + }, + { + "epoch": 0.3587112661604761, + "grad_norm": 1.7119903025213221, + "learning_rate": 7.424478695786746e-06, + "loss": 0.1846, + "step": 5244 + }, + { + "epoch": 0.35877967029208563, + "grad_norm": 1.894750466396009, + "learning_rate": 7.423509824987242e-06, + "loss": 0.3586, + "step": 5245 + }, + { + "epoch": 0.3588480744236952, + "grad_norm": 2.2721294887578343, + "learning_rate": 7.422540835230173e-06, + "loss": 0.6321, + "step": 5246 + }, + { + "epoch": 0.35891647855530473, + "grad_norm": 1.717857937797397, + "learning_rate": 7.4215717265631e-06, + "loss": 0.2721, + "step": 5247 + }, + { + "epoch": 0.3589848826869143, + "grad_norm": 1.8782976358382977, + "learning_rate": 7.420602499033595e-06, + "loss": 0.2862, + "step": 5248 + }, + { + "epoch": 0.35905328681852383, + "grad_norm": 1.456163961300093, + "learning_rate": 7.41963315268923e-06, + "loss": 0.3564, + "step": 5249 + }, + { + "epoch": 0.3591216909501334, + "grad_norm": 2.706812862157781, + "learning_rate": 7.418663687577583e-06, + "loss": 0.7278, + "step": 5250 + }, + { + "epoch": 0.35919009508174293, + "grad_norm": 1.8470453619444092, + "learning_rate": 7.417694103746245e-06, + "loss": 0.5529, + "step": 5251 + }, + { + "epoch": 0.3592584992133525, + "grad_norm": 2.1987093848857437, + "learning_rate": 7.416724401242805e-06, + "loss": 0.5866, + "step": 5252 + }, + { + "epoch": 0.35932690334496203, + "grad_norm": 1.3693945046641502, + "learning_rate": 7.41575458011486e-06, + "loss": 0.1614, + "step": 5253 + }, + { + "epoch": 0.35939530747657156, + "grad_norm": 1.799741262642028, + "learning_rate": 7.4147846404100155e-06, + "loss": 0.3571, + "step": 5254 + }, + { + "epoch": 0.35946371160818114, + "grad_norm": 2.204524380376258, + "learning_rate": 7.413814582175881e-06, + "loss": 0.5625, + "step": 5255 + }, + { + "epoch": 0.35953211573979066, + "grad_norm": 2.259349438481811, + "learning_rate": 7.41284440546007e-06, + "loss": 0.6693, + "step": 5256 + }, + { + "epoch": 0.35960051987140024, + "grad_norm": 2.4534665981496637, + "learning_rate": 7.411874110310204e-06, + "loss": 0.277, + "step": 5257 + }, + { + "epoch": 0.35966892400300976, + "grad_norm": 1.381533316645062, + "learning_rate": 7.41090369677391e-06, + "loss": 0.1867, + "step": 5258 + }, + { + "epoch": 0.35973732813461934, + "grad_norm": 1.6650686222260407, + "learning_rate": 7.409933164898819e-06, + "loss": 0.345, + "step": 5259 + }, + { + "epoch": 0.35980573226622886, + "grad_norm": 1.9129267845919322, + "learning_rate": 7.408962514732572e-06, + "loss": 0.4013, + "step": 5260 + }, + { + "epoch": 0.35987413639783844, + "grad_norm": 2.463981417118323, + "learning_rate": 7.407991746322812e-06, + "loss": 0.7545, + "step": 5261 + }, + { + "epoch": 0.35994254052944796, + "grad_norm": 2.092586293577537, + "learning_rate": 7.40702085971719e-06, + "loss": 0.1791, + "step": 5262 + }, + { + "epoch": 0.36001094466105754, + "grad_norm": 2.397041883962875, + "learning_rate": 7.40604985496336e-06, + "loss": 0.5625, + "step": 5263 + }, + { + "epoch": 0.36007934879266706, + "grad_norm": 1.9470735682872482, + "learning_rate": 7.4050787321089835e-06, + "loss": 0.4946, + "step": 5264 + }, + { + "epoch": 0.36014775292427664, + "grad_norm": 2.3334914832220286, + "learning_rate": 7.4041074912017295e-06, + "loss": 0.4947, + "step": 5265 + }, + { + "epoch": 0.36021615705588617, + "grad_norm": 2.1501672162689105, + "learning_rate": 7.403136132289272e-06, + "loss": 0.4425, + "step": 5266 + }, + { + "epoch": 0.36028456118749574, + "grad_norm": 2.2578624799007785, + "learning_rate": 7.402164655419286e-06, + "loss": 0.6273, + "step": 5267 + }, + { + "epoch": 0.36035296531910527, + "grad_norm": 1.8726544809472327, + "learning_rate": 7.40119306063946e-06, + "loss": 0.3312, + "step": 5268 + }, + { + "epoch": 0.36042136945071485, + "grad_norm": 1.988395103087733, + "learning_rate": 7.4002213479974824e-06, + "loss": 0.5123, + "step": 5269 + }, + { + "epoch": 0.36048977358232437, + "grad_norm": 1.3284963782771786, + "learning_rate": 7.399249517541052e-06, + "loss": 0.1415, + "step": 5270 + }, + { + "epoch": 0.36055817771393395, + "grad_norm": 1.7678101613874038, + "learning_rate": 7.398277569317869e-06, + "loss": 0.2695, + "step": 5271 + }, + { + "epoch": 0.36062658184554347, + "grad_norm": 1.9495673011332018, + "learning_rate": 7.397305503375641e-06, + "loss": 0.3859, + "step": 5272 + }, + { + "epoch": 0.360694985977153, + "grad_norm": 1.9188931135304625, + "learning_rate": 7.396333319762083e-06, + "loss": 0.3778, + "step": 5273 + }, + { + "epoch": 0.36076339010876257, + "grad_norm": 1.5548180424693854, + "learning_rate": 7.395361018524913e-06, + "loss": 0.1954, + "step": 5274 + }, + { + "epoch": 0.3608317942403721, + "grad_norm": 2.532529195700079, + "learning_rate": 7.394388599711859e-06, + "loss": 0.4837, + "step": 5275 + }, + { + "epoch": 0.3609001983719817, + "grad_norm": 1.6146657604793937, + "learning_rate": 7.393416063370649e-06, + "loss": 0.1965, + "step": 5276 + }, + { + "epoch": 0.3609686025035912, + "grad_norm": 1.5863965443611352, + "learning_rate": 7.3924434095490205e-06, + "loss": 0.1924, + "step": 5277 + }, + { + "epoch": 0.3610370066352008, + "grad_norm": 2.2187684152403078, + "learning_rate": 7.391470638294717e-06, + "loss": 0.5398, + "step": 5278 + }, + { + "epoch": 0.3611054107668103, + "grad_norm": 2.4434255321154463, + "learning_rate": 7.390497749655487e-06, + "loss": 0.4928, + "step": 5279 + }, + { + "epoch": 0.3611738148984199, + "grad_norm": 1.4453425636257768, + "learning_rate": 7.389524743679082e-06, + "loss": 0.2721, + "step": 5280 + }, + { + "epoch": 0.3612422190300294, + "grad_norm": 2.011007661047522, + "learning_rate": 7.388551620413265e-06, + "loss": 0.4384, + "step": 5281 + }, + { + "epoch": 0.361310623161639, + "grad_norm": 1.8919815682225114, + "learning_rate": 7.3875783799058e-06, + "loss": 0.2125, + "step": 5282 + }, + { + "epoch": 0.3613790272932485, + "grad_norm": 1.9538076886100464, + "learning_rate": 7.386605022204458e-06, + "loss": 0.4352, + "step": 5283 + }, + { + "epoch": 0.3614474314248581, + "grad_norm": 2.392653682839145, + "learning_rate": 7.385631547357017e-06, + "loss": 0.7945, + "step": 5284 + }, + { + "epoch": 0.3615158355564676, + "grad_norm": 1.861802151853277, + "learning_rate": 7.384657955411261e-06, + "loss": 0.429, + "step": 5285 + }, + { + "epoch": 0.3615842396880772, + "grad_norm": 1.523423150801939, + "learning_rate": 7.383684246414976e-06, + "loss": 0.1553, + "step": 5286 + }, + { + "epoch": 0.3616526438196867, + "grad_norm": 1.7757643378711534, + "learning_rate": 7.382710420415957e-06, + "loss": 0.3927, + "step": 5287 + }, + { + "epoch": 0.3617210479512963, + "grad_norm": 2.116958802881442, + "learning_rate": 7.381736477462006e-06, + "loss": 0.4238, + "step": 5288 + }, + { + "epoch": 0.3617894520829058, + "grad_norm": 2.01761243113389, + "learning_rate": 7.380762417600926e-06, + "loss": 0.3705, + "step": 5289 + }, + { + "epoch": 0.3618578562145154, + "grad_norm": 1.578365130614861, + "learning_rate": 7.37978824088053e-06, + "loss": 0.2552, + "step": 5290 + }, + { + "epoch": 0.3619262603461249, + "grad_norm": 2.1873490553957167, + "learning_rate": 7.3788139473486365e-06, + "loss": 0.589, + "step": 5291 + }, + { + "epoch": 0.36199466447773443, + "grad_norm": 1.676183125518828, + "learning_rate": 7.377839537053066e-06, + "loss": 0.2379, + "step": 5292 + }, + { + "epoch": 0.362063068609344, + "grad_norm": 2.277360265027422, + "learning_rate": 7.37686501004165e-06, + "loss": 0.7862, + "step": 5293 + }, + { + "epoch": 0.36213147274095353, + "grad_norm": 1.3391368825277217, + "learning_rate": 7.375890366362219e-06, + "loss": 0.3065, + "step": 5294 + }, + { + "epoch": 0.3621998768725631, + "grad_norm": 2.0753876217763474, + "learning_rate": 7.374915606062618e-06, + "loss": 0.2975, + "step": 5295 + }, + { + "epoch": 0.36226828100417263, + "grad_norm": 2.0967522028059444, + "learning_rate": 7.3739407291906905e-06, + "loss": 0.5206, + "step": 5296 + }, + { + "epoch": 0.3623366851357822, + "grad_norm": 2.7289706419417143, + "learning_rate": 7.3729657357942885e-06, + "loss": 0.6393, + "step": 5297 + }, + { + "epoch": 0.36240508926739173, + "grad_norm": 2.01407791987066, + "learning_rate": 7.371990625921268e-06, + "loss": 0.3157, + "step": 5298 + }, + { + "epoch": 0.3624734933990013, + "grad_norm": 2.1798012406401037, + "learning_rate": 7.371015399619494e-06, + "loss": 0.6065, + "step": 5299 + }, + { + "epoch": 0.36254189753061083, + "grad_norm": 1.3713750387671082, + "learning_rate": 7.370040056936835e-06, + "loss": 0.2071, + "step": 5300 + }, + { + "epoch": 0.3626103016622204, + "grad_norm": 2.215306154721741, + "learning_rate": 7.369064597921165e-06, + "loss": 0.455, + "step": 5301 + }, + { + "epoch": 0.36267870579382994, + "grad_norm": 1.7710575401262818, + "learning_rate": 7.368089022620365e-06, + "loss": 0.3266, + "step": 5302 + }, + { + "epoch": 0.3627471099254395, + "grad_norm": 2.672449947982214, + "learning_rate": 7.367113331082318e-06, + "loss": 0.2927, + "step": 5303 + }, + { + "epoch": 0.36281551405704904, + "grad_norm": 1.9161961632772284, + "learning_rate": 7.366137523354921e-06, + "loss": 0.3469, + "step": 5304 + }, + { + "epoch": 0.3628839181886586, + "grad_norm": 1.534721016763508, + "learning_rate": 7.365161599486067e-06, + "loss": 0.2677, + "step": 5305 + }, + { + "epoch": 0.36295232232026814, + "grad_norm": 1.581992911827946, + "learning_rate": 7.36418555952366e-06, + "loss": 0.287, + "step": 5306 + }, + { + "epoch": 0.3630207264518777, + "grad_norm": 2.5135408321389243, + "learning_rate": 7.36320940351561e-06, + "loss": 0.4105, + "step": 5307 + }, + { + "epoch": 0.36308913058348724, + "grad_norm": 1.9042190535856478, + "learning_rate": 7.36223313150983e-06, + "loss": 0.3918, + "step": 5308 + }, + { + "epoch": 0.3631575347150968, + "grad_norm": 1.8701898585609555, + "learning_rate": 7.361256743554241e-06, + "loss": 0.3094, + "step": 5309 + }, + { + "epoch": 0.36322593884670634, + "grad_norm": 1.8182718520197385, + "learning_rate": 7.360280239696769e-06, + "loss": 0.3991, + "step": 5310 + }, + { + "epoch": 0.36329434297831587, + "grad_norm": 2.435204825285509, + "learning_rate": 7.359303619985346e-06, + "loss": 0.5226, + "step": 5311 + }, + { + "epoch": 0.36336274710992544, + "grad_norm": 1.9078206798681159, + "learning_rate": 7.358326884467908e-06, + "loss": 0.2335, + "step": 5312 + }, + { + "epoch": 0.36343115124153497, + "grad_norm": 2.070988886483332, + "learning_rate": 7.357350033192397e-06, + "loss": 0.6299, + "step": 5313 + }, + { + "epoch": 0.36349955537314455, + "grad_norm": 1.9819487960920328, + "learning_rate": 7.356373066206763e-06, + "loss": 0.5801, + "step": 5314 + }, + { + "epoch": 0.36356795950475407, + "grad_norm": 1.8709875805385008, + "learning_rate": 7.3553959835589615e-06, + "loss": 0.3621, + "step": 5315 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 1.4972814322736174, + "learning_rate": 7.3544187852969495e-06, + "loss": 0.2372, + "step": 5316 + }, + { + "epoch": 0.36370476776797317, + "grad_norm": 1.5824171231475577, + "learning_rate": 7.3534414714686955e-06, + "loss": 0.2944, + "step": 5317 + }, + { + "epoch": 0.36377317189958275, + "grad_norm": 1.7891523513552914, + "learning_rate": 7.352464042122169e-06, + "loss": 0.3309, + "step": 5318 + }, + { + "epoch": 0.36384157603119227, + "grad_norm": 1.5960584966171132, + "learning_rate": 7.351486497305347e-06, + "loss": 0.2232, + "step": 5319 + }, + { + "epoch": 0.36390998016280185, + "grad_norm": 2.1027619013011285, + "learning_rate": 7.3505088370662125e-06, + "loss": 0.2286, + "step": 5320 + }, + { + "epoch": 0.3639783842944114, + "grad_norm": 1.7992455229131734, + "learning_rate": 7.349531061452754e-06, + "loss": 0.182, + "step": 5321 + }, + { + "epoch": 0.36404678842602095, + "grad_norm": 1.9538469582415368, + "learning_rate": 7.348553170512965e-06, + "loss": 0.5436, + "step": 5322 + }, + { + "epoch": 0.3641151925576305, + "grad_norm": 1.933019167367687, + "learning_rate": 7.347575164294845e-06, + "loss": 0.4407, + "step": 5323 + }, + { + "epoch": 0.36418359668924005, + "grad_norm": 1.9119926393623605, + "learning_rate": 7.3465970428463995e-06, + "loss": 0.3673, + "step": 5324 + }, + { + "epoch": 0.3642520008208496, + "grad_norm": 1.967696566535605, + "learning_rate": 7.3456188062156405e-06, + "loss": 0.388, + "step": 5325 + }, + { + "epoch": 0.36432040495245915, + "grad_norm": 1.3889055659034633, + "learning_rate": 7.344640454450584e-06, + "loss": 0.2292, + "step": 5326 + }, + { + "epoch": 0.3643888090840687, + "grad_norm": 1.4330760790331156, + "learning_rate": 7.343661987599251e-06, + "loss": 0.2184, + "step": 5327 + }, + { + "epoch": 0.3644572132156782, + "grad_norm": 2.040915097037891, + "learning_rate": 7.342683405709671e-06, + "loss": 0.6181, + "step": 5328 + }, + { + "epoch": 0.3645256173472878, + "grad_norm": 1.632674439539105, + "learning_rate": 7.3417047088298755e-06, + "loss": 0.2994, + "step": 5329 + }, + { + "epoch": 0.3645940214788973, + "grad_norm": 2.1262609548303337, + "learning_rate": 7.3407258970079055e-06, + "loss": 0.2141, + "step": 5330 + }, + { + "epoch": 0.3646624256105069, + "grad_norm": 1.814504041665675, + "learning_rate": 7.3397469702918055e-06, + "loss": 0.4637, + "step": 5331 + }, + { + "epoch": 0.3647308297421164, + "grad_norm": 2.0441242181639643, + "learning_rate": 7.338767928729625e-06, + "loss": 0.2578, + "step": 5332 + }, + { + "epoch": 0.364799233873726, + "grad_norm": 2.0000612884682436, + "learning_rate": 7.3377887723694205e-06, + "loss": 0.3306, + "step": 5333 + }, + { + "epoch": 0.3648676380053355, + "grad_norm": 1.7924774676998978, + "learning_rate": 7.336809501259254e-06, + "loss": 0.3181, + "step": 5334 + }, + { + "epoch": 0.3649360421369451, + "grad_norm": 1.3244490901386252, + "learning_rate": 7.335830115447195e-06, + "loss": 0.181, + "step": 5335 + }, + { + "epoch": 0.3650044462685546, + "grad_norm": 2.3307561087871473, + "learning_rate": 7.334850614981313e-06, + "loss": 0.5068, + "step": 5336 + }, + { + "epoch": 0.3650728504001642, + "grad_norm": 2.124233545895081, + "learning_rate": 7.333870999909687e-06, + "loss": 0.6245, + "step": 5337 + }, + { + "epoch": 0.3651412545317737, + "grad_norm": 1.5513775589466698, + "learning_rate": 7.332891270280403e-06, + "loss": 0.3154, + "step": 5338 + }, + { + "epoch": 0.3652096586633833, + "grad_norm": 1.324141013415304, + "learning_rate": 7.33191142614155e-06, + "loss": 0.2007, + "step": 5339 + }, + { + "epoch": 0.3652780627949928, + "grad_norm": 1.6872931459590506, + "learning_rate": 7.330931467541222e-06, + "loss": 0.2838, + "step": 5340 + }, + { + "epoch": 0.3653464669266024, + "grad_norm": 2.1705736268211506, + "learning_rate": 7.329951394527524e-06, + "loss": 0.3001, + "step": 5341 + }, + { + "epoch": 0.3654148710582119, + "grad_norm": 2.384323579026257, + "learning_rate": 7.328971207148561e-06, + "loss": 0.4999, + "step": 5342 + }, + { + "epoch": 0.3654832751898215, + "grad_norm": 1.185866141654008, + "learning_rate": 7.327990905452442e-06, + "loss": 0.2182, + "step": 5343 + }, + { + "epoch": 0.365551679321431, + "grad_norm": 1.8291383078693286, + "learning_rate": 7.3270104894872885e-06, + "loss": 0.2302, + "step": 5344 + }, + { + "epoch": 0.3656200834530406, + "grad_norm": 3.464467024939259, + "learning_rate": 7.326029959301224e-06, + "loss": 0.2463, + "step": 5345 + }, + { + "epoch": 0.3656884875846501, + "grad_norm": 2.3337426607111564, + "learning_rate": 7.325049314942375e-06, + "loss": 0.4668, + "step": 5346 + }, + { + "epoch": 0.36575689171625964, + "grad_norm": 1.3961227749121914, + "learning_rate": 7.32406855645888e-06, + "loss": 0.1373, + "step": 5347 + }, + { + "epoch": 0.3658252958478692, + "grad_norm": 1.8382780982110873, + "learning_rate": 7.323087683898877e-06, + "loss": 0.5463, + "step": 5348 + }, + { + "epoch": 0.36589369997947874, + "grad_norm": 2.3499868443303455, + "learning_rate": 7.322106697310512e-06, + "loss": 0.4443, + "step": 5349 + }, + { + "epoch": 0.3659621041110883, + "grad_norm": 1.656908453204673, + "learning_rate": 7.321125596741936e-06, + "loss": 0.4122, + "step": 5350 + }, + { + "epoch": 0.36603050824269784, + "grad_norm": 2.0282550814134837, + "learning_rate": 7.3201443822413085e-06, + "loss": 0.4145, + "step": 5351 + }, + { + "epoch": 0.3660989123743074, + "grad_norm": 2.2890002141296635, + "learning_rate": 7.319163053856791e-06, + "loss": 0.7607, + "step": 5352 + }, + { + "epoch": 0.36616731650591694, + "grad_norm": 1.9417312549591446, + "learning_rate": 7.31818161163655e-06, + "loss": 0.4702, + "step": 5353 + }, + { + "epoch": 0.3662357206375265, + "grad_norm": 2.1697689040246497, + "learning_rate": 7.3172000556287635e-06, + "loss": 0.4685, + "step": 5354 + }, + { + "epoch": 0.36630412476913604, + "grad_norm": 2.069485862813138, + "learning_rate": 7.316218385881607e-06, + "loss": 0.4843, + "step": 5355 + }, + { + "epoch": 0.3663725289007456, + "grad_norm": 1.8973067805582289, + "learning_rate": 7.315236602443268e-06, + "loss": 0.1623, + "step": 5356 + }, + { + "epoch": 0.36644093303235514, + "grad_norm": 1.7201694847448128, + "learning_rate": 7.314254705361936e-06, + "loss": 0.4107, + "step": 5357 + }, + { + "epoch": 0.3665093371639647, + "grad_norm": 1.0567661464688125, + "learning_rate": 7.313272694685807e-06, + "loss": 0.1518, + "step": 5358 + }, + { + "epoch": 0.36657774129557424, + "grad_norm": 1.9665428383955308, + "learning_rate": 7.3122905704630845e-06, + "loss": 0.454, + "step": 5359 + }, + { + "epoch": 0.3666461454271838, + "grad_norm": 1.833931975992311, + "learning_rate": 7.3113083327419745e-06, + "loss": 0.4408, + "step": 5360 + }, + { + "epoch": 0.36671454955879335, + "grad_norm": 1.7777748674958302, + "learning_rate": 7.31032598157069e-06, + "loss": 0.4582, + "step": 5361 + }, + { + "epoch": 0.3667829536904029, + "grad_norm": 1.9963058052969467, + "learning_rate": 7.30934351699745e-06, + "loss": 0.4701, + "step": 5362 + }, + { + "epoch": 0.36685135782201245, + "grad_norm": 1.9507989612977643, + "learning_rate": 7.308360939070479e-06, + "loss": 0.5221, + "step": 5363 + }, + { + "epoch": 0.366919761953622, + "grad_norm": 1.696297886572393, + "learning_rate": 7.307378247838006e-06, + "loss": 0.1602, + "step": 5364 + }, + { + "epoch": 0.36698816608523155, + "grad_norm": 2.033385914437975, + "learning_rate": 7.306395443348267e-06, + "loss": 0.3965, + "step": 5365 + }, + { + "epoch": 0.36705657021684107, + "grad_norm": 1.5191415664618926, + "learning_rate": 7.3054125256495e-06, + "loss": 0.2508, + "step": 5366 + }, + { + "epoch": 0.36712497434845065, + "grad_norm": 2.0187455905063056, + "learning_rate": 7.304429494789956e-06, + "loss": 0.2824, + "step": 5367 + }, + { + "epoch": 0.3671933784800602, + "grad_norm": 2.0854713726363037, + "learning_rate": 7.303446350817883e-06, + "loss": 0.4975, + "step": 5368 + }, + { + "epoch": 0.36726178261166975, + "grad_norm": 1.4655391898788224, + "learning_rate": 7.30246309378154e-06, + "loss": 0.1538, + "step": 5369 + }, + { + "epoch": 0.3673301867432793, + "grad_norm": 1.6368773026248076, + "learning_rate": 7.30147972372919e-06, + "loss": 0.3189, + "step": 5370 + }, + { + "epoch": 0.36739859087488885, + "grad_norm": 2.481037170881652, + "learning_rate": 7.300496240709102e-06, + "loss": 0.6723, + "step": 5371 + }, + { + "epoch": 0.3674669950064984, + "grad_norm": 1.9539954458359812, + "learning_rate": 7.299512644769549e-06, + "loss": 0.3332, + "step": 5372 + }, + { + "epoch": 0.36753539913810795, + "grad_norm": 1.8718481673290102, + "learning_rate": 7.298528935958811e-06, + "loss": 0.4306, + "step": 5373 + }, + { + "epoch": 0.3676038032697175, + "grad_norm": 1.949979874311418, + "learning_rate": 7.297545114325173e-06, + "loss": 0.4798, + "step": 5374 + }, + { + "epoch": 0.36767220740132706, + "grad_norm": 1.7175333745565045, + "learning_rate": 7.296561179916928e-06, + "loss": 0.3855, + "step": 5375 + }, + { + "epoch": 0.3677406115329366, + "grad_norm": 1.6845931187067231, + "learning_rate": 7.295577132782368e-06, + "loss": 0.3122, + "step": 5376 + }, + { + "epoch": 0.36780901566454616, + "grad_norm": 1.8315037209829428, + "learning_rate": 7.2945929729697984e-06, + "loss": 0.4337, + "step": 5377 + }, + { + "epoch": 0.3678774197961557, + "grad_norm": 2.3351981227230776, + "learning_rate": 7.293608700527525e-06, + "loss": 0.5626, + "step": 5378 + }, + { + "epoch": 0.36794582392776526, + "grad_norm": 2.081754845026608, + "learning_rate": 7.29262431550386e-06, + "loss": 0.5641, + "step": 5379 + }, + { + "epoch": 0.3680142280593748, + "grad_norm": 1.987149435441531, + "learning_rate": 7.291639817947122e-06, + "loss": 0.5026, + "step": 5380 + }, + { + "epoch": 0.36808263219098436, + "grad_norm": 1.4617624707516148, + "learning_rate": 7.290655207905639e-06, + "loss": 0.2571, + "step": 5381 + }, + { + "epoch": 0.3681510363225939, + "grad_norm": 1.9448872526557501, + "learning_rate": 7.289670485427733e-06, + "loss": 0.4793, + "step": 5382 + }, + { + "epoch": 0.36821944045420346, + "grad_norm": 2.2723888915296397, + "learning_rate": 7.288685650561744e-06, + "loss": 0.5113, + "step": 5383 + }, + { + "epoch": 0.368287844585813, + "grad_norm": 2.1374448836063626, + "learning_rate": 7.287700703356013e-06, + "loss": 0.3867, + "step": 5384 + }, + { + "epoch": 0.3683562487174225, + "grad_norm": 1.7717590539662427, + "learning_rate": 7.286715643858884e-06, + "loss": 0.2791, + "step": 5385 + }, + { + "epoch": 0.3684246528490321, + "grad_norm": 1.693562258981843, + "learning_rate": 7.285730472118707e-06, + "loss": 0.3297, + "step": 5386 + }, + { + "epoch": 0.3684930569806416, + "grad_norm": 2.254905637788295, + "learning_rate": 7.2847451881838425e-06, + "loss": 0.4315, + "step": 5387 + }, + { + "epoch": 0.3685614611122512, + "grad_norm": 2.220184674969338, + "learning_rate": 7.2837597921026505e-06, + "loss": 0.7324, + "step": 5388 + }, + { + "epoch": 0.3686298652438607, + "grad_norm": 1.4050683699675661, + "learning_rate": 7.2827742839235e-06, + "loss": 0.2706, + "step": 5389 + }, + { + "epoch": 0.3686982693754703, + "grad_norm": 1.8090451915042771, + "learning_rate": 7.281788663694763e-06, + "loss": 0.3675, + "step": 5390 + }, + { + "epoch": 0.3687666735070798, + "grad_norm": 1.9110566067011998, + "learning_rate": 7.280802931464821e-06, + "loss": 0.2495, + "step": 5391 + }, + { + "epoch": 0.3688350776386894, + "grad_norm": 1.7026999975708554, + "learning_rate": 7.279817087282057e-06, + "loss": 0.3695, + "step": 5392 + }, + { + "epoch": 0.3689034817702989, + "grad_norm": 1.8138594626291675, + "learning_rate": 7.278831131194861e-06, + "loss": 0.3322, + "step": 5393 + }, + { + "epoch": 0.3689718859019085, + "grad_norm": 2.2781473217565247, + "learning_rate": 7.2778450632516295e-06, + "loss": 0.5486, + "step": 5394 + }, + { + "epoch": 0.369040290033518, + "grad_norm": 3.614820212408553, + "learning_rate": 7.276858883500761e-06, + "loss": 0.5154, + "step": 5395 + }, + { + "epoch": 0.3691086941651276, + "grad_norm": 2.0991197094909406, + "learning_rate": 7.275872591990665e-06, + "loss": 0.3667, + "step": 5396 + }, + { + "epoch": 0.3691770982967371, + "grad_norm": 2.458092743744159, + "learning_rate": 7.274886188769752e-06, + "loss": 0.5265, + "step": 5397 + }, + { + "epoch": 0.3692455024283467, + "grad_norm": 2.220222870791446, + "learning_rate": 7.27389967388644e-06, + "loss": 0.6668, + "step": 5398 + }, + { + "epoch": 0.3693139065599562, + "grad_norm": 1.36251250975448, + "learning_rate": 7.27291304738915e-06, + "loss": 0.2302, + "step": 5399 + }, + { + "epoch": 0.3693823106915658, + "grad_norm": 2.2600625814839406, + "learning_rate": 7.271926309326313e-06, + "loss": 0.52, + "step": 5400 + }, + { + "epoch": 0.3694507148231753, + "grad_norm": 1.8989809145468048, + "learning_rate": 7.2709394597463625e-06, + "loss": 0.3188, + "step": 5401 + }, + { + "epoch": 0.36951911895478484, + "grad_norm": 1.2255663349666082, + "learning_rate": 7.269952498697734e-06, + "loss": 0.1191, + "step": 5402 + }, + { + "epoch": 0.3695875230863944, + "grad_norm": 1.60854413519699, + "learning_rate": 7.2689654262288776e-06, + "loss": 0.235, + "step": 5403 + }, + { + "epoch": 0.36965592721800394, + "grad_norm": 1.6808375480169064, + "learning_rate": 7.267978242388242e-06, + "loss": 0.34, + "step": 5404 + }, + { + "epoch": 0.3697243313496135, + "grad_norm": 2.2505735355119243, + "learning_rate": 7.26699094722428e-06, + "loss": 0.5185, + "step": 5405 + }, + { + "epoch": 0.36979273548122304, + "grad_norm": 2.2749489940961944, + "learning_rate": 7.266003540785458e-06, + "loss": 0.326, + "step": 5406 + }, + { + "epoch": 0.3698611396128326, + "grad_norm": 1.897540378919946, + "learning_rate": 7.265016023120238e-06, + "loss": 0.3411, + "step": 5407 + }, + { + "epoch": 0.36992954374444215, + "grad_norm": 1.7554162116479846, + "learning_rate": 7.264028394277094e-06, + "loss": 0.2082, + "step": 5408 + }, + { + "epoch": 0.3699979478760517, + "grad_norm": 2.362227856134499, + "learning_rate": 7.263040654304503e-06, + "loss": 0.5891, + "step": 5409 + }, + { + "epoch": 0.37006635200766125, + "grad_norm": 2.1811606785102815, + "learning_rate": 7.26205280325095e-06, + "loss": 0.3894, + "step": 5410 + }, + { + "epoch": 0.3701347561392708, + "grad_norm": 2.2616004879986793, + "learning_rate": 7.261064841164922e-06, + "loss": 0.4661, + "step": 5411 + }, + { + "epoch": 0.37020316027088035, + "grad_norm": 1.8929197477452353, + "learning_rate": 7.260076768094912e-06, + "loss": 0.483, + "step": 5412 + }, + { + "epoch": 0.3702715644024899, + "grad_norm": 1.6928218749675614, + "learning_rate": 7.25908858408942e-06, + "loss": 0.2828, + "step": 5413 + }, + { + "epoch": 0.37033996853409945, + "grad_norm": 1.9673037439350147, + "learning_rate": 7.2581002891969535e-06, + "loss": 0.5265, + "step": 5414 + }, + { + "epoch": 0.37040837266570903, + "grad_norm": 2.320996360314937, + "learning_rate": 7.257111883466018e-06, + "loss": 0.3084, + "step": 5415 + }, + { + "epoch": 0.37047677679731855, + "grad_norm": 2.054977309526349, + "learning_rate": 7.256123366945134e-06, + "loss": 0.476, + "step": 5416 + }, + { + "epoch": 0.37054518092892813, + "grad_norm": 1.5886921988491702, + "learning_rate": 7.25513473968282e-06, + "loss": 0.2248, + "step": 5417 + }, + { + "epoch": 0.37061358506053765, + "grad_norm": 1.8662808982555092, + "learning_rate": 7.254146001727602e-06, + "loss": 0.4611, + "step": 5418 + }, + { + "epoch": 0.37068198919214723, + "grad_norm": 2.243824942407808, + "learning_rate": 7.253157153128012e-06, + "loss": 0.4874, + "step": 5419 + }, + { + "epoch": 0.37075039332375676, + "grad_norm": 2.319095105968909, + "learning_rate": 7.252168193932592e-06, + "loss": 0.5899, + "step": 5420 + }, + { + "epoch": 0.3708187974553663, + "grad_norm": 2.720116896501609, + "learning_rate": 7.25117912418988e-06, + "loss": 0.6919, + "step": 5421 + }, + { + "epoch": 0.37088720158697586, + "grad_norm": 2.416881372242634, + "learning_rate": 7.250189943948424e-06, + "loss": 0.7425, + "step": 5422 + }, + { + "epoch": 0.3709556057185854, + "grad_norm": 1.7176315786872696, + "learning_rate": 7.24920065325678e-06, + "loss": 0.4185, + "step": 5423 + }, + { + "epoch": 0.37102400985019496, + "grad_norm": 2.3953493410326834, + "learning_rate": 7.248211252163509e-06, + "loss": 0.679, + "step": 5424 + }, + { + "epoch": 0.3710924139818045, + "grad_norm": 2.418631504418963, + "learning_rate": 7.24722174071717e-06, + "loss": 0.6396, + "step": 5425 + }, + { + "epoch": 0.37116081811341406, + "grad_norm": 2.42524192330864, + "learning_rate": 7.246232118966338e-06, + "loss": 0.9516, + "step": 5426 + }, + { + "epoch": 0.3712292222450236, + "grad_norm": 2.1981936754036524, + "learning_rate": 7.2452423869595874e-06, + "loss": 0.457, + "step": 5427 + }, + { + "epoch": 0.37129762637663316, + "grad_norm": 1.5234842831201871, + "learning_rate": 7.244252544745497e-06, + "loss": 0.3077, + "step": 5428 + }, + { + "epoch": 0.3713660305082427, + "grad_norm": 1.6870857962258807, + "learning_rate": 7.243262592372655e-06, + "loss": 0.358, + "step": 5429 + }, + { + "epoch": 0.37143443463985226, + "grad_norm": 1.4916050658035647, + "learning_rate": 7.242272529889652e-06, + "loss": 0.2762, + "step": 5430 + }, + { + "epoch": 0.3715028387714618, + "grad_norm": 1.3913131093674664, + "learning_rate": 7.241282357345086e-06, + "loss": 0.159, + "step": 5431 + }, + { + "epoch": 0.37157124290307136, + "grad_norm": 1.5217337789556191, + "learning_rate": 7.240292074787558e-06, + "loss": 0.2164, + "step": 5432 + }, + { + "epoch": 0.3716396470346809, + "grad_norm": 2.2700148246192837, + "learning_rate": 7.239301682265677e-06, + "loss": 0.4659, + "step": 5433 + }, + { + "epoch": 0.37170805116629047, + "grad_norm": 2.5429686386647474, + "learning_rate": 7.238311179828056e-06, + "loss": 0.5107, + "step": 5434 + }, + { + "epoch": 0.3717764552979, + "grad_norm": 2.00895733237493, + "learning_rate": 7.237320567523314e-06, + "loss": 0.4985, + "step": 5435 + }, + { + "epoch": 0.37184485942950957, + "grad_norm": 1.9712557173871597, + "learning_rate": 7.236329845400074e-06, + "loss": 0.2068, + "step": 5436 + }, + { + "epoch": 0.3719132635611191, + "grad_norm": 2.41390299964561, + "learning_rate": 7.235339013506967e-06, + "loss": 0.7124, + "step": 5437 + }, + { + "epoch": 0.37198166769272867, + "grad_norm": 1.9082621601386476, + "learning_rate": 7.234348071892625e-06, + "loss": 0.1893, + "step": 5438 + }, + { + "epoch": 0.3720500718243382, + "grad_norm": 2.2030994637003625, + "learning_rate": 7.233357020605692e-06, + "loss": 0.3574, + "step": 5439 + }, + { + "epoch": 0.3721184759559477, + "grad_norm": 1.6482439560286852, + "learning_rate": 7.232365859694811e-06, + "loss": 0.1947, + "step": 5440 + }, + { + "epoch": 0.3721868800875573, + "grad_norm": 2.296993430970439, + "learning_rate": 7.231374589208634e-06, + "loss": 0.5809, + "step": 5441 + }, + { + "epoch": 0.3722552842191668, + "grad_norm": 1.95958734476601, + "learning_rate": 7.230383209195816e-06, + "loss": 0.3284, + "step": 5442 + }, + { + "epoch": 0.3723236883507764, + "grad_norm": 1.426894414421269, + "learning_rate": 7.22939171970502e-06, + "loss": 0.2165, + "step": 5443 + }, + { + "epoch": 0.3723920924823859, + "grad_norm": 1.9850653044680604, + "learning_rate": 7.228400120784913e-06, + "loss": 0.441, + "step": 5444 + }, + { + "epoch": 0.3724604966139955, + "grad_norm": 1.754448125176169, + "learning_rate": 7.227408412484167e-06, + "loss": 0.2493, + "step": 5445 + }, + { + "epoch": 0.372528900745605, + "grad_norm": 2.1696258191881164, + "learning_rate": 7.226416594851461e-06, + "loss": 0.3672, + "step": 5446 + }, + { + "epoch": 0.3725973048772146, + "grad_norm": 1.9065767539716558, + "learning_rate": 7.225424667935477e-06, + "loss": 0.3219, + "step": 5447 + }, + { + "epoch": 0.3726657090088241, + "grad_norm": 1.9067136599310137, + "learning_rate": 7.224432631784902e-06, + "loss": 0.5142, + "step": 5448 + }, + { + "epoch": 0.3727341131404337, + "grad_norm": 1.6231293686009998, + "learning_rate": 7.2234404864484335e-06, + "loss": 0.311, + "step": 5449 + }, + { + "epoch": 0.3728025172720432, + "grad_norm": 2.428043385708792, + "learning_rate": 7.222448231974769e-06, + "loss": 0.6354, + "step": 5450 + }, + { + "epoch": 0.3728709214036528, + "grad_norm": 2.0351670570599127, + "learning_rate": 7.2214558684126124e-06, + "loss": 0.5591, + "step": 5451 + }, + { + "epoch": 0.3729393255352623, + "grad_norm": 1.7648258025095636, + "learning_rate": 7.220463395810675e-06, + "loss": 0.2495, + "step": 5452 + }, + { + "epoch": 0.3730077296668719, + "grad_norm": 1.7990127949145684, + "learning_rate": 7.219470814217672e-06, + "loss": 0.2315, + "step": 5453 + }, + { + "epoch": 0.3730761337984814, + "grad_norm": 2.0777330172283843, + "learning_rate": 7.218478123682323e-06, + "loss": 0.438, + "step": 5454 + }, + { + "epoch": 0.373144537930091, + "grad_norm": 2.411098858519692, + "learning_rate": 7.217485324253355e-06, + "loss": 0.517, + "step": 5455 + }, + { + "epoch": 0.3732129420617005, + "grad_norm": 6.53421759377053, + "learning_rate": 7.216492415979499e-06, + "loss": 0.3096, + "step": 5456 + }, + { + "epoch": 0.37328134619331005, + "grad_norm": 2.147525017906806, + "learning_rate": 7.215499398909493e-06, + "loss": 0.5025, + "step": 5457 + }, + { + "epoch": 0.3733497503249196, + "grad_norm": 1.9384785419136914, + "learning_rate": 7.214506273092075e-06, + "loss": 0.5256, + "step": 5458 + }, + { + "epoch": 0.37341815445652915, + "grad_norm": 2.742011322709015, + "learning_rate": 7.213513038575999e-06, + "loss": 0.5544, + "step": 5459 + }, + { + "epoch": 0.37348655858813873, + "grad_norm": 2.0857877213213882, + "learning_rate": 7.212519695410013e-06, + "loss": 0.5408, + "step": 5460 + }, + { + "epoch": 0.37355496271974825, + "grad_norm": 2.2259962198357304, + "learning_rate": 7.211526243642874e-06, + "loss": 0.6836, + "step": 5461 + }, + { + "epoch": 0.37362336685135783, + "grad_norm": 1.609594080189058, + "learning_rate": 7.2105326833233505e-06, + "loss": 0.2311, + "step": 5462 + }, + { + "epoch": 0.37369177098296735, + "grad_norm": 1.7570550261398825, + "learning_rate": 7.209539014500207e-06, + "loss": 0.366, + "step": 5463 + }, + { + "epoch": 0.37376017511457693, + "grad_norm": 1.8119663200971998, + "learning_rate": 7.20854523722222e-06, + "loss": 0.508, + "step": 5464 + }, + { + "epoch": 0.37382857924618645, + "grad_norm": 2.481299277808382, + "learning_rate": 7.207551351538167e-06, + "loss": 0.5394, + "step": 5465 + }, + { + "epoch": 0.37389698337779603, + "grad_norm": 2.3344263943091943, + "learning_rate": 7.206557357496833e-06, + "loss": 0.7312, + "step": 5466 + }, + { + "epoch": 0.37396538750940556, + "grad_norm": 2.3532156630448275, + "learning_rate": 7.205563255147009e-06, + "loss": 0.6105, + "step": 5467 + }, + { + "epoch": 0.37403379164101513, + "grad_norm": 2.175435489048753, + "learning_rate": 7.204569044537489e-06, + "loss": 0.5995, + "step": 5468 + }, + { + "epoch": 0.37410219577262466, + "grad_norm": 1.8450840552625964, + "learning_rate": 7.2035747257170766e-06, + "loss": 0.3505, + "step": 5469 + }, + { + "epoch": 0.37417059990423424, + "grad_norm": 2.236660675243572, + "learning_rate": 7.202580298734575e-06, + "loss": 0.6374, + "step": 5470 + }, + { + "epoch": 0.37423900403584376, + "grad_norm": 2.0542305848066604, + "learning_rate": 7.201585763638794e-06, + "loss": 0.4732, + "step": 5471 + }, + { + "epoch": 0.37430740816745334, + "grad_norm": 1.9059007864157314, + "learning_rate": 7.200591120478555e-06, + "loss": 0.3218, + "step": 5472 + }, + { + "epoch": 0.37437581229906286, + "grad_norm": 2.102296055961908, + "learning_rate": 7.199596369302676e-06, + "loss": 0.5306, + "step": 5473 + }, + { + "epoch": 0.37444421643067244, + "grad_norm": 2.359810694197984, + "learning_rate": 7.198601510159986e-06, + "loss": 0.5595, + "step": 5474 + }, + { + "epoch": 0.37451262056228196, + "grad_norm": 2.479633052350751, + "learning_rate": 7.197606543099317e-06, + "loss": 0.6487, + "step": 5475 + }, + { + "epoch": 0.3745810246938915, + "grad_norm": 1.3337273692237295, + "learning_rate": 7.196611468169507e-06, + "loss": 0.2096, + "step": 5476 + }, + { + "epoch": 0.37464942882550106, + "grad_norm": 2.084025867910702, + "learning_rate": 7.195616285419398e-06, + "loss": 0.5823, + "step": 5477 + }, + { + "epoch": 0.3747178329571106, + "grad_norm": 2.477031133939644, + "learning_rate": 7.194620994897838e-06, + "loss": 0.4055, + "step": 5478 + }, + { + "epoch": 0.37478623708872016, + "grad_norm": 2.115724635348879, + "learning_rate": 7.193625596653684e-06, + "loss": 0.4955, + "step": 5479 + }, + { + "epoch": 0.3748546412203297, + "grad_norm": 1.639185010115372, + "learning_rate": 7.192630090735793e-06, + "loss": 0.2194, + "step": 5480 + }, + { + "epoch": 0.37492304535193927, + "grad_norm": 4.446251473371611, + "learning_rate": 7.191634477193028e-06, + "loss": 0.4311, + "step": 5481 + }, + { + "epoch": 0.3749914494835488, + "grad_norm": 1.7850582797089174, + "learning_rate": 7.190638756074258e-06, + "loss": 0.3878, + "step": 5482 + }, + { + "epoch": 0.37505985361515837, + "grad_norm": 1.8067205586663952, + "learning_rate": 7.189642927428362e-06, + "loss": 0.3401, + "step": 5483 + }, + { + "epoch": 0.3751282577467679, + "grad_norm": 1.8635629964737572, + "learning_rate": 7.188646991304218e-06, + "loss": 0.275, + "step": 5484 + }, + { + "epoch": 0.37519666187837747, + "grad_norm": 2.202432293581916, + "learning_rate": 7.187650947750708e-06, + "loss": 0.449, + "step": 5485 + }, + { + "epoch": 0.375265066009987, + "grad_norm": 1.6628710865022587, + "learning_rate": 7.186654796816728e-06, + "loss": 0.1857, + "step": 5486 + }, + { + "epoch": 0.37533347014159657, + "grad_norm": 1.9751027714270522, + "learning_rate": 7.185658538551169e-06, + "loss": 0.4445, + "step": 5487 + }, + { + "epoch": 0.3754018742732061, + "grad_norm": 1.6939688688945256, + "learning_rate": 7.184662173002936e-06, + "loss": 0.3286, + "step": 5488 + }, + { + "epoch": 0.37547027840481567, + "grad_norm": 1.459606717111395, + "learning_rate": 7.183665700220933e-06, + "loss": 0.2037, + "step": 5489 + }, + { + "epoch": 0.3755386825364252, + "grad_norm": 1.9130276509977127, + "learning_rate": 7.182669120254074e-06, + "loss": 0.4178, + "step": 5490 + }, + { + "epoch": 0.3756070866680348, + "grad_norm": 7.812830147443116, + "learning_rate": 7.181672433151273e-06, + "loss": 0.212, + "step": 5491 + }, + { + "epoch": 0.3756754907996443, + "grad_norm": 1.7624731161556246, + "learning_rate": 7.180675638961454e-06, + "loss": 0.365, + "step": 5492 + }, + { + "epoch": 0.3757438949312539, + "grad_norm": 2.70955758057081, + "learning_rate": 7.179678737733545e-06, + "loss": 0.4018, + "step": 5493 + }, + { + "epoch": 0.3758122990628634, + "grad_norm": 1.6341733326679515, + "learning_rate": 7.178681729516479e-06, + "loss": 0.2846, + "step": 5494 + }, + { + "epoch": 0.3758807031944729, + "grad_norm": 2.127662239231561, + "learning_rate": 7.1776846143591905e-06, + "loss": 0.232, + "step": 5495 + }, + { + "epoch": 0.3759491073260825, + "grad_norm": 1.9553475475355793, + "learning_rate": 7.1766873923106276e-06, + "loss": 0.2045, + "step": 5496 + }, + { + "epoch": 0.376017511457692, + "grad_norm": 2.932486044345634, + "learning_rate": 7.175690063419734e-06, + "loss": 0.3587, + "step": 5497 + }, + { + "epoch": 0.3760859155893016, + "grad_norm": 1.7481428365100906, + "learning_rate": 7.1746926277354666e-06, + "loss": 0.306, + "step": 5498 + }, + { + "epoch": 0.3761543197209111, + "grad_norm": 3.6831671264455736, + "learning_rate": 7.173695085306785e-06, + "loss": 0.5606, + "step": 5499 + }, + { + "epoch": 0.3762227238525207, + "grad_norm": 1.6120420458048434, + "learning_rate": 7.172697436182652e-06, + "loss": 0.2431, + "step": 5500 + }, + { + "epoch": 0.3762911279841302, + "grad_norm": 1.9483407514301485, + "learning_rate": 7.171699680412034e-06, + "loss": 0.3571, + "step": 5501 + }, + { + "epoch": 0.3763595321157398, + "grad_norm": 2.0601071290751967, + "learning_rate": 7.1707018180439104e-06, + "loss": 0.3568, + "step": 5502 + }, + { + "epoch": 0.3764279362473493, + "grad_norm": 1.8647957330985494, + "learning_rate": 7.169703849127261e-06, + "loss": 0.4742, + "step": 5503 + }, + { + "epoch": 0.3764963403789589, + "grad_norm": 1.8706164975838186, + "learning_rate": 7.1687057737110665e-06, + "loss": 0.4597, + "step": 5504 + }, + { + "epoch": 0.3765647445105684, + "grad_norm": 2.0523820968606064, + "learning_rate": 7.167707591844322e-06, + "loss": 0.5193, + "step": 5505 + }, + { + "epoch": 0.376633148642178, + "grad_norm": 1.962970925581708, + "learning_rate": 7.16670930357602e-06, + "loss": 0.6002, + "step": 5506 + }, + { + "epoch": 0.37670155277378753, + "grad_norm": 1.7676592413061716, + "learning_rate": 7.1657109089551635e-06, + "loss": 0.3334, + "step": 5507 + }, + { + "epoch": 0.3767699569053971, + "grad_norm": 1.6392206539368075, + "learning_rate": 7.164712408030756e-06, + "loss": 0.1882, + "step": 5508 + }, + { + "epoch": 0.37683836103700663, + "grad_norm": 2.07911590205704, + "learning_rate": 7.163713800851812e-06, + "loss": 0.3141, + "step": 5509 + }, + { + "epoch": 0.3769067651686162, + "grad_norm": 2.832545561293568, + "learning_rate": 7.162715087467346e-06, + "loss": 0.6037, + "step": 5510 + }, + { + "epoch": 0.37697516930022573, + "grad_norm": 1.4934385183418282, + "learning_rate": 7.161716267926377e-06, + "loss": 0.3263, + "step": 5511 + }, + { + "epoch": 0.3770435734318353, + "grad_norm": 2.5396453012495326, + "learning_rate": 7.160717342277937e-06, + "loss": 0.5932, + "step": 5512 + }, + { + "epoch": 0.37711197756344483, + "grad_norm": 1.625839220724668, + "learning_rate": 7.159718310571057e-06, + "loss": 0.2825, + "step": 5513 + }, + { + "epoch": 0.37718038169505436, + "grad_norm": 2.2516214965249217, + "learning_rate": 7.158719172854771e-06, + "loss": 0.622, + "step": 5514 + }, + { + "epoch": 0.37724878582666393, + "grad_norm": 2.259513132006738, + "learning_rate": 7.157719929178125e-06, + "loss": 0.5327, + "step": 5515 + }, + { + "epoch": 0.37731718995827346, + "grad_norm": 1.9381203888730945, + "learning_rate": 7.156720579590165e-06, + "loss": 0.4466, + "step": 5516 + }, + { + "epoch": 0.37738559408988304, + "grad_norm": 2.2826909778999736, + "learning_rate": 7.155721124139944e-06, + "loss": 0.5456, + "step": 5517 + }, + { + "epoch": 0.37745399822149256, + "grad_norm": 1.9829186742506175, + "learning_rate": 7.15472156287652e-06, + "loss": 0.5336, + "step": 5518 + }, + { + "epoch": 0.37752240235310214, + "grad_norm": 1.6475503803387839, + "learning_rate": 7.1537218958489575e-06, + "loss": 0.3705, + "step": 5519 + }, + { + "epoch": 0.37759080648471166, + "grad_norm": 2.0072380349919983, + "learning_rate": 7.152722123106325e-06, + "loss": 0.4171, + "step": 5520 + }, + { + "epoch": 0.37765921061632124, + "grad_norm": 2.3814519844096242, + "learning_rate": 7.151722244697693e-06, + "loss": 0.6402, + "step": 5521 + }, + { + "epoch": 0.37772761474793076, + "grad_norm": 1.6112224486405908, + "learning_rate": 7.1507222606721425e-06, + "loss": 0.1566, + "step": 5522 + }, + { + "epoch": 0.37779601887954034, + "grad_norm": 2.2102297367838646, + "learning_rate": 7.14972217107876e-06, + "loss": 0.485, + "step": 5523 + }, + { + "epoch": 0.37786442301114986, + "grad_norm": 1.4981549459017942, + "learning_rate": 7.148721975966632e-06, + "loss": 0.2009, + "step": 5524 + }, + { + "epoch": 0.37793282714275944, + "grad_norm": 3.106478971834417, + "learning_rate": 7.147721675384852e-06, + "loss": 0.6633, + "step": 5525 + }, + { + "epoch": 0.37800123127436897, + "grad_norm": 2.3559799651773976, + "learning_rate": 7.146721269382523e-06, + "loss": 0.7238, + "step": 5526 + }, + { + "epoch": 0.37806963540597854, + "grad_norm": 1.7629840984114193, + "learning_rate": 7.145720758008745e-06, + "loss": 0.3547, + "step": 5527 + }, + { + "epoch": 0.37813803953758807, + "grad_norm": 1.9989209915246131, + "learning_rate": 7.144720141312632e-06, + "loss": 0.4351, + "step": 5528 + }, + { + "epoch": 0.37820644366919765, + "grad_norm": 2.1515707265160557, + "learning_rate": 7.143719419343298e-06, + "loss": 0.5322, + "step": 5529 + }, + { + "epoch": 0.37827484780080717, + "grad_norm": 1.9888535146309767, + "learning_rate": 7.142718592149863e-06, + "loss": 0.4159, + "step": 5530 + }, + { + "epoch": 0.3783432519324167, + "grad_norm": 1.7161781359127803, + "learning_rate": 7.141717659781452e-06, + "loss": 0.3502, + "step": 5531 + }, + { + "epoch": 0.37841165606402627, + "grad_norm": 1.2418690819160125, + "learning_rate": 7.140716622287196e-06, + "loss": 0.205, + "step": 5532 + }, + { + "epoch": 0.3784800601956358, + "grad_norm": 2.1820738045612655, + "learning_rate": 7.139715479716232e-06, + "loss": 0.726, + "step": 5533 + }, + { + "epoch": 0.37854846432724537, + "grad_norm": 2.4145060752070204, + "learning_rate": 7.138714232117697e-06, + "loss": 0.5726, + "step": 5534 + }, + { + "epoch": 0.3786168684588549, + "grad_norm": 2.300625363494712, + "learning_rate": 7.137712879540742e-06, + "loss": 0.496, + "step": 5535 + }, + { + "epoch": 0.3786852725904645, + "grad_norm": 1.8434960849135562, + "learning_rate": 7.136711422034516e-06, + "loss": 0.3542, + "step": 5536 + }, + { + "epoch": 0.378753676722074, + "grad_norm": 1.730034431393268, + "learning_rate": 7.135709859648175e-06, + "loss": 0.3851, + "step": 5537 + }, + { + "epoch": 0.3788220808536836, + "grad_norm": 2.0186156700604174, + "learning_rate": 7.1347081924308794e-06, + "loss": 0.5806, + "step": 5538 + }, + { + "epoch": 0.3788904849852931, + "grad_norm": 2.0480977184700904, + "learning_rate": 7.133706420431799e-06, + "loss": 0.58, + "step": 5539 + }, + { + "epoch": 0.3789588891169027, + "grad_norm": 3.019258750868101, + "learning_rate": 7.132704543700103e-06, + "loss": 0.3476, + "step": 5540 + }, + { + "epoch": 0.3790272932485122, + "grad_norm": 1.8096260683926484, + "learning_rate": 7.131702562284969e-06, + "loss": 0.258, + "step": 5541 + }, + { + "epoch": 0.3790956973801218, + "grad_norm": 2.0102793645833468, + "learning_rate": 7.130700476235579e-06, + "loss": 0.4363, + "step": 5542 + }, + { + "epoch": 0.3791641015117313, + "grad_norm": 2.141061560811907, + "learning_rate": 7.1296982856011225e-06, + "loss": 0.4287, + "step": 5543 + }, + { + "epoch": 0.3792325056433409, + "grad_norm": 1.7313974484682204, + "learning_rate": 7.128695990430788e-06, + "loss": 0.3415, + "step": 5544 + }, + { + "epoch": 0.3793009097749504, + "grad_norm": 2.2779852550852446, + "learning_rate": 7.127693590773777e-06, + "loss": 0.623, + "step": 5545 + }, + { + "epoch": 0.37936931390656, + "grad_norm": 1.9615176015324747, + "learning_rate": 7.126691086679289e-06, + "loss": 0.3908, + "step": 5546 + }, + { + "epoch": 0.3794377180381695, + "grad_norm": 1.516423004260263, + "learning_rate": 7.125688478196532e-06, + "loss": 0.2505, + "step": 5547 + }, + { + "epoch": 0.3795061221697791, + "grad_norm": 2.529758555573028, + "learning_rate": 7.12468576537472e-06, + "loss": 0.3721, + "step": 5548 + }, + { + "epoch": 0.3795745263013886, + "grad_norm": 2.154000553284665, + "learning_rate": 7.123682948263071e-06, + "loss": 0.3871, + "step": 5549 + }, + { + "epoch": 0.3796429304329981, + "grad_norm": 1.8746341070160122, + "learning_rate": 7.1226800269108075e-06, + "loss": 0.324, + "step": 5550 + }, + { + "epoch": 0.3797113345646077, + "grad_norm": 1.9084017014819616, + "learning_rate": 7.121677001367159e-06, + "loss": 0.4047, + "step": 5551 + }, + { + "epoch": 0.37977973869621723, + "grad_norm": 1.6359362313485168, + "learning_rate": 7.120673871681356e-06, + "loss": 0.2455, + "step": 5552 + }, + { + "epoch": 0.3798481428278268, + "grad_norm": 1.2434139884486075, + "learning_rate": 7.1196706379026405e-06, + "loss": 0.1461, + "step": 5553 + }, + { + "epoch": 0.37991654695943633, + "grad_norm": 2.5210659646180824, + "learning_rate": 7.1186673000802535e-06, + "loss": 0.5094, + "step": 5554 + }, + { + "epoch": 0.3799849510910459, + "grad_norm": 1.6944475208664502, + "learning_rate": 7.117663858263444e-06, + "loss": 0.3361, + "step": 5555 + }, + { + "epoch": 0.38005335522265543, + "grad_norm": 2.090402075795356, + "learning_rate": 7.116660312501466e-06, + "loss": 0.4736, + "step": 5556 + }, + { + "epoch": 0.380121759354265, + "grad_norm": 2.1004471637706588, + "learning_rate": 7.11565666284358e-06, + "loss": 0.5564, + "step": 5557 + }, + { + "epoch": 0.38019016348587453, + "grad_norm": 1.5206214480117537, + "learning_rate": 7.114652909339048e-06, + "loss": 0.2601, + "step": 5558 + }, + { + "epoch": 0.3802585676174841, + "grad_norm": 2.463256019271911, + "learning_rate": 7.1136490520371394e-06, + "loss": 0.304, + "step": 5559 + }, + { + "epoch": 0.38032697174909363, + "grad_norm": 1.5899903279061307, + "learning_rate": 7.112645090987129e-06, + "loss": 0.2012, + "step": 5560 + }, + { + "epoch": 0.3803953758807032, + "grad_norm": 2.4119725980087807, + "learning_rate": 7.111641026238294e-06, + "loss": 0.658, + "step": 5561 + }, + { + "epoch": 0.38046378001231274, + "grad_norm": 1.8386768195930874, + "learning_rate": 7.110636857839923e-06, + "loss": 0.4683, + "step": 5562 + }, + { + "epoch": 0.3805321841439223, + "grad_norm": 2.2500858290514487, + "learning_rate": 7.1096325858413015e-06, + "loss": 0.3392, + "step": 5563 + }, + { + "epoch": 0.38060058827553184, + "grad_norm": 3.031614532378078, + "learning_rate": 7.108628210291726e-06, + "loss": 0.3122, + "step": 5564 + }, + { + "epoch": 0.3806689924071414, + "grad_norm": 1.8610460161652003, + "learning_rate": 7.107623731240496e-06, + "loss": 0.2731, + "step": 5565 + }, + { + "epoch": 0.38073739653875094, + "grad_norm": 2.1146617582808336, + "learning_rate": 7.106619148736914e-06, + "loss": 0.3066, + "step": 5566 + }, + { + "epoch": 0.3808058006703605, + "grad_norm": 2.2806395210290984, + "learning_rate": 7.105614462830292e-06, + "loss": 0.6344, + "step": 5567 + }, + { + "epoch": 0.38087420480197004, + "grad_norm": 1.9103090034052235, + "learning_rate": 7.104609673569944e-06, + "loss": 0.4714, + "step": 5568 + }, + { + "epoch": 0.38094260893357956, + "grad_norm": 1.7239434932135316, + "learning_rate": 7.103604781005191e-06, + "loss": 0.3531, + "step": 5569 + }, + { + "epoch": 0.38101101306518914, + "grad_norm": 2.3712558848016942, + "learning_rate": 7.102599785185357e-06, + "loss": 0.4493, + "step": 5570 + }, + { + "epoch": 0.38107941719679866, + "grad_norm": 1.8125926561349823, + "learning_rate": 7.101594686159771e-06, + "loss": 0.5434, + "step": 5571 + }, + { + "epoch": 0.38114782132840824, + "grad_norm": 1.7913897299389527, + "learning_rate": 7.100589483977771e-06, + "loss": 0.4028, + "step": 5572 + }, + { + "epoch": 0.38121622546001777, + "grad_norm": 2.036516419644921, + "learning_rate": 7.099584178688694e-06, + "loss": 0.3962, + "step": 5573 + }, + { + "epoch": 0.38128462959162734, + "grad_norm": 2.1898616168095897, + "learning_rate": 7.098578770341889e-06, + "loss": 0.5552, + "step": 5574 + }, + { + "epoch": 0.38135303372323687, + "grad_norm": 2.2432541187694524, + "learning_rate": 7.097573258986702e-06, + "loss": 0.5225, + "step": 5575 + }, + { + "epoch": 0.38142143785484645, + "grad_norm": 1.8393751709938293, + "learning_rate": 7.09656764467249e-06, + "loss": 0.4567, + "step": 5576 + }, + { + "epoch": 0.38148984198645597, + "grad_norm": 1.7309647574743272, + "learning_rate": 7.095561927448614e-06, + "loss": 0.2063, + "step": 5577 + }, + { + "epoch": 0.38155824611806555, + "grad_norm": 2.097513691268676, + "learning_rate": 7.094556107364441e-06, + "loss": 0.3144, + "step": 5578 + }, + { + "epoch": 0.38162665024967507, + "grad_norm": 2.202574116572596, + "learning_rate": 7.093550184469339e-06, + "loss": 0.5086, + "step": 5579 + }, + { + "epoch": 0.38169505438128465, + "grad_norm": 2.3636745094154765, + "learning_rate": 7.092544158812683e-06, + "loss": 0.704, + "step": 5580 + }, + { + "epoch": 0.38176345851289417, + "grad_norm": 1.8535349958430603, + "learning_rate": 7.091538030443855e-06, + "loss": 0.3055, + "step": 5581 + }, + { + "epoch": 0.38183186264450375, + "grad_norm": 2.1462056444991284, + "learning_rate": 7.0905317994122424e-06, + "loss": 0.6034, + "step": 5582 + }, + { + "epoch": 0.3819002667761133, + "grad_norm": 1.5928838470567732, + "learning_rate": 7.089525465767233e-06, + "loss": 0.3418, + "step": 5583 + }, + { + "epoch": 0.38196867090772285, + "grad_norm": 1.8018878507337823, + "learning_rate": 7.088519029558223e-06, + "loss": 0.2796, + "step": 5584 + }, + { + "epoch": 0.3820370750393324, + "grad_norm": 2.0409565602202377, + "learning_rate": 7.087512490834615e-06, + "loss": 0.1967, + "step": 5585 + }, + { + "epoch": 0.3821054791709419, + "grad_norm": 1.716924933820764, + "learning_rate": 7.086505849645812e-06, + "loss": 0.4748, + "step": 5586 + }, + { + "epoch": 0.3821738833025515, + "grad_norm": 1.9512479841721635, + "learning_rate": 7.085499106041226e-06, + "loss": 0.4832, + "step": 5587 + }, + { + "epoch": 0.382242287434161, + "grad_norm": 1.5825169865470936, + "learning_rate": 7.084492260070275e-06, + "loss": 0.3272, + "step": 5588 + }, + { + "epoch": 0.3823106915657706, + "grad_norm": 2.007524692376416, + "learning_rate": 7.083485311782376e-06, + "loss": 0.4381, + "step": 5589 + }, + { + "epoch": 0.3823790956973801, + "grad_norm": 1.8105640883795135, + "learning_rate": 7.0824782612269584e-06, + "loss": 0.3387, + "step": 5590 + }, + { + "epoch": 0.3824474998289897, + "grad_norm": 2.336483103847167, + "learning_rate": 7.08147110845345e-06, + "loss": 0.4315, + "step": 5591 + }, + { + "epoch": 0.3825159039605992, + "grad_norm": 1.906664163318745, + "learning_rate": 7.0804638535112905e-06, + "loss": 0.2525, + "step": 5592 + }, + { + "epoch": 0.3825843080922088, + "grad_norm": 1.4059293354854692, + "learning_rate": 7.079456496449917e-06, + "loss": 0.2014, + "step": 5593 + }, + { + "epoch": 0.3826527122238183, + "grad_norm": 2.6545946909918428, + "learning_rate": 7.078449037318777e-06, + "loss": 0.63, + "step": 5594 + }, + { + "epoch": 0.3827211163554279, + "grad_norm": 2.122019613110635, + "learning_rate": 7.0774414761673215e-06, + "loss": 0.3507, + "step": 5595 + }, + { + "epoch": 0.3827895204870374, + "grad_norm": 1.8549950045071093, + "learning_rate": 7.076433813045007e-06, + "loss": 0.3678, + "step": 5596 + }, + { + "epoch": 0.382857924618647, + "grad_norm": 1.531683991685824, + "learning_rate": 7.075426048001294e-06, + "loss": 0.2787, + "step": 5597 + }, + { + "epoch": 0.3829263287502565, + "grad_norm": 2.9729300634772446, + "learning_rate": 7.074418181085648e-06, + "loss": 0.6024, + "step": 5598 + }, + { + "epoch": 0.3829947328818661, + "grad_norm": 2.038124307107676, + "learning_rate": 7.073410212347541e-06, + "loss": 0.4852, + "step": 5599 + }, + { + "epoch": 0.3830631370134756, + "grad_norm": 1.8412208586806196, + "learning_rate": 7.072402141836448e-06, + "loss": 0.3876, + "step": 5600 + }, + { + "epoch": 0.3831315411450852, + "grad_norm": 1.7825754863985146, + "learning_rate": 7.071393969601849e-06, + "loss": 0.379, + "step": 5601 + }, + { + "epoch": 0.3831999452766947, + "grad_norm": 2.130108163308066, + "learning_rate": 7.070385695693233e-06, + "loss": 0.1535, + "step": 5602 + }, + { + "epoch": 0.3832683494083043, + "grad_norm": 1.9366030462679333, + "learning_rate": 7.06937732016009e-06, + "loss": 0.5708, + "step": 5603 + }, + { + "epoch": 0.3833367535399138, + "grad_norm": 1.6538004034496732, + "learning_rate": 7.0683688430519135e-06, + "loss": 0.2431, + "step": 5604 + }, + { + "epoch": 0.38340515767152333, + "grad_norm": 3.4988047568454506, + "learning_rate": 7.067360264418208e-06, + "loss": 0.676, + "step": 5605 + }, + { + "epoch": 0.3834735618031329, + "grad_norm": 1.2406054985281683, + "learning_rate": 7.066351584308475e-06, + "loss": 0.1948, + "step": 5606 + }, + { + "epoch": 0.38354196593474243, + "grad_norm": 1.4235121555620014, + "learning_rate": 7.065342802772229e-06, + "loss": 0.2129, + "step": 5607 + }, + { + "epoch": 0.383610370066352, + "grad_norm": 2.081080092405277, + "learning_rate": 7.064333919858986e-06, + "loss": 0.6581, + "step": 5608 + }, + { + "epoch": 0.38367877419796154, + "grad_norm": 1.9222339938374629, + "learning_rate": 7.0633249356182656e-06, + "loss": 0.4517, + "step": 5609 + }, + { + "epoch": 0.3837471783295711, + "grad_norm": 2.015201679400028, + "learning_rate": 7.062315850099592e-06, + "loss": 0.3408, + "step": 5610 + }, + { + "epoch": 0.38381558246118064, + "grad_norm": 2.194811943792149, + "learning_rate": 7.0613066633524985e-06, + "loss": 0.2121, + "step": 5611 + }, + { + "epoch": 0.3838839865927902, + "grad_norm": 1.8660385116058429, + "learning_rate": 7.0602973754265195e-06, + "loss": 0.3085, + "step": 5612 + }, + { + "epoch": 0.38395239072439974, + "grad_norm": 2.0335021481599944, + "learning_rate": 7.0592879863711975e-06, + "loss": 0.4357, + "step": 5613 + }, + { + "epoch": 0.3840207948560093, + "grad_norm": 2.782085936002688, + "learning_rate": 7.0582784962360755e-06, + "loss": 0.654, + "step": 5614 + }, + { + "epoch": 0.38408919898761884, + "grad_norm": 2.0133070296094, + "learning_rate": 7.0572689050707064e-06, + "loss": 0.4249, + "step": 5615 + }, + { + "epoch": 0.3841576031192284, + "grad_norm": 1.6378447544653585, + "learning_rate": 7.056259212924644e-06, + "loss": 0.3491, + "step": 5616 + }, + { + "epoch": 0.38422600725083794, + "grad_norm": 1.580853070604004, + "learning_rate": 7.055249419847451e-06, + "loss": 0.3366, + "step": 5617 + }, + { + "epoch": 0.3842944113824475, + "grad_norm": 2.6286183918084367, + "learning_rate": 7.054239525888692e-06, + "loss": 0.3573, + "step": 5618 + }, + { + "epoch": 0.38436281551405704, + "grad_norm": 2.8147509997669573, + "learning_rate": 7.053229531097937e-06, + "loss": 0.7656, + "step": 5619 + }, + { + "epoch": 0.3844312196456666, + "grad_norm": 2.510253095969882, + "learning_rate": 7.052219435524761e-06, + "loss": 0.33, + "step": 5620 + }, + { + "epoch": 0.38449962377727614, + "grad_norm": 2.6401478770863154, + "learning_rate": 7.051209239218745e-06, + "loss": 0.5781, + "step": 5621 + }, + { + "epoch": 0.3845680279088857, + "grad_norm": 2.1429503542631685, + "learning_rate": 7.050198942229474e-06, + "loss": 0.2813, + "step": 5622 + }, + { + "epoch": 0.38463643204049525, + "grad_norm": 2.165556254403575, + "learning_rate": 7.04918854460654e-06, + "loss": 0.3066, + "step": 5623 + }, + { + "epoch": 0.38470483617210477, + "grad_norm": 1.7211789481070632, + "learning_rate": 7.048178046399536e-06, + "loss": 0.3623, + "step": 5624 + }, + { + "epoch": 0.38477324030371435, + "grad_norm": 2.3431149957646284, + "learning_rate": 7.047167447658064e-06, + "loss": 0.3772, + "step": 5625 + }, + { + "epoch": 0.38484164443532387, + "grad_norm": 1.5855817078596026, + "learning_rate": 7.0461567484317265e-06, + "loss": 0.1538, + "step": 5626 + }, + { + "epoch": 0.38491004856693345, + "grad_norm": 2.3830191663326676, + "learning_rate": 7.045145948770136e-06, + "loss": 0.1564, + "step": 5627 + }, + { + "epoch": 0.38497845269854297, + "grad_norm": 2.400223607459841, + "learning_rate": 7.0441350487229055e-06, + "loss": 0.2464, + "step": 5628 + }, + { + "epoch": 0.38504685683015255, + "grad_norm": 2.096829238909612, + "learning_rate": 7.043124048339657e-06, + "loss": 0.6131, + "step": 5629 + }, + { + "epoch": 0.3851152609617621, + "grad_norm": 2.5818869280626164, + "learning_rate": 7.042112947670012e-06, + "loss": 0.2771, + "step": 5630 + }, + { + "epoch": 0.38518366509337165, + "grad_norm": 2.4927492674513836, + "learning_rate": 7.041101746763603e-06, + "loss": 0.7459, + "step": 5631 + }, + { + "epoch": 0.3852520692249812, + "grad_norm": 2.1598815329392353, + "learning_rate": 7.0400904456700654e-06, + "loss": 0.5406, + "step": 5632 + }, + { + "epoch": 0.38532047335659075, + "grad_norm": 1.9489153547524658, + "learning_rate": 7.0390790444390365e-06, + "loss": 0.4311, + "step": 5633 + }, + { + "epoch": 0.3853888774882003, + "grad_norm": 1.9637656695410315, + "learning_rate": 7.038067543120162e-06, + "loss": 0.2475, + "step": 5634 + }, + { + "epoch": 0.38545728161980986, + "grad_norm": 2.4638975778853225, + "learning_rate": 7.03705594176309e-06, + "loss": 0.3375, + "step": 5635 + }, + { + "epoch": 0.3855256857514194, + "grad_norm": 2.577059947724707, + "learning_rate": 7.036044240417474e-06, + "loss": 0.377, + "step": 5636 + }, + { + "epoch": 0.38559408988302896, + "grad_norm": 2.2365856039088614, + "learning_rate": 7.035032439132977e-06, + "loss": 0.3677, + "step": 5637 + }, + { + "epoch": 0.3856624940146385, + "grad_norm": 1.9456735911613992, + "learning_rate": 7.034020537959261e-06, + "loss": 0.5604, + "step": 5638 + }, + { + "epoch": 0.38573089814624806, + "grad_norm": 1.117306666254258, + "learning_rate": 7.033008536945994e-06, + "loss": 0.1758, + "step": 5639 + }, + { + "epoch": 0.3857993022778576, + "grad_norm": 1.9831007372949554, + "learning_rate": 7.03199643614285e-06, + "loss": 0.3735, + "step": 5640 + }, + { + "epoch": 0.38586770640946716, + "grad_norm": 2.0389773654338574, + "learning_rate": 7.030984235599509e-06, + "loss": 0.4521, + "step": 5641 + }, + { + "epoch": 0.3859361105410767, + "grad_norm": 2.3119544726075123, + "learning_rate": 7.029971935365655e-06, + "loss": 0.2778, + "step": 5642 + }, + { + "epoch": 0.3860045146726862, + "grad_norm": 2.6661464225906006, + "learning_rate": 7.028959535490974e-06, + "loss": 0.3177, + "step": 5643 + }, + { + "epoch": 0.3860729188042958, + "grad_norm": 1.632431989804141, + "learning_rate": 7.027947036025164e-06, + "loss": 0.3179, + "step": 5644 + }, + { + "epoch": 0.3861413229359053, + "grad_norm": 2.0190152960913945, + "learning_rate": 7.026934437017918e-06, + "loss": 0.478, + "step": 5645 + }, + { + "epoch": 0.3862097270675149, + "grad_norm": 1.7050544064195003, + "learning_rate": 7.025921738518943e-06, + "loss": 0.353, + "step": 5646 + }, + { + "epoch": 0.3862781311991244, + "grad_norm": 1.9784740959546312, + "learning_rate": 7.024908940577946e-06, + "loss": 0.5544, + "step": 5647 + }, + { + "epoch": 0.386346535330734, + "grad_norm": 2.3446885327897817, + "learning_rate": 7.02389604324464e-06, + "loss": 0.7149, + "step": 5648 + }, + { + "epoch": 0.3864149394623435, + "grad_norm": 1.8232811104819127, + "learning_rate": 7.022883046568743e-06, + "loss": 0.3613, + "step": 5649 + }, + { + "epoch": 0.3864833435939531, + "grad_norm": 2.144971979431719, + "learning_rate": 7.021869950599977e-06, + "loss": 0.4832, + "step": 5650 + }, + { + "epoch": 0.3865517477255626, + "grad_norm": 2.3894137854624984, + "learning_rate": 7.020856755388071e-06, + "loss": 0.6647, + "step": 5651 + }, + { + "epoch": 0.3866201518571722, + "grad_norm": 1.8154437170774227, + "learning_rate": 7.019843460982756e-06, + "loss": 0.2612, + "step": 5652 + }, + { + "epoch": 0.3866885559887817, + "grad_norm": 1.838464693902251, + "learning_rate": 7.018830067433771e-06, + "loss": 0.3835, + "step": 5653 + }, + { + "epoch": 0.3867569601203913, + "grad_norm": 1.5160115378328902, + "learning_rate": 7.017816574790858e-06, + "loss": 0.264, + "step": 5654 + }, + { + "epoch": 0.3868253642520008, + "grad_norm": 1.8164903559747687, + "learning_rate": 7.016802983103763e-06, + "loss": 0.3953, + "step": 5655 + }, + { + "epoch": 0.3868937683836104, + "grad_norm": 1.900527051398184, + "learning_rate": 7.015789292422237e-06, + "loss": 0.4022, + "step": 5656 + }, + { + "epoch": 0.3869621725152199, + "grad_norm": 1.9557043896022883, + "learning_rate": 7.0147755027960395e-06, + "loss": 0.4201, + "step": 5657 + }, + { + "epoch": 0.3870305766468295, + "grad_norm": 1.845772272661505, + "learning_rate": 7.013761614274932e-06, + "loss": 0.5112, + "step": 5658 + }, + { + "epoch": 0.387098980778439, + "grad_norm": 1.831828726369841, + "learning_rate": 7.0127476269086796e-06, + "loss": 0.4163, + "step": 5659 + }, + { + "epoch": 0.38716738491004854, + "grad_norm": 1.5573354005785929, + "learning_rate": 7.011733540747054e-06, + "loss": 0.3241, + "step": 5660 + }, + { + "epoch": 0.3872357890416581, + "grad_norm": 2.593795494461926, + "learning_rate": 7.010719355839831e-06, + "loss": 0.4541, + "step": 5661 + }, + { + "epoch": 0.38730419317326764, + "grad_norm": 1.8541022800310354, + "learning_rate": 7.009705072236795e-06, + "loss": 0.1957, + "step": 5662 + }, + { + "epoch": 0.3873725973048772, + "grad_norm": 1.1949265299441705, + "learning_rate": 7.008690689987726e-06, + "loss": 0.1128, + "step": 5663 + }, + { + "epoch": 0.38744100143648674, + "grad_norm": 1.749808535149567, + "learning_rate": 7.00767620914242e-06, + "loss": 0.2044, + "step": 5664 + }, + { + "epoch": 0.3875094055680963, + "grad_norm": 2.101731137596596, + "learning_rate": 7.006661629750669e-06, + "loss": 0.5676, + "step": 5665 + }, + { + "epoch": 0.38757780969970584, + "grad_norm": 2.3901778656921695, + "learning_rate": 7.005646951862276e-06, + "loss": 0.5714, + "step": 5666 + }, + { + "epoch": 0.3876462138313154, + "grad_norm": 2.313077107149892, + "learning_rate": 7.004632175527046e-06, + "loss": 0.5358, + "step": 5667 + }, + { + "epoch": 0.38771461796292495, + "grad_norm": 1.7731447398324323, + "learning_rate": 7.0036173007947895e-06, + "loss": 0.2822, + "step": 5668 + }, + { + "epoch": 0.3877830220945345, + "grad_norm": 2.781553498328454, + "learning_rate": 7.002602327715318e-06, + "loss": 0.3268, + "step": 5669 + }, + { + "epoch": 0.38785142622614405, + "grad_norm": 1.89549724294869, + "learning_rate": 7.001587256338454e-06, + "loss": 0.3497, + "step": 5670 + }, + { + "epoch": 0.3879198303577536, + "grad_norm": 2.0013193681144643, + "learning_rate": 7.0005720867140235e-06, + "loss": 0.3326, + "step": 5671 + }, + { + "epoch": 0.38798823448936315, + "grad_norm": 2.0322868506871314, + "learning_rate": 6.999556818891855e-06, + "loss": 0.3221, + "step": 5672 + }, + { + "epoch": 0.3880566386209727, + "grad_norm": 1.3411285229525631, + "learning_rate": 6.99854145292178e-06, + "loss": 0.1455, + "step": 5673 + }, + { + "epoch": 0.38812504275258225, + "grad_norm": 2.273614830625621, + "learning_rate": 6.997525988853641e-06, + "loss": 0.5011, + "step": 5674 + }, + { + "epoch": 0.38819344688419183, + "grad_norm": 1.8808023557011955, + "learning_rate": 6.99651042673728e-06, + "loss": 0.5002, + "step": 5675 + }, + { + "epoch": 0.38826185101580135, + "grad_norm": 1.9389982045324197, + "learning_rate": 6.995494766622545e-06, + "loss": 0.4726, + "step": 5676 + }, + { + "epoch": 0.38833025514741093, + "grad_norm": 1.892489333771514, + "learning_rate": 6.994479008559293e-06, + "loss": 0.5194, + "step": 5677 + }, + { + "epoch": 0.38839865927902045, + "grad_norm": 2.0974127203658828, + "learning_rate": 6.9934631525973795e-06, + "loss": 0.5139, + "step": 5678 + }, + { + "epoch": 0.38846706341063, + "grad_norm": 1.9079157164799638, + "learning_rate": 6.992447198786666e-06, + "loss": 0.503, + "step": 5679 + }, + { + "epoch": 0.38853546754223955, + "grad_norm": 2.2721906500427127, + "learning_rate": 6.991431147177023e-06, + "loss": 0.6475, + "step": 5680 + }, + { + "epoch": 0.3886038716738491, + "grad_norm": 1.8582922884623594, + "learning_rate": 6.990414997818325e-06, + "loss": 0.2746, + "step": 5681 + }, + { + "epoch": 0.38867227580545866, + "grad_norm": 2.023675087195513, + "learning_rate": 6.989398750760447e-06, + "loss": 0.4428, + "step": 5682 + }, + { + "epoch": 0.3887406799370682, + "grad_norm": 1.6204838812013758, + "learning_rate": 6.9883824060532705e-06, + "loss": 0.3104, + "step": 5683 + }, + { + "epoch": 0.38880908406867776, + "grad_norm": 1.7455890830811482, + "learning_rate": 6.987365963746684e-06, + "loss": 0.3834, + "step": 5684 + }, + { + "epoch": 0.3888774882002873, + "grad_norm": 1.6588844993209522, + "learning_rate": 6.9863494238905795e-06, + "loss": 0.3847, + "step": 5685 + }, + { + "epoch": 0.38894589233189686, + "grad_norm": 1.858136056808746, + "learning_rate": 6.985332786534852e-06, + "loss": 0.2793, + "step": 5686 + }, + { + "epoch": 0.3890142964635064, + "grad_norm": 1.822143984022976, + "learning_rate": 6.984316051729405e-06, + "loss": 0.4447, + "step": 5687 + }, + { + "epoch": 0.38908270059511596, + "grad_norm": 1.6595503655135377, + "learning_rate": 6.983299219524144e-06, + "loss": 0.3121, + "step": 5688 + }, + { + "epoch": 0.3891511047267255, + "grad_norm": 1.93878853158888, + "learning_rate": 6.98228228996898e-06, + "loss": 0.4429, + "step": 5689 + }, + { + "epoch": 0.38921950885833506, + "grad_norm": 1.9648350967611043, + "learning_rate": 6.981265263113829e-06, + "loss": 0.3857, + "step": 5690 + }, + { + "epoch": 0.3892879129899446, + "grad_norm": 1.4242655836891294, + "learning_rate": 6.980248139008612e-06, + "loss": 0.1941, + "step": 5691 + }, + { + "epoch": 0.38935631712155416, + "grad_norm": 1.9107461888141646, + "learning_rate": 6.979230917703254e-06, + "loss": 0.3463, + "step": 5692 + }, + { + "epoch": 0.3894247212531637, + "grad_norm": 1.558507228543588, + "learning_rate": 6.978213599247683e-06, + "loss": 0.3218, + "step": 5693 + }, + { + "epoch": 0.38949312538477326, + "grad_norm": 1.9491289993447665, + "learning_rate": 6.977196183691839e-06, + "loss": 0.3947, + "step": 5694 + }, + { + "epoch": 0.3895615295163828, + "grad_norm": 2.147555548234815, + "learning_rate": 6.976178671085656e-06, + "loss": 0.2309, + "step": 5695 + }, + { + "epoch": 0.38962993364799237, + "grad_norm": 1.9344904506168628, + "learning_rate": 6.975161061479081e-06, + "loss": 0.498, + "step": 5696 + }, + { + "epoch": 0.3896983377796019, + "grad_norm": 2.411195416601472, + "learning_rate": 6.974143354922064e-06, + "loss": 0.6443, + "step": 5697 + }, + { + "epoch": 0.3897667419112114, + "grad_norm": 2.2518410435598386, + "learning_rate": 6.973125551464558e-06, + "loss": 0.2204, + "step": 5698 + }, + { + "epoch": 0.389835146042821, + "grad_norm": 2.7736112029791355, + "learning_rate": 6.972107651156521e-06, + "loss": 0.6214, + "step": 5699 + }, + { + "epoch": 0.3899035501744305, + "grad_norm": 2.0427535437878936, + "learning_rate": 6.9710896540479175e-06, + "loss": 0.4142, + "step": 5700 + }, + { + "epoch": 0.3899719543060401, + "grad_norm": 2.2327753022541055, + "learning_rate": 6.970071560188716e-06, + "loss": 0.2228, + "step": 5701 + }, + { + "epoch": 0.3900403584376496, + "grad_norm": 2.243974199755734, + "learning_rate": 6.969053369628889e-06, + "loss": 0.6341, + "step": 5702 + }, + { + "epoch": 0.3901087625692592, + "grad_norm": 1.7694425244855754, + "learning_rate": 6.968035082418414e-06, + "loss": 0.2427, + "step": 5703 + }, + { + "epoch": 0.3901771667008687, + "grad_norm": 1.6575635928113748, + "learning_rate": 6.967016698607274e-06, + "loss": 0.296, + "step": 5704 + }, + { + "epoch": 0.3902455708324783, + "grad_norm": 1.8380958531730094, + "learning_rate": 6.965998218245456e-06, + "loss": 0.4763, + "step": 5705 + }, + { + "epoch": 0.3903139749640878, + "grad_norm": 2.2266182072680993, + "learning_rate": 6.96497964138295e-06, + "loss": 0.3228, + "step": 5706 + }, + { + "epoch": 0.3903823790956974, + "grad_norm": 1.5228819078526574, + "learning_rate": 6.963960968069756e-06, + "loss": 0.3256, + "step": 5707 + }, + { + "epoch": 0.3904507832273069, + "grad_norm": 1.9355923505843398, + "learning_rate": 6.962942198355874e-06, + "loss": 0.3762, + "step": 5708 + }, + { + "epoch": 0.3905191873589165, + "grad_norm": 2.108176410075332, + "learning_rate": 6.96192333229131e-06, + "loss": 0.476, + "step": 5709 + }, + { + "epoch": 0.390587591490526, + "grad_norm": 2.1641155091827127, + "learning_rate": 6.960904369926073e-06, + "loss": 0.4336, + "step": 5710 + }, + { + "epoch": 0.3906559956221356, + "grad_norm": 2.0418634907737014, + "learning_rate": 6.959885311310182e-06, + "loss": 0.2445, + "step": 5711 + }, + { + "epoch": 0.3907243997537451, + "grad_norm": 1.7422081343108837, + "learning_rate": 6.958866156493654e-06, + "loss": 0.4653, + "step": 5712 + }, + { + "epoch": 0.3907928038853547, + "grad_norm": 2.318359207886045, + "learning_rate": 6.957846905526517e-06, + "loss": 0.529, + "step": 5713 + }, + { + "epoch": 0.3908612080169642, + "grad_norm": 1.7863312116237977, + "learning_rate": 6.956827558458801e-06, + "loss": 0.4916, + "step": 5714 + }, + { + "epoch": 0.39092961214857375, + "grad_norm": 1.4036562841966453, + "learning_rate": 6.955808115340538e-06, + "loss": 0.2753, + "step": 5715 + }, + { + "epoch": 0.3909980162801833, + "grad_norm": 2.1754924404424574, + "learning_rate": 6.954788576221767e-06, + "loss": 0.5413, + "step": 5716 + }, + { + "epoch": 0.39106642041179285, + "grad_norm": 1.8704572602469551, + "learning_rate": 6.953768941152535e-06, + "loss": 0.5218, + "step": 5717 + }, + { + "epoch": 0.3911348245434024, + "grad_norm": 1.810003668273586, + "learning_rate": 6.952749210182888e-06, + "loss": 0.4109, + "step": 5718 + }, + { + "epoch": 0.39120322867501195, + "grad_norm": 2.498024598142362, + "learning_rate": 6.9517293833628785e-06, + "loss": 0.5034, + "step": 5719 + }, + { + "epoch": 0.3912716328066215, + "grad_norm": 2.5309512144518593, + "learning_rate": 6.950709460742566e-06, + "loss": 0.4386, + "step": 5720 + }, + { + "epoch": 0.39134003693823105, + "grad_norm": 1.9918733461299527, + "learning_rate": 6.949689442372016e-06, + "loss": 0.4977, + "step": 5721 + }, + { + "epoch": 0.39140844106984063, + "grad_norm": 1.7989246702845687, + "learning_rate": 6.948669328301291e-06, + "loss": 0.4055, + "step": 5722 + }, + { + "epoch": 0.39147684520145015, + "grad_norm": 1.9576494569163536, + "learning_rate": 6.9476491185804665e-06, + "loss": 0.5539, + "step": 5723 + }, + { + "epoch": 0.39154524933305973, + "grad_norm": 2.127956248411293, + "learning_rate": 6.946628813259618e-06, + "loss": 0.6256, + "step": 5724 + }, + { + "epoch": 0.39161365346466925, + "grad_norm": 2.0648928221847864, + "learning_rate": 6.945608412388827e-06, + "loss": 0.5856, + "step": 5725 + }, + { + "epoch": 0.39168205759627883, + "grad_norm": 2.2975963316222026, + "learning_rate": 6.944587916018178e-06, + "loss": 0.5836, + "step": 5726 + }, + { + "epoch": 0.39175046172788835, + "grad_norm": 1.4455947445739215, + "learning_rate": 6.9435673241977665e-06, + "loss": 0.224, + "step": 5727 + }, + { + "epoch": 0.39181886585949793, + "grad_norm": 2.5799151303411922, + "learning_rate": 6.9425466369776854e-06, + "loss": 0.7359, + "step": 5728 + }, + { + "epoch": 0.39188726999110746, + "grad_norm": 1.8880978000781603, + "learning_rate": 6.941525854408033e-06, + "loss": 0.4301, + "step": 5729 + }, + { + "epoch": 0.39195567412271703, + "grad_norm": 1.6815656718387404, + "learning_rate": 6.940504976538916e-06, + "loss": 0.2121, + "step": 5730 + }, + { + "epoch": 0.39202407825432656, + "grad_norm": 1.3131475752742356, + "learning_rate": 6.939484003420447e-06, + "loss": 0.2128, + "step": 5731 + }, + { + "epoch": 0.39209248238593614, + "grad_norm": 1.2667965235255316, + "learning_rate": 6.9384629351027346e-06, + "loss": 0.1746, + "step": 5732 + }, + { + "epoch": 0.39216088651754566, + "grad_norm": 1.833328548248144, + "learning_rate": 6.937441771635902e-06, + "loss": 0.143, + "step": 5733 + }, + { + "epoch": 0.3922292906491552, + "grad_norm": 2.203162970790605, + "learning_rate": 6.936420513070071e-06, + "loss": 0.4883, + "step": 5734 + }, + { + "epoch": 0.39229769478076476, + "grad_norm": 1.5482099340839361, + "learning_rate": 6.9353991594553706e-06, + "loss": 0.3254, + "step": 5735 + }, + { + "epoch": 0.3923660989123743, + "grad_norm": 1.9510017832562632, + "learning_rate": 6.934377710841933e-06, + "loss": 0.447, + "step": 5736 + }, + { + "epoch": 0.39243450304398386, + "grad_norm": 1.8534781368793771, + "learning_rate": 6.9333561672798964e-06, + "loss": 0.3477, + "step": 5737 + }, + { + "epoch": 0.3925029071755934, + "grad_norm": 2.095425910070905, + "learning_rate": 6.932334528819404e-06, + "loss": 0.5044, + "step": 5738 + }, + { + "epoch": 0.39257131130720296, + "grad_norm": 1.7438592245903088, + "learning_rate": 6.931312795510601e-06, + "loss": 0.2356, + "step": 5739 + }, + { + "epoch": 0.3926397154388125, + "grad_norm": 1.1901151501303868, + "learning_rate": 6.930290967403639e-06, + "loss": 0.1019, + "step": 5740 + }, + { + "epoch": 0.39270811957042207, + "grad_norm": 2.521535310001785, + "learning_rate": 6.929269044548676e-06, + "loss": 0.4803, + "step": 5741 + }, + { + "epoch": 0.3927765237020316, + "grad_norm": 2.432239026814474, + "learning_rate": 6.928247026995871e-06, + "loss": 0.4427, + "step": 5742 + }, + { + "epoch": 0.39284492783364117, + "grad_norm": 2.4997231687816295, + "learning_rate": 6.927224914795392e-06, + "loss": 0.2221, + "step": 5743 + }, + { + "epoch": 0.3929133319652507, + "grad_norm": 1.552015058778063, + "learning_rate": 6.926202707997407e-06, + "loss": 0.217, + "step": 5744 + }, + { + "epoch": 0.39298173609686027, + "grad_norm": 1.6438218441916586, + "learning_rate": 6.925180406652092e-06, + "loss": 0.3498, + "step": 5745 + }, + { + "epoch": 0.3930501402284698, + "grad_norm": 1.686206427789562, + "learning_rate": 6.924158010809625e-06, + "loss": 0.2967, + "step": 5746 + }, + { + "epoch": 0.39311854436007937, + "grad_norm": 2.041441260678866, + "learning_rate": 6.9231355205201925e-06, + "loss": 0.5015, + "step": 5747 + }, + { + "epoch": 0.3931869484916889, + "grad_norm": 1.418976766995096, + "learning_rate": 6.922112935833981e-06, + "loss": 0.1686, + "step": 5748 + }, + { + "epoch": 0.39325535262329847, + "grad_norm": 1.6794733665060306, + "learning_rate": 6.921090256801184e-06, + "loss": 0.2831, + "step": 5749 + }, + { + "epoch": 0.393323756754908, + "grad_norm": 1.7371650712606355, + "learning_rate": 6.920067483472001e-06, + "loss": 0.3375, + "step": 5750 + }, + { + "epoch": 0.3933921608865176, + "grad_norm": 2.005968614629023, + "learning_rate": 6.919044615896635e-06, + "loss": 0.3805, + "step": 5751 + }, + { + "epoch": 0.3934605650181271, + "grad_norm": 2.5042702325508377, + "learning_rate": 6.918021654125291e-06, + "loss": 0.7502, + "step": 5752 + }, + { + "epoch": 0.3935289691497366, + "grad_norm": 1.491173364631843, + "learning_rate": 6.916998598208185e-06, + "loss": 0.2167, + "step": 5753 + }, + { + "epoch": 0.3935973732813462, + "grad_norm": 1.765714887845926, + "learning_rate": 6.915975448195528e-06, + "loss": 0.4214, + "step": 5754 + }, + { + "epoch": 0.3936657774129557, + "grad_norm": 2.051260334549092, + "learning_rate": 6.9149522041375435e-06, + "loss": 0.4366, + "step": 5755 + }, + { + "epoch": 0.3937341815445653, + "grad_norm": 2.0834806561037795, + "learning_rate": 6.913928866084458e-06, + "loss": 0.5169, + "step": 5756 + }, + { + "epoch": 0.3938025856761748, + "grad_norm": 2.1058319960370917, + "learning_rate": 6.912905434086503e-06, + "loss": 0.2774, + "step": 5757 + }, + { + "epoch": 0.3938709898077844, + "grad_norm": 2.622705353568954, + "learning_rate": 6.91188190819391e-06, + "loss": 0.6906, + "step": 5758 + }, + { + "epoch": 0.3939393939393939, + "grad_norm": 1.0274593592818233, + "learning_rate": 6.9108582884569206e-06, + "loss": 0.1357, + "step": 5759 + }, + { + "epoch": 0.3940077980710035, + "grad_norm": 1.8080388035487134, + "learning_rate": 6.909834574925781e-06, + "loss": 0.2754, + "step": 5760 + }, + { + "epoch": 0.394076202202613, + "grad_norm": 1.9832275125395225, + "learning_rate": 6.908810767650738e-06, + "loss": 0.4578, + "step": 5761 + }, + { + "epoch": 0.3941446063342226, + "grad_norm": 1.510610035872295, + "learning_rate": 6.9077868666820425e-06, + "loss": 0.2464, + "step": 5762 + }, + { + "epoch": 0.3942130104658321, + "grad_norm": 2.172462128447954, + "learning_rate": 6.906762872069956e-06, + "loss": 0.2286, + "step": 5763 + }, + { + "epoch": 0.3942814145974417, + "grad_norm": 2.2536596497183523, + "learning_rate": 6.9057387838647406e-06, + "loss": 0.4671, + "step": 5764 + }, + { + "epoch": 0.3943498187290512, + "grad_norm": 1.2598266054520004, + "learning_rate": 6.904714602116662e-06, + "loss": 0.2299, + "step": 5765 + }, + { + "epoch": 0.3944182228606608, + "grad_norm": 2.027203233439355, + "learning_rate": 6.903690326875993e-06, + "loss": 0.3517, + "step": 5766 + }, + { + "epoch": 0.39448662699227033, + "grad_norm": 2.9224161881045214, + "learning_rate": 6.902665958193012e-06, + "loss": 0.6856, + "step": 5767 + }, + { + "epoch": 0.3945550311238799, + "grad_norm": 1.5090419668434725, + "learning_rate": 6.901641496117997e-06, + "loss": 0.1846, + "step": 5768 + }, + { + "epoch": 0.39462343525548943, + "grad_norm": 1.5628745059378937, + "learning_rate": 6.900616940701233e-06, + "loss": 0.1843, + "step": 5769 + }, + { + "epoch": 0.394691839387099, + "grad_norm": 2.220373298623322, + "learning_rate": 6.899592291993012e-06, + "loss": 0.6271, + "step": 5770 + }, + { + "epoch": 0.39476024351870853, + "grad_norm": 1.5424836373968325, + "learning_rate": 6.89856755004363e-06, + "loss": 0.2209, + "step": 5771 + }, + { + "epoch": 0.39482864765031805, + "grad_norm": 2.153113090941737, + "learning_rate": 6.897542714903383e-06, + "loss": 0.4096, + "step": 5772 + }, + { + "epoch": 0.39489705178192763, + "grad_norm": 1.7571504066537154, + "learning_rate": 6.8965177866225776e-06, + "loss": 0.4781, + "step": 5773 + }, + { + "epoch": 0.39496545591353716, + "grad_norm": 2.349830030447779, + "learning_rate": 6.895492765251519e-06, + "loss": 0.2695, + "step": 5774 + }, + { + "epoch": 0.39503386004514673, + "grad_norm": 2.150295757865509, + "learning_rate": 6.894467650840523e-06, + "loss": 0.4762, + "step": 5775 + }, + { + "epoch": 0.39510226417675626, + "grad_norm": 2.250577051287032, + "learning_rate": 6.893442443439907e-06, + "loss": 0.6986, + "step": 5776 + }, + { + "epoch": 0.39517066830836584, + "grad_norm": 1.6378007557998695, + "learning_rate": 6.892417143099993e-06, + "loss": 0.2482, + "step": 5777 + }, + { + "epoch": 0.39523907243997536, + "grad_norm": 1.9126069860204036, + "learning_rate": 6.891391749871106e-06, + "loss": 0.3884, + "step": 5778 + }, + { + "epoch": 0.39530747657158494, + "grad_norm": 1.7316258525153734, + "learning_rate": 6.890366263803579e-06, + "loss": 0.2715, + "step": 5779 + }, + { + "epoch": 0.39537588070319446, + "grad_norm": 2.4248293269699595, + "learning_rate": 6.889340684947748e-06, + "loss": 0.4424, + "step": 5780 + }, + { + "epoch": 0.39544428483480404, + "grad_norm": 2.2525750895187238, + "learning_rate": 6.888315013353952e-06, + "loss": 0.568, + "step": 5781 + }, + { + "epoch": 0.39551268896641356, + "grad_norm": 2.1124970925610187, + "learning_rate": 6.887289249072536e-06, + "loss": 0.5315, + "step": 5782 + }, + { + "epoch": 0.39558109309802314, + "grad_norm": 1.7991712079411215, + "learning_rate": 6.8862633921538515e-06, + "loss": 0.4274, + "step": 5783 + }, + { + "epoch": 0.39564949722963266, + "grad_norm": 2.2592133792671154, + "learning_rate": 6.885237442648249e-06, + "loss": 0.5011, + "step": 5784 + }, + { + "epoch": 0.39571790136124224, + "grad_norm": 2.2396844323574903, + "learning_rate": 6.88421140060609e-06, + "loss": 0.6171, + "step": 5785 + }, + { + "epoch": 0.39578630549285176, + "grad_norm": 1.5039191753277892, + "learning_rate": 6.8831852660777374e-06, + "loss": 0.2345, + "step": 5786 + }, + { + "epoch": 0.39585470962446134, + "grad_norm": 2.2152343889914836, + "learning_rate": 6.8821590391135584e-06, + "loss": 0.6167, + "step": 5787 + }, + { + "epoch": 0.39592311375607087, + "grad_norm": 1.9831566862838386, + "learning_rate": 6.881132719763924e-06, + "loss": 0.5262, + "step": 5788 + }, + { + "epoch": 0.3959915178876804, + "grad_norm": 2.0370726102033205, + "learning_rate": 6.880106308079211e-06, + "loss": 0.2197, + "step": 5789 + }, + { + "epoch": 0.39605992201928997, + "grad_norm": 1.4736548157320208, + "learning_rate": 6.8790798041098025e-06, + "loss": 0.1913, + "step": 5790 + }, + { + "epoch": 0.3961283261508995, + "grad_norm": 1.6113842463112285, + "learning_rate": 6.8780532079060835e-06, + "loss": 0.219, + "step": 5791 + }, + { + "epoch": 0.39619673028250907, + "grad_norm": 1.6690003557319102, + "learning_rate": 6.877026519518444e-06, + "loss": 0.3539, + "step": 5792 + }, + { + "epoch": 0.3962651344141186, + "grad_norm": 1.745589663560714, + "learning_rate": 6.875999738997278e-06, + "loss": 0.3429, + "step": 5793 + }, + { + "epoch": 0.39633353854572817, + "grad_norm": 2.188251601252912, + "learning_rate": 6.874972866392986e-06, + "loss": 0.3881, + "step": 5794 + }, + { + "epoch": 0.3964019426773377, + "grad_norm": 2.3451678628562185, + "learning_rate": 6.873945901755972e-06, + "loss": 0.5653, + "step": 5795 + }, + { + "epoch": 0.39647034680894727, + "grad_norm": 1.7979946297118765, + "learning_rate": 6.8729188451366446e-06, + "loss": 0.3801, + "step": 5796 + }, + { + "epoch": 0.3965387509405568, + "grad_norm": 2.5885547635049653, + "learning_rate": 6.8718916965854156e-06, + "loss": 0.5682, + "step": 5797 + }, + { + "epoch": 0.3966071550721664, + "grad_norm": 2.6327425157884177, + "learning_rate": 6.870864456152702e-06, + "loss": 0.4554, + "step": 5798 + }, + { + "epoch": 0.3966755592037759, + "grad_norm": 2.2656459083907774, + "learning_rate": 6.869837123888926e-06, + "loss": 0.2723, + "step": 5799 + }, + { + "epoch": 0.3967439633353855, + "grad_norm": 2.1357511564676055, + "learning_rate": 6.868809699844516e-06, + "loss": 0.4929, + "step": 5800 + }, + { + "epoch": 0.396812367466995, + "grad_norm": 1.6377930529549307, + "learning_rate": 6.8677821840699025e-06, + "loss": 0.2699, + "step": 5801 + }, + { + "epoch": 0.3968807715986046, + "grad_norm": 1.9802627381087363, + "learning_rate": 6.866754576615517e-06, + "loss": 0.4832, + "step": 5802 + }, + { + "epoch": 0.3969491757302141, + "grad_norm": 1.825786819253637, + "learning_rate": 6.8657268775318056e-06, + "loss": 0.5789, + "step": 5803 + }, + { + "epoch": 0.3970175798618237, + "grad_norm": 2.1838516328740694, + "learning_rate": 6.864699086869207e-06, + "loss": 0.691, + "step": 5804 + }, + { + "epoch": 0.3970859839934332, + "grad_norm": 1.5865339674947128, + "learning_rate": 6.863671204678174e-06, + "loss": 0.2395, + "step": 5805 + }, + { + "epoch": 0.3971543881250428, + "grad_norm": 1.6744050085654385, + "learning_rate": 6.862643231009158e-06, + "loss": 0.3212, + "step": 5806 + }, + { + "epoch": 0.3972227922566523, + "grad_norm": 1.7925711670917368, + "learning_rate": 6.861615165912618e-06, + "loss": 0.3979, + "step": 5807 + }, + { + "epoch": 0.3972911963882618, + "grad_norm": 1.6834771100417842, + "learning_rate": 6.860587009439017e-06, + "loss": 0.3068, + "step": 5808 + }, + { + "epoch": 0.3973596005198714, + "grad_norm": 2.1546056704235346, + "learning_rate": 6.85955876163882e-06, + "loss": 0.4285, + "step": 5809 + }, + { + "epoch": 0.3974280046514809, + "grad_norm": 1.9361737926821372, + "learning_rate": 6.858530422562499e-06, + "loss": 0.3463, + "step": 5810 + }, + { + "epoch": 0.3974964087830905, + "grad_norm": 2.4443645528961695, + "learning_rate": 6.857501992260529e-06, + "loss": 0.3983, + "step": 5811 + }, + { + "epoch": 0.3975648129147, + "grad_norm": 1.3843619023084506, + "learning_rate": 6.856473470783394e-06, + "loss": 0.1799, + "step": 5812 + }, + { + "epoch": 0.3976332170463096, + "grad_norm": 1.8271540728858804, + "learning_rate": 6.8554448581815746e-06, + "loss": 0.2829, + "step": 5813 + }, + { + "epoch": 0.39770162117791913, + "grad_norm": 1.6740174559229417, + "learning_rate": 6.854416154505562e-06, + "loss": 0.3184, + "step": 5814 + }, + { + "epoch": 0.3977700253095287, + "grad_norm": 1.3073022931896812, + "learning_rate": 6.85338735980585e-06, + "loss": 0.2252, + "step": 5815 + }, + { + "epoch": 0.39783842944113823, + "grad_norm": 1.730867163473387, + "learning_rate": 6.8523584741329365e-06, + "loss": 0.3025, + "step": 5816 + }, + { + "epoch": 0.3979068335727478, + "grad_norm": 2.1240005088533223, + "learning_rate": 6.8513294975373235e-06, + "loss": 0.3365, + "step": 5817 + }, + { + "epoch": 0.39797523770435733, + "grad_norm": 1.832642154468648, + "learning_rate": 6.850300430069519e-06, + "loss": 0.5462, + "step": 5818 + }, + { + "epoch": 0.3980436418359669, + "grad_norm": 2.1739968322655816, + "learning_rate": 6.849271271780034e-06, + "loss": 0.5151, + "step": 5819 + }, + { + "epoch": 0.39811204596757643, + "grad_norm": 1.6059464419216751, + "learning_rate": 6.848242022719386e-06, + "loss": 0.3105, + "step": 5820 + }, + { + "epoch": 0.398180450099186, + "grad_norm": 2.6563253672510223, + "learning_rate": 6.8472126829380936e-06, + "loss": 0.4265, + "step": 5821 + }, + { + "epoch": 0.39824885423079553, + "grad_norm": 1.2091994487150202, + "learning_rate": 6.8461832524866835e-06, + "loss": 0.1573, + "step": 5822 + }, + { + "epoch": 0.3983172583624051, + "grad_norm": 2.685595197763047, + "learning_rate": 6.845153731415683e-06, + "loss": 0.3545, + "step": 5823 + }, + { + "epoch": 0.39838566249401464, + "grad_norm": 1.5346065197013923, + "learning_rate": 6.844124119775628e-06, + "loss": 0.2554, + "step": 5824 + }, + { + "epoch": 0.3984540666256242, + "grad_norm": 2.3339522100171526, + "learning_rate": 6.843094417617055e-06, + "loss": 0.7569, + "step": 5825 + }, + { + "epoch": 0.39852247075723374, + "grad_norm": 2.1320792254637664, + "learning_rate": 6.84206462499051e-06, + "loss": 0.6124, + "step": 5826 + }, + { + "epoch": 0.39859087488884326, + "grad_norm": 2.5487820208241945, + "learning_rate": 6.841034741946537e-06, + "loss": 0.6449, + "step": 5827 + }, + { + "epoch": 0.39865927902045284, + "grad_norm": 1.5230903132061437, + "learning_rate": 6.8400047685356874e-06, + "loss": 0.2454, + "step": 5828 + }, + { + "epoch": 0.39872768315206236, + "grad_norm": 1.6199524504253038, + "learning_rate": 6.838974704808519e-06, + "loss": 0.1134, + "step": 5829 + }, + { + "epoch": 0.39879608728367194, + "grad_norm": 3.064820228114944, + "learning_rate": 6.837944550815591e-06, + "loss": 0.4627, + "step": 5830 + }, + { + "epoch": 0.39886449141528146, + "grad_norm": 1.7073778328749534, + "learning_rate": 6.836914306607471e-06, + "loss": 0.3242, + "step": 5831 + }, + { + "epoch": 0.39893289554689104, + "grad_norm": 2.1801703338725864, + "learning_rate": 6.8358839722347255e-06, + "loss": 0.2769, + "step": 5832 + }, + { + "epoch": 0.39900129967850057, + "grad_norm": 2.6438593384157705, + "learning_rate": 6.834853547747931e-06, + "loss": 0.786, + "step": 5833 + }, + { + "epoch": 0.39906970381011014, + "grad_norm": 2.2257602517934045, + "learning_rate": 6.833823033197662e-06, + "loss": 0.1863, + "step": 5834 + }, + { + "epoch": 0.39913810794171967, + "grad_norm": 2.205949288561128, + "learning_rate": 6.832792428634502e-06, + "loss": 0.609, + "step": 5835 + }, + { + "epoch": 0.39920651207332924, + "grad_norm": 2.310389206215449, + "learning_rate": 6.831761734109042e-06, + "loss": 0.4259, + "step": 5836 + }, + { + "epoch": 0.39927491620493877, + "grad_norm": 1.4733576078387634, + "learning_rate": 6.830730949671869e-06, + "loss": 0.3308, + "step": 5837 + }, + { + "epoch": 0.39934332033654835, + "grad_norm": 1.3487385370688374, + "learning_rate": 6.82970007537358e-06, + "loss": 0.1987, + "step": 5838 + }, + { + "epoch": 0.39941172446815787, + "grad_norm": 1.810291403942701, + "learning_rate": 6.828669111264776e-06, + "loss": 0.1745, + "step": 5839 + }, + { + "epoch": 0.39948012859976745, + "grad_norm": 2.3124501764239804, + "learning_rate": 6.827638057396062e-06, + "loss": 0.3894, + "step": 5840 + }, + { + "epoch": 0.39954853273137697, + "grad_norm": 1.6513964776183006, + "learning_rate": 6.826606913818046e-06, + "loss": 0.3017, + "step": 5841 + }, + { + "epoch": 0.39961693686298655, + "grad_norm": 1.4331541855702108, + "learning_rate": 6.8255756805813435e-06, + "loss": 0.2678, + "step": 5842 + }, + { + "epoch": 0.39968534099459607, + "grad_norm": 1.9171677138373142, + "learning_rate": 6.82454435773657e-06, + "loss": 0.3797, + "step": 5843 + }, + { + "epoch": 0.3997537451262056, + "grad_norm": 1.9060643430102338, + "learning_rate": 6.823512945334348e-06, + "loss": 0.4776, + "step": 5844 + }, + { + "epoch": 0.3998221492578152, + "grad_norm": 2.034411165835941, + "learning_rate": 6.822481443425306e-06, + "loss": 0.1895, + "step": 5845 + }, + { + "epoch": 0.3998905533894247, + "grad_norm": 2.310963919529191, + "learning_rate": 6.821449852060074e-06, + "loss": 0.7285, + "step": 5846 + }, + { + "epoch": 0.3999589575210343, + "grad_norm": 2.1005787230810666, + "learning_rate": 6.820418171289288e-06, + "loss": 0.6468, + "step": 5847 + }, + { + "epoch": 0.4000273616526438, + "grad_norm": 1.9723954585454193, + "learning_rate": 6.819386401163585e-06, + "loss": 0.2707, + "step": 5848 + }, + { + "epoch": 0.4000957657842534, + "grad_norm": 2.158783971721307, + "learning_rate": 6.8183545417336135e-06, + "loss": 0.4696, + "step": 5849 + }, + { + "epoch": 0.4001641699158629, + "grad_norm": 1.879246992116411, + "learning_rate": 6.8173225930500216e-06, + "loss": 0.2806, + "step": 5850 + }, + { + "epoch": 0.4002325740474725, + "grad_norm": 1.7287080776168102, + "learning_rate": 6.81629055516346e-06, + "loss": 0.2252, + "step": 5851 + }, + { + "epoch": 0.400300978179082, + "grad_norm": 1.6405516199322805, + "learning_rate": 6.815258428124587e-06, + "loss": 0.2922, + "step": 5852 + }, + { + "epoch": 0.4003693823106916, + "grad_norm": 2.212611003154768, + "learning_rate": 6.814226211984065e-06, + "loss": 0.6766, + "step": 5853 + }, + { + "epoch": 0.4004377864423011, + "grad_norm": 1.8647028320440384, + "learning_rate": 6.813193906792559e-06, + "loss": 0.1924, + "step": 5854 + }, + { + "epoch": 0.4005061905739107, + "grad_norm": 2.3628915512731434, + "learning_rate": 6.81216151260074e-06, + "loss": 0.5029, + "step": 5855 + }, + { + "epoch": 0.4005745947055202, + "grad_norm": 1.901368461024598, + "learning_rate": 6.811129029459285e-06, + "loss": 0.5033, + "step": 5856 + }, + { + "epoch": 0.4006429988371298, + "grad_norm": 1.8101828434472043, + "learning_rate": 6.810096457418869e-06, + "loss": 0.3779, + "step": 5857 + }, + { + "epoch": 0.4007114029687393, + "grad_norm": 1.9857737499950576, + "learning_rate": 6.809063796530178e-06, + "loss": 0.3601, + "step": 5858 + }, + { + "epoch": 0.4007798071003489, + "grad_norm": 2.426618704848047, + "learning_rate": 6.8080310468439015e-06, + "loss": 0.8224, + "step": 5859 + }, + { + "epoch": 0.4008482112319584, + "grad_norm": 1.7090678080219839, + "learning_rate": 6.806998208410731e-06, + "loss": 0.3376, + "step": 5860 + }, + { + "epoch": 0.400916615363568, + "grad_norm": 1.8602808731145493, + "learning_rate": 6.805965281281362e-06, + "loss": 0.1438, + "step": 5861 + }, + { + "epoch": 0.4009850194951775, + "grad_norm": 2.3626262009181014, + "learning_rate": 6.804932265506494e-06, + "loss": 0.4851, + "step": 5862 + }, + { + "epoch": 0.40105342362678703, + "grad_norm": 2.0970788157367863, + "learning_rate": 6.8038991611368365e-06, + "loss": 0.3844, + "step": 5863 + }, + { + "epoch": 0.4011218277583966, + "grad_norm": 1.6721493727604566, + "learning_rate": 6.802865968223096e-06, + "loss": 0.1862, + "step": 5864 + }, + { + "epoch": 0.40119023189000613, + "grad_norm": 2.514770955184364, + "learning_rate": 6.801832686815989e-06, + "loss": 0.4945, + "step": 5865 + }, + { + "epoch": 0.4012586360216157, + "grad_norm": 2.0331748205554545, + "learning_rate": 6.800799316966233e-06, + "loss": 0.4832, + "step": 5866 + }, + { + "epoch": 0.40132704015322523, + "grad_norm": 2.0787163445207857, + "learning_rate": 6.7997658587245496e-06, + "loss": 0.5336, + "step": 5867 + }, + { + "epoch": 0.4013954442848348, + "grad_norm": 2.081190293689989, + "learning_rate": 6.7987323121416655e-06, + "loss": 0.4724, + "step": 5868 + }, + { + "epoch": 0.40146384841644434, + "grad_norm": 2.222676670939355, + "learning_rate": 6.797698677268316e-06, + "loss": 0.7085, + "step": 5869 + }, + { + "epoch": 0.4015322525480539, + "grad_norm": 1.875972352626352, + "learning_rate": 6.796664954155234e-06, + "loss": 0.1792, + "step": 5870 + }, + { + "epoch": 0.40160065667966344, + "grad_norm": 1.875407500490057, + "learning_rate": 6.79563114285316e-06, + "loss": 0.2757, + "step": 5871 + }, + { + "epoch": 0.401669060811273, + "grad_norm": 2.2036919709721268, + "learning_rate": 6.794597243412838e-06, + "loss": 0.5474, + "step": 5872 + }, + { + "epoch": 0.40173746494288254, + "grad_norm": 1.709572125319186, + "learning_rate": 6.793563255885017e-06, + "loss": 0.2885, + "step": 5873 + }, + { + "epoch": 0.4018058690744921, + "grad_norm": 1.305113300350694, + "learning_rate": 6.7925291803204505e-06, + "loss": 0.191, + "step": 5874 + }, + { + "epoch": 0.40187427320610164, + "grad_norm": 1.8859228741980212, + "learning_rate": 6.791495016769896e-06, + "loss": 0.4384, + "step": 5875 + }, + { + "epoch": 0.4019426773377112, + "grad_norm": 2.260402355646868, + "learning_rate": 6.7904607652841155e-06, + "loss": 0.3343, + "step": 5876 + }, + { + "epoch": 0.40201108146932074, + "grad_norm": 1.4753378489768336, + "learning_rate": 6.789426425913874e-06, + "loss": 0.2554, + "step": 5877 + }, + { + "epoch": 0.4020794856009303, + "grad_norm": 2.298681411663742, + "learning_rate": 6.7883919987099424e-06, + "loss": 0.5739, + "step": 5878 + }, + { + "epoch": 0.40214788973253984, + "grad_norm": 1.476569014237178, + "learning_rate": 6.787357483723096e-06, + "loss": 0.1655, + "step": 5879 + }, + { + "epoch": 0.4022162938641494, + "grad_norm": 1.7579363122357547, + "learning_rate": 6.786322881004114e-06, + "loss": 0.3318, + "step": 5880 + }, + { + "epoch": 0.40228469799575894, + "grad_norm": 1.735747962376309, + "learning_rate": 6.785288190603777e-06, + "loss": 0.3858, + "step": 5881 + }, + { + "epoch": 0.40235310212736847, + "grad_norm": 1.7335244877075644, + "learning_rate": 6.784253412572875e-06, + "loss": 0.4783, + "step": 5882 + }, + { + "epoch": 0.40242150625897805, + "grad_norm": 2.0120182332936616, + "learning_rate": 6.7832185469622005e-06, + "loss": 0.364, + "step": 5883 + }, + { + "epoch": 0.40248991039058757, + "grad_norm": 2.6271305974408907, + "learning_rate": 6.782183593822547e-06, + "loss": 0.4662, + "step": 5884 + }, + { + "epoch": 0.40255831452219715, + "grad_norm": 1.6772631315340456, + "learning_rate": 6.781148553204718e-06, + "loss": 0.373, + "step": 5885 + }, + { + "epoch": 0.40262671865380667, + "grad_norm": 1.999255481366143, + "learning_rate": 6.780113425159517e-06, + "loss": 0.4736, + "step": 5886 + }, + { + "epoch": 0.40269512278541625, + "grad_norm": 1.5768502298935358, + "learning_rate": 6.779078209737753e-06, + "loss": 0.3063, + "step": 5887 + }, + { + "epoch": 0.40276352691702577, + "grad_norm": 2.7953495682059857, + "learning_rate": 6.778042906990238e-06, + "loss": 0.1859, + "step": 5888 + }, + { + "epoch": 0.40283193104863535, + "grad_norm": 2.0871695690981173, + "learning_rate": 6.777007516967792e-06, + "loss": 0.4721, + "step": 5889 + }, + { + "epoch": 0.4029003351802449, + "grad_norm": 1.9516910438870683, + "learning_rate": 6.775972039721237e-06, + "loss": 0.4072, + "step": 5890 + }, + { + "epoch": 0.40296873931185445, + "grad_norm": 2.0151522362908585, + "learning_rate": 6.774936475301396e-06, + "loss": 0.3656, + "step": 5891 + }, + { + "epoch": 0.403037143443464, + "grad_norm": 1.4821654840628218, + "learning_rate": 6.773900823759103e-06, + "loss": 0.2123, + "step": 5892 + }, + { + "epoch": 0.40310554757507355, + "grad_norm": 1.9003959138175237, + "learning_rate": 6.772865085145192e-06, + "loss": 0.3453, + "step": 5893 + }, + { + "epoch": 0.4031739517066831, + "grad_norm": 2.1505556935387213, + "learning_rate": 6.7718292595105e-06, + "loss": 0.4633, + "step": 5894 + }, + { + "epoch": 0.40324235583829265, + "grad_norm": 1.5589168185070539, + "learning_rate": 6.770793346905873e-06, + "loss": 0.1411, + "step": 5895 + }, + { + "epoch": 0.4033107599699022, + "grad_norm": 1.8067051602771032, + "learning_rate": 6.769757347382158e-06, + "loss": 0.321, + "step": 5896 + }, + { + "epoch": 0.40337916410151176, + "grad_norm": 2.6437920195033646, + "learning_rate": 6.768721260990205e-06, + "loss": 0.3592, + "step": 5897 + }, + { + "epoch": 0.4034475682331213, + "grad_norm": 1.7201202695751159, + "learning_rate": 6.767685087780871e-06, + "loss": 0.217, + "step": 5898 + }, + { + "epoch": 0.40351597236473086, + "grad_norm": 2.4996145845392603, + "learning_rate": 6.766648827805019e-06, + "loss": 0.3678, + "step": 5899 + }, + { + "epoch": 0.4035843764963404, + "grad_norm": 2.496707130489602, + "learning_rate": 6.765612481113511e-06, + "loss": 0.2344, + "step": 5900 + }, + { + "epoch": 0.4036527806279499, + "grad_norm": 2.214920432833793, + "learning_rate": 6.764576047757214e-06, + "loss": 0.4011, + "step": 5901 + }, + { + "epoch": 0.4037211847595595, + "grad_norm": 2.2569148777386294, + "learning_rate": 6.763539527787007e-06, + "loss": 0.5023, + "step": 5902 + }, + { + "epoch": 0.403789588891169, + "grad_norm": 2.40026918481185, + "learning_rate": 6.762502921253761e-06, + "loss": 0.7466, + "step": 5903 + }, + { + "epoch": 0.4038579930227786, + "grad_norm": 1.638658680368255, + "learning_rate": 6.761466228208361e-06, + "loss": 0.3791, + "step": 5904 + }, + { + "epoch": 0.4039263971543881, + "grad_norm": 2.319084279110124, + "learning_rate": 6.760429448701694e-06, + "loss": 0.274, + "step": 5905 + }, + { + "epoch": 0.4039948012859977, + "grad_norm": 2.10475800997203, + "learning_rate": 6.7593925827846486e-06, + "loss": 0.5676, + "step": 5906 + }, + { + "epoch": 0.4040632054176072, + "grad_norm": 2.1128474105757213, + "learning_rate": 6.758355630508118e-06, + "loss": 0.4002, + "step": 5907 + }, + { + "epoch": 0.4041316095492168, + "grad_norm": 2.982785345786194, + "learning_rate": 6.7573185919230015e-06, + "loss": 0.43, + "step": 5908 + }, + { + "epoch": 0.4042000136808263, + "grad_norm": 1.8677736383869015, + "learning_rate": 6.7562814670802045e-06, + "loss": 0.3455, + "step": 5909 + }, + { + "epoch": 0.4042684178124359, + "grad_norm": 1.2811912081185737, + "learning_rate": 6.755244256030631e-06, + "loss": 0.2667, + "step": 5910 + }, + { + "epoch": 0.4043368219440454, + "grad_norm": 2.1639088488336014, + "learning_rate": 6.754206958825194e-06, + "loss": 0.1595, + "step": 5911 + }, + { + "epoch": 0.404405226075655, + "grad_norm": 2.337546008468138, + "learning_rate": 6.753169575514807e-06, + "loss": 0.5484, + "step": 5912 + }, + { + "epoch": 0.4044736302072645, + "grad_norm": 2.388198013277731, + "learning_rate": 6.752132106150393e-06, + "loss": 0.7294, + "step": 5913 + }, + { + "epoch": 0.4045420343388741, + "grad_norm": 2.0634869179418414, + "learning_rate": 6.751094550782873e-06, + "loss": 0.3726, + "step": 5914 + }, + { + "epoch": 0.4046104384704836, + "grad_norm": 2.05507343714429, + "learning_rate": 6.750056909463177e-06, + "loss": 0.3995, + "step": 5915 + }, + { + "epoch": 0.4046788426020932, + "grad_norm": 1.4205166621714689, + "learning_rate": 6.749019182242238e-06, + "loss": 0.2226, + "step": 5916 + }, + { + "epoch": 0.4047472467337027, + "grad_norm": 2.0714642038357747, + "learning_rate": 6.74798136917099e-06, + "loss": 0.5022, + "step": 5917 + }, + { + "epoch": 0.40481565086531224, + "grad_norm": 2.2517794023455364, + "learning_rate": 6.746943470300376e-06, + "loss": 0.5615, + "step": 5918 + }, + { + "epoch": 0.4048840549969218, + "grad_norm": 2.2385417926398365, + "learning_rate": 6.745905485681341e-06, + "loss": 0.5147, + "step": 5919 + }, + { + "epoch": 0.40495245912853134, + "grad_norm": 5.875050016961344, + "learning_rate": 6.744867415364833e-06, + "loss": 0.4174, + "step": 5920 + }, + { + "epoch": 0.4050208632601409, + "grad_norm": 2.433351807790892, + "learning_rate": 6.743829259401806e-06, + "loss": 0.7284, + "step": 5921 + }, + { + "epoch": 0.40508926739175044, + "grad_norm": 2.2926721072306653, + "learning_rate": 6.74279101784322e-06, + "loss": 0.4779, + "step": 5922 + }, + { + "epoch": 0.40515767152336, + "grad_norm": 2.5688132345579486, + "learning_rate": 6.741752690740033e-06, + "loss": 0.1998, + "step": 5923 + }, + { + "epoch": 0.40522607565496954, + "grad_norm": 1.650409355283011, + "learning_rate": 6.740714278143215e-06, + "loss": 0.3203, + "step": 5924 + }, + { + "epoch": 0.4052944797865791, + "grad_norm": 1.7442016056929925, + "learning_rate": 6.7396757801037334e-06, + "loss": 0.3268, + "step": 5925 + }, + { + "epoch": 0.40536288391818864, + "grad_norm": 2.3792356441562132, + "learning_rate": 6.738637196672564e-06, + "loss": 0.7553, + "step": 5926 + }, + { + "epoch": 0.4054312880497982, + "grad_norm": 1.510004476231136, + "learning_rate": 6.737598527900685e-06, + "loss": 0.2989, + "step": 5927 + }, + { + "epoch": 0.40549969218140774, + "grad_norm": 2.2433126862862, + "learning_rate": 6.736559773839079e-06, + "loss": 0.5901, + "step": 5928 + }, + { + "epoch": 0.4055680963130173, + "grad_norm": 1.6711806655901542, + "learning_rate": 6.735520934538734e-06, + "loss": 0.3482, + "step": 5929 + }, + { + "epoch": 0.40563650044462685, + "grad_norm": 2.539741384273259, + "learning_rate": 6.73448201005064e-06, + "loss": 0.8254, + "step": 5930 + }, + { + "epoch": 0.4057049045762364, + "grad_norm": 2.1394248856064544, + "learning_rate": 6.733443000425794e-06, + "loss": 0.7004, + "step": 5931 + }, + { + "epoch": 0.40577330870784595, + "grad_norm": 1.7928025910421337, + "learning_rate": 6.732403905715197e-06, + "loss": 0.2812, + "step": 5932 + }, + { + "epoch": 0.4058417128394555, + "grad_norm": 2.047360573750222, + "learning_rate": 6.731364725969847e-06, + "loss": 0.4553, + "step": 5933 + }, + { + "epoch": 0.40591011697106505, + "grad_norm": 2.173709670795563, + "learning_rate": 6.730325461240756e-06, + "loss": 0.5903, + "step": 5934 + }, + { + "epoch": 0.4059785211026746, + "grad_norm": 1.6636389721717852, + "learning_rate": 6.729286111578938e-06, + "loss": 0.3603, + "step": 5935 + }, + { + "epoch": 0.40604692523428415, + "grad_norm": 2.5007998170553694, + "learning_rate": 6.728246677035406e-06, + "loss": 0.4332, + "step": 5936 + }, + { + "epoch": 0.4061153293658937, + "grad_norm": 2.22099110758301, + "learning_rate": 6.7272071576611805e-06, + "loss": 0.4453, + "step": 5937 + }, + { + "epoch": 0.40618373349750325, + "grad_norm": 2.1389909340789837, + "learning_rate": 6.726167553507288e-06, + "loss": 0.4385, + "step": 5938 + }, + { + "epoch": 0.4062521376291128, + "grad_norm": 1.756632807440452, + "learning_rate": 6.725127864624757e-06, + "loss": 0.3159, + "step": 5939 + }, + { + "epoch": 0.40632054176072235, + "grad_norm": 1.9846125670774795, + "learning_rate": 6.724088091064617e-06, + "loss": 0.5028, + "step": 5940 + }, + { + "epoch": 0.4063889458923319, + "grad_norm": 1.4694430359763735, + "learning_rate": 6.723048232877911e-06, + "loss": 0.3047, + "step": 5941 + }, + { + "epoch": 0.40645735002394145, + "grad_norm": 2.120626099460001, + "learning_rate": 6.722008290115677e-06, + "loss": 0.6943, + "step": 5942 + }, + { + "epoch": 0.406525754155551, + "grad_norm": 1.7937530098866588, + "learning_rate": 6.72096826282896e-06, + "loss": 0.3581, + "step": 5943 + }, + { + "epoch": 0.40659415828716056, + "grad_norm": 2.5697823125490906, + "learning_rate": 6.719928151068809e-06, + "loss": 0.8178, + "step": 5944 + }, + { + "epoch": 0.4066625624187701, + "grad_norm": 1.7245948721771622, + "learning_rate": 6.7188879548862805e-06, + "loss": 0.3212, + "step": 5945 + }, + { + "epoch": 0.40673096655037966, + "grad_norm": 1.2894210576851697, + "learning_rate": 6.71784767433243e-06, + "loss": 0.1633, + "step": 5946 + }, + { + "epoch": 0.4067993706819892, + "grad_norm": 1.6036864548638916, + "learning_rate": 6.716807309458321e-06, + "loss": 0.1151, + "step": 5947 + }, + { + "epoch": 0.40686777481359876, + "grad_norm": 2.2042532730454645, + "learning_rate": 6.715766860315017e-06, + "loss": 0.4188, + "step": 5948 + }, + { + "epoch": 0.4069361789452083, + "grad_norm": 1.9413999446098738, + "learning_rate": 6.7147263269535925e-06, + "loss": 0.4426, + "step": 5949 + }, + { + "epoch": 0.40700458307681786, + "grad_norm": 2.100869160121841, + "learning_rate": 6.713685709425118e-06, + "loss": 0.5881, + "step": 5950 + }, + { + "epoch": 0.4070729872084274, + "grad_norm": 1.8481902143307465, + "learning_rate": 6.712645007780674e-06, + "loss": 0.4449, + "step": 5951 + }, + { + "epoch": 0.40714139134003696, + "grad_norm": 1.613348989473125, + "learning_rate": 6.7116042220713436e-06, + "loss": 0.2558, + "step": 5952 + }, + { + "epoch": 0.4072097954716465, + "grad_norm": 2.204455792141919, + "learning_rate": 6.710563352348212e-06, + "loss": 0.4146, + "step": 5953 + }, + { + "epoch": 0.40727819960325606, + "grad_norm": 1.3515386248558046, + "learning_rate": 6.70952239866237e-06, + "loss": 0.1922, + "step": 5954 + }, + { + "epoch": 0.4073466037348656, + "grad_norm": 2.333427164768306, + "learning_rate": 6.708481361064915e-06, + "loss": 0.4171, + "step": 5955 + }, + { + "epoch": 0.4074150078664751, + "grad_norm": 0.9767647877035839, + "learning_rate": 6.707440239606943e-06, + "loss": 0.0707, + "step": 5956 + }, + { + "epoch": 0.4074834119980847, + "grad_norm": 2.0714115968076836, + "learning_rate": 6.70639903433956e-06, + "loss": 0.5336, + "step": 5957 + }, + { + "epoch": 0.4075518161296942, + "grad_norm": 1.9135368131314023, + "learning_rate": 6.705357745313872e-06, + "loss": 0.6077, + "step": 5958 + }, + { + "epoch": 0.4076202202613038, + "grad_norm": 2.135919756281948, + "learning_rate": 6.70431637258099e-06, + "loss": 0.5319, + "step": 5959 + }, + { + "epoch": 0.4076886243929133, + "grad_norm": 2.0055442792863696, + "learning_rate": 6.703274916192031e-06, + "loss": 0.2484, + "step": 5960 + }, + { + "epoch": 0.4077570285245229, + "grad_norm": 1.8706086670786721, + "learning_rate": 6.702233376198115e-06, + "loss": 0.4976, + "step": 5961 + }, + { + "epoch": 0.4078254326561324, + "grad_norm": 2.3644319932599274, + "learning_rate": 6.701191752650365e-06, + "loss": 0.4378, + "step": 5962 + }, + { + "epoch": 0.407893836787742, + "grad_norm": 1.8988352974962392, + "learning_rate": 6.700150045599907e-06, + "loss": 0.4006, + "step": 5963 + }, + { + "epoch": 0.4079622409193515, + "grad_norm": 1.8045845002345513, + "learning_rate": 6.699108255097876e-06, + "loss": 0.3616, + "step": 5964 + }, + { + "epoch": 0.4080306450509611, + "grad_norm": 2.026204261204719, + "learning_rate": 6.698066381195407e-06, + "loss": 0.5464, + "step": 5965 + }, + { + "epoch": 0.4080990491825706, + "grad_norm": 1.4362337895325996, + "learning_rate": 6.6970244239436385e-06, + "loss": 0.1577, + "step": 5966 + }, + { + "epoch": 0.4081674533141802, + "grad_norm": 2.158352904494917, + "learning_rate": 6.695982383393718e-06, + "loss": 0.439, + "step": 5967 + }, + { + "epoch": 0.4082358574457897, + "grad_norm": 1.3841509376146444, + "learning_rate": 6.694940259596793e-06, + "loss": 0.2304, + "step": 5968 + }, + { + "epoch": 0.4083042615773993, + "grad_norm": 1.647106084951954, + "learning_rate": 6.693898052604015e-06, + "loss": 0.2253, + "step": 5969 + }, + { + "epoch": 0.4083726657090088, + "grad_norm": 2.1311671649239368, + "learning_rate": 6.692855762466539e-06, + "loss": 0.2772, + "step": 5970 + }, + { + "epoch": 0.4084410698406184, + "grad_norm": 2.7077625712498326, + "learning_rate": 6.691813389235529e-06, + "loss": 0.4551, + "step": 5971 + }, + { + "epoch": 0.4085094739722279, + "grad_norm": 1.8591057919110354, + "learning_rate": 6.690770932962148e-06, + "loss": 0.4239, + "step": 5972 + }, + { + "epoch": 0.40857787810383744, + "grad_norm": 1.8997870738469016, + "learning_rate": 6.6897283936975635e-06, + "loss": 0.4671, + "step": 5973 + }, + { + "epoch": 0.408646282235447, + "grad_norm": 1.9625053426189956, + "learning_rate": 6.688685771492952e-06, + "loss": 0.3004, + "step": 5974 + }, + { + "epoch": 0.40871468636705655, + "grad_norm": 2.2430356445222683, + "learning_rate": 6.687643066399487e-06, + "loss": 0.2294, + "step": 5975 + }, + { + "epoch": 0.4087830904986661, + "grad_norm": 2.4195282302655303, + "learning_rate": 6.686600278468352e-06, + "loss": 0.6073, + "step": 5976 + }, + { + "epoch": 0.40885149463027565, + "grad_norm": 1.9923031642472622, + "learning_rate": 6.685557407750729e-06, + "loss": 0.4158, + "step": 5977 + }, + { + "epoch": 0.4089198987618852, + "grad_norm": 1.802152642096012, + "learning_rate": 6.68451445429781e-06, + "loss": 0.2201, + "step": 5978 + }, + { + "epoch": 0.40898830289349475, + "grad_norm": 2.5349269463473143, + "learning_rate": 6.683471418160787e-06, + "loss": 0.4269, + "step": 5979 + }, + { + "epoch": 0.4090567070251043, + "grad_norm": 2.110609992857569, + "learning_rate": 6.6824282993908574e-06, + "loss": 0.6743, + "step": 5980 + }, + { + "epoch": 0.40912511115671385, + "grad_norm": 2.4380278260385335, + "learning_rate": 6.681385098039224e-06, + "loss": 0.3287, + "step": 5981 + }, + { + "epoch": 0.40919351528832343, + "grad_norm": 2.2879608224289734, + "learning_rate": 6.680341814157088e-06, + "loss": 0.5187, + "step": 5982 + }, + { + "epoch": 0.40926191941993295, + "grad_norm": 2.0948377032906516, + "learning_rate": 6.679298447795663e-06, + "loss": 0.6118, + "step": 5983 + }, + { + "epoch": 0.40933032355154253, + "grad_norm": 1.831538839800825, + "learning_rate": 6.67825499900616e-06, + "loss": 0.1799, + "step": 5984 + }, + { + "epoch": 0.40939872768315205, + "grad_norm": 1.898567544541076, + "learning_rate": 6.677211467839799e-06, + "loss": 0.4481, + "step": 5985 + }, + { + "epoch": 0.40946713181476163, + "grad_norm": 2.1191089258591846, + "learning_rate": 6.676167854347798e-06, + "loss": 0.347, + "step": 5986 + }, + { + "epoch": 0.40953553594637115, + "grad_norm": 2.1560548300111964, + "learning_rate": 6.675124158581385e-06, + "loss": 0.5881, + "step": 5987 + }, + { + "epoch": 0.40960394007798073, + "grad_norm": 2.1073756668073473, + "learning_rate": 6.674080380591791e-06, + "loss": 0.2508, + "step": 5988 + }, + { + "epoch": 0.40967234420959026, + "grad_norm": 1.8854036269255117, + "learning_rate": 6.673036520430247e-06, + "loss": 0.4216, + "step": 5989 + }, + { + "epoch": 0.40974074834119983, + "grad_norm": 1.3066365394471393, + "learning_rate": 6.67199257814799e-06, + "loss": 0.1293, + "step": 5990 + }, + { + "epoch": 0.40980915247280936, + "grad_norm": 2.3122659642124916, + "learning_rate": 6.670948553796266e-06, + "loss": 0.37, + "step": 5991 + }, + { + "epoch": 0.4098775566044189, + "grad_norm": 2.355349415908256, + "learning_rate": 6.669904447426317e-06, + "loss": 0.5396, + "step": 5992 + }, + { + "epoch": 0.40994596073602846, + "grad_norm": 1.964310738710674, + "learning_rate": 6.668860259089391e-06, + "loss": 0.4274, + "step": 5993 + }, + { + "epoch": 0.410014364867638, + "grad_norm": 2.2311164162822417, + "learning_rate": 6.667815988836747e-06, + "loss": 0.4311, + "step": 5994 + }, + { + "epoch": 0.41008276899924756, + "grad_norm": 2.0656717650043994, + "learning_rate": 6.666771636719642e-06, + "loss": 0.5067, + "step": 5995 + }, + { + "epoch": 0.4101511731308571, + "grad_norm": 1.6630917928147966, + "learning_rate": 6.6657272027893345e-06, + "loss": 0.2325, + "step": 5996 + }, + { + "epoch": 0.41021957726246666, + "grad_norm": 1.5803935345615345, + "learning_rate": 6.664682687097092e-06, + "loss": 0.3116, + "step": 5997 + }, + { + "epoch": 0.4102879813940762, + "grad_norm": 1.7685813312008563, + "learning_rate": 6.6636380896941855e-06, + "loss": 0.3673, + "step": 5998 + }, + { + "epoch": 0.41035638552568576, + "grad_norm": 2.1921818494553995, + "learning_rate": 6.662593410631888e-06, + "loss": 0.6376, + "step": 5999 + }, + { + "epoch": 0.4104247896572953, + "grad_norm": 2.044410532330898, + "learning_rate": 6.661548649961477e-06, + "loss": 0.6124, + "step": 6000 + }, + { + "epoch": 0.41049319378890486, + "grad_norm": 1.8610526218246308, + "learning_rate": 6.660503807734234e-06, + "loss": 0.4625, + "step": 6001 + }, + { + "epoch": 0.4105615979205144, + "grad_norm": 2.100927583159533, + "learning_rate": 6.659458884001447e-06, + "loss": 0.5712, + "step": 6002 + }, + { + "epoch": 0.41063000205212397, + "grad_norm": 1.5849467504292993, + "learning_rate": 6.658413878814403e-06, + "loss": 0.2329, + "step": 6003 + }, + { + "epoch": 0.4106984061837335, + "grad_norm": 1.3125078649512292, + "learning_rate": 6.657368792224398e-06, + "loss": 0.2489, + "step": 6004 + }, + { + "epoch": 0.41076681031534307, + "grad_norm": 1.9228332658151965, + "learning_rate": 6.65632362428273e-06, + "loss": 0.4494, + "step": 6005 + }, + { + "epoch": 0.4108352144469526, + "grad_norm": 2.175821399831926, + "learning_rate": 6.655278375040699e-06, + "loss": 0.3999, + "step": 6006 + }, + { + "epoch": 0.41090361857856217, + "grad_norm": 2.131764912699973, + "learning_rate": 6.6542330445496115e-06, + "loss": 0.3518, + "step": 6007 + }, + { + "epoch": 0.4109720227101717, + "grad_norm": 2.4358288831884414, + "learning_rate": 6.65318763286078e-06, + "loss": 0.7584, + "step": 6008 + }, + { + "epoch": 0.41104042684178127, + "grad_norm": 2.1091533897689896, + "learning_rate": 6.652142140025517e-06, + "loss": 0.466, + "step": 6009 + }, + { + "epoch": 0.4111088309733908, + "grad_norm": 2.44173495711974, + "learning_rate": 6.651096566095137e-06, + "loss": 0.5409, + "step": 6010 + }, + { + "epoch": 0.4111772351050003, + "grad_norm": 2.2484451491309954, + "learning_rate": 6.650050911120967e-06, + "loss": 0.5233, + "step": 6011 + }, + { + "epoch": 0.4112456392366099, + "grad_norm": 1.8087434809685248, + "learning_rate": 6.649005175154329e-06, + "loss": 0.3353, + "step": 6012 + }, + { + "epoch": 0.4113140433682194, + "grad_norm": 1.7678337881752946, + "learning_rate": 6.647959358246552e-06, + "loss": 0.409, + "step": 6013 + }, + { + "epoch": 0.411382447499829, + "grad_norm": 2.1137141914199518, + "learning_rate": 6.646913460448975e-06, + "loss": 0.5801, + "step": 6014 + }, + { + "epoch": 0.4114508516314385, + "grad_norm": 2.0832129876822294, + "learning_rate": 6.645867481812931e-06, + "loss": 0.3696, + "step": 6015 + }, + { + "epoch": 0.4115192557630481, + "grad_norm": 2.2115886794165, + "learning_rate": 6.644821422389763e-06, + "loss": 0.4392, + "step": 6016 + }, + { + "epoch": 0.4115876598946576, + "grad_norm": 2.03588737270919, + "learning_rate": 6.643775282230818e-06, + "loss": 0.3043, + "step": 6017 + }, + { + "epoch": 0.4116560640262672, + "grad_norm": 1.805345072816429, + "learning_rate": 6.642729061387443e-06, + "loss": 0.1754, + "step": 6018 + }, + { + "epoch": 0.4117244681578767, + "grad_norm": 1.6581486849166274, + "learning_rate": 6.641682759910993e-06, + "loss": 0.3162, + "step": 6019 + }, + { + "epoch": 0.4117928722894863, + "grad_norm": 1.9881338803099888, + "learning_rate": 6.640636377852827e-06, + "loss": 0.2552, + "step": 6020 + }, + { + "epoch": 0.4118612764210958, + "grad_norm": 2.0252110798049334, + "learning_rate": 6.6395899152643055e-06, + "loss": 0.5037, + "step": 6021 + }, + { + "epoch": 0.4119296805527054, + "grad_norm": 1.1696379289918453, + "learning_rate": 6.638543372196791e-06, + "loss": 0.234, + "step": 6022 + }, + { + "epoch": 0.4119980846843149, + "grad_norm": 2.137045708002371, + "learning_rate": 6.637496748701656e-06, + "loss": 0.3461, + "step": 6023 + }, + { + "epoch": 0.4120664888159245, + "grad_norm": 1.781549068909506, + "learning_rate": 6.636450044830275e-06, + "loss": 0.3657, + "step": 6024 + }, + { + "epoch": 0.412134892947534, + "grad_norm": 2.3335443888023057, + "learning_rate": 6.635403260634022e-06, + "loss": 0.4824, + "step": 6025 + }, + { + "epoch": 0.4122032970791436, + "grad_norm": 2.286811577933198, + "learning_rate": 6.63435639616428e-06, + "loss": 0.5829, + "step": 6026 + }, + { + "epoch": 0.4122717012107531, + "grad_norm": 2.697117430574893, + "learning_rate": 6.633309451472433e-06, + "loss": 0.6902, + "step": 6027 + }, + { + "epoch": 0.4123401053423627, + "grad_norm": 1.953821656913748, + "learning_rate": 6.632262426609872e-06, + "loss": 0.3766, + "step": 6028 + }, + { + "epoch": 0.41240850947397223, + "grad_norm": 2.5127304080627835, + "learning_rate": 6.631215321627989e-06, + "loss": 0.6933, + "step": 6029 + }, + { + "epoch": 0.41247691360558175, + "grad_norm": 1.803764055396243, + "learning_rate": 6.630168136578181e-06, + "loss": 0.3353, + "step": 6030 + }, + { + "epoch": 0.41254531773719133, + "grad_norm": 2.7772635114322073, + "learning_rate": 6.629120871511848e-06, + "loss": 0.6347, + "step": 6031 + }, + { + "epoch": 0.41261372186880085, + "grad_norm": 2.580206279474046, + "learning_rate": 6.6280735264803964e-06, + "loss": 0.7499, + "step": 6032 + }, + { + "epoch": 0.41268212600041043, + "grad_norm": 2.2881790773888397, + "learning_rate": 6.627026101535233e-06, + "loss": 0.782, + "step": 6033 + }, + { + "epoch": 0.41275053013201995, + "grad_norm": 1.5880812827274189, + "learning_rate": 6.625978596727774e-06, + "loss": 0.1952, + "step": 6034 + }, + { + "epoch": 0.41281893426362953, + "grad_norm": 2.081006773205091, + "learning_rate": 6.6249310121094326e-06, + "loss": 0.4926, + "step": 6035 + }, + { + "epoch": 0.41288733839523906, + "grad_norm": 2.105624283405204, + "learning_rate": 6.62388334773163e-06, + "loss": 0.3484, + "step": 6036 + }, + { + "epoch": 0.41295574252684863, + "grad_norm": 2.0199510739321664, + "learning_rate": 6.62283560364579e-06, + "loss": 0.4284, + "step": 6037 + }, + { + "epoch": 0.41302414665845816, + "grad_norm": 5.078291670191525, + "learning_rate": 6.621787779903343e-06, + "loss": 0.381, + "step": 6038 + }, + { + "epoch": 0.41309255079006774, + "grad_norm": 1.3515504246651897, + "learning_rate": 6.620739876555721e-06, + "loss": 0.1821, + "step": 6039 + }, + { + "epoch": 0.41316095492167726, + "grad_norm": 2.304576741806128, + "learning_rate": 6.61969189365436e-06, + "loss": 0.3724, + "step": 6040 + }, + { + "epoch": 0.41322935905328684, + "grad_norm": 1.7876809712145951, + "learning_rate": 6.618643831250699e-06, + "loss": 0.2659, + "step": 6041 + }, + { + "epoch": 0.41329776318489636, + "grad_norm": 1.4575155082404305, + "learning_rate": 6.61759568939618e-06, + "loss": 0.2273, + "step": 6042 + }, + { + "epoch": 0.41336616731650594, + "grad_norm": 1.405336613192673, + "learning_rate": 6.616547468142255e-06, + "loss": 0.1662, + "step": 6043 + }, + { + "epoch": 0.41343457144811546, + "grad_norm": 1.8003933503255911, + "learning_rate": 6.6154991675403745e-06, + "loss": 0.3044, + "step": 6044 + }, + { + "epoch": 0.41350297557972504, + "grad_norm": 1.5483057865859136, + "learning_rate": 6.614450787641994e-06, + "loss": 0.3356, + "step": 6045 + }, + { + "epoch": 0.41357137971133456, + "grad_norm": 1.8470058506151135, + "learning_rate": 6.613402328498571e-06, + "loss": 0.3763, + "step": 6046 + }, + { + "epoch": 0.4136397838429441, + "grad_norm": 1.5727404144096868, + "learning_rate": 6.612353790161572e-06, + "loss": 0.3188, + "step": 6047 + }, + { + "epoch": 0.41370818797455367, + "grad_norm": 2.105826947205429, + "learning_rate": 6.611305172682463e-06, + "loss": 0.5653, + "step": 6048 + }, + { + "epoch": 0.4137765921061632, + "grad_norm": 2.5779898521373066, + "learning_rate": 6.610256476112714e-06, + "loss": 0.632, + "step": 6049 + }, + { + "epoch": 0.41384499623777277, + "grad_norm": 1.727469184631007, + "learning_rate": 6.609207700503802e-06, + "loss": 0.1817, + "step": 6050 + }, + { + "epoch": 0.4139134003693823, + "grad_norm": 1.9330636771768104, + "learning_rate": 6.608158845907206e-06, + "loss": 0.3222, + "step": 6051 + }, + { + "epoch": 0.41398180450099187, + "grad_norm": 2.769242141117737, + "learning_rate": 6.607109912374407e-06, + "loss": 0.9312, + "step": 6052 + }, + { + "epoch": 0.4140502086326014, + "grad_norm": 1.9814775895770083, + "learning_rate": 6.606060899956893e-06, + "loss": 0.3706, + "step": 6053 + }, + { + "epoch": 0.41411861276421097, + "grad_norm": 2.231193084151159, + "learning_rate": 6.605011808706157e-06, + "loss": 0.4774, + "step": 6054 + }, + { + "epoch": 0.4141870168958205, + "grad_norm": 2.2389349257548057, + "learning_rate": 6.6039626386736885e-06, + "loss": 0.1766, + "step": 6055 + }, + { + "epoch": 0.41425542102743007, + "grad_norm": 2.410349759289053, + "learning_rate": 6.6029133899109874e-06, + "loss": 0.3783, + "step": 6056 + }, + { + "epoch": 0.4143238251590396, + "grad_norm": 2.024730843446437, + "learning_rate": 6.6018640624695575e-06, + "loss": 0.2512, + "step": 6057 + }, + { + "epoch": 0.41439222929064917, + "grad_norm": 1.9571579760791882, + "learning_rate": 6.600814656400904e-06, + "loss": 0.4248, + "step": 6058 + }, + { + "epoch": 0.4144606334222587, + "grad_norm": 2.48461840894669, + "learning_rate": 6.599765171756538e-06, + "loss": 0.5582, + "step": 6059 + }, + { + "epoch": 0.4145290375538683, + "grad_norm": 1.95254514097424, + "learning_rate": 6.598715608587972e-06, + "loss": 0.4256, + "step": 6060 + }, + { + "epoch": 0.4145974416854778, + "grad_norm": 1.8000357940545477, + "learning_rate": 6.597665966946723e-06, + "loss": 0.3232, + "step": 6061 + }, + { + "epoch": 0.4146658458170874, + "grad_norm": 1.9600202687344312, + "learning_rate": 6.596616246884315e-06, + "loss": 0.383, + "step": 6062 + }, + { + "epoch": 0.4147342499486969, + "grad_norm": 2.3281739633006953, + "learning_rate": 6.5955664484522684e-06, + "loss": 0.7069, + "step": 6063 + }, + { + "epoch": 0.4148026540803065, + "grad_norm": 1.9514299514141182, + "learning_rate": 6.5945165717021185e-06, + "loss": 0.5404, + "step": 6064 + }, + { + "epoch": 0.414871058211916, + "grad_norm": 1.8296363484621698, + "learning_rate": 6.593466616685393e-06, + "loss": 0.3887, + "step": 6065 + }, + { + "epoch": 0.4149394623435255, + "grad_norm": 2.5698956496487444, + "learning_rate": 6.592416583453632e-06, + "loss": 0.7013, + "step": 6066 + }, + { + "epoch": 0.4150078664751351, + "grad_norm": 2.1598333148418587, + "learning_rate": 6.591366472058377e-06, + "loss": 0.6286, + "step": 6067 + }, + { + "epoch": 0.4150762706067446, + "grad_norm": 2.975034556869349, + "learning_rate": 6.59031628255117e-06, + "loss": 0.3743, + "step": 6068 + }, + { + "epoch": 0.4151446747383542, + "grad_norm": 2.233387862406899, + "learning_rate": 6.58926601498356e-06, + "loss": 0.3904, + "step": 6069 + }, + { + "epoch": 0.4152130788699637, + "grad_norm": 2.0143642030876943, + "learning_rate": 6.5882156694071e-06, + "loss": 0.5401, + "step": 6070 + }, + { + "epoch": 0.4152814830015733, + "grad_norm": 2.7732412671796167, + "learning_rate": 6.587165245873345e-06, + "loss": 0.4842, + "step": 6071 + }, + { + "epoch": 0.4153498871331828, + "grad_norm": 1.6417630471904407, + "learning_rate": 6.586114744433855e-06, + "loss": 0.3012, + "step": 6072 + }, + { + "epoch": 0.4154182912647924, + "grad_norm": 1.912596417471217, + "learning_rate": 6.585064165140195e-06, + "loss": 0.4495, + "step": 6073 + }, + { + "epoch": 0.41548669539640193, + "grad_norm": 1.9815550088011415, + "learning_rate": 6.584013508043932e-06, + "loss": 0.2249, + "step": 6074 + }, + { + "epoch": 0.4155550995280115, + "grad_norm": 1.9085898617530344, + "learning_rate": 6.582962773196635e-06, + "loss": 0.2347, + "step": 6075 + }, + { + "epoch": 0.41562350365962103, + "grad_norm": 2.139306790242135, + "learning_rate": 6.581911960649882e-06, + "loss": 0.5578, + "step": 6076 + }, + { + "epoch": 0.4156919077912306, + "grad_norm": 1.6897088028263525, + "learning_rate": 6.580861070455251e-06, + "loss": 0.2225, + "step": 6077 + }, + { + "epoch": 0.41576031192284013, + "grad_norm": 2.116424159005547, + "learning_rate": 6.579810102664326e-06, + "loss": 0.4973, + "step": 6078 + }, + { + "epoch": 0.4158287160544497, + "grad_norm": 1.704800953573066, + "learning_rate": 6.578759057328691e-06, + "loss": 0.4421, + "step": 6079 + }, + { + "epoch": 0.41589712018605923, + "grad_norm": 2.7001136367337644, + "learning_rate": 6.577707934499938e-06, + "loss": 1.1, + "step": 6080 + }, + { + "epoch": 0.4159655243176688, + "grad_norm": 1.8101176027692516, + "learning_rate": 6.57665673422966e-06, + "loss": 0.4594, + "step": 6081 + }, + { + "epoch": 0.41603392844927833, + "grad_norm": 2.223383337072907, + "learning_rate": 6.575605456569456e-06, + "loss": 0.202, + "step": 6082 + }, + { + "epoch": 0.4161023325808879, + "grad_norm": 1.8286996329562255, + "learning_rate": 6.574554101570929e-06, + "loss": 0.1279, + "step": 6083 + }, + { + "epoch": 0.41617073671249744, + "grad_norm": 2.216766755491819, + "learning_rate": 6.573502669285684e-06, + "loss": 0.4004, + "step": 6084 + }, + { + "epoch": 0.41623914084410696, + "grad_norm": 1.8942030651919248, + "learning_rate": 6.572451159765327e-06, + "loss": 0.3143, + "step": 6085 + }, + { + "epoch": 0.41630754497571654, + "grad_norm": 1.7180078031048518, + "learning_rate": 6.5713995730614755e-06, + "loss": 0.362, + "step": 6086 + }, + { + "epoch": 0.41637594910732606, + "grad_norm": 1.436140389268764, + "learning_rate": 6.570347909225745e-06, + "loss": 0.2707, + "step": 6087 + }, + { + "epoch": 0.41644435323893564, + "grad_norm": 2.0318944752269386, + "learning_rate": 6.569296168309755e-06, + "loss": 0.4698, + "step": 6088 + }, + { + "epoch": 0.41651275737054516, + "grad_norm": 1.7027233157550268, + "learning_rate": 6.5682443503651316e-06, + "loss": 0.3329, + "step": 6089 + }, + { + "epoch": 0.41658116150215474, + "grad_norm": 2.33931408078967, + "learning_rate": 6.567192455443502e-06, + "loss": 0.4892, + "step": 6090 + }, + { + "epoch": 0.41664956563376426, + "grad_norm": 2.6321093811776106, + "learning_rate": 6.566140483596499e-06, + "loss": 0.4691, + "step": 6091 + }, + { + "epoch": 0.41671796976537384, + "grad_norm": 1.8982760788535236, + "learning_rate": 6.565088434875757e-06, + "loss": 0.3914, + "step": 6092 + }, + { + "epoch": 0.41678637389698336, + "grad_norm": 2.0485427192184313, + "learning_rate": 6.564036309332919e-06, + "loss": 0.3312, + "step": 6093 + }, + { + "epoch": 0.41685477802859294, + "grad_norm": 2.458259518102435, + "learning_rate": 6.5629841070196255e-06, + "loss": 0.4809, + "step": 6094 + }, + { + "epoch": 0.41692318216020247, + "grad_norm": 2.4634221567491266, + "learning_rate": 6.5619318279875245e-06, + "loss": 0.6764, + "step": 6095 + }, + { + "epoch": 0.41699158629181204, + "grad_norm": 1.4090923684098113, + "learning_rate": 6.560879472288266e-06, + "loss": 0.1561, + "step": 6096 + }, + { + "epoch": 0.41705999042342157, + "grad_norm": 2.2132674634553235, + "learning_rate": 6.5598270399735075e-06, + "loss": 0.2699, + "step": 6097 + }, + { + "epoch": 0.41712839455503115, + "grad_norm": 2.6368114426885674, + "learning_rate": 6.558774531094905e-06, + "loss": 0.3565, + "step": 6098 + }, + { + "epoch": 0.41719679868664067, + "grad_norm": 1.6467849212781611, + "learning_rate": 6.55772194570412e-06, + "loss": 0.3118, + "step": 6099 + }, + { + "epoch": 0.41726520281825025, + "grad_norm": 1.8243427111965884, + "learning_rate": 6.5566692838528215e-06, + "loss": 0.3459, + "step": 6100 + }, + { + "epoch": 0.41733360694985977, + "grad_norm": 1.6573500601273534, + "learning_rate": 6.555616545592675e-06, + "loss": 0.2711, + "step": 6101 + }, + { + "epoch": 0.41740201108146935, + "grad_norm": 1.7642028913223518, + "learning_rate": 6.554563730975357e-06, + "loss": 0.3561, + "step": 6102 + }, + { + "epoch": 0.41747041521307887, + "grad_norm": 2.077760994373655, + "learning_rate": 6.553510840052546e-06, + "loss": 0.5703, + "step": 6103 + }, + { + "epoch": 0.4175388193446884, + "grad_norm": 1.4120968017440467, + "learning_rate": 6.55245787287592e-06, + "loss": 0.177, + "step": 6104 + }, + { + "epoch": 0.417607223476298, + "grad_norm": 2.2927755277672723, + "learning_rate": 6.551404829497165e-06, + "loss": 0.6861, + "step": 6105 + }, + { + "epoch": 0.4176756276079075, + "grad_norm": 2.4014850167139246, + "learning_rate": 6.550351709967968e-06, + "loss": 0.5569, + "step": 6106 + }, + { + "epoch": 0.4177440317395171, + "grad_norm": 2.0968963198423625, + "learning_rate": 6.549298514340026e-06, + "loss": 0.5364, + "step": 6107 + }, + { + "epoch": 0.4178124358711266, + "grad_norm": 0.9620524302073757, + "learning_rate": 6.548245242665029e-06, + "loss": 0.0982, + "step": 6108 + }, + { + "epoch": 0.4178808400027362, + "grad_norm": 1.9603635130279797, + "learning_rate": 6.547191894994679e-06, + "loss": 0.4861, + "step": 6109 + }, + { + "epoch": 0.4179492441343457, + "grad_norm": 1.8777248491046001, + "learning_rate": 6.54613847138068e-06, + "loss": 0.3962, + "step": 6110 + }, + { + "epoch": 0.4180176482659553, + "grad_norm": 2.1184675246397213, + "learning_rate": 6.545084971874738e-06, + "loss": 0.6672, + "step": 6111 + }, + { + "epoch": 0.4180860523975648, + "grad_norm": 1.9739645670152401, + "learning_rate": 6.544031396528564e-06, + "loss": 0.331, + "step": 6112 + }, + { + "epoch": 0.4181544565291744, + "grad_norm": 1.9566678301747105, + "learning_rate": 6.5429777453938745e-06, + "loss": 0.3112, + "step": 6113 + }, + { + "epoch": 0.4182228606607839, + "grad_norm": 2.0281651951373108, + "learning_rate": 6.541924018522386e-06, + "loss": 0.3873, + "step": 6114 + }, + { + "epoch": 0.4182912647923935, + "grad_norm": 2.6944418675835293, + "learning_rate": 6.5408702159658196e-06, + "loss": 0.4135, + "step": 6115 + }, + { + "epoch": 0.418359668924003, + "grad_norm": 2.291180779503558, + "learning_rate": 6.539816337775904e-06, + "loss": 0.1401, + "step": 6116 + }, + { + "epoch": 0.4184280730556126, + "grad_norm": 3.17669936137188, + "learning_rate": 6.5387623840043655e-06, + "loss": 0.3949, + "step": 6117 + }, + { + "epoch": 0.4184964771872221, + "grad_norm": 2.934262647702118, + "learning_rate": 6.537708354702939e-06, + "loss": 0.4513, + "step": 6118 + }, + { + "epoch": 0.4185648813188317, + "grad_norm": 3.706324100879371, + "learning_rate": 6.536654249923361e-06, + "loss": 0.5009, + "step": 6119 + }, + { + "epoch": 0.4186332854504412, + "grad_norm": 1.7689876268718574, + "learning_rate": 6.535600069717372e-06, + "loss": 0.2328, + "step": 6120 + }, + { + "epoch": 0.41870168958205073, + "grad_norm": 1.9035857726475325, + "learning_rate": 6.534545814136716e-06, + "loss": 0.3897, + "step": 6121 + }, + { + "epoch": 0.4187700937136603, + "grad_norm": 2.9291567766765114, + "learning_rate": 6.533491483233141e-06, + "loss": 0.5859, + "step": 6122 + }, + { + "epoch": 0.41883849784526983, + "grad_norm": 1.8563716378158732, + "learning_rate": 6.5324370770584004e-06, + "loss": 0.4679, + "step": 6123 + }, + { + "epoch": 0.4189069019768794, + "grad_norm": 2.4953616804461003, + "learning_rate": 6.531382595664246e-06, + "loss": 0.5489, + "step": 6124 + }, + { + "epoch": 0.41897530610848893, + "grad_norm": 1.9723235841252214, + "learning_rate": 6.530328039102441e-06, + "loss": 0.596, + "step": 6125 + }, + { + "epoch": 0.4190437102400985, + "grad_norm": 2.099266123547679, + "learning_rate": 6.529273407424743e-06, + "loss": 0.4512, + "step": 6126 + }, + { + "epoch": 0.41911211437170803, + "grad_norm": 1.823486569193013, + "learning_rate": 6.528218700682924e-06, + "loss": 0.3685, + "step": 6127 + }, + { + "epoch": 0.4191805185033176, + "grad_norm": 1.4937116717764203, + "learning_rate": 6.527163918928749e-06, + "loss": 0.3131, + "step": 6128 + }, + { + "epoch": 0.41924892263492713, + "grad_norm": 2.4453399096040513, + "learning_rate": 6.526109062213996e-06, + "loss": 0.3485, + "step": 6129 + }, + { + "epoch": 0.4193173267665367, + "grad_norm": 4.013434636513689, + "learning_rate": 6.525054130590441e-06, + "loss": 0.4159, + "step": 6130 + }, + { + "epoch": 0.41938573089814624, + "grad_norm": 1.9865425916811046, + "learning_rate": 6.523999124109865e-06, + "loss": 0.3893, + "step": 6131 + }, + { + "epoch": 0.4194541350297558, + "grad_norm": 1.654942194110652, + "learning_rate": 6.5229440428240505e-06, + "loss": 0.4284, + "step": 6132 + }, + { + "epoch": 0.41952253916136534, + "grad_norm": 1.9279502933248513, + "learning_rate": 6.52188888678479e-06, + "loss": 0.4041, + "step": 6133 + }, + { + "epoch": 0.4195909432929749, + "grad_norm": 1.7966551915773532, + "learning_rate": 6.520833656043874e-06, + "loss": 0.2606, + "step": 6134 + }, + { + "epoch": 0.41965934742458444, + "grad_norm": 2.7761075907830404, + "learning_rate": 6.519778350653097e-06, + "loss": 0.4388, + "step": 6135 + }, + { + "epoch": 0.419727751556194, + "grad_norm": 1.9584698823911955, + "learning_rate": 6.518722970664259e-06, + "loss": 0.2341, + "step": 6136 + }, + { + "epoch": 0.41979615568780354, + "grad_norm": 1.9302320291211013, + "learning_rate": 6.517667516129167e-06, + "loss": 0.5771, + "step": 6137 + }, + { + "epoch": 0.4198645598194131, + "grad_norm": 1.4636359767008635, + "learning_rate": 6.516611987099621e-06, + "loss": 0.1741, + "step": 6138 + }, + { + "epoch": 0.41993296395102264, + "grad_norm": 1.8652645807666506, + "learning_rate": 6.515556383627437e-06, + "loss": 0.5456, + "step": 6139 + }, + { + "epoch": 0.42000136808263216, + "grad_norm": 1.5573379409787333, + "learning_rate": 6.514500705764428e-06, + "loss": 0.3118, + "step": 6140 + }, + { + "epoch": 0.42006977221424174, + "grad_norm": 2.020791359757197, + "learning_rate": 6.513444953562409e-06, + "loss": 0.4795, + "step": 6141 + }, + { + "epoch": 0.42013817634585127, + "grad_norm": 2.354488909168089, + "learning_rate": 6.512389127073203e-06, + "loss": 0.6471, + "step": 6142 + }, + { + "epoch": 0.42020658047746084, + "grad_norm": 2.601032329069484, + "learning_rate": 6.511333226348637e-06, + "loss": 0.7059, + "step": 6143 + }, + { + "epoch": 0.42027498460907037, + "grad_norm": 2.0388668010238806, + "learning_rate": 6.5102772514405375e-06, + "loss": 0.405, + "step": 6144 + }, + { + "epoch": 0.42034338874067995, + "grad_norm": 2.05751527987423, + "learning_rate": 6.509221202400737e-06, + "loss": 0.468, + "step": 6145 + }, + { + "epoch": 0.42041179287228947, + "grad_norm": 2.086683236451549, + "learning_rate": 6.5081650792810725e-06, + "loss": 0.3333, + "step": 6146 + }, + { + "epoch": 0.42048019700389905, + "grad_norm": 1.8093808729112675, + "learning_rate": 6.507108882133383e-06, + "loss": 0.4558, + "step": 6147 + }, + { + "epoch": 0.42054860113550857, + "grad_norm": 2.1616906757413576, + "learning_rate": 6.506052611009511e-06, + "loss": 0.6164, + "step": 6148 + }, + { + "epoch": 0.42061700526711815, + "grad_norm": 1.6119718664452856, + "learning_rate": 6.504996265961305e-06, + "loss": 0.3087, + "step": 6149 + }, + { + "epoch": 0.42068540939872767, + "grad_norm": 1.845535246465198, + "learning_rate": 6.503939847040616e-06, + "loss": 0.2776, + "step": 6150 + }, + { + "epoch": 0.42075381353033725, + "grad_norm": 2.216362602280106, + "learning_rate": 6.5028833542992944e-06, + "loss": 0.5041, + "step": 6151 + }, + { + "epoch": 0.4208222176619468, + "grad_norm": 1.7709716719465132, + "learning_rate": 6.5018267877892e-06, + "loss": 0.2214, + "step": 6152 + }, + { + "epoch": 0.42089062179355635, + "grad_norm": 1.8317441573020539, + "learning_rate": 6.500770147562198e-06, + "loss": 0.5299, + "step": 6153 + }, + { + "epoch": 0.4209590259251659, + "grad_norm": 2.378779340594139, + "learning_rate": 6.499713433670148e-06, + "loss": 0.5523, + "step": 6154 + }, + { + "epoch": 0.42102743005677545, + "grad_norm": 1.830724581653466, + "learning_rate": 6.4986566461649205e-06, + "loss": 0.1988, + "step": 6155 + }, + { + "epoch": 0.421095834188385, + "grad_norm": 2.1290224096123103, + "learning_rate": 6.497599785098388e-06, + "loss": 0.3538, + "step": 6156 + }, + { + "epoch": 0.42116423831999455, + "grad_norm": 1.8834511595731176, + "learning_rate": 6.496542850522426e-06, + "loss": 0.3122, + "step": 6157 + }, + { + "epoch": 0.4212326424516041, + "grad_norm": 1.7376053213884015, + "learning_rate": 6.495485842488913e-06, + "loss": 0.376, + "step": 6158 + }, + { + "epoch": 0.4213010465832136, + "grad_norm": 2.1350783723655646, + "learning_rate": 6.494428761049736e-06, + "loss": 0.3223, + "step": 6159 + }, + { + "epoch": 0.4213694507148232, + "grad_norm": 2.2536289203139344, + "learning_rate": 6.493371606256777e-06, + "loss": 0.2939, + "step": 6160 + }, + { + "epoch": 0.4214378548464327, + "grad_norm": 2.1023130671921164, + "learning_rate": 6.492314378161929e-06, + "loss": 0.4181, + "step": 6161 + }, + { + "epoch": 0.4215062589780423, + "grad_norm": 2.1255013211124383, + "learning_rate": 6.491257076817085e-06, + "loss": 0.4555, + "step": 6162 + }, + { + "epoch": 0.4215746631096518, + "grad_norm": 1.8218352370724051, + "learning_rate": 6.490199702274142e-06, + "loss": 0.3726, + "step": 6163 + }, + { + "epoch": 0.4216430672412614, + "grad_norm": 1.997671030442657, + "learning_rate": 6.489142254585001e-06, + "loss": 0.4904, + "step": 6164 + }, + { + "epoch": 0.4217114713728709, + "grad_norm": 2.4769364083758796, + "learning_rate": 6.488084733801567e-06, + "loss": 0.2843, + "step": 6165 + }, + { + "epoch": 0.4217798755044805, + "grad_norm": 1.9821291552380582, + "learning_rate": 6.48702713997575e-06, + "loss": 0.5717, + "step": 6166 + }, + { + "epoch": 0.42184827963609, + "grad_norm": 1.904096410910477, + "learning_rate": 6.48596947315946e-06, + "loss": 0.4521, + "step": 6167 + }, + { + "epoch": 0.4219166837676996, + "grad_norm": 1.8110773077014284, + "learning_rate": 6.484911733404612e-06, + "loss": 0.2721, + "step": 6168 + }, + { + "epoch": 0.4219850878993091, + "grad_norm": 2.1870371737641476, + "learning_rate": 6.4838539207631254e-06, + "loss": 0.3556, + "step": 6169 + }, + { + "epoch": 0.4220534920309187, + "grad_norm": 1.9173602325611372, + "learning_rate": 6.482796035286922e-06, + "loss": 0.5268, + "step": 6170 + }, + { + "epoch": 0.4221218961625282, + "grad_norm": 1.7612292927187343, + "learning_rate": 6.48173807702793e-06, + "loss": 0.3049, + "step": 6171 + }, + { + "epoch": 0.4221903002941378, + "grad_norm": 1.6642093145948076, + "learning_rate": 6.480680046038078e-06, + "loss": 0.2508, + "step": 6172 + }, + { + "epoch": 0.4222587044257473, + "grad_norm": 1.976059845951522, + "learning_rate": 6.4796219423692995e-06, + "loss": 0.373, + "step": 6173 + }, + { + "epoch": 0.4223271085573569, + "grad_norm": 2.09737493095532, + "learning_rate": 6.47856376607353e-06, + "loss": 0.5122, + "step": 6174 + }, + { + "epoch": 0.4223955126889664, + "grad_norm": 1.7257843982880223, + "learning_rate": 6.477505517202711e-06, + "loss": 0.2811, + "step": 6175 + }, + { + "epoch": 0.42246391682057594, + "grad_norm": 2.4332639614650424, + "learning_rate": 6.476447195808788e-06, + "loss": 0.4443, + "step": 6176 + }, + { + "epoch": 0.4225323209521855, + "grad_norm": 1.6787681769553668, + "learning_rate": 6.475388801943706e-06, + "loss": 0.1513, + "step": 6177 + }, + { + "epoch": 0.42260072508379504, + "grad_norm": 2.0982457474549903, + "learning_rate": 6.4743303356594165e-06, + "loss": 0.4502, + "step": 6178 + }, + { + "epoch": 0.4226691292154046, + "grad_norm": 1.9556681936302887, + "learning_rate": 6.473271797007876e-06, + "loss": 0.3335, + "step": 6179 + }, + { + "epoch": 0.42273753334701414, + "grad_norm": 2.8363178618118923, + "learning_rate": 6.472213186041039e-06, + "loss": 0.5137, + "step": 6180 + }, + { + "epoch": 0.4228059374786237, + "grad_norm": 2.036594548917831, + "learning_rate": 6.471154502810872e-06, + "loss": 0.3855, + "step": 6181 + }, + { + "epoch": 0.42287434161023324, + "grad_norm": 2.1089093365420206, + "learning_rate": 6.470095747369336e-06, + "loss": 0.3859, + "step": 6182 + }, + { + "epoch": 0.4229427457418428, + "grad_norm": 2.4865829564924202, + "learning_rate": 6.469036919768403e-06, + "loss": 0.8708, + "step": 6183 + }, + { + "epoch": 0.42301114987345234, + "grad_norm": 2.26366543048932, + "learning_rate": 6.467978020060044e-06, + "loss": 0.629, + "step": 6184 + }, + { + "epoch": 0.4230795540050619, + "grad_norm": 2.199130983108936, + "learning_rate": 6.466919048296234e-06, + "loss": 0.5463, + "step": 6185 + }, + { + "epoch": 0.42314795813667144, + "grad_norm": 2.2973820068374566, + "learning_rate": 6.465860004528955e-06, + "loss": 0.5731, + "step": 6186 + }, + { + "epoch": 0.423216362268281, + "grad_norm": 1.677567694647609, + "learning_rate": 6.464800888810188e-06, + "loss": 0.2202, + "step": 6187 + }, + { + "epoch": 0.42328476639989054, + "grad_norm": 2.0727615374277497, + "learning_rate": 6.463741701191918e-06, + "loss": 0.3934, + "step": 6188 + }, + { + "epoch": 0.4233531705315001, + "grad_norm": 2.0086538244260357, + "learning_rate": 6.4626824417261394e-06, + "loss": 0.513, + "step": 6189 + }, + { + "epoch": 0.42342157466310965, + "grad_norm": 1.7195041215863294, + "learning_rate": 6.461623110464841e-06, + "loss": 0.4193, + "step": 6190 + }, + { + "epoch": 0.4234899787947192, + "grad_norm": 2.0721983543644424, + "learning_rate": 6.460563707460024e-06, + "loss": 0.5062, + "step": 6191 + }, + { + "epoch": 0.42355838292632875, + "grad_norm": 1.176335463966186, + "learning_rate": 6.459504232763687e-06, + "loss": 0.1584, + "step": 6192 + }, + { + "epoch": 0.4236267870579383, + "grad_norm": 1.6792154014630687, + "learning_rate": 6.458444686427834e-06, + "loss": 0.2243, + "step": 6193 + }, + { + "epoch": 0.42369519118954785, + "grad_norm": 1.9004236592855053, + "learning_rate": 6.457385068504472e-06, + "loss": 0.2787, + "step": 6194 + }, + { + "epoch": 0.42376359532115737, + "grad_norm": 2.1639724946843293, + "learning_rate": 6.456325379045615e-06, + "loss": 0.5537, + "step": 6195 + }, + { + "epoch": 0.42383199945276695, + "grad_norm": 2.009452647166879, + "learning_rate": 6.455265618103276e-06, + "loss": 0.5477, + "step": 6196 + }, + { + "epoch": 0.4239004035843765, + "grad_norm": 1.9587564040981882, + "learning_rate": 6.4542057857294725e-06, + "loss": 0.5187, + "step": 6197 + }, + { + "epoch": 0.42396880771598605, + "grad_norm": 1.8108608463313798, + "learning_rate": 6.453145881976226e-06, + "loss": 0.2378, + "step": 6198 + }, + { + "epoch": 0.4240372118475956, + "grad_norm": 1.7439800179961142, + "learning_rate": 6.452085906895564e-06, + "loss": 0.2866, + "step": 6199 + }, + { + "epoch": 0.42410561597920515, + "grad_norm": 1.844693354746455, + "learning_rate": 6.451025860539512e-06, + "loss": 0.4389, + "step": 6200 + }, + { + "epoch": 0.4241740201108147, + "grad_norm": 2.531028166810893, + "learning_rate": 6.449965742960104e-06, + "loss": 0.327, + "step": 6201 + }, + { + "epoch": 0.42424242424242425, + "grad_norm": 2.1652582012078896, + "learning_rate": 6.448905554209376e-06, + "loss": 0.445, + "step": 6202 + }, + { + "epoch": 0.4243108283740338, + "grad_norm": 2.0601309840215705, + "learning_rate": 6.447845294339368e-06, + "loss": 0.3441, + "step": 6203 + }, + { + "epoch": 0.42437923250564336, + "grad_norm": 1.9031705117225501, + "learning_rate": 6.446784963402118e-06, + "loss": 0.1742, + "step": 6204 + }, + { + "epoch": 0.4244476366372529, + "grad_norm": 2.628650692539147, + "learning_rate": 6.4457245614496775e-06, + "loss": 0.278, + "step": 6205 + }, + { + "epoch": 0.42451604076886246, + "grad_norm": 2.357071010518501, + "learning_rate": 6.4446640885340946e-06, + "loss": 0.6631, + "step": 6206 + }, + { + "epoch": 0.424584444900472, + "grad_norm": 1.5570692838664133, + "learning_rate": 6.443603544707423e-06, + "loss": 0.2684, + "step": 6207 + }, + { + "epoch": 0.42465284903208156, + "grad_norm": 2.457015330106895, + "learning_rate": 6.442542930021716e-06, + "loss": 0.6458, + "step": 6208 + }, + { + "epoch": 0.4247212531636911, + "grad_norm": 1.3401131817805663, + "learning_rate": 6.441482244529038e-06, + "loss": 0.2362, + "step": 6209 + }, + { + "epoch": 0.42478965729530066, + "grad_norm": 1.9227922390707775, + "learning_rate": 6.440421488281449e-06, + "loss": 0.3036, + "step": 6210 + }, + { + "epoch": 0.4248580614269102, + "grad_norm": 1.5256410129496736, + "learning_rate": 6.439360661331018e-06, + "loss": 0.1864, + "step": 6211 + }, + { + "epoch": 0.42492646555851976, + "grad_norm": 2.025121551415374, + "learning_rate": 6.438299763729816e-06, + "loss": 0.4516, + "step": 6212 + }, + { + "epoch": 0.4249948696901293, + "grad_norm": 1.4764325902193758, + "learning_rate": 6.437238795529917e-06, + "loss": 0.2436, + "step": 6213 + }, + { + "epoch": 0.4250632738217388, + "grad_norm": 1.9445827601450822, + "learning_rate": 6.436177756783395e-06, + "loss": 0.4285, + "step": 6214 + }, + { + "epoch": 0.4251316779533484, + "grad_norm": 2.1698842976747423, + "learning_rate": 6.4351166475423335e-06, + "loss": 0.4457, + "step": 6215 + }, + { + "epoch": 0.4252000820849579, + "grad_norm": 2.3765740699661992, + "learning_rate": 6.434055467858819e-06, + "loss": 0.5105, + "step": 6216 + }, + { + "epoch": 0.4252684862165675, + "grad_norm": 2.2189066892018663, + "learning_rate": 6.432994217784935e-06, + "loss": 0.3922, + "step": 6217 + }, + { + "epoch": 0.425336890348177, + "grad_norm": 1.6026707164516505, + "learning_rate": 6.431932897372777e-06, + "loss": 0.2603, + "step": 6218 + }, + { + "epoch": 0.4254052944797866, + "grad_norm": 1.6785851565157655, + "learning_rate": 6.430871506674437e-06, + "loss": 0.4552, + "step": 6219 + }, + { + "epoch": 0.4254736986113961, + "grad_norm": 1.9900157609991151, + "learning_rate": 6.429810045742013e-06, + "loss": 0.5841, + "step": 6220 + }, + { + "epoch": 0.4255421027430057, + "grad_norm": 1.8210866370731122, + "learning_rate": 6.428748514627608e-06, + "loss": 0.3761, + "step": 6221 + }, + { + "epoch": 0.4256105068746152, + "grad_norm": 1.599336621851493, + "learning_rate": 6.427686913383326e-06, + "loss": 0.3076, + "step": 6222 + }, + { + "epoch": 0.4256789110062248, + "grad_norm": 2.157875781888831, + "learning_rate": 6.426625242061277e-06, + "loss": 0.5513, + "step": 6223 + }, + { + "epoch": 0.4257473151378343, + "grad_norm": 2.246347389827146, + "learning_rate": 6.425563500713572e-06, + "loss": 0.55, + "step": 6224 + }, + { + "epoch": 0.4258157192694439, + "grad_norm": 2.0957722203648825, + "learning_rate": 6.424501689392325e-06, + "loss": 0.3071, + "step": 6225 + }, + { + "epoch": 0.4258841234010534, + "grad_norm": 2.516874009211051, + "learning_rate": 6.423439808149659e-06, + "loss": 0.4882, + "step": 6226 + }, + { + "epoch": 0.425952527532663, + "grad_norm": 1.8422153418218996, + "learning_rate": 6.422377857037691e-06, + "loss": 0.3174, + "step": 6227 + }, + { + "epoch": 0.4260209316642725, + "grad_norm": 2.5645521192771232, + "learning_rate": 6.421315836108549e-06, + "loss": 0.44, + "step": 6228 + }, + { + "epoch": 0.4260893357958821, + "grad_norm": 1.6481116727652516, + "learning_rate": 6.420253745414365e-06, + "loss": 0.2322, + "step": 6229 + }, + { + "epoch": 0.4261577399274916, + "grad_norm": 1.949489589978895, + "learning_rate": 6.4191915850072664e-06, + "loss": 0.3359, + "step": 6230 + }, + { + "epoch": 0.4262261440591012, + "grad_norm": 1.8115559700977664, + "learning_rate": 6.418129354939391e-06, + "loss": 0.3568, + "step": 6231 + }, + { + "epoch": 0.4262945481907107, + "grad_norm": 2.476380855986819, + "learning_rate": 6.417067055262881e-06, + "loss": 0.3147, + "step": 6232 + }, + { + "epoch": 0.42636295232232024, + "grad_norm": 2.0885330724816824, + "learning_rate": 6.416004686029876e-06, + "loss": 0.177, + "step": 6233 + }, + { + "epoch": 0.4264313564539298, + "grad_norm": 2.0597278467798064, + "learning_rate": 6.414942247292523e-06, + "loss": 0.4002, + "step": 6234 + }, + { + "epoch": 0.42649976058553934, + "grad_norm": 1.671181312039697, + "learning_rate": 6.413879739102972e-06, + "loss": 0.3048, + "step": 6235 + }, + { + "epoch": 0.4265681647171489, + "grad_norm": 1.864011349579789, + "learning_rate": 6.412817161513376e-06, + "loss": 0.4625, + "step": 6236 + }, + { + "epoch": 0.42663656884875845, + "grad_norm": 1.9168218536024353, + "learning_rate": 6.411754514575891e-06, + "loss": 0.3669, + "step": 6237 + }, + { + "epoch": 0.426704972980368, + "grad_norm": 2.189860582508144, + "learning_rate": 6.410691798342678e-06, + "loss": 0.4534, + "step": 6238 + }, + { + "epoch": 0.42677337711197755, + "grad_norm": 2.1073102806603004, + "learning_rate": 6.409629012865899e-06, + "loss": 0.6453, + "step": 6239 + }, + { + "epoch": 0.4268417812435871, + "grad_norm": 1.5145686194357983, + "learning_rate": 6.40856615819772e-06, + "loss": 0.2655, + "step": 6240 + }, + { + "epoch": 0.42691018537519665, + "grad_norm": 2.371991963072183, + "learning_rate": 6.407503234390312e-06, + "loss": 0.5951, + "step": 6241 + }, + { + "epoch": 0.4269785895068062, + "grad_norm": 1.922638264726565, + "learning_rate": 6.4064402414958505e-06, + "loss": 0.4008, + "step": 6242 + }, + { + "epoch": 0.42704699363841575, + "grad_norm": 2.2062945832835035, + "learning_rate": 6.405377179566508e-06, + "loss": 0.2546, + "step": 6243 + }, + { + "epoch": 0.42711539777002533, + "grad_norm": 2.104974674578793, + "learning_rate": 6.4043140486544684e-06, + "loss": 0.4569, + "step": 6244 + }, + { + "epoch": 0.42718380190163485, + "grad_norm": 2.185001121063512, + "learning_rate": 6.403250848811913e-06, + "loss": 0.3578, + "step": 6245 + }, + { + "epoch": 0.42725220603324443, + "grad_norm": 2.084783895868868, + "learning_rate": 6.402187580091031e-06, + "loss": 0.3741, + "step": 6246 + }, + { + "epoch": 0.42732061016485395, + "grad_norm": 2.4327491470251528, + "learning_rate": 6.40112424254401e-06, + "loss": 0.4928, + "step": 6247 + }, + { + "epoch": 0.42738901429646353, + "grad_norm": 2.575539870749463, + "learning_rate": 6.400060836223047e-06, + "loss": 0.4689, + "step": 6248 + }, + { + "epoch": 0.42745741842807305, + "grad_norm": 2.04185883840535, + "learning_rate": 6.398997361180337e-06, + "loss": 0.3953, + "step": 6249 + }, + { + "epoch": 0.4275258225596826, + "grad_norm": 2.3572318909961223, + "learning_rate": 6.397933817468079e-06, + "loss": 0.6113, + "step": 6250 + }, + { + "epoch": 0.42759422669129216, + "grad_norm": 2.3160183599478095, + "learning_rate": 6.3968702051384805e-06, + "loss": 0.3474, + "step": 6251 + }, + { + "epoch": 0.4276626308229017, + "grad_norm": 1.9907658497192464, + "learning_rate": 6.395806524243746e-06, + "loss": 0.4489, + "step": 6252 + }, + { + "epoch": 0.42773103495451126, + "grad_norm": 2.469609780829403, + "learning_rate": 6.394742774836088e-06, + "loss": 0.8301, + "step": 6253 + }, + { + "epoch": 0.4277994390861208, + "grad_norm": 2.159922726858282, + "learning_rate": 6.393678956967718e-06, + "loss": 0.4262, + "step": 6254 + }, + { + "epoch": 0.42786784321773036, + "grad_norm": 1.9424874348031531, + "learning_rate": 6.3926150706908556e-06, + "loss": 0.5922, + "step": 6255 + }, + { + "epoch": 0.4279362473493399, + "grad_norm": 2.3368855533223827, + "learning_rate": 6.391551116057721e-06, + "loss": 0.4768, + "step": 6256 + }, + { + "epoch": 0.42800465148094946, + "grad_norm": 2.3670242114764766, + "learning_rate": 6.390487093120537e-06, + "loss": 0.5248, + "step": 6257 + }, + { + "epoch": 0.428073055612559, + "grad_norm": 1.821216972499315, + "learning_rate": 6.389423001931533e-06, + "loss": 0.3352, + "step": 6258 + }, + { + "epoch": 0.42814145974416856, + "grad_norm": 2.437449943811009, + "learning_rate": 6.388358842542939e-06, + "loss": 0.372, + "step": 6259 + }, + { + "epoch": 0.4282098638757781, + "grad_norm": 2.1981326583900267, + "learning_rate": 6.387294615006987e-06, + "loss": 0.4855, + "step": 6260 + }, + { + "epoch": 0.42827826800738766, + "grad_norm": 2.3188269962266785, + "learning_rate": 6.386230319375915e-06, + "loss": 0.6875, + "step": 6261 + }, + { + "epoch": 0.4283466721389972, + "grad_norm": 1.688442404177002, + "learning_rate": 6.385165955701968e-06, + "loss": 0.1812, + "step": 6262 + }, + { + "epoch": 0.42841507627060676, + "grad_norm": 1.7795070854821964, + "learning_rate": 6.384101524037385e-06, + "loss": 0.3527, + "step": 6263 + }, + { + "epoch": 0.4284834804022163, + "grad_norm": 1.9906743979201176, + "learning_rate": 6.383037024434416e-06, + "loss": 0.4735, + "step": 6264 + }, + { + "epoch": 0.42855188453382587, + "grad_norm": 1.913107069131493, + "learning_rate": 6.3819724569453126e-06, + "loss": 0.4871, + "step": 6265 + }, + { + "epoch": 0.4286202886654354, + "grad_norm": 1.3430528745328072, + "learning_rate": 6.380907821622327e-06, + "loss": 0.2362, + "step": 6266 + }, + { + "epoch": 0.42868869279704497, + "grad_norm": 1.9338485395776188, + "learning_rate": 6.379843118517714e-06, + "loss": 0.1934, + "step": 6267 + }, + { + "epoch": 0.4287570969286545, + "grad_norm": 1.77597235286142, + "learning_rate": 6.378778347683742e-06, + "loss": 0.5303, + "step": 6268 + }, + { + "epoch": 0.428825501060264, + "grad_norm": 2.2224008282686585, + "learning_rate": 6.377713509172667e-06, + "loss": 0.4303, + "step": 6269 + }, + { + "epoch": 0.4288939051918736, + "grad_norm": 1.3206988734278693, + "learning_rate": 6.3766486030367615e-06, + "loss": 0.1451, + "step": 6270 + }, + { + "epoch": 0.4289623093234831, + "grad_norm": 1.842565301361651, + "learning_rate": 6.375583629328296e-06, + "loss": 0.3007, + "step": 6271 + }, + { + "epoch": 0.4290307134550927, + "grad_norm": 2.173797312447019, + "learning_rate": 6.374518588099543e-06, + "loss": 0.3828, + "step": 6272 + }, + { + "epoch": 0.4290991175867022, + "grad_norm": 2.5825974136584486, + "learning_rate": 6.373453479402779e-06, + "loss": 0.5036, + "step": 6273 + }, + { + "epoch": 0.4291675217183118, + "grad_norm": 2.281677480385927, + "learning_rate": 6.372388303290286e-06, + "loss": 0.5814, + "step": 6274 + }, + { + "epoch": 0.4292359258499213, + "grad_norm": 2.041398690520289, + "learning_rate": 6.3713230598143495e-06, + "loss": 0.1998, + "step": 6275 + }, + { + "epoch": 0.4293043299815309, + "grad_norm": 2.051292726221681, + "learning_rate": 6.370257749027253e-06, + "loss": 0.3788, + "step": 6276 + }, + { + "epoch": 0.4293727341131404, + "grad_norm": 2.3663304385726476, + "learning_rate": 6.369192370981291e-06, + "loss": 0.552, + "step": 6277 + }, + { + "epoch": 0.42944113824475, + "grad_norm": 2.3215048709319572, + "learning_rate": 6.368126925728755e-06, + "loss": 0.4816, + "step": 6278 + }, + { + "epoch": 0.4295095423763595, + "grad_norm": 2.1496437501177788, + "learning_rate": 6.367061413321942e-06, + "loss": 0.3804, + "step": 6279 + }, + { + "epoch": 0.4295779465079691, + "grad_norm": 1.8269018662337249, + "learning_rate": 6.365995833813154e-06, + "loss": 0.5245, + "step": 6280 + }, + { + "epoch": 0.4296463506395786, + "grad_norm": 1.710204973567984, + "learning_rate": 6.364930187254695e-06, + "loss": 0.3135, + "step": 6281 + }, + { + "epoch": 0.4297147547711882, + "grad_norm": 2.2131966684797932, + "learning_rate": 6.363864473698871e-06, + "loss": 0.4511, + "step": 6282 + }, + { + "epoch": 0.4297831589027977, + "grad_norm": 2.2477047856314334, + "learning_rate": 6.362798693197991e-06, + "loss": 0.4969, + "step": 6283 + }, + { + "epoch": 0.4298515630344073, + "grad_norm": 1.581440283046629, + "learning_rate": 6.361732845804371e-06, + "loss": 0.2707, + "step": 6284 + }, + { + "epoch": 0.4299199671660168, + "grad_norm": 2.0373523453879954, + "learning_rate": 6.360666931570329e-06, + "loss": 0.1987, + "step": 6285 + }, + { + "epoch": 0.4299883712976264, + "grad_norm": 2.8743634244814724, + "learning_rate": 6.359600950548182e-06, + "loss": 0.5062, + "step": 6286 + }, + { + "epoch": 0.4300567754292359, + "grad_norm": 2.141509165860248, + "learning_rate": 6.358534902790255e-06, + "loss": 0.5467, + "step": 6287 + }, + { + "epoch": 0.43012517956084545, + "grad_norm": 1.7374032921601141, + "learning_rate": 6.357468788348875e-06, + "loss": 0.4389, + "step": 6288 + }, + { + "epoch": 0.43019358369245503, + "grad_norm": 1.8844110975714459, + "learning_rate": 6.35640260727637e-06, + "loss": 0.3646, + "step": 6289 + }, + { + "epoch": 0.43026198782406455, + "grad_norm": 2.1136070994467895, + "learning_rate": 6.355336359625076e-06, + "loss": 0.5353, + "step": 6290 + }, + { + "epoch": 0.43033039195567413, + "grad_norm": 11.565246013666812, + "learning_rate": 6.354270045447331e-06, + "loss": 0.1957, + "step": 6291 + }, + { + "epoch": 0.43039879608728365, + "grad_norm": 2.2268939557902785, + "learning_rate": 6.35320366479547e-06, + "loss": 0.472, + "step": 6292 + }, + { + "epoch": 0.43046720021889323, + "grad_norm": 1.8215627449318195, + "learning_rate": 6.3521372177218384e-06, + "loss": 0.157, + "step": 6293 + }, + { + "epoch": 0.43053560435050275, + "grad_norm": 2.548528482647876, + "learning_rate": 6.351070704278782e-06, + "loss": 0.5359, + "step": 6294 + }, + { + "epoch": 0.43060400848211233, + "grad_norm": 1.7049610976076794, + "learning_rate": 6.350004124518654e-06, + "loss": 0.1734, + "step": 6295 + }, + { + "epoch": 0.43067241261372186, + "grad_norm": 2.881992755942006, + "learning_rate": 6.3489374784938025e-06, + "loss": 0.3756, + "step": 6296 + }, + { + "epoch": 0.43074081674533143, + "grad_norm": 2.434872947762116, + "learning_rate": 6.3478707662565865e-06, + "loss": 0.5826, + "step": 6297 + }, + { + "epoch": 0.43080922087694096, + "grad_norm": 1.8715233120563224, + "learning_rate": 6.346803987859364e-06, + "loss": 0.4203, + "step": 6298 + }, + { + "epoch": 0.43087762500855054, + "grad_norm": 1.9893729130251308, + "learning_rate": 6.3457371433544975e-06, + "loss": 0.3891, + "step": 6299 + }, + { + "epoch": 0.43094602914016006, + "grad_norm": 2.08846088540174, + "learning_rate": 6.344670232794354e-06, + "loss": 0.4293, + "step": 6300 + }, + { + "epoch": 0.43101443327176964, + "grad_norm": 2.433763547089642, + "learning_rate": 6.343603256231303e-06, + "loss": 0.3125, + "step": 6301 + }, + { + "epoch": 0.43108283740337916, + "grad_norm": 1.7894073741121401, + "learning_rate": 6.342536213717716e-06, + "loss": 0.4098, + "step": 6302 + }, + { + "epoch": 0.43115124153498874, + "grad_norm": 1.7103816091736943, + "learning_rate": 6.341469105305967e-06, + "loss": 0.2017, + "step": 6303 + }, + { + "epoch": 0.43121964566659826, + "grad_norm": 2.3662920255853575, + "learning_rate": 6.3404019310484365e-06, + "loss": 0.2448, + "step": 6304 + }, + { + "epoch": 0.4312880497982078, + "grad_norm": 2.420148603333786, + "learning_rate": 6.339334690997508e-06, + "loss": 0.6642, + "step": 6305 + }, + { + "epoch": 0.43135645392981736, + "grad_norm": 2.3016728096999004, + "learning_rate": 6.338267385205565e-06, + "loss": 0.4178, + "step": 6306 + }, + { + "epoch": 0.4314248580614269, + "grad_norm": 2.699791992307778, + "learning_rate": 6.337200013724996e-06, + "loss": 0.9471, + "step": 6307 + }, + { + "epoch": 0.43149326219303646, + "grad_norm": 1.8974657122463086, + "learning_rate": 6.336132576608194e-06, + "loss": 0.4509, + "step": 6308 + }, + { + "epoch": 0.431561666324646, + "grad_norm": 1.6998183233991215, + "learning_rate": 6.335065073907551e-06, + "loss": 0.2437, + "step": 6309 + }, + { + "epoch": 0.43163007045625557, + "grad_norm": 2.059576545514171, + "learning_rate": 6.333997505675468e-06, + "loss": 0.4657, + "step": 6310 + }, + { + "epoch": 0.4316984745878651, + "grad_norm": 2.028134289266182, + "learning_rate": 6.332929871964346e-06, + "loss": 0.3245, + "step": 6311 + }, + { + "epoch": 0.43176687871947467, + "grad_norm": 2.8615862576959743, + "learning_rate": 6.33186217282659e-06, + "loss": 0.4274, + "step": 6312 + }, + { + "epoch": 0.4318352828510842, + "grad_norm": 2.031017704665589, + "learning_rate": 6.3307944083146055e-06, + "loss": 0.3583, + "step": 6313 + }, + { + "epoch": 0.43190368698269377, + "grad_norm": 1.7105570069592502, + "learning_rate": 6.329726578480805e-06, + "loss": 0.3481, + "step": 6314 + }, + { + "epoch": 0.4319720911143033, + "grad_norm": 1.8933408792551312, + "learning_rate": 6.328658683377604e-06, + "loss": 0.3867, + "step": 6315 + }, + { + "epoch": 0.43204049524591287, + "grad_norm": 2.1110347081637566, + "learning_rate": 6.3275907230574175e-06, + "loss": 0.5231, + "step": 6316 + }, + { + "epoch": 0.4321088993775224, + "grad_norm": 1.6225849967964214, + "learning_rate": 6.326522697572669e-06, + "loss": 0.2894, + "step": 6317 + }, + { + "epoch": 0.43217730350913197, + "grad_norm": 2.2921733483915854, + "learning_rate": 6.325454606975781e-06, + "loss": 0.5445, + "step": 6318 + }, + { + "epoch": 0.4322457076407415, + "grad_norm": 2.7884666383377916, + "learning_rate": 6.324386451319179e-06, + "loss": 0.3047, + "step": 6319 + }, + { + "epoch": 0.4323141117723511, + "grad_norm": 2.0068310405736427, + "learning_rate": 6.323318230655294e-06, + "loss": 0.3879, + "step": 6320 + }, + { + "epoch": 0.4323825159039606, + "grad_norm": 1.8951725533087895, + "learning_rate": 6.322249945036562e-06, + "loss": 0.4723, + "step": 6321 + }, + { + "epoch": 0.4324509200355702, + "grad_norm": 2.610448776244509, + "learning_rate": 6.3211815945154175e-06, + "loss": 0.5042, + "step": 6322 + }, + { + "epoch": 0.4325193241671797, + "grad_norm": 2.529092752475346, + "learning_rate": 6.320113179144298e-06, + "loss": 0.7204, + "step": 6323 + }, + { + "epoch": 0.4325877282987892, + "grad_norm": 1.4219508412853517, + "learning_rate": 6.319044698975651e-06, + "loss": 0.1773, + "step": 6324 + }, + { + "epoch": 0.4326561324303988, + "grad_norm": 1.844744858416273, + "learning_rate": 6.31797615406192e-06, + "loss": 0.4302, + "step": 6325 + }, + { + "epoch": 0.4327245365620083, + "grad_norm": 2.3696622973891133, + "learning_rate": 6.316907544455555e-06, + "loss": 0.1873, + "step": 6326 + }, + { + "epoch": 0.4327929406936179, + "grad_norm": 2.631901518335502, + "learning_rate": 6.315838870209008e-06, + "loss": 0.4017, + "step": 6327 + }, + { + "epoch": 0.4328613448252274, + "grad_norm": 2.96880099604124, + "learning_rate": 6.314770131374737e-06, + "loss": 0.6726, + "step": 6328 + }, + { + "epoch": 0.432929748956837, + "grad_norm": 1.2160815191995604, + "learning_rate": 6.3137013280051965e-06, + "loss": 0.1596, + "step": 6329 + }, + { + "epoch": 0.4329981530884465, + "grad_norm": 2.314180887685366, + "learning_rate": 6.312632460152851e-06, + "loss": 0.3508, + "step": 6330 + }, + { + "epoch": 0.4330665572200561, + "grad_norm": 1.7399864491121395, + "learning_rate": 6.311563527870167e-06, + "loss": 0.4383, + "step": 6331 + }, + { + "epoch": 0.4331349613516656, + "grad_norm": 1.9314369966040745, + "learning_rate": 6.3104945312096124e-06, + "loss": 0.3588, + "step": 6332 + }, + { + "epoch": 0.4332033654832752, + "grad_norm": 1.8631764148514074, + "learning_rate": 6.309425470223656e-06, + "loss": 0.1556, + "step": 6333 + }, + { + "epoch": 0.4332717696148847, + "grad_norm": 1.2972270878297352, + "learning_rate": 6.308356344964774e-06, + "loss": 0.154, + "step": 6334 + }, + { + "epoch": 0.4333401737464943, + "grad_norm": 1.94514036374496, + "learning_rate": 6.307287155485446e-06, + "loss": 0.3122, + "step": 6335 + }, + { + "epoch": 0.43340857787810383, + "grad_norm": 1.5083616620462994, + "learning_rate": 6.306217901838152e-06, + "loss": 0.267, + "step": 6336 + }, + { + "epoch": 0.4334769820097134, + "grad_norm": 1.8105015116320244, + "learning_rate": 6.305148584075376e-06, + "loss": 0.3988, + "step": 6337 + }, + { + "epoch": 0.43354538614132293, + "grad_norm": 2.863090638897783, + "learning_rate": 6.304079202249605e-06, + "loss": 0.6518, + "step": 6338 + }, + { + "epoch": 0.4336137902729325, + "grad_norm": 1.7678068150062582, + "learning_rate": 6.303009756413327e-06, + "loss": 0.2499, + "step": 6339 + }, + { + "epoch": 0.43368219440454203, + "grad_norm": 2.3352620095443535, + "learning_rate": 6.301940246619039e-06, + "loss": 0.5106, + "step": 6340 + }, + { + "epoch": 0.4337505985361516, + "grad_norm": 1.7357200227096818, + "learning_rate": 6.300870672919237e-06, + "loss": 0.3813, + "step": 6341 + }, + { + "epoch": 0.43381900266776113, + "grad_norm": 1.854715077774452, + "learning_rate": 6.299801035366422e-06, + "loss": 0.4711, + "step": 6342 + }, + { + "epoch": 0.43388740679937066, + "grad_norm": 2.3063566596739027, + "learning_rate": 6.2987313340130925e-06, + "loss": 0.3439, + "step": 6343 + }, + { + "epoch": 0.43395581093098023, + "grad_norm": 2.1696694653901725, + "learning_rate": 6.297661568911759e-06, + "loss": 0.5307, + "step": 6344 + }, + { + "epoch": 0.43402421506258976, + "grad_norm": 1.9129826672163128, + "learning_rate": 6.2965917401149315e-06, + "loss": 0.3738, + "step": 6345 + }, + { + "epoch": 0.43409261919419934, + "grad_norm": 2.019484248250764, + "learning_rate": 6.295521847675117e-06, + "loss": 0.3716, + "step": 6346 + }, + { + "epoch": 0.43416102332580886, + "grad_norm": 1.3614478599072035, + "learning_rate": 6.294451891644837e-06, + "loss": 0.2061, + "step": 6347 + }, + { + "epoch": 0.43422942745741844, + "grad_norm": 2.4603230753859937, + "learning_rate": 6.293381872076606e-06, + "loss": 0.2873, + "step": 6348 + }, + { + "epoch": 0.43429783158902796, + "grad_norm": 1.5856150642312274, + "learning_rate": 6.292311789022947e-06, + "loss": 0.2497, + "step": 6349 + }, + { + "epoch": 0.43436623572063754, + "grad_norm": 1.751308394826036, + "learning_rate": 6.291241642536383e-06, + "loss": 0.3691, + "step": 6350 + }, + { + "epoch": 0.43443463985224706, + "grad_norm": 1.3069189656167337, + "learning_rate": 6.2901714326694475e-06, + "loss": 0.2705, + "step": 6351 + }, + { + "epoch": 0.43450304398385664, + "grad_norm": 1.5141347829494336, + "learning_rate": 6.289101159474666e-06, + "loss": 0.2508, + "step": 6352 + }, + { + "epoch": 0.43457144811546616, + "grad_norm": 2.0753548271355857, + "learning_rate": 6.288030823004574e-06, + "loss": 0.3007, + "step": 6353 + }, + { + "epoch": 0.43463985224707574, + "grad_norm": 1.9685925920320908, + "learning_rate": 6.28696042331171e-06, + "loss": 0.3709, + "step": 6354 + }, + { + "epoch": 0.43470825637868526, + "grad_norm": 2.2870739185637303, + "learning_rate": 6.285889960448614e-06, + "loss": 0.7082, + "step": 6355 + }, + { + "epoch": 0.43477666051029484, + "grad_norm": 2.7700357932573807, + "learning_rate": 6.2848194344678285e-06, + "loss": 0.5228, + "step": 6356 + }, + { + "epoch": 0.43484506464190437, + "grad_norm": 2.157276724581421, + "learning_rate": 6.283748845421902e-06, + "loss": 0.3767, + "step": 6357 + }, + { + "epoch": 0.43491346877351394, + "grad_norm": 1.8398599441690329, + "learning_rate": 6.282678193363383e-06, + "loss": 0.3251, + "step": 6358 + }, + { + "epoch": 0.43498187290512347, + "grad_norm": 2.3980007581661895, + "learning_rate": 6.281607478344823e-06, + "loss": 0.3856, + "step": 6359 + }, + { + "epoch": 0.43505027703673305, + "grad_norm": 2.9786232950730054, + "learning_rate": 6.280536700418779e-06, + "loss": 0.427, + "step": 6360 + }, + { + "epoch": 0.43511868116834257, + "grad_norm": 2.258374002455284, + "learning_rate": 6.279465859637812e-06, + "loss": 0.2649, + "step": 6361 + }, + { + "epoch": 0.4351870852999521, + "grad_norm": 2.3035943914269086, + "learning_rate": 6.27839495605448e-06, + "loss": 0.2756, + "step": 6362 + }, + { + "epoch": 0.43525548943156167, + "grad_norm": 2.355529075849596, + "learning_rate": 6.2773239897213524e-06, + "loss": 0.6689, + "step": 6363 + }, + { + "epoch": 0.4353238935631712, + "grad_norm": 1.6967827007840346, + "learning_rate": 6.2762529606909925e-06, + "loss": 0.2208, + "step": 6364 + }, + { + "epoch": 0.43539229769478077, + "grad_norm": 2.0224014376623987, + "learning_rate": 6.2751818690159774e-06, + "loss": 0.318, + "step": 6365 + }, + { + "epoch": 0.4354607018263903, + "grad_norm": 2.3762963922859224, + "learning_rate": 6.274110714748876e-06, + "loss": 0.5174, + "step": 6366 + }, + { + "epoch": 0.4355291059579999, + "grad_norm": 2.1954164853701976, + "learning_rate": 6.2730394979422695e-06, + "loss": 0.7328, + "step": 6367 + }, + { + "epoch": 0.4355975100896094, + "grad_norm": 1.8215245153498338, + "learning_rate": 6.2719682186487366e-06, + "loss": 0.2964, + "step": 6368 + }, + { + "epoch": 0.435665914221219, + "grad_norm": 1.952675019880889, + "learning_rate": 6.27089687692086e-06, + "loss": 0.5855, + "step": 6369 + }, + { + "epoch": 0.4357343183528285, + "grad_norm": 2.5449720247469845, + "learning_rate": 6.2698254728112295e-06, + "loss": 0.59, + "step": 6370 + }, + { + "epoch": 0.4358027224844381, + "grad_norm": 3.638451709321487, + "learning_rate": 6.268754006372433e-06, + "loss": 0.3407, + "step": 6371 + }, + { + "epoch": 0.4358711266160476, + "grad_norm": 1.9488049375584844, + "learning_rate": 6.267682477657061e-06, + "loss": 0.5218, + "step": 6372 + }, + { + "epoch": 0.4359395307476572, + "grad_norm": 1.436442753479332, + "learning_rate": 6.266610886717712e-06, + "loss": 0.2256, + "step": 6373 + }, + { + "epoch": 0.4360079348792667, + "grad_norm": 1.9229340155395078, + "learning_rate": 6.265539233606985e-06, + "loss": 0.5665, + "step": 6374 + }, + { + "epoch": 0.4360763390108763, + "grad_norm": 1.6403342488918813, + "learning_rate": 6.264467518377483e-06, + "loss": 0.2059, + "step": 6375 + }, + { + "epoch": 0.4361447431424858, + "grad_norm": 2.18074502892864, + "learning_rate": 6.263395741081806e-06, + "loss": 0.6138, + "step": 6376 + }, + { + "epoch": 0.4362131472740954, + "grad_norm": 1.8742828984669828, + "learning_rate": 6.262323901772566e-06, + "loss": 0.2858, + "step": 6377 + }, + { + "epoch": 0.4362815514057049, + "grad_norm": 2.115352521645403, + "learning_rate": 6.2612520005023744e-06, + "loss": 0.5718, + "step": 6378 + }, + { + "epoch": 0.4363499555373144, + "grad_norm": 2.4418838949263613, + "learning_rate": 6.260180037323843e-06, + "loss": 0.7182, + "step": 6379 + }, + { + "epoch": 0.436418359668924, + "grad_norm": 1.5693158550200001, + "learning_rate": 6.259108012289591e-06, + "loss": 0.3408, + "step": 6380 + }, + { + "epoch": 0.43648676380053353, + "grad_norm": 1.9067764337022055, + "learning_rate": 6.2580359254522375e-06, + "loss": 0.2423, + "step": 6381 + }, + { + "epoch": 0.4365551679321431, + "grad_norm": 1.9843026996567463, + "learning_rate": 6.256963776864404e-06, + "loss": 0.3267, + "step": 6382 + }, + { + "epoch": 0.43662357206375263, + "grad_norm": 1.7797756451220488, + "learning_rate": 6.25589156657872e-06, + "loss": 0.3184, + "step": 6383 + }, + { + "epoch": 0.4366919761953622, + "grad_norm": 2.0262706099326278, + "learning_rate": 6.254819294647814e-06, + "loss": 0.2194, + "step": 6384 + }, + { + "epoch": 0.43676038032697173, + "grad_norm": 2.8561542747450077, + "learning_rate": 6.253746961124316e-06, + "loss": 0.4796, + "step": 6385 + }, + { + "epoch": 0.4368287844585813, + "grad_norm": 2.1998424695792442, + "learning_rate": 6.252674566060863e-06, + "loss": 0.5835, + "step": 6386 + }, + { + "epoch": 0.43689718859019083, + "grad_norm": 2.517250063335946, + "learning_rate": 6.251602109510095e-06, + "loss": 0.2786, + "step": 6387 + }, + { + "epoch": 0.4369655927218004, + "grad_norm": 2.0576342639477017, + "learning_rate": 6.250529591524649e-06, + "loss": 0.4187, + "step": 6388 + }, + { + "epoch": 0.43703399685340993, + "grad_norm": 3.2783580933531016, + "learning_rate": 6.249457012157173e-06, + "loss": 0.4225, + "step": 6389 + }, + { + "epoch": 0.4371024009850195, + "grad_norm": 2.0106088335634165, + "learning_rate": 6.248384371460313e-06, + "loss": 0.445, + "step": 6390 + }, + { + "epoch": 0.43717080511662904, + "grad_norm": 2.3049156270027975, + "learning_rate": 6.2473116694867206e-06, + "loss": 0.7305, + "step": 6391 + }, + { + "epoch": 0.4372392092482386, + "grad_norm": 2.354415490425316, + "learning_rate": 6.246238906289047e-06, + "loss": 0.4949, + "step": 6392 + }, + { + "epoch": 0.43730761337984814, + "grad_norm": 2.605975815189074, + "learning_rate": 6.24516608191995e-06, + "loss": 0.5569, + "step": 6393 + }, + { + "epoch": 0.4373760175114577, + "grad_norm": 1.903944798714423, + "learning_rate": 6.2440931964320905e-06, + "loss": 0.4211, + "step": 6394 + }, + { + "epoch": 0.43744442164306724, + "grad_norm": 1.8684048295598097, + "learning_rate": 6.243020249878129e-06, + "loss": 0.2294, + "step": 6395 + }, + { + "epoch": 0.4375128257746768, + "grad_norm": 1.750978851742373, + "learning_rate": 6.241947242310729e-06, + "loss": 0.4602, + "step": 6396 + }, + { + "epoch": 0.43758122990628634, + "grad_norm": 1.7292090151285202, + "learning_rate": 6.240874173782563e-06, + "loss": 0.4414, + "step": 6397 + }, + { + "epoch": 0.43764963403789586, + "grad_norm": 2.1364951879787286, + "learning_rate": 6.2398010443463e-06, + "loss": 0.4536, + "step": 6398 + }, + { + "epoch": 0.43771803816950544, + "grad_norm": 3.54845178736469, + "learning_rate": 6.238727854054614e-06, + "loss": 0.3498, + "step": 6399 + }, + { + "epoch": 0.43778644230111496, + "grad_norm": 1.923827075473867, + "learning_rate": 6.237654602960184e-06, + "loss": 0.2662, + "step": 6400 + }, + { + "epoch": 0.43785484643272454, + "grad_norm": 1.9858027525777735, + "learning_rate": 6.23658129111569e-06, + "loss": 0.5142, + "step": 6401 + }, + { + "epoch": 0.43792325056433407, + "grad_norm": 2.5644881165463964, + "learning_rate": 6.235507918573813e-06, + "loss": 0.6119, + "step": 6402 + }, + { + "epoch": 0.43799165469594364, + "grad_norm": 2.3614142349356535, + "learning_rate": 6.234434485387242e-06, + "loss": 0.3738, + "step": 6403 + }, + { + "epoch": 0.43806005882755317, + "grad_norm": 2.041844549182628, + "learning_rate": 6.233360991608666e-06, + "loss": 0.3511, + "step": 6404 + }, + { + "epoch": 0.43812846295916275, + "grad_norm": 2.340706960569854, + "learning_rate": 6.232287437290777e-06, + "loss": 0.4644, + "step": 6405 + }, + { + "epoch": 0.43819686709077227, + "grad_norm": 2.096172042106794, + "learning_rate": 6.231213822486268e-06, + "loss": 0.6279, + "step": 6406 + }, + { + "epoch": 0.43826527122238185, + "grad_norm": 1.6587765334651163, + "learning_rate": 6.230140147247842e-06, + "loss": 0.2523, + "step": 6407 + }, + { + "epoch": 0.43833367535399137, + "grad_norm": 1.9393236015226067, + "learning_rate": 6.229066411628195e-06, + "loss": 0.2236, + "step": 6408 + }, + { + "epoch": 0.43840207948560095, + "grad_norm": 1.921766146236581, + "learning_rate": 6.227992615680033e-06, + "loss": 0.2618, + "step": 6409 + }, + { + "epoch": 0.43847048361721047, + "grad_norm": 2.5306632221300314, + "learning_rate": 6.226918759456065e-06, + "loss": 0.6603, + "step": 6410 + }, + { + "epoch": 0.43853888774882005, + "grad_norm": 1.6280098284953666, + "learning_rate": 6.2258448430090005e-06, + "loss": 0.3614, + "step": 6411 + }, + { + "epoch": 0.4386072918804296, + "grad_norm": 1.837531732752305, + "learning_rate": 6.2247708663915495e-06, + "loss": 0.3793, + "step": 6412 + }, + { + "epoch": 0.43867569601203915, + "grad_norm": 2.186620804670515, + "learning_rate": 6.223696829656429e-06, + "loss": 0.4492, + "step": 6413 + }, + { + "epoch": 0.4387441001436487, + "grad_norm": 1.5210647543018059, + "learning_rate": 6.222622732856362e-06, + "loss": 0.3451, + "step": 6414 + }, + { + "epoch": 0.43881250427525825, + "grad_norm": 1.7048781781325346, + "learning_rate": 6.221548576044065e-06, + "loss": 0.353, + "step": 6415 + }, + { + "epoch": 0.4388809084068678, + "grad_norm": 1.7638479421607112, + "learning_rate": 6.2204743592722676e-06, + "loss": 0.3827, + "step": 6416 + }, + { + "epoch": 0.4389493125384773, + "grad_norm": 1.5492438904400345, + "learning_rate": 6.219400082593694e-06, + "loss": 0.2962, + "step": 6417 + }, + { + "epoch": 0.4390177166700869, + "grad_norm": 1.760051191829036, + "learning_rate": 6.218325746061075e-06, + "loss": 0.3576, + "step": 6418 + }, + { + "epoch": 0.4390861208016964, + "grad_norm": 1.4703256140275094, + "learning_rate": 6.217251349727145e-06, + "loss": 0.2463, + "step": 6419 + }, + { + "epoch": 0.439154524933306, + "grad_norm": 2.481727797197475, + "learning_rate": 6.216176893644641e-06, + "loss": 0.4377, + "step": 6420 + }, + { + "epoch": 0.4392229290649155, + "grad_norm": 1.6237120889036525, + "learning_rate": 6.2151023778663034e-06, + "loss": 0.2121, + "step": 6421 + }, + { + "epoch": 0.4392913331965251, + "grad_norm": 1.4895332506959287, + "learning_rate": 6.214027802444872e-06, + "loss": 0.1939, + "step": 6422 + }, + { + "epoch": 0.4393597373281346, + "grad_norm": 1.8753266010348633, + "learning_rate": 6.2129531674330936e-06, + "loss": 0.4586, + "step": 6423 + }, + { + "epoch": 0.4394281414597442, + "grad_norm": 2.220350614968217, + "learning_rate": 6.211878472883717e-06, + "loss": 0.6088, + "step": 6424 + }, + { + "epoch": 0.4394965455913537, + "grad_norm": 2.3405180136316015, + "learning_rate": 6.210803718849492e-06, + "loss": 0.5198, + "step": 6425 + }, + { + "epoch": 0.4395649497229633, + "grad_norm": 1.8931393967185277, + "learning_rate": 6.209728905383176e-06, + "loss": 0.3747, + "step": 6426 + }, + { + "epoch": 0.4396333538545728, + "grad_norm": 1.785274664459398, + "learning_rate": 6.2086540325375225e-06, + "loss": 0.298, + "step": 6427 + }, + { + "epoch": 0.4397017579861824, + "grad_norm": 2.2332534342649404, + "learning_rate": 6.207579100365291e-06, + "loss": 0.5244, + "step": 6428 + }, + { + "epoch": 0.4397701621177919, + "grad_norm": 1.708128343084074, + "learning_rate": 6.206504108919246e-06, + "loss": 0.1098, + "step": 6429 + }, + { + "epoch": 0.4398385662494015, + "grad_norm": 2.001843682996052, + "learning_rate": 6.205429058252155e-06, + "loss": 0.47, + "step": 6430 + }, + { + "epoch": 0.439906970381011, + "grad_norm": 2.273471687820155, + "learning_rate": 6.204353948416784e-06, + "loss": 0.3171, + "step": 6431 + }, + { + "epoch": 0.4399753745126206, + "grad_norm": 1.7083226584957747, + "learning_rate": 6.203278779465905e-06, + "loss": 0.3563, + "step": 6432 + }, + { + "epoch": 0.4400437786442301, + "grad_norm": 1.8481507637594092, + "learning_rate": 6.202203551452293e-06, + "loss": 0.2606, + "step": 6433 + }, + { + "epoch": 0.44011218277583963, + "grad_norm": 2.6589004672605685, + "learning_rate": 6.201128264428726e-06, + "loss": 0.6154, + "step": 6434 + }, + { + "epoch": 0.4401805869074492, + "grad_norm": 2.113860257237922, + "learning_rate": 6.200052918447983e-06, + "loss": 0.3044, + "step": 6435 + }, + { + "epoch": 0.44024899103905873, + "grad_norm": 2.724472983169807, + "learning_rate": 6.198977513562849e-06, + "loss": 0.7442, + "step": 6436 + }, + { + "epoch": 0.4403173951706683, + "grad_norm": 1.8504740195540352, + "learning_rate": 6.1979020498261075e-06, + "loss": 0.3663, + "step": 6437 + }, + { + "epoch": 0.44038579930227784, + "grad_norm": 2.0468563959863277, + "learning_rate": 6.196826527290549e-06, + "loss": 0.5403, + "step": 6438 + }, + { + "epoch": 0.4404542034338874, + "grad_norm": 2.203310329677385, + "learning_rate": 6.195750946008965e-06, + "loss": 0.5851, + "step": 6439 + }, + { + "epoch": 0.44052260756549694, + "grad_norm": 2.204249636062621, + "learning_rate": 6.19467530603415e-06, + "loss": 0.5235, + "step": 6440 + }, + { + "epoch": 0.4405910116971065, + "grad_norm": 2.243062021594246, + "learning_rate": 6.193599607418903e-06, + "loss": 0.484, + "step": 6441 + }, + { + "epoch": 0.44065941582871604, + "grad_norm": 1.99397043964682, + "learning_rate": 6.192523850216023e-06, + "loss": 0.3442, + "step": 6442 + }, + { + "epoch": 0.4407278199603256, + "grad_norm": 1.8523012047768304, + "learning_rate": 6.191448034478314e-06, + "loss": 0.3924, + "step": 6443 + }, + { + "epoch": 0.44079622409193514, + "grad_norm": 1.7307282160724697, + "learning_rate": 6.190372160258582e-06, + "loss": 0.3667, + "step": 6444 + }, + { + "epoch": 0.4408646282235447, + "grad_norm": 2.361572606356748, + "learning_rate": 6.189296227609634e-06, + "loss": 0.611, + "step": 6445 + }, + { + "epoch": 0.44093303235515424, + "grad_norm": 1.830030495625335, + "learning_rate": 6.188220236584288e-06, + "loss": 0.317, + "step": 6446 + }, + { + "epoch": 0.4410014364867638, + "grad_norm": 2.585452307968527, + "learning_rate": 6.187144187235353e-06, + "loss": 0.4567, + "step": 6447 + }, + { + "epoch": 0.44106984061837334, + "grad_norm": 2.3501865822702737, + "learning_rate": 6.186068079615648e-06, + "loss": 0.4878, + "step": 6448 + }, + { + "epoch": 0.4411382447499829, + "grad_norm": 1.7052894321522114, + "learning_rate": 6.1849919137779945e-06, + "loss": 0.4763, + "step": 6449 + }, + { + "epoch": 0.44120664888159244, + "grad_norm": 2.021144805251548, + "learning_rate": 6.183915689775216e-06, + "loss": 0.5374, + "step": 6450 + }, + { + "epoch": 0.441275053013202, + "grad_norm": 1.4761232802666677, + "learning_rate": 6.182839407660139e-06, + "loss": 0.1895, + "step": 6451 + }, + { + "epoch": 0.44134345714481155, + "grad_norm": 1.9129639801362257, + "learning_rate": 6.181763067485591e-06, + "loss": 0.4892, + "step": 6452 + }, + { + "epoch": 0.44141186127642107, + "grad_norm": 2.003444841591892, + "learning_rate": 6.180686669304404e-06, + "loss": 0.6876, + "step": 6453 + }, + { + "epoch": 0.44148026540803065, + "grad_norm": 2.404162826953733, + "learning_rate": 6.179610213169416e-06, + "loss": 0.4358, + "step": 6454 + }, + { + "epoch": 0.44154866953964017, + "grad_norm": 2.048485486350904, + "learning_rate": 6.178533699133461e-06, + "loss": 0.5761, + "step": 6455 + }, + { + "epoch": 0.44161707367124975, + "grad_norm": 1.8264732566035609, + "learning_rate": 6.177457127249382e-06, + "loss": 0.2005, + "step": 6456 + }, + { + "epoch": 0.44168547780285927, + "grad_norm": 1.5580363412200753, + "learning_rate": 6.1763804975700216e-06, + "loss": 0.2557, + "step": 6457 + }, + { + "epoch": 0.44175388193446885, + "grad_norm": 1.522577478280567, + "learning_rate": 6.175303810148223e-06, + "loss": 0.1984, + "step": 6458 + }, + { + "epoch": 0.4418222860660784, + "grad_norm": 2.196277067118616, + "learning_rate": 6.1742270650368395e-06, + "loss": 0.484, + "step": 6459 + }, + { + "epoch": 0.44189069019768795, + "grad_norm": 1.9634635354914007, + "learning_rate": 6.173150262288723e-06, + "loss": 0.4761, + "step": 6460 + }, + { + "epoch": 0.4419590943292975, + "grad_norm": 1.9851077033678295, + "learning_rate": 6.1720734019567255e-06, + "loss": 0.5403, + "step": 6461 + }, + { + "epoch": 0.44202749846090705, + "grad_norm": 1.7562482189020105, + "learning_rate": 6.170996484093704e-06, + "loss": 0.3914, + "step": 6462 + }, + { + "epoch": 0.4420959025925166, + "grad_norm": 1.8387799558851021, + "learning_rate": 6.169919508752522e-06, + "loss": 0.281, + "step": 6463 + }, + { + "epoch": 0.44216430672412615, + "grad_norm": 1.627971332843466, + "learning_rate": 6.168842475986041e-06, + "loss": 0.164, + "step": 6464 + }, + { + "epoch": 0.4422327108557357, + "grad_norm": 1.4275710799196253, + "learning_rate": 6.167765385847125e-06, + "loss": 0.2101, + "step": 6465 + }, + { + "epoch": 0.44230111498734526, + "grad_norm": 1.494458756829831, + "learning_rate": 6.166688238388647e-06, + "loss": 0.3504, + "step": 6466 + }, + { + "epoch": 0.4423695191189548, + "grad_norm": 1.758521441443639, + "learning_rate": 6.165611033663475e-06, + "loss": 0.3987, + "step": 6467 + }, + { + "epoch": 0.44243792325056436, + "grad_norm": 1.867564478530926, + "learning_rate": 6.1645337717244866e-06, + "loss": 0.4102, + "step": 6468 + }, + { + "epoch": 0.4425063273821739, + "grad_norm": 2.023780130959623, + "learning_rate": 6.163456452624557e-06, + "loss": 0.4268, + "step": 6469 + }, + { + "epoch": 0.44257473151378346, + "grad_norm": 2.5717513295124133, + "learning_rate": 6.1623790764165644e-06, + "loss": 0.274, + "step": 6470 + }, + { + "epoch": 0.442643135645393, + "grad_norm": 1.5350316800349304, + "learning_rate": 6.161301643153394e-06, + "loss": 0.1739, + "step": 6471 + }, + { + "epoch": 0.4427115397770025, + "grad_norm": 2.3153261429421668, + "learning_rate": 6.160224152887932e-06, + "loss": 0.5974, + "step": 6472 + }, + { + "epoch": 0.4427799439086121, + "grad_norm": 2.141246837595816, + "learning_rate": 6.159146605673066e-06, + "loss": 0.1801, + "step": 6473 + }, + { + "epoch": 0.4428483480402216, + "grad_norm": 2.0438097285954187, + "learning_rate": 6.158069001561689e-06, + "loss": 0.4211, + "step": 6474 + }, + { + "epoch": 0.4429167521718312, + "grad_norm": 2.065336657293508, + "learning_rate": 6.156991340606689e-06, + "loss": 0.3466, + "step": 6475 + }, + { + "epoch": 0.4429851563034407, + "grad_norm": 1.930928742283477, + "learning_rate": 6.15591362286097e-06, + "loss": 0.4069, + "step": 6476 + }, + { + "epoch": 0.4430535604350503, + "grad_norm": 2.0089782418059694, + "learning_rate": 6.154835848377427e-06, + "loss": 0.2706, + "step": 6477 + }, + { + "epoch": 0.4431219645666598, + "grad_norm": 2.3971904936179875, + "learning_rate": 6.153758017208965e-06, + "loss": 0.3573, + "step": 6478 + }, + { + "epoch": 0.4431903686982694, + "grad_norm": 1.5757095079445442, + "learning_rate": 6.152680129408488e-06, + "loss": 0.2286, + "step": 6479 + }, + { + "epoch": 0.4432587728298789, + "grad_norm": 1.8704311236828188, + "learning_rate": 6.1516021850289045e-06, + "loss": 0.2512, + "step": 6480 + }, + { + "epoch": 0.4433271769614885, + "grad_norm": 3.453464560405729, + "learning_rate": 6.1505241841231225e-06, + "loss": 0.3138, + "step": 6481 + }, + { + "epoch": 0.443395581093098, + "grad_norm": 1.6711251725376441, + "learning_rate": 6.149446126744059e-06, + "loss": 0.318, + "step": 6482 + }, + { + "epoch": 0.4434639852247076, + "grad_norm": 2.2649204474072246, + "learning_rate": 6.148368012944629e-06, + "loss": 0.3024, + "step": 6483 + }, + { + "epoch": 0.4435323893563171, + "grad_norm": 2.3885680087972, + "learning_rate": 6.147289842777753e-06, + "loss": 0.3247, + "step": 6484 + }, + { + "epoch": 0.4436007934879267, + "grad_norm": 1.992019399394803, + "learning_rate": 6.146211616296349e-06, + "loss": 0.4346, + "step": 6485 + }, + { + "epoch": 0.4436691976195362, + "grad_norm": 2.497527139266128, + "learning_rate": 6.145133333553345e-06, + "loss": 0.1703, + "step": 6486 + }, + { + "epoch": 0.4437376017511458, + "grad_norm": 3.0516043223119294, + "learning_rate": 6.144054994601667e-06, + "loss": 0.461, + "step": 6487 + }, + { + "epoch": 0.4438060058827553, + "grad_norm": 2.09199404032161, + "learning_rate": 6.1429765994942435e-06, + "loss": 0.5505, + "step": 6488 + }, + { + "epoch": 0.4438744100143649, + "grad_norm": 1.7690153232976071, + "learning_rate": 6.141898148284012e-06, + "loss": 0.2934, + "step": 6489 + }, + { + "epoch": 0.4439428141459744, + "grad_norm": 1.963463592410567, + "learning_rate": 6.140819641023905e-06, + "loss": 0.4804, + "step": 6490 + }, + { + "epoch": 0.44401121827758394, + "grad_norm": 2.1641340416055237, + "learning_rate": 6.139741077766858e-06, + "loss": 0.4752, + "step": 6491 + }, + { + "epoch": 0.4440796224091935, + "grad_norm": 2.80099340723318, + "learning_rate": 6.138662458565818e-06, + "loss": 0.2689, + "step": 6492 + }, + { + "epoch": 0.44414802654080304, + "grad_norm": 2.0148907773171496, + "learning_rate": 6.1375837834737265e-06, + "loss": 0.1733, + "step": 6493 + }, + { + "epoch": 0.4442164306724126, + "grad_norm": 3.635501530849075, + "learning_rate": 6.136505052543528e-06, + "loss": 0.3992, + "step": 6494 + }, + { + "epoch": 0.44428483480402214, + "grad_norm": 2.4192476233829114, + "learning_rate": 6.1354262658281725e-06, + "loss": 0.5761, + "step": 6495 + }, + { + "epoch": 0.4443532389356317, + "grad_norm": 2.037568930179579, + "learning_rate": 6.1343474233806156e-06, + "loss": 0.4151, + "step": 6496 + }, + { + "epoch": 0.44442164306724125, + "grad_norm": 1.5074559714243998, + "learning_rate": 6.133268525253807e-06, + "loss": 0.2793, + "step": 6497 + }, + { + "epoch": 0.4444900471988508, + "grad_norm": 1.871323529590861, + "learning_rate": 6.132189571500707e-06, + "loss": 0.2599, + "step": 6498 + }, + { + "epoch": 0.44455845133046035, + "grad_norm": 1.8334868490542255, + "learning_rate": 6.1311105621742775e-06, + "loss": 0.2004, + "step": 6499 + }, + { + "epoch": 0.4446268554620699, + "grad_norm": 2.004490854455586, + "learning_rate": 6.130031497327479e-06, + "loss": 0.5308, + "step": 6500 + }, + { + "epoch": 0.44469525959367945, + "grad_norm": 3.0149461940848084, + "learning_rate": 6.128952377013276e-06, + "loss": 0.6948, + "step": 6501 + }, + { + "epoch": 0.444763663725289, + "grad_norm": 2.5694430918661344, + "learning_rate": 6.1278732012846385e-06, + "loss": 0.5456, + "step": 6502 + }, + { + "epoch": 0.44483206785689855, + "grad_norm": 1.8547193117956666, + "learning_rate": 6.12679397019454e-06, + "loss": 0.2115, + "step": 6503 + }, + { + "epoch": 0.44490047198850813, + "grad_norm": 1.8397992527596223, + "learning_rate": 6.1257146837959514e-06, + "loss": 0.4821, + "step": 6504 + }, + { + "epoch": 0.44496887612011765, + "grad_norm": 1.6657762389022241, + "learning_rate": 6.12463534214185e-06, + "loss": 0.2313, + "step": 6505 + }, + { + "epoch": 0.44503728025172723, + "grad_norm": 1.2425334976370528, + "learning_rate": 6.123555945285215e-06, + "loss": 0.2883, + "step": 6506 + }, + { + "epoch": 0.44510568438333675, + "grad_norm": 2.5857641156204196, + "learning_rate": 6.122476493279029e-06, + "loss": 0.7705, + "step": 6507 + }, + { + "epoch": 0.4451740885149463, + "grad_norm": 2.271737436382606, + "learning_rate": 6.1213969861762754e-06, + "loss": 0.312, + "step": 6508 + }, + { + "epoch": 0.44524249264655585, + "grad_norm": 1.805283897903646, + "learning_rate": 6.120317424029944e-06, + "loss": 0.1982, + "step": 6509 + }, + { + "epoch": 0.4453108967781654, + "grad_norm": 2.0505900411930833, + "learning_rate": 6.1192378068930225e-06, + "loss": 0.4717, + "step": 6510 + }, + { + "epoch": 0.44537930090977496, + "grad_norm": 2.1730888185733828, + "learning_rate": 6.118158134818505e-06, + "loss": 0.4796, + "step": 6511 + }, + { + "epoch": 0.4454477050413845, + "grad_norm": 2.406493218007568, + "learning_rate": 6.117078407859386e-06, + "loss": 0.4426, + "step": 6512 + }, + { + "epoch": 0.44551610917299406, + "grad_norm": 1.5765522116258264, + "learning_rate": 6.115998626068666e-06, + "loss": 0.2366, + "step": 6513 + }, + { + "epoch": 0.4455845133046036, + "grad_norm": 1.9697445446398327, + "learning_rate": 6.114918789499344e-06, + "loss": 0.2572, + "step": 6514 + }, + { + "epoch": 0.44565291743621316, + "grad_norm": 4.410589935089834, + "learning_rate": 6.113838898204421e-06, + "loss": 0.4519, + "step": 6515 + }, + { + "epoch": 0.4457213215678227, + "grad_norm": 1.7803934112316782, + "learning_rate": 6.11275895223691e-06, + "loss": 0.3905, + "step": 6516 + }, + { + "epoch": 0.44578972569943226, + "grad_norm": 2.0603461217487475, + "learning_rate": 6.111678951649814e-06, + "loss": 0.3783, + "step": 6517 + }, + { + "epoch": 0.4458581298310418, + "grad_norm": 1.284582401269213, + "learning_rate": 6.110598896496147e-06, + "loss": 0.1492, + "step": 6518 + }, + { + "epoch": 0.44592653396265136, + "grad_norm": 2.140177902810638, + "learning_rate": 6.109518786828924e-06, + "loss": 0.5799, + "step": 6519 + }, + { + "epoch": 0.4459949380942609, + "grad_norm": 2.4145346491538, + "learning_rate": 6.1084386227011615e-06, + "loss": 0.4208, + "step": 6520 + }, + { + "epoch": 0.44606334222587046, + "grad_norm": 2.159822619302449, + "learning_rate": 6.1073584041658775e-06, + "loss": 0.311, + "step": 6521 + }, + { + "epoch": 0.44613174635748, + "grad_norm": 2.2469700770628784, + "learning_rate": 6.106278131276095e-06, + "loss": 0.4792, + "step": 6522 + }, + { + "epoch": 0.44620015048908956, + "grad_norm": 1.7408464128285341, + "learning_rate": 6.105197804084841e-06, + "loss": 0.2163, + "step": 6523 + }, + { + "epoch": 0.4462685546206991, + "grad_norm": 1.6380724077540407, + "learning_rate": 6.104117422645141e-06, + "loss": 0.2407, + "step": 6524 + }, + { + "epoch": 0.44633695875230867, + "grad_norm": 2.690174092026842, + "learning_rate": 6.103036987010028e-06, + "loss": 0.3036, + "step": 6525 + }, + { + "epoch": 0.4464053628839182, + "grad_norm": 2.119918842078029, + "learning_rate": 6.101956497232533e-06, + "loss": 0.5018, + "step": 6526 + }, + { + "epoch": 0.4464737670155277, + "grad_norm": 1.6000234326494451, + "learning_rate": 6.10087595336569e-06, + "loss": 0.3962, + "step": 6527 + }, + { + "epoch": 0.4465421711471373, + "grad_norm": 1.9201314196710555, + "learning_rate": 6.09979535546254e-06, + "loss": 0.1839, + "step": 6528 + }, + { + "epoch": 0.4466105752787468, + "grad_norm": 2.0805258900283583, + "learning_rate": 6.098714703576123e-06, + "loss": 0.3635, + "step": 6529 + }, + { + "epoch": 0.4466789794103564, + "grad_norm": 1.8452221965416173, + "learning_rate": 6.097633997759483e-06, + "loss": 0.3562, + "step": 6530 + }, + { + "epoch": 0.4467473835419659, + "grad_norm": 3.6104137692858997, + "learning_rate": 6.096553238065666e-06, + "loss": 0.6348, + "step": 6531 + }, + { + "epoch": 0.4468157876735755, + "grad_norm": 1.8337890147197886, + "learning_rate": 6.095472424547719e-06, + "loss": 0.5025, + "step": 6532 + }, + { + "epoch": 0.446884191805185, + "grad_norm": 2.4277615748973997, + "learning_rate": 6.0943915572586985e-06, + "loss": 0.3619, + "step": 6533 + }, + { + "epoch": 0.4469525959367946, + "grad_norm": 2.5841671222355047, + "learning_rate": 6.093310636251653e-06, + "loss": 0.4634, + "step": 6534 + }, + { + "epoch": 0.4470210000684041, + "grad_norm": 1.973937421166223, + "learning_rate": 6.092229661579644e-06, + "loss": 0.4796, + "step": 6535 + }, + { + "epoch": 0.4470894042000137, + "grad_norm": 2.552536907891291, + "learning_rate": 6.0911486332957286e-06, + "loss": 0.3373, + "step": 6536 + }, + { + "epoch": 0.4471578083316232, + "grad_norm": 2.007960673998632, + "learning_rate": 6.090067551452967e-06, + "loss": 0.4031, + "step": 6537 + }, + { + "epoch": 0.4472262124632328, + "grad_norm": 2.328060747706492, + "learning_rate": 6.088986416104428e-06, + "loss": 0.4657, + "step": 6538 + }, + { + "epoch": 0.4472946165948423, + "grad_norm": 2.1605011662910667, + "learning_rate": 6.087905227303177e-06, + "loss": 0.6124, + "step": 6539 + }, + { + "epoch": 0.4473630207264519, + "grad_norm": 2.0687238103816528, + "learning_rate": 6.0868239851022834e-06, + "loss": 0.2345, + "step": 6540 + }, + { + "epoch": 0.4474314248580614, + "grad_norm": 2.070238781462263, + "learning_rate": 6.085742689554821e-06, + "loss": 0.4136, + "step": 6541 + }, + { + "epoch": 0.447499828989671, + "grad_norm": 1.5384282775639673, + "learning_rate": 6.084661340713864e-06, + "loss": 0.3327, + "step": 6542 + }, + { + "epoch": 0.4475682331212805, + "grad_norm": 2.8933666733353194, + "learning_rate": 6.083579938632491e-06, + "loss": 0.484, + "step": 6543 + }, + { + "epoch": 0.4476366372528901, + "grad_norm": 1.3140923853865283, + "learning_rate": 6.0824984833637815e-06, + "loss": 0.1986, + "step": 6544 + }, + { + "epoch": 0.4477050413844996, + "grad_norm": 1.59753462041658, + "learning_rate": 6.081416974960819e-06, + "loss": 0.286, + "step": 6545 + }, + { + "epoch": 0.44777344551610915, + "grad_norm": 2.0394999228564203, + "learning_rate": 6.08033541347669e-06, + "loss": 0.6181, + "step": 6546 + }, + { + "epoch": 0.4478418496477187, + "grad_norm": 2.2088620449634155, + "learning_rate": 6.0792537989644816e-06, + "loss": 0.414, + "step": 6547 + }, + { + "epoch": 0.44791025377932825, + "grad_norm": 2.199950081356457, + "learning_rate": 6.078172131477285e-06, + "loss": 0.5935, + "step": 6548 + }, + { + "epoch": 0.4479786579109378, + "grad_norm": 2.2133591535533856, + "learning_rate": 6.077090411068194e-06, + "loss": 0.331, + "step": 6549 + }, + { + "epoch": 0.44804706204254735, + "grad_norm": 2.1851613793092985, + "learning_rate": 6.076008637790305e-06, + "loss": 0.3832, + "step": 6550 + }, + { + "epoch": 0.44811546617415693, + "grad_norm": 2.171813387648662, + "learning_rate": 6.074926811696714e-06, + "loss": 0.5002, + "step": 6551 + }, + { + "epoch": 0.44818387030576645, + "grad_norm": 2.0568338576902105, + "learning_rate": 6.0738449328405245e-06, + "loss": 0.3827, + "step": 6552 + }, + { + "epoch": 0.44825227443737603, + "grad_norm": 2.0277036042470042, + "learning_rate": 6.072763001274843e-06, + "loss": 0.5193, + "step": 6553 + }, + { + "epoch": 0.44832067856898555, + "grad_norm": 2.241205432361499, + "learning_rate": 6.071681017052771e-06, + "loss": 0.6363, + "step": 6554 + }, + { + "epoch": 0.44838908270059513, + "grad_norm": 1.7415424230360521, + "learning_rate": 6.07059898022742e-06, + "loss": 0.3126, + "step": 6555 + }, + { + "epoch": 0.44845748683220465, + "grad_norm": 2.404632659569344, + "learning_rate": 6.069516890851901e-06, + "loss": 0.6722, + "step": 6556 + }, + { + "epoch": 0.44852589096381423, + "grad_norm": 2.0412179289948016, + "learning_rate": 6.068434748979328e-06, + "loss": 0.3014, + "step": 6557 + }, + { + "epoch": 0.44859429509542376, + "grad_norm": 2.1202570492627597, + "learning_rate": 6.067352554662819e-06, + "loss": 0.2702, + "step": 6558 + }, + { + "epoch": 0.44866269922703333, + "grad_norm": 2.076025479485364, + "learning_rate": 6.066270307955492e-06, + "loss": 0.4564, + "step": 6559 + }, + { + "epoch": 0.44873110335864286, + "grad_norm": 1.8751898589967269, + "learning_rate": 6.0651880089104695e-06, + "loss": 0.3315, + "step": 6560 + }, + { + "epoch": 0.44879950749025244, + "grad_norm": 1.8175209890698953, + "learning_rate": 6.0641056575808755e-06, + "loss": 0.336, + "step": 6561 + }, + { + "epoch": 0.44886791162186196, + "grad_norm": 3.151412750756427, + "learning_rate": 6.063023254019837e-06, + "loss": 0.5621, + "step": 6562 + }, + { + "epoch": 0.4489363157534715, + "grad_norm": 2.0649665547749794, + "learning_rate": 6.061940798280484e-06, + "loss": 0.3508, + "step": 6563 + }, + { + "epoch": 0.44900471988508106, + "grad_norm": 2.1599541029756018, + "learning_rate": 6.060858290415948e-06, + "loss": 0.3123, + "step": 6564 + }, + { + "epoch": 0.4490731240166906, + "grad_norm": 1.7244686531553666, + "learning_rate": 6.059775730479367e-06, + "loss": 0.2649, + "step": 6565 + }, + { + "epoch": 0.44914152814830016, + "grad_norm": 1.7076229012899957, + "learning_rate": 6.058693118523874e-06, + "loss": 0.374, + "step": 6566 + }, + { + "epoch": 0.4492099322799097, + "grad_norm": 1.6600608865612232, + "learning_rate": 6.057610454602609e-06, + "loss": 0.3436, + "step": 6567 + }, + { + "epoch": 0.44927833641151926, + "grad_norm": 2.6520664255502737, + "learning_rate": 6.056527738768716e-06, + "loss": 0.3724, + "step": 6568 + }, + { + "epoch": 0.4493467405431288, + "grad_norm": 1.2689471250817328, + "learning_rate": 6.055444971075341e-06, + "loss": 0.2214, + "step": 6569 + }, + { + "epoch": 0.44941514467473836, + "grad_norm": 1.7119318019472272, + "learning_rate": 6.0543621515756305e-06, + "loss": 0.3288, + "step": 6570 + }, + { + "epoch": 0.4494835488063479, + "grad_norm": 2.024274019806688, + "learning_rate": 6.053279280322732e-06, + "loss": 0.5237, + "step": 6571 + }, + { + "epoch": 0.44955195293795747, + "grad_norm": 2.4034777221220582, + "learning_rate": 6.0521963573698015e-06, + "loss": 0.6092, + "step": 6572 + }, + { + "epoch": 0.449620357069567, + "grad_norm": 2.7507861054195337, + "learning_rate": 6.051113382769993e-06, + "loss": 0.4861, + "step": 6573 + }, + { + "epoch": 0.44968876120117657, + "grad_norm": 2.0441898321800376, + "learning_rate": 6.050030356576464e-06, + "loss": 0.4259, + "step": 6574 + }, + { + "epoch": 0.4497571653327861, + "grad_norm": 1.3962862821277295, + "learning_rate": 6.048947278842374e-06, + "loss": 0.162, + "step": 6575 + }, + { + "epoch": 0.44982556946439567, + "grad_norm": 1.8738713125691737, + "learning_rate": 6.047864149620886e-06, + "loss": 0.4004, + "step": 6576 + }, + { + "epoch": 0.4498939735960052, + "grad_norm": 1.8187810365906125, + "learning_rate": 6.046780968965166e-06, + "loss": 0.2122, + "step": 6577 + }, + { + "epoch": 0.44996237772761477, + "grad_norm": 3.0803401934708288, + "learning_rate": 6.045697736928381e-06, + "loss": 0.7595, + "step": 6578 + }, + { + "epoch": 0.4500307818592243, + "grad_norm": 2.144005320111389, + "learning_rate": 6.044614453563702e-06, + "loss": 0.3382, + "step": 6579 + }, + { + "epoch": 0.45009918599083387, + "grad_norm": 2.224568769370724, + "learning_rate": 6.0435311189242995e-06, + "loss": 0.6736, + "step": 6580 + }, + { + "epoch": 0.4501675901224434, + "grad_norm": 2.07895081109274, + "learning_rate": 6.0424477330633515e-06, + "loss": 0.4703, + "step": 6581 + }, + { + "epoch": 0.4502359942540529, + "grad_norm": 1.8832700497901798, + "learning_rate": 6.041364296034036e-06, + "loss": 0.3082, + "step": 6582 + }, + { + "epoch": 0.4503043983856625, + "grad_norm": 2.503215664806312, + "learning_rate": 6.040280807889531e-06, + "loss": 0.5501, + "step": 6583 + }, + { + "epoch": 0.450372802517272, + "grad_norm": 1.8813641109583568, + "learning_rate": 6.039197268683021e-06, + "loss": 0.3871, + "step": 6584 + }, + { + "epoch": 0.4504412066488816, + "grad_norm": 2.0583239963129243, + "learning_rate": 6.038113678467691e-06, + "loss": 0.3908, + "step": 6585 + }, + { + "epoch": 0.4505096107804911, + "grad_norm": 2.255013628535734, + "learning_rate": 6.037030037296727e-06, + "loss": 0.6653, + "step": 6586 + }, + { + "epoch": 0.4505780149121007, + "grad_norm": 1.7819412170237057, + "learning_rate": 6.035946345223323e-06, + "loss": 0.2483, + "step": 6587 + }, + { + "epoch": 0.4506464190437102, + "grad_norm": 2.320734932846352, + "learning_rate": 6.034862602300669e-06, + "loss": 0.5382, + "step": 6588 + }, + { + "epoch": 0.4507148231753198, + "grad_norm": 1.8038550351861808, + "learning_rate": 6.0337788085819615e-06, + "loss": 0.4185, + "step": 6589 + }, + { + "epoch": 0.4507832273069293, + "grad_norm": 2.266294709241709, + "learning_rate": 6.032694964120397e-06, + "loss": 0.7487, + "step": 6590 + }, + { + "epoch": 0.4508516314385389, + "grad_norm": 1.897389456226635, + "learning_rate": 6.031611068969177e-06, + "loss": 0.3573, + "step": 6591 + }, + { + "epoch": 0.4509200355701484, + "grad_norm": 2.115766724216244, + "learning_rate": 6.030527123181505e-06, + "loss": 0.6933, + "step": 6592 + }, + { + "epoch": 0.450988439701758, + "grad_norm": 2.2750727681165386, + "learning_rate": 6.029443126810586e-06, + "loss": 0.6996, + "step": 6593 + }, + { + "epoch": 0.4510568438333675, + "grad_norm": 1.5449706165133699, + "learning_rate": 6.028359079909626e-06, + "loss": 0.3024, + "step": 6594 + }, + { + "epoch": 0.4511252479649771, + "grad_norm": 2.0473126606013454, + "learning_rate": 6.027274982531837e-06, + "loss": 0.5517, + "step": 6595 + }, + { + "epoch": 0.45119365209658663, + "grad_norm": 2.047140097013244, + "learning_rate": 6.02619083473043e-06, + "loss": 0.5194, + "step": 6596 + }, + { + "epoch": 0.4512620562281962, + "grad_norm": 1.6329441815989068, + "learning_rate": 6.025106636558622e-06, + "loss": 0.2131, + "step": 6597 + }, + { + "epoch": 0.45133046035980573, + "grad_norm": 2.10299469315988, + "learning_rate": 6.024022388069632e-06, + "loss": 0.5632, + "step": 6598 + }, + { + "epoch": 0.4513988644914153, + "grad_norm": 2.14064720587803, + "learning_rate": 6.022938089316677e-06, + "loss": 0.3849, + "step": 6599 + }, + { + "epoch": 0.45146726862302483, + "grad_norm": 2.043151832677429, + "learning_rate": 6.021853740352978e-06, + "loss": 0.4064, + "step": 6600 + }, + { + "epoch": 0.45153567275463435, + "grad_norm": 2.2185002978945643, + "learning_rate": 6.020769341231766e-06, + "loss": 0.5535, + "step": 6601 + }, + { + "epoch": 0.45160407688624393, + "grad_norm": 2.225237991163462, + "learning_rate": 6.0196848920062655e-06, + "loss": 0.4737, + "step": 6602 + }, + { + "epoch": 0.45167248101785346, + "grad_norm": 1.6681052694801042, + "learning_rate": 6.018600392729706e-06, + "loss": 0.4015, + "step": 6603 + }, + { + "epoch": 0.45174088514946303, + "grad_norm": 2.246541000286682, + "learning_rate": 6.01751584345532e-06, + "loss": 0.3773, + "step": 6604 + }, + { + "epoch": 0.45180928928107256, + "grad_norm": 1.9276373973133862, + "learning_rate": 6.016431244236345e-06, + "loss": 0.3508, + "step": 6605 + }, + { + "epoch": 0.45187769341268214, + "grad_norm": 2.5141011059098393, + "learning_rate": 6.0153465951260135e-06, + "loss": 0.8777, + "step": 6606 + }, + { + "epoch": 0.45194609754429166, + "grad_norm": 2.8647821247597633, + "learning_rate": 6.014261896177569e-06, + "loss": 0.3279, + "step": 6607 + }, + { + "epoch": 0.45201450167590124, + "grad_norm": 1.9417926415332236, + "learning_rate": 6.013177147444253e-06, + "loss": 0.2848, + "step": 6608 + }, + { + "epoch": 0.45208290580751076, + "grad_norm": 1.7334233241913972, + "learning_rate": 6.01209234897931e-06, + "loss": 0.4372, + "step": 6609 + }, + { + "epoch": 0.45215130993912034, + "grad_norm": 2.1044640028055293, + "learning_rate": 6.011007500835987e-06, + "loss": 0.5255, + "step": 6610 + }, + { + "epoch": 0.45221971407072986, + "grad_norm": 2.2402757243403686, + "learning_rate": 6.009922603067533e-06, + "loss": 0.316, + "step": 6611 + }, + { + "epoch": 0.45228811820233944, + "grad_norm": 1.9969174986482312, + "learning_rate": 6.008837655727202e-06, + "loss": 0.4225, + "step": 6612 + }, + { + "epoch": 0.45235652233394896, + "grad_norm": 2.577962685301653, + "learning_rate": 6.007752658868248e-06, + "loss": 0.8284, + "step": 6613 + }, + { + "epoch": 0.45242492646555854, + "grad_norm": 2.4385928919508295, + "learning_rate": 6.006667612543925e-06, + "loss": 0.5055, + "step": 6614 + }, + { + "epoch": 0.45249333059716806, + "grad_norm": 2.1088430581648576, + "learning_rate": 6.005582516807495e-06, + "loss": 0.4059, + "step": 6615 + }, + { + "epoch": 0.45256173472877764, + "grad_norm": 1.6624043715013694, + "learning_rate": 6.00449737171222e-06, + "loss": 0.2604, + "step": 6616 + }, + { + "epoch": 0.45263013886038717, + "grad_norm": 2.911028819512533, + "learning_rate": 6.003412177311363e-06, + "loss": 0.6948, + "step": 6617 + }, + { + "epoch": 0.45269854299199674, + "grad_norm": 2.2525301026239117, + "learning_rate": 6.0023269336581915e-06, + "loss": 0.4956, + "step": 6618 + }, + { + "epoch": 0.45276694712360627, + "grad_norm": 2.264244560098278, + "learning_rate": 6.001241640805973e-06, + "loss": 0.6306, + "step": 6619 + }, + { + "epoch": 0.4528353512552158, + "grad_norm": 1.67186771819053, + "learning_rate": 6.00015629880798e-06, + "loss": 0.2767, + "step": 6620 + }, + { + "epoch": 0.45290375538682537, + "grad_norm": 2.195064035113758, + "learning_rate": 5.999070907717485e-06, + "loss": 0.3115, + "step": 6621 + }, + { + "epoch": 0.4529721595184349, + "grad_norm": 3.2855526824417285, + "learning_rate": 5.997985467587767e-06, + "loss": 0.1258, + "step": 6622 + }, + { + "epoch": 0.45304056365004447, + "grad_norm": 2.117364249646408, + "learning_rate": 5.996899978472103e-06, + "loss": 0.4404, + "step": 6623 + }, + { + "epoch": 0.453108967781654, + "grad_norm": 2.17453049163246, + "learning_rate": 5.995814440423774e-06, + "loss": 0.3602, + "step": 6624 + }, + { + "epoch": 0.45317737191326357, + "grad_norm": 2.8847519158184163, + "learning_rate": 5.994728853496063e-06, + "loss": 0.4646, + "step": 6625 + }, + { + "epoch": 0.4532457760448731, + "grad_norm": 1.9182352475240754, + "learning_rate": 5.993643217742257e-06, + "loss": 0.3151, + "step": 6626 + }, + { + "epoch": 0.4533141801764827, + "grad_norm": 1.9514012473979443, + "learning_rate": 5.992557533215642e-06, + "loss": 0.39, + "step": 6627 + }, + { + "epoch": 0.4533825843080922, + "grad_norm": 2.3662202451540977, + "learning_rate": 5.991471799969511e-06, + "loss": 0.4405, + "step": 6628 + }, + { + "epoch": 0.4534509884397018, + "grad_norm": 2.2345426736620615, + "learning_rate": 5.990386018057158e-06, + "loss": 0.6481, + "step": 6629 + }, + { + "epoch": 0.4535193925713113, + "grad_norm": 1.8152496100833646, + "learning_rate": 5.989300187531874e-06, + "loss": 0.4612, + "step": 6630 + }, + { + "epoch": 0.4535877967029209, + "grad_norm": 2.0202152082732696, + "learning_rate": 5.988214308446961e-06, + "loss": 0.32, + "step": 6631 + }, + { + "epoch": 0.4536562008345304, + "grad_norm": 2.334492584188141, + "learning_rate": 5.987128380855719e-06, + "loss": 0.4165, + "step": 6632 + }, + { + "epoch": 0.45372460496614, + "grad_norm": 1.6601008776237824, + "learning_rate": 5.9860424048114474e-06, + "loss": 0.1754, + "step": 6633 + }, + { + "epoch": 0.4537930090977495, + "grad_norm": 1.6074686982205582, + "learning_rate": 5.984956380367454e-06, + "loss": 0.1038, + "step": 6634 + }, + { + "epoch": 0.4538614132293591, + "grad_norm": 2.093505894964012, + "learning_rate": 5.983870307577044e-06, + "loss": 0.4778, + "step": 6635 + }, + { + "epoch": 0.4539298173609686, + "grad_norm": 2.415190090789602, + "learning_rate": 5.9827841864935286e-06, + "loss": 0.3449, + "step": 6636 + }, + { + "epoch": 0.4539982214925781, + "grad_norm": 1.8933074970823376, + "learning_rate": 5.98169801717022e-06, + "loss": 0.2973, + "step": 6637 + }, + { + "epoch": 0.4540666256241877, + "grad_norm": 2.469883911864268, + "learning_rate": 5.9806117996604325e-06, + "loss": 0.4434, + "step": 6638 + }, + { + "epoch": 0.4541350297557972, + "grad_norm": 6.220556370607536, + "learning_rate": 5.9795255340174825e-06, + "loss": 0.5968, + "step": 6639 + }, + { + "epoch": 0.4542034338874068, + "grad_norm": 1.636827201254612, + "learning_rate": 5.9784392202946885e-06, + "loss": 0.2826, + "step": 6640 + }, + { + "epoch": 0.4542718380190163, + "grad_norm": 1.730250831384849, + "learning_rate": 5.977352858545372e-06, + "loss": 0.1952, + "step": 6641 + }, + { + "epoch": 0.4543402421506259, + "grad_norm": 1.9473309482036438, + "learning_rate": 5.97626644882286e-06, + "loss": 0.4478, + "step": 6642 + }, + { + "epoch": 0.45440864628223543, + "grad_norm": 2.405585767049932, + "learning_rate": 5.975179991180474e-06, + "loss": 0.6304, + "step": 6643 + }, + { + "epoch": 0.454477050413845, + "grad_norm": 2.20094690129402, + "learning_rate": 5.974093485671546e-06, + "loss": 0.4717, + "step": 6644 + }, + { + "epoch": 0.45454545454545453, + "grad_norm": 2.3142735150014007, + "learning_rate": 5.9730069323494066e-06, + "loss": 0.178, + "step": 6645 + }, + { + "epoch": 0.4546138586770641, + "grad_norm": 1.8660144792329252, + "learning_rate": 5.971920331267387e-06, + "loss": 0.3919, + "step": 6646 + }, + { + "epoch": 0.45468226280867363, + "grad_norm": 1.5603442961747682, + "learning_rate": 5.970833682478824e-06, + "loss": 0.3013, + "step": 6647 + }, + { + "epoch": 0.4547506669402832, + "grad_norm": 2.276829021773926, + "learning_rate": 5.9697469860370574e-06, + "loss": 0.2936, + "step": 6648 + }, + { + "epoch": 0.45481907107189273, + "grad_norm": 1.6867836032006267, + "learning_rate": 5.968660241995426e-06, + "loss": 0.2294, + "step": 6649 + }, + { + "epoch": 0.4548874752035023, + "grad_norm": 1.969145099413086, + "learning_rate": 5.96757345040727e-06, + "loss": 0.5377, + "step": 6650 + }, + { + "epoch": 0.45495587933511183, + "grad_norm": 1.5287853800870141, + "learning_rate": 5.966486611325937e-06, + "loss": 0.1575, + "step": 6651 + }, + { + "epoch": 0.4550242834667214, + "grad_norm": 2.1842710710395283, + "learning_rate": 5.965399724804776e-06, + "loss": 0.5001, + "step": 6652 + }, + { + "epoch": 0.45509268759833094, + "grad_norm": 1.9120719134486477, + "learning_rate": 5.964312790897133e-06, + "loss": 0.2218, + "step": 6653 + }, + { + "epoch": 0.4551610917299405, + "grad_norm": 4.725461645634056, + "learning_rate": 5.963225809656362e-06, + "loss": 0.289, + "step": 6654 + }, + { + "epoch": 0.45522949586155004, + "grad_norm": 2.222035205607659, + "learning_rate": 5.9621387811358175e-06, + "loss": 0.3767, + "step": 6655 + }, + { + "epoch": 0.45529789999315956, + "grad_norm": 2.2413121352276417, + "learning_rate": 5.961051705388855e-06, + "loss": 0.2766, + "step": 6656 + }, + { + "epoch": 0.45536630412476914, + "grad_norm": 1.7553059784719005, + "learning_rate": 5.959964582468834e-06, + "loss": 0.3108, + "step": 6657 + }, + { + "epoch": 0.45543470825637866, + "grad_norm": 2.4282790606460436, + "learning_rate": 5.958877412429116e-06, + "loss": 0.2425, + "step": 6658 + }, + { + "epoch": 0.45550311238798824, + "grad_norm": 1.8545093304573448, + "learning_rate": 5.957790195323064e-06, + "loss": 0.3656, + "step": 6659 + }, + { + "epoch": 0.45557151651959776, + "grad_norm": 2.399503247439887, + "learning_rate": 5.956702931204043e-06, + "loss": 0.3162, + "step": 6660 + }, + { + "epoch": 0.45563992065120734, + "grad_norm": 1.7620716714176181, + "learning_rate": 5.955615620125423e-06, + "loss": 0.4065, + "step": 6661 + }, + { + "epoch": 0.45570832478281686, + "grad_norm": 2.217603299843186, + "learning_rate": 5.954528262140574e-06, + "loss": 0.3356, + "step": 6662 + }, + { + "epoch": 0.45577672891442644, + "grad_norm": 2.252674023789043, + "learning_rate": 5.953440857302869e-06, + "loss": 0.6906, + "step": 6663 + }, + { + "epoch": 0.45584513304603597, + "grad_norm": 1.6719358869856198, + "learning_rate": 5.952353405665683e-06, + "loss": 0.2732, + "step": 6664 + }, + { + "epoch": 0.45591353717764554, + "grad_norm": 1.7079323914999018, + "learning_rate": 5.951265907282392e-06, + "loss": 0.2591, + "step": 6665 + }, + { + "epoch": 0.45598194130925507, + "grad_norm": 3.517196876076157, + "learning_rate": 5.950178362206376e-06, + "loss": 0.3378, + "step": 6666 + }, + { + "epoch": 0.45605034544086465, + "grad_norm": 2.278205401044442, + "learning_rate": 5.949090770491019e-06, + "loss": 0.7181, + "step": 6667 + }, + { + "epoch": 0.45611874957247417, + "grad_norm": 2.5768867177648263, + "learning_rate": 5.948003132189703e-06, + "loss": 0.6291, + "step": 6668 + }, + { + "epoch": 0.45618715370408375, + "grad_norm": 2.378812456082822, + "learning_rate": 5.946915447355814e-06, + "loss": 0.3611, + "step": 6669 + }, + { + "epoch": 0.45625555783569327, + "grad_norm": 1.816561212389923, + "learning_rate": 5.945827716042745e-06, + "loss": 0.2738, + "step": 6670 + }, + { + "epoch": 0.45632396196730285, + "grad_norm": 1.701514720858953, + "learning_rate": 5.944739938303883e-06, + "loss": 0.2054, + "step": 6671 + }, + { + "epoch": 0.45639236609891237, + "grad_norm": 1.8977430078015585, + "learning_rate": 5.943652114192624e-06, + "loss": 0.5165, + "step": 6672 + }, + { + "epoch": 0.45646077023052195, + "grad_norm": 1.8719472192961117, + "learning_rate": 5.94256424376236e-06, + "loss": 0.2172, + "step": 6673 + }, + { + "epoch": 0.4565291743621315, + "grad_norm": 1.4195409129920806, + "learning_rate": 5.941476327066493e-06, + "loss": 0.2638, + "step": 6674 + }, + { + "epoch": 0.456597578493741, + "grad_norm": 2.2392609401170307, + "learning_rate": 5.940388364158421e-06, + "loss": 0.1619, + "step": 6675 + }, + { + "epoch": 0.4566659826253506, + "grad_norm": 1.8484834267077352, + "learning_rate": 5.939300355091546e-06, + "loss": 0.3178, + "step": 6676 + }, + { + "epoch": 0.4567343867569601, + "grad_norm": 2.397288603328322, + "learning_rate": 5.9382122999192755e-06, + "loss": 0.4148, + "step": 6677 + }, + { + "epoch": 0.4568027908885697, + "grad_norm": 1.6237206558148516, + "learning_rate": 5.937124198695015e-06, + "loss": 0.3007, + "step": 6678 + }, + { + "epoch": 0.4568711950201792, + "grad_norm": 1.8428134054712506, + "learning_rate": 5.936036051472173e-06, + "loss": 0.3375, + "step": 6679 + }, + { + "epoch": 0.4569395991517888, + "grad_norm": 2.0356363693004136, + "learning_rate": 5.93494785830416e-06, + "loss": 0.3558, + "step": 6680 + }, + { + "epoch": 0.4570080032833983, + "grad_norm": 1.3048533260015545, + "learning_rate": 5.933859619244393e-06, + "loss": 0.1026, + "step": 6681 + }, + { + "epoch": 0.4570764074150079, + "grad_norm": 1.8592125076328927, + "learning_rate": 5.932771334346286e-06, + "loss": 0.3218, + "step": 6682 + }, + { + "epoch": 0.4571448115466174, + "grad_norm": 2.225825281400576, + "learning_rate": 5.931683003663258e-06, + "loss": 0.3364, + "step": 6683 + }, + { + "epoch": 0.457213215678227, + "grad_norm": 2.9903839741665355, + "learning_rate": 5.930594627248728e-06, + "loss": 0.3518, + "step": 6684 + }, + { + "epoch": 0.4572816198098365, + "grad_norm": 1.2976916855504488, + "learning_rate": 5.92950620515612e-06, + "loss": 0.1521, + "step": 6685 + }, + { + "epoch": 0.4573500239414461, + "grad_norm": 1.908549867828985, + "learning_rate": 5.928417737438859e-06, + "loss": 0.4416, + "step": 6686 + }, + { + "epoch": 0.4574184280730556, + "grad_norm": 2.193769374174984, + "learning_rate": 5.927329224150373e-06, + "loss": 0.4395, + "step": 6687 + }, + { + "epoch": 0.4574868322046652, + "grad_norm": 3.4091005734104383, + "learning_rate": 5.926240665344091e-06, + "loss": 0.2979, + "step": 6688 + }, + { + "epoch": 0.4575552363362747, + "grad_norm": 2.4466962951257294, + "learning_rate": 5.925152061073443e-06, + "loss": 0.5479, + "step": 6689 + }, + { + "epoch": 0.4576236404678843, + "grad_norm": 1.923298463669225, + "learning_rate": 5.924063411391864e-06, + "loss": 0.4096, + "step": 6690 + }, + { + "epoch": 0.4576920445994938, + "grad_norm": 2.1643536506104395, + "learning_rate": 5.922974716352792e-06, + "loss": 0.703, + "step": 6691 + }, + { + "epoch": 0.45776044873110333, + "grad_norm": 1.76595921855607, + "learning_rate": 5.921885976009664e-06, + "loss": 0.4762, + "step": 6692 + }, + { + "epoch": 0.4578288528627129, + "grad_norm": 2.6040481286091217, + "learning_rate": 5.920797190415919e-06, + "loss": 0.6102, + "step": 6693 + }, + { + "epoch": 0.45789725699432243, + "grad_norm": 2.0634155119978503, + "learning_rate": 5.9197083596250025e-06, + "loss": 0.4818, + "step": 6694 + }, + { + "epoch": 0.457965661125932, + "grad_norm": 1.9323171003849364, + "learning_rate": 5.918619483690357e-06, + "loss": 0.357, + "step": 6695 + }, + { + "epoch": 0.45803406525754153, + "grad_norm": 2.049664380349839, + "learning_rate": 5.9175305626654335e-06, + "loss": 0.3635, + "step": 6696 + }, + { + "epoch": 0.4581024693891511, + "grad_norm": 1.7268349704389812, + "learning_rate": 5.916441596603678e-06, + "loss": 0.2711, + "step": 6697 + }, + { + "epoch": 0.45817087352076064, + "grad_norm": 2.3919911189982916, + "learning_rate": 5.915352585558545e-06, + "loss": 0.5387, + "step": 6698 + }, + { + "epoch": 0.4582392776523702, + "grad_norm": 1.957749362545521, + "learning_rate": 5.914263529583485e-06, + "loss": 0.2908, + "step": 6699 + }, + { + "epoch": 0.45830768178397974, + "grad_norm": 1.5820209267361875, + "learning_rate": 5.913174428731956e-06, + "loss": 0.3563, + "step": 6700 + }, + { + "epoch": 0.4583760859155893, + "grad_norm": 1.9686730839069997, + "learning_rate": 5.9120852830574184e-06, + "loss": 0.4151, + "step": 6701 + }, + { + "epoch": 0.45844449004719884, + "grad_norm": 2.119211784609541, + "learning_rate": 5.910996092613329e-06, + "loss": 0.6291, + "step": 6702 + }, + { + "epoch": 0.4585128941788084, + "grad_norm": 2.3017315480704648, + "learning_rate": 5.909906857453153e-06, + "loss": 0.504, + "step": 6703 + }, + { + "epoch": 0.45858129831041794, + "grad_norm": 2.1488169491076268, + "learning_rate": 5.908817577630353e-06, + "loss": 0.5828, + "step": 6704 + }, + { + "epoch": 0.4586497024420275, + "grad_norm": 1.3649124207932604, + "learning_rate": 5.907728253198399e-06, + "loss": 0.2023, + "step": 6705 + }, + { + "epoch": 0.45871810657363704, + "grad_norm": 2.4861765273655734, + "learning_rate": 5.906638884210758e-06, + "loss": 0.3259, + "step": 6706 + }, + { + "epoch": 0.4587865107052466, + "grad_norm": 2.838457222945297, + "learning_rate": 5.9055494707209035e-06, + "loss": 0.4167, + "step": 6707 + }, + { + "epoch": 0.45885491483685614, + "grad_norm": 1.9792658709877144, + "learning_rate": 5.90446001278231e-06, + "loss": 0.1364, + "step": 6708 + }, + { + "epoch": 0.4589233189684657, + "grad_norm": 2.089853382979679, + "learning_rate": 5.903370510448447e-06, + "loss": 0.6936, + "step": 6709 + }, + { + "epoch": 0.45899172310007524, + "grad_norm": 2.0681055541808866, + "learning_rate": 5.9022809637728e-06, + "loss": 0.3616, + "step": 6710 + }, + { + "epoch": 0.45906012723168477, + "grad_norm": 2.138934027921773, + "learning_rate": 5.901191372808845e-06, + "loss": 0.6192, + "step": 6711 + }, + { + "epoch": 0.45912853136329435, + "grad_norm": 2.3223259278754225, + "learning_rate": 5.900101737610068e-06, + "loss": 0.2666, + "step": 6712 + }, + { + "epoch": 0.45919693549490387, + "grad_norm": 2.3624381650169215, + "learning_rate": 5.899012058229948e-06, + "loss": 0.3483, + "step": 6713 + }, + { + "epoch": 0.45926533962651345, + "grad_norm": 2.502901842882985, + "learning_rate": 5.897922334721977e-06, + "loss": 0.6594, + "step": 6714 + }, + { + "epoch": 0.45933374375812297, + "grad_norm": 2.314799505783603, + "learning_rate": 5.896832567139641e-06, + "loss": 0.6411, + "step": 6715 + }, + { + "epoch": 0.45940214788973255, + "grad_norm": 2.3418291867582592, + "learning_rate": 5.895742755536431e-06, + "loss": 0.5888, + "step": 6716 + }, + { + "epoch": 0.45947055202134207, + "grad_norm": 1.698093427533758, + "learning_rate": 5.894652899965843e-06, + "loss": 0.3163, + "step": 6717 + }, + { + "epoch": 0.45953895615295165, + "grad_norm": 2.6725399907530623, + "learning_rate": 5.893563000481369e-06, + "loss": 0.3703, + "step": 6718 + }, + { + "epoch": 0.4596073602845612, + "grad_norm": 2.0221500056408432, + "learning_rate": 5.892473057136508e-06, + "loss": 0.4903, + "step": 6719 + }, + { + "epoch": 0.45967576441617075, + "grad_norm": 2.711915724647021, + "learning_rate": 5.89138306998476e-06, + "loss": 0.3578, + "step": 6720 + }, + { + "epoch": 0.4597441685477803, + "grad_norm": 2.0943777936633903, + "learning_rate": 5.890293039079626e-06, + "loss": 0.2563, + "step": 6721 + }, + { + "epoch": 0.45981257267938985, + "grad_norm": 1.9977123804602062, + "learning_rate": 5.889202964474611e-06, + "loss": 0.2417, + "step": 6722 + }, + { + "epoch": 0.4598809768109994, + "grad_norm": 2.1781476368541925, + "learning_rate": 5.888112846223221e-06, + "loss": 0.6576, + "step": 6723 + }, + { + "epoch": 0.45994938094260895, + "grad_norm": 1.8615535827699783, + "learning_rate": 5.887022684378964e-06, + "loss": 0.343, + "step": 6724 + }, + { + "epoch": 0.4600177850742185, + "grad_norm": 1.8487274057053518, + "learning_rate": 5.8859324789953485e-06, + "loss": 0.4295, + "step": 6725 + }, + { + "epoch": 0.46008618920582806, + "grad_norm": 2.213554350124747, + "learning_rate": 5.884842230125889e-06, + "loss": 0.3706, + "step": 6726 + }, + { + "epoch": 0.4601545933374376, + "grad_norm": 2.353478573076145, + "learning_rate": 5.883751937824101e-06, + "loss": 0.7599, + "step": 6727 + }, + { + "epoch": 0.46022299746904716, + "grad_norm": 1.9824570901683856, + "learning_rate": 5.882661602143502e-06, + "loss": 0.2866, + "step": 6728 + }, + { + "epoch": 0.4602914016006567, + "grad_norm": 2.0965688327006236, + "learning_rate": 5.8815712231376065e-06, + "loss": 0.3111, + "step": 6729 + }, + { + "epoch": 0.4603598057322662, + "grad_norm": 2.290495843615838, + "learning_rate": 5.880480800859939e-06, + "loss": 0.6014, + "step": 6730 + }, + { + "epoch": 0.4604282098638758, + "grad_norm": 2.42753518009328, + "learning_rate": 5.879390335364023e-06, + "loss": 0.4796, + "step": 6731 + }, + { + "epoch": 0.4604966139954853, + "grad_norm": 3.10101390378136, + "learning_rate": 5.8782998267033805e-06, + "loss": 0.4169, + "step": 6732 + }, + { + "epoch": 0.4605650181270949, + "grad_norm": 2.2089958764003637, + "learning_rate": 5.877209274931544e-06, + "loss": 0.313, + "step": 6733 + }, + { + "epoch": 0.4606334222587044, + "grad_norm": 2.215668721433477, + "learning_rate": 5.87611868010204e-06, + "loss": 0.4091, + "step": 6734 + }, + { + "epoch": 0.460701826390314, + "grad_norm": 2.3386872670329937, + "learning_rate": 5.875028042268398e-06, + "loss": 0.3811, + "step": 6735 + }, + { + "epoch": 0.4607702305219235, + "grad_norm": 2.7122712715679764, + "learning_rate": 5.873937361484156e-06, + "loss": 0.6582, + "step": 6736 + }, + { + "epoch": 0.4608386346535331, + "grad_norm": 2.0233453449782246, + "learning_rate": 5.872846637802848e-06, + "loss": 0.4664, + "step": 6737 + }, + { + "epoch": 0.4609070387851426, + "grad_norm": 2.1934454434798965, + "learning_rate": 5.871755871278012e-06, + "loss": 0.6329, + "step": 6738 + }, + { + "epoch": 0.4609754429167522, + "grad_norm": 1.9180609531343522, + "learning_rate": 5.870665061963188e-06, + "loss": 0.5031, + "step": 6739 + }, + { + "epoch": 0.4610438470483617, + "grad_norm": 2.19959118333738, + "learning_rate": 5.869574209911918e-06, + "loss": 0.5928, + "step": 6740 + }, + { + "epoch": 0.4611122511799713, + "grad_norm": 2.204954760524199, + "learning_rate": 5.868483315177748e-06, + "loss": 0.7372, + "step": 6741 + }, + { + "epoch": 0.4611806553115808, + "grad_norm": 2.0156382035407012, + "learning_rate": 5.86739237781422e-06, + "loss": 0.3888, + "step": 6742 + }, + { + "epoch": 0.4612490594431904, + "grad_norm": 3.570283551589169, + "learning_rate": 5.866301397874889e-06, + "loss": 0.4862, + "step": 6743 + }, + { + "epoch": 0.4613174635747999, + "grad_norm": 2.263949968767731, + "learning_rate": 5.8652103754133015e-06, + "loss": 0.2712, + "step": 6744 + }, + { + "epoch": 0.4613858677064095, + "grad_norm": 1.7559190003791423, + "learning_rate": 5.86411931048301e-06, + "loss": 0.3388, + "step": 6745 + }, + { + "epoch": 0.461454271838019, + "grad_norm": 2.0164819455635574, + "learning_rate": 5.863028203137569e-06, + "loss": 0.5211, + "step": 6746 + }, + { + "epoch": 0.4615226759696286, + "grad_norm": 2.2953827610259627, + "learning_rate": 5.861937053430538e-06, + "loss": 0.5167, + "step": 6747 + }, + { + "epoch": 0.4615910801012381, + "grad_norm": 1.689421220158433, + "learning_rate": 5.860845861415473e-06, + "loss": 0.2563, + "step": 6748 + }, + { + "epoch": 0.46165948423284764, + "grad_norm": 2.069047195719051, + "learning_rate": 5.859754627145936e-06, + "loss": 0.4529, + "step": 6749 + }, + { + "epoch": 0.4617278883644572, + "grad_norm": 2.2569857076801676, + "learning_rate": 5.85866335067549e-06, + "loss": 0.6406, + "step": 6750 + }, + { + "epoch": 0.46179629249606674, + "grad_norm": 2.0067713167873276, + "learning_rate": 5.857572032057701e-06, + "loss": 0.4167, + "step": 6751 + }, + { + "epoch": 0.4618646966276763, + "grad_norm": 2.421660518378084, + "learning_rate": 5.856480671346135e-06, + "loss": 0.3413, + "step": 6752 + }, + { + "epoch": 0.46193310075928584, + "grad_norm": 1.9258058093750836, + "learning_rate": 5.855389268594363e-06, + "loss": 0.4612, + "step": 6753 + }, + { + "epoch": 0.4620015048908954, + "grad_norm": 2.6719637738922972, + "learning_rate": 5.854297823855954e-06, + "loss": 0.4233, + "step": 6754 + }, + { + "epoch": 0.46206990902250494, + "grad_norm": 2.2125838808684097, + "learning_rate": 5.853206337184483e-06, + "loss": 0.3843, + "step": 6755 + }, + { + "epoch": 0.4621383131541145, + "grad_norm": 1.9934486100072175, + "learning_rate": 5.8521148086335245e-06, + "loss": 0.4149, + "step": 6756 + }, + { + "epoch": 0.46220671728572404, + "grad_norm": 1.6781922538252136, + "learning_rate": 5.851023238256657e-06, + "loss": 0.1781, + "step": 6757 + }, + { + "epoch": 0.4622751214173336, + "grad_norm": 1.751075495038188, + "learning_rate": 5.849931626107459e-06, + "loss": 0.4353, + "step": 6758 + }, + { + "epoch": 0.46234352554894315, + "grad_norm": 2.259146998866393, + "learning_rate": 5.848839972239512e-06, + "loss": 0.4773, + "step": 6759 + }, + { + "epoch": 0.4624119296805527, + "grad_norm": 1.8052529113982723, + "learning_rate": 5.847748276706399e-06, + "loss": 0.3693, + "step": 6760 + }, + { + "epoch": 0.46248033381216225, + "grad_norm": 1.3516137604005782, + "learning_rate": 5.846656539561709e-06, + "loss": 0.1757, + "step": 6761 + }, + { + "epoch": 0.4625487379437718, + "grad_norm": 1.9474070085414186, + "learning_rate": 5.845564760859027e-06, + "loss": 0.4253, + "step": 6762 + }, + { + "epoch": 0.46261714207538135, + "grad_norm": 5.036144328330741, + "learning_rate": 5.844472940651944e-06, + "loss": 0.463, + "step": 6763 + }, + { + "epoch": 0.4626855462069909, + "grad_norm": 1.90720765885274, + "learning_rate": 5.843381078994051e-06, + "loss": 0.3387, + "step": 6764 + }, + { + "epoch": 0.46275395033860045, + "grad_norm": 2.49865975217446, + "learning_rate": 5.842289175938941e-06, + "loss": 0.2862, + "step": 6765 + }, + { + "epoch": 0.46282235447021, + "grad_norm": 1.7516046005144073, + "learning_rate": 5.8411972315402095e-06, + "loss": 0.3352, + "step": 6766 + }, + { + "epoch": 0.46289075860181955, + "grad_norm": 1.1960322700013535, + "learning_rate": 5.840105245851459e-06, + "loss": 0.1621, + "step": 6767 + }, + { + "epoch": 0.4629591627334291, + "grad_norm": 2.603806286989538, + "learning_rate": 5.839013218926284e-06, + "loss": 0.5519, + "step": 6768 + }, + { + "epoch": 0.46302756686503865, + "grad_norm": 1.9079859247825592, + "learning_rate": 5.837921150818291e-06, + "loss": 0.2329, + "step": 6769 + }, + { + "epoch": 0.4630959709966482, + "grad_norm": 2.2117533846467397, + "learning_rate": 5.836829041581079e-06, + "loss": 0.4439, + "step": 6770 + }, + { + "epoch": 0.46316437512825775, + "grad_norm": 2.0533920889491832, + "learning_rate": 5.835736891268258e-06, + "loss": 0.6112, + "step": 6771 + }, + { + "epoch": 0.4632327792598673, + "grad_norm": 1.9717963469003676, + "learning_rate": 5.834644699933434e-06, + "loss": 0.5141, + "step": 6772 + }, + { + "epoch": 0.46330118339147686, + "grad_norm": 2.0748641595842816, + "learning_rate": 5.833552467630219e-06, + "loss": 0.4998, + "step": 6773 + }, + { + "epoch": 0.4633695875230864, + "grad_norm": 1.249530113833141, + "learning_rate": 5.832460194412222e-06, + "loss": 0.225, + "step": 6774 + }, + { + "epoch": 0.46343799165469596, + "grad_norm": 2.3685537430578996, + "learning_rate": 5.831367880333059e-06, + "loss": 0.6674, + "step": 6775 + }, + { + "epoch": 0.4635063957863055, + "grad_norm": 1.3995188136802998, + "learning_rate": 5.8302755254463465e-06, + "loss": 0.2002, + "step": 6776 + }, + { + "epoch": 0.46357479991791506, + "grad_norm": 1.7492616492288742, + "learning_rate": 5.829183129805702e-06, + "loss": 0.4132, + "step": 6777 + }, + { + "epoch": 0.4636432040495246, + "grad_norm": 2.2246062401807234, + "learning_rate": 5.828090693464745e-06, + "loss": 0.5923, + "step": 6778 + }, + { + "epoch": 0.46371160818113416, + "grad_norm": 1.8414211031444596, + "learning_rate": 5.826998216477097e-06, + "loss": 0.3011, + "step": 6779 + }, + { + "epoch": 0.4637800123127437, + "grad_norm": 2.1273623756716855, + "learning_rate": 5.825905698896385e-06, + "loss": 0.5153, + "step": 6780 + }, + { + "epoch": 0.46384841644435326, + "grad_norm": 2.0657453147025295, + "learning_rate": 5.824813140776232e-06, + "loss": 0.3386, + "step": 6781 + }, + { + "epoch": 0.4639168205759628, + "grad_norm": 2.0531643998707216, + "learning_rate": 5.823720542170267e-06, + "loss": 0.3307, + "step": 6782 + }, + { + "epoch": 0.46398522470757236, + "grad_norm": 2.626621789847416, + "learning_rate": 5.822627903132121e-06, + "loss": 0.8227, + "step": 6783 + }, + { + "epoch": 0.4640536288391819, + "grad_norm": 2.6108835165674154, + "learning_rate": 5.821535223715423e-06, + "loss": 0.8354, + "step": 6784 + }, + { + "epoch": 0.4641220329707914, + "grad_norm": 1.8755266046121863, + "learning_rate": 5.82044250397381e-06, + "loss": 0.1792, + "step": 6785 + }, + { + "epoch": 0.464190437102401, + "grad_norm": 2.0120615250487908, + "learning_rate": 5.819349743960918e-06, + "loss": 0.4925, + "step": 6786 + }, + { + "epoch": 0.4642588412340105, + "grad_norm": 2.5165525260650905, + "learning_rate": 5.818256943730384e-06, + "loss": 0.5113, + "step": 6787 + }, + { + "epoch": 0.4643272453656201, + "grad_norm": 2.6519035850546846, + "learning_rate": 5.817164103335846e-06, + "loss": 0.8366, + "step": 6788 + }, + { + "epoch": 0.4643956494972296, + "grad_norm": 1.8146331582818527, + "learning_rate": 5.816071222830949e-06, + "loss": 0.4504, + "step": 6789 + }, + { + "epoch": 0.4644640536288392, + "grad_norm": 2.1733773236202234, + "learning_rate": 5.814978302269337e-06, + "loss": 0.5731, + "step": 6790 + }, + { + "epoch": 0.4645324577604487, + "grad_norm": 1.8158056312107858, + "learning_rate": 5.813885341704652e-06, + "loss": 0.4485, + "step": 6791 + }, + { + "epoch": 0.4646008618920583, + "grad_norm": 1.2626366273440746, + "learning_rate": 5.812792341190545e-06, + "loss": 0.1772, + "step": 6792 + }, + { + "epoch": 0.4646692660236678, + "grad_norm": 1.970208033061871, + "learning_rate": 5.811699300780666e-06, + "loss": 0.2667, + "step": 6793 + }, + { + "epoch": 0.4647376701552774, + "grad_norm": 2.0919478546280192, + "learning_rate": 5.810606220528664e-06, + "loss": 0.5089, + "step": 6794 + }, + { + "epoch": 0.4648060742868869, + "grad_norm": 1.8166237403525605, + "learning_rate": 5.809513100488196e-06, + "loss": 0.4814, + "step": 6795 + }, + { + "epoch": 0.4648744784184965, + "grad_norm": 2.1497642443278497, + "learning_rate": 5.808419940712915e-06, + "loss": 0.2377, + "step": 6796 + }, + { + "epoch": 0.464942882550106, + "grad_norm": 1.804617269396251, + "learning_rate": 5.80732674125648e-06, + "loss": 0.4941, + "step": 6797 + }, + { + "epoch": 0.4650112866817156, + "grad_norm": 1.873966794651332, + "learning_rate": 5.806233502172549e-06, + "loss": 0.1996, + "step": 6798 + }, + { + "epoch": 0.4650796908133251, + "grad_norm": 2.187371113931156, + "learning_rate": 5.805140223514785e-06, + "loss": 0.4195, + "step": 6799 + }, + { + "epoch": 0.4651480949449347, + "grad_norm": 2.20559653852113, + "learning_rate": 5.804046905336851e-06, + "loss": 0.2379, + "step": 6800 + }, + { + "epoch": 0.4652164990765442, + "grad_norm": 2.218384091333104, + "learning_rate": 5.8029535476924125e-06, + "loss": 0.4922, + "step": 6801 + }, + { + "epoch": 0.4652849032081538, + "grad_norm": 2.092727681993504, + "learning_rate": 5.801860150635136e-06, + "loss": 0.4736, + "step": 6802 + }, + { + "epoch": 0.4653533073397633, + "grad_norm": 2.5429596644413572, + "learning_rate": 5.800766714218691e-06, + "loss": 0.5924, + "step": 6803 + }, + { + "epoch": 0.46542171147137285, + "grad_norm": 2.1114821117613554, + "learning_rate": 5.799673238496748e-06, + "loss": 0.4768, + "step": 6804 + }, + { + "epoch": 0.4654901156029824, + "grad_norm": 2.33554854711156, + "learning_rate": 5.798579723522983e-06, + "loss": 0.3295, + "step": 6805 + }, + { + "epoch": 0.46555851973459195, + "grad_norm": 2.221919103444933, + "learning_rate": 5.797486169351067e-06, + "loss": 0.4823, + "step": 6806 + }, + { + "epoch": 0.4656269238662015, + "grad_norm": 2.219649428081793, + "learning_rate": 5.79639257603468e-06, + "loss": 0.5349, + "step": 6807 + }, + { + "epoch": 0.46569532799781105, + "grad_norm": 1.5824758870274935, + "learning_rate": 5.795298943627498e-06, + "loss": 0.2125, + "step": 6808 + }, + { + "epoch": 0.4657637321294206, + "grad_norm": 2.398397594455422, + "learning_rate": 5.794205272183206e-06, + "loss": 0.4118, + "step": 6809 + }, + { + "epoch": 0.46583213626103015, + "grad_norm": 2.041844976109271, + "learning_rate": 5.793111561755483e-06, + "loss": 0.4486, + "step": 6810 + }, + { + "epoch": 0.46590054039263973, + "grad_norm": 1.8713055731663426, + "learning_rate": 5.7920178123980144e-06, + "loss": 0.3259, + "step": 6811 + }, + { + "epoch": 0.46596894452424925, + "grad_norm": 1.7582242928404304, + "learning_rate": 5.790924024164486e-06, + "loss": 0.2535, + "step": 6812 + }, + { + "epoch": 0.46603734865585883, + "grad_norm": 1.8247698782493875, + "learning_rate": 5.789830197108588e-06, + "loss": 0.3707, + "step": 6813 + }, + { + "epoch": 0.46610575278746835, + "grad_norm": 1.9551722592491245, + "learning_rate": 5.788736331284009e-06, + "loss": 0.452, + "step": 6814 + }, + { + "epoch": 0.46617415691907793, + "grad_norm": 2.488835379373245, + "learning_rate": 5.7876424267444435e-06, + "loss": 0.2717, + "step": 6815 + }, + { + "epoch": 0.46624256105068745, + "grad_norm": 1.4462542532905278, + "learning_rate": 5.786548483543584e-06, + "loss": 0.2624, + "step": 6816 + }, + { + "epoch": 0.46631096518229703, + "grad_norm": 1.9856116428952604, + "learning_rate": 5.785454501735126e-06, + "loss": 0.3716, + "step": 6817 + }, + { + "epoch": 0.46637936931390656, + "grad_norm": 2.388446812854236, + "learning_rate": 5.784360481372768e-06, + "loss": 0.4832, + "step": 6818 + }, + { + "epoch": 0.46644777344551613, + "grad_norm": 2.0513476579585124, + "learning_rate": 5.783266422510211e-06, + "loss": 0.3622, + "step": 6819 + }, + { + "epoch": 0.46651617757712566, + "grad_norm": 2.2401025495513727, + "learning_rate": 5.782172325201155e-06, + "loss": 0.5468, + "step": 6820 + }, + { + "epoch": 0.4665845817087352, + "grad_norm": 1.77228292616937, + "learning_rate": 5.781078189499304e-06, + "loss": 0.3678, + "step": 6821 + }, + { + "epoch": 0.46665298584034476, + "grad_norm": 2.031297125636478, + "learning_rate": 5.779984015458364e-06, + "loss": 0.413, + "step": 6822 + }, + { + "epoch": 0.4667213899719543, + "grad_norm": 2.278577761405322, + "learning_rate": 5.7788898031320425e-06, + "loss": 0.3883, + "step": 6823 + }, + { + "epoch": 0.46678979410356386, + "grad_norm": 2.1947720225474514, + "learning_rate": 5.777795552574048e-06, + "loss": 0.1716, + "step": 6824 + }, + { + "epoch": 0.4668581982351734, + "grad_norm": 1.7897176126151875, + "learning_rate": 5.77670126383809e-06, + "loss": 0.3384, + "step": 6825 + }, + { + "epoch": 0.46692660236678296, + "grad_norm": 1.6312103556240498, + "learning_rate": 5.775606936977886e-06, + "loss": 0.2583, + "step": 6826 + }, + { + "epoch": 0.4669950064983925, + "grad_norm": 2.350967699531594, + "learning_rate": 5.774512572047147e-06, + "loss": 0.3744, + "step": 6827 + }, + { + "epoch": 0.46706341063000206, + "grad_norm": 1.9809298006874878, + "learning_rate": 5.7734181690995905e-06, + "loss": 0.2852, + "step": 6828 + }, + { + "epoch": 0.4671318147616116, + "grad_norm": 2.3464625111642925, + "learning_rate": 5.772323728188935e-06, + "loss": 0.3259, + "step": 6829 + }, + { + "epoch": 0.46720021889322116, + "grad_norm": 1.7249523612368785, + "learning_rate": 5.771229249368904e-06, + "loss": 0.379, + "step": 6830 + }, + { + "epoch": 0.4672686230248307, + "grad_norm": 1.663761219098607, + "learning_rate": 5.7701347326932145e-06, + "loss": 0.2681, + "step": 6831 + }, + { + "epoch": 0.46733702715644027, + "grad_norm": 1.4637411316489788, + "learning_rate": 5.7690401782155945e-06, + "loss": 0.3397, + "step": 6832 + }, + { + "epoch": 0.4674054312880498, + "grad_norm": 1.95002014183377, + "learning_rate": 5.767945585989769e-06, + "loss": 0.3968, + "step": 6833 + }, + { + "epoch": 0.46747383541965937, + "grad_norm": 1.8537970470958491, + "learning_rate": 5.766850956069465e-06, + "loss": 0.3381, + "step": 6834 + }, + { + "epoch": 0.4675422395512689, + "grad_norm": 1.8564583398160213, + "learning_rate": 5.765756288508414e-06, + "loss": 0.4115, + "step": 6835 + }, + { + "epoch": 0.46761064368287847, + "grad_norm": 2.258662014499975, + "learning_rate": 5.764661583360347e-06, + "loss": 0.4974, + "step": 6836 + }, + { + "epoch": 0.467679047814488, + "grad_norm": 2.7514092454141714, + "learning_rate": 5.763566840678995e-06, + "loss": 0.7778, + "step": 6837 + }, + { + "epoch": 0.46774745194609757, + "grad_norm": 1.920303221479182, + "learning_rate": 5.762472060518096e-06, + "loss": 0.557, + "step": 6838 + }, + { + "epoch": 0.4678158560777071, + "grad_norm": 2.131837894650241, + "learning_rate": 5.761377242931386e-06, + "loss": 0.6297, + "step": 6839 + }, + { + "epoch": 0.4678842602093166, + "grad_norm": 2.2245719779256414, + "learning_rate": 5.760282387972607e-06, + "loss": 0.4087, + "step": 6840 + }, + { + "epoch": 0.4679526643409262, + "grad_norm": 2.6264413668392965, + "learning_rate": 5.7591874956954944e-06, + "loss": 0.6966, + "step": 6841 + }, + { + "epoch": 0.4680210684725357, + "grad_norm": 1.23967806887519, + "learning_rate": 5.7580925661537936e-06, + "loss": 0.222, + "step": 6842 + }, + { + "epoch": 0.4680894726041453, + "grad_norm": 2.1156834330863816, + "learning_rate": 5.756997599401249e-06, + "loss": 0.5252, + "step": 6843 + }, + { + "epoch": 0.4681578767357548, + "grad_norm": 2.059907562730448, + "learning_rate": 5.755902595491607e-06, + "loss": 0.4231, + "step": 6844 + }, + { + "epoch": 0.4682262808673644, + "grad_norm": 2.3130254535132133, + "learning_rate": 5.7548075544786156e-06, + "loss": 0.5731, + "step": 6845 + }, + { + "epoch": 0.4682946849989739, + "grad_norm": 1.7104178950397884, + "learning_rate": 5.753712476416025e-06, + "loss": 0.2792, + "step": 6846 + }, + { + "epoch": 0.4683630891305835, + "grad_norm": 4.7949896471629785, + "learning_rate": 5.7526173613575855e-06, + "loss": 0.2336, + "step": 6847 + }, + { + "epoch": 0.468431493262193, + "grad_norm": 2.078759254382208, + "learning_rate": 5.751522209357051e-06, + "loss": 0.4271, + "step": 6848 + }, + { + "epoch": 0.4684998973938026, + "grad_norm": 2.4207716674358957, + "learning_rate": 5.750427020468178e-06, + "loss": 0.662, + "step": 6849 + }, + { + "epoch": 0.4685683015254121, + "grad_norm": 1.5023010084169215, + "learning_rate": 5.749331794744723e-06, + "loss": 0.1972, + "step": 6850 + }, + { + "epoch": 0.4686367056570217, + "grad_norm": 1.9810665101303402, + "learning_rate": 5.748236532240446e-06, + "loss": 0.4379, + "step": 6851 + }, + { + "epoch": 0.4687051097886312, + "grad_norm": 2.0717213964988423, + "learning_rate": 5.747141233009106e-06, + "loss": 0.382, + "step": 6852 + }, + { + "epoch": 0.4687735139202408, + "grad_norm": 1.7550664369426152, + "learning_rate": 5.746045897104467e-06, + "loss": 0.3228, + "step": 6853 + }, + { + "epoch": 0.4688419180518503, + "grad_norm": 4.206973274702814, + "learning_rate": 5.744950524580291e-06, + "loss": 0.5509, + "step": 6854 + }, + { + "epoch": 0.4689103221834599, + "grad_norm": 1.8120233139123503, + "learning_rate": 5.743855115490347e-06, + "loss": 0.3093, + "step": 6855 + }, + { + "epoch": 0.4689787263150694, + "grad_norm": 2.278260960727631, + "learning_rate": 5.7427596698884025e-06, + "loss": 0.4924, + "step": 6856 + }, + { + "epoch": 0.469047130446679, + "grad_norm": 2.0892243367566135, + "learning_rate": 5.7416641878282265e-06, + "loss": 0.1742, + "step": 6857 + }, + { + "epoch": 0.46911553457828853, + "grad_norm": 2.223521612215658, + "learning_rate": 5.74056866936359e-06, + "loss": 0.3904, + "step": 6858 + }, + { + "epoch": 0.46918393870989805, + "grad_norm": 2.107925213514129, + "learning_rate": 5.739473114548266e-06, + "loss": 0.514, + "step": 6859 + }, + { + "epoch": 0.46925234284150763, + "grad_norm": 1.7364643845875123, + "learning_rate": 5.738377523436033e-06, + "loss": 0.2376, + "step": 6860 + }, + { + "epoch": 0.46932074697311715, + "grad_norm": 2.113056174401057, + "learning_rate": 5.737281896080664e-06, + "loss": 0.2629, + "step": 6861 + }, + { + "epoch": 0.46938915110472673, + "grad_norm": 2.1061001611155374, + "learning_rate": 5.73618623253594e-06, + "loss": 0.4025, + "step": 6862 + }, + { + "epoch": 0.46945755523633625, + "grad_norm": 2.6851099331601187, + "learning_rate": 5.735090532855641e-06, + "loss": 0.297, + "step": 6863 + }, + { + "epoch": 0.46952595936794583, + "grad_norm": 2.056176625047026, + "learning_rate": 5.733994797093548e-06, + "loss": 0.3541, + "step": 6864 + }, + { + "epoch": 0.46959436349955536, + "grad_norm": 1.9096428194782122, + "learning_rate": 5.732899025303445e-06, + "loss": 0.2728, + "step": 6865 + }, + { + "epoch": 0.46966276763116493, + "grad_norm": 1.8590807842241037, + "learning_rate": 5.731803217539121e-06, + "loss": 0.2932, + "step": 6866 + }, + { + "epoch": 0.46973117176277446, + "grad_norm": 1.6461901509319279, + "learning_rate": 5.73070737385436e-06, + "loss": 0.2648, + "step": 6867 + }, + { + "epoch": 0.46979957589438404, + "grad_norm": 1.5979669940260348, + "learning_rate": 5.729611494302954e-06, + "loss": 0.1716, + "step": 6868 + }, + { + "epoch": 0.46986798002599356, + "grad_norm": 2.0170836382779456, + "learning_rate": 5.72851557893869e-06, + "loss": 0.4206, + "step": 6869 + }, + { + "epoch": 0.46993638415760314, + "grad_norm": 1.8308027902905055, + "learning_rate": 5.727419627815366e-06, + "loss": 0.2695, + "step": 6870 + }, + { + "epoch": 0.47000478828921266, + "grad_norm": 1.7322218907476385, + "learning_rate": 5.726323640986772e-06, + "loss": 0.3745, + "step": 6871 + }, + { + "epoch": 0.47007319242082224, + "grad_norm": 2.0514987598636716, + "learning_rate": 5.725227618506708e-06, + "loss": 0.4601, + "step": 6872 + }, + { + "epoch": 0.47014159655243176, + "grad_norm": 2.142394343990327, + "learning_rate": 5.724131560428969e-06, + "loss": 0.4059, + "step": 6873 + }, + { + "epoch": 0.47021000068404134, + "grad_norm": 2.0913293246520475, + "learning_rate": 5.723035466807358e-06, + "loss": 0.4184, + "step": 6874 + }, + { + "epoch": 0.47027840481565086, + "grad_norm": 1.2727150447417894, + "learning_rate": 5.7219393376956725e-06, + "loss": 0.2246, + "step": 6875 + }, + { + "epoch": 0.47034680894726044, + "grad_norm": 1.658172005045718, + "learning_rate": 5.72084317314772e-06, + "loss": 0.3688, + "step": 6876 + }, + { + "epoch": 0.47041521307886996, + "grad_norm": 2.2217926063766202, + "learning_rate": 5.719746973217302e-06, + "loss": 0.5918, + "step": 6877 + }, + { + "epoch": 0.4704836172104795, + "grad_norm": 2.3971870530075474, + "learning_rate": 5.718650737958227e-06, + "loss": 0.5519, + "step": 6878 + }, + { + "epoch": 0.47055202134208907, + "grad_norm": 2.0622609317825997, + "learning_rate": 5.7175544674243044e-06, + "loss": 0.4181, + "step": 6879 + }, + { + "epoch": 0.4706204254736986, + "grad_norm": 2.183632210544633, + "learning_rate": 5.716458161669344e-06, + "loss": 0.4766, + "step": 6880 + }, + { + "epoch": 0.47068882960530817, + "grad_norm": 1.9179511215114595, + "learning_rate": 5.715361820747156e-06, + "loss": 0.2078, + "step": 6881 + }, + { + "epoch": 0.4707572337369177, + "grad_norm": 2.3118724100263925, + "learning_rate": 5.714265444711556e-06, + "loss": 0.543, + "step": 6882 + }, + { + "epoch": 0.47082563786852727, + "grad_norm": 1.828741487253783, + "learning_rate": 5.713169033616359e-06, + "loss": 0.2765, + "step": 6883 + }, + { + "epoch": 0.4708940420001368, + "grad_norm": 2.614698171310013, + "learning_rate": 5.712072587515381e-06, + "loss": 0.2924, + "step": 6884 + }, + { + "epoch": 0.47096244613174637, + "grad_norm": 2.249507946278692, + "learning_rate": 5.710976106462444e-06, + "loss": 0.3301, + "step": 6885 + }, + { + "epoch": 0.4710308502633559, + "grad_norm": 2.5101300164284623, + "learning_rate": 5.709879590511365e-06, + "loss": 0.2493, + "step": 6886 + }, + { + "epoch": 0.47109925439496547, + "grad_norm": 2.4499294278140233, + "learning_rate": 5.708783039715967e-06, + "loss": 0.2414, + "step": 6887 + }, + { + "epoch": 0.471167658526575, + "grad_norm": 1.6172370707244517, + "learning_rate": 5.707686454130074e-06, + "loss": 0.2987, + "step": 6888 + }, + { + "epoch": 0.4712360626581846, + "grad_norm": 1.483377116227756, + "learning_rate": 5.706589833807515e-06, + "loss": 0.1848, + "step": 6889 + }, + { + "epoch": 0.4713044667897941, + "grad_norm": 1.7240196982542295, + "learning_rate": 5.705493178802115e-06, + "loss": 0.4256, + "step": 6890 + }, + { + "epoch": 0.4713728709214037, + "grad_norm": 1.7590490400036893, + "learning_rate": 5.704396489167701e-06, + "loss": 0.2722, + "step": 6891 + }, + { + "epoch": 0.4714412750530132, + "grad_norm": 2.1212860111614673, + "learning_rate": 5.703299764958108e-06, + "loss": 0.5873, + "step": 6892 + }, + { + "epoch": 0.4715096791846228, + "grad_norm": 2.0961654238509007, + "learning_rate": 5.7022030062271636e-06, + "loss": 0.3605, + "step": 6893 + }, + { + "epoch": 0.4715780833162323, + "grad_norm": 2.539885138248065, + "learning_rate": 5.701106213028705e-06, + "loss": 0.2467, + "step": 6894 + }, + { + "epoch": 0.4716464874478418, + "grad_norm": 2.069231859715951, + "learning_rate": 5.700009385416571e-06, + "loss": 0.3497, + "step": 6895 + }, + { + "epoch": 0.4717148915794514, + "grad_norm": 1.7695424414274488, + "learning_rate": 5.6989125234445944e-06, + "loss": 0.3995, + "step": 6896 + }, + { + "epoch": 0.4717832957110609, + "grad_norm": 1.4882295153980065, + "learning_rate": 5.697815627166614e-06, + "loss": 0.3287, + "step": 6897 + }, + { + "epoch": 0.4718516998426705, + "grad_norm": 1.7997137948162205, + "learning_rate": 5.696718696636475e-06, + "loss": 0.4578, + "step": 6898 + }, + { + "epoch": 0.47192010397428, + "grad_norm": 2.239775862661241, + "learning_rate": 5.695621731908018e-06, + "loss": 0.5469, + "step": 6899 + }, + { + "epoch": 0.4719885081058896, + "grad_norm": 1.6168152753949248, + "learning_rate": 5.694524733035088e-06, + "loss": 0.2313, + "step": 6900 + }, + { + "epoch": 0.4720569122374991, + "grad_norm": 1.9049540280817492, + "learning_rate": 5.693427700071527e-06, + "loss": 0.2975, + "step": 6901 + }, + { + "epoch": 0.4721253163691087, + "grad_norm": 2.344901844544706, + "learning_rate": 5.692330633071189e-06, + "loss": 0.498, + "step": 6902 + }, + { + "epoch": 0.47219372050071823, + "grad_norm": 2.1979082007192368, + "learning_rate": 5.691233532087919e-06, + "loss": 0.5496, + "step": 6903 + }, + { + "epoch": 0.4722621246323278, + "grad_norm": 2.208941984570952, + "learning_rate": 5.690136397175569e-06, + "loss": 0.2356, + "step": 6904 + }, + { + "epoch": 0.47233052876393733, + "grad_norm": 1.4575527321566015, + "learning_rate": 5.689039228387993e-06, + "loss": 0.2851, + "step": 6905 + }, + { + "epoch": 0.4723989328955469, + "grad_norm": 2.088256090257373, + "learning_rate": 5.6879420257790455e-06, + "loss": 0.4792, + "step": 6906 + }, + { + "epoch": 0.47246733702715643, + "grad_norm": 2.346797100404798, + "learning_rate": 5.686844789402579e-06, + "loss": 0.3465, + "step": 6907 + }, + { + "epoch": 0.472535741158766, + "grad_norm": 1.7284285162404474, + "learning_rate": 5.685747519312454e-06, + "loss": 0.3042, + "step": 6908 + }, + { + "epoch": 0.47260414529037553, + "grad_norm": 2.0698434550297438, + "learning_rate": 5.68465021556253e-06, + "loss": 0.3588, + "step": 6909 + }, + { + "epoch": 0.4726725494219851, + "grad_norm": 2.1571566776337714, + "learning_rate": 5.683552878206668e-06, + "loss": 0.4953, + "step": 6910 + }, + { + "epoch": 0.47274095355359463, + "grad_norm": 1.9524098078683108, + "learning_rate": 5.682455507298729e-06, + "loss": 0.2359, + "step": 6911 + }, + { + "epoch": 0.4728093576852042, + "grad_norm": 2.9562799665936366, + "learning_rate": 5.681358102892578e-06, + "loss": 0.4508, + "step": 6912 + }, + { + "epoch": 0.47287776181681374, + "grad_norm": 1.564758740028617, + "learning_rate": 5.68026066504208e-06, + "loss": 0.3342, + "step": 6913 + }, + { + "epoch": 0.47294616594842326, + "grad_norm": 1.9263654445658487, + "learning_rate": 5.679163193801105e-06, + "loss": 0.4464, + "step": 6914 + }, + { + "epoch": 0.47301457008003284, + "grad_norm": 1.776459165669631, + "learning_rate": 5.678065689223521e-06, + "loss": 0.3093, + "step": 6915 + }, + { + "epoch": 0.47308297421164236, + "grad_norm": 2.352292524177188, + "learning_rate": 5.676968151363198e-06, + "loss": 0.401, + "step": 6916 + }, + { + "epoch": 0.47315137834325194, + "grad_norm": 1.755054739298116, + "learning_rate": 5.675870580274009e-06, + "loss": 0.3234, + "step": 6917 + }, + { + "epoch": 0.47321978247486146, + "grad_norm": 2.20522996432063, + "learning_rate": 5.674772976009828e-06, + "loss": 0.3301, + "step": 6918 + }, + { + "epoch": 0.47328818660647104, + "grad_norm": 1.4453820611349408, + "learning_rate": 5.6736753386245315e-06, + "loss": 0.2097, + "step": 6919 + }, + { + "epoch": 0.47335659073808056, + "grad_norm": 2.356682440995198, + "learning_rate": 5.672577668171997e-06, + "loss": 0.3516, + "step": 6920 + }, + { + "epoch": 0.47342499486969014, + "grad_norm": 2.263436687848015, + "learning_rate": 5.671479964706101e-06, + "loss": 0.4072, + "step": 6921 + }, + { + "epoch": 0.47349339900129966, + "grad_norm": 2.59148027316046, + "learning_rate": 5.6703822282807265e-06, + "loss": 0.2339, + "step": 6922 + }, + { + "epoch": 0.47356180313290924, + "grad_norm": 2.258814431001442, + "learning_rate": 5.669284458949755e-06, + "loss": 0.6224, + "step": 6923 + }, + { + "epoch": 0.47363020726451877, + "grad_norm": 1.879292226391952, + "learning_rate": 5.66818665676707e-06, + "loss": 0.42, + "step": 6924 + }, + { + "epoch": 0.47369861139612834, + "grad_norm": 1.983037313758287, + "learning_rate": 5.667088821786557e-06, + "loss": 0.447, + "step": 6925 + }, + { + "epoch": 0.47376701552773787, + "grad_norm": 1.8258766792505357, + "learning_rate": 5.665990954062105e-06, + "loss": 0.3075, + "step": 6926 + }, + { + "epoch": 0.47383541965934745, + "grad_norm": 1.5368660391274453, + "learning_rate": 5.664893053647599e-06, + "loss": 0.2844, + "step": 6927 + }, + { + "epoch": 0.47390382379095697, + "grad_norm": 2.2052896565425457, + "learning_rate": 5.663795120596931e-06, + "loss": 0.2037, + "step": 6928 + }, + { + "epoch": 0.47397222792256655, + "grad_norm": 2.117406439633971, + "learning_rate": 5.662697154963995e-06, + "loss": 0.4316, + "step": 6929 + }, + { + "epoch": 0.47404063205417607, + "grad_norm": 1.843824312361228, + "learning_rate": 5.6615991568026815e-06, + "loss": 0.4522, + "step": 6930 + }, + { + "epoch": 0.47410903618578565, + "grad_norm": 1.9562317248639471, + "learning_rate": 5.660501126166886e-06, + "loss": 0.509, + "step": 6931 + }, + { + "epoch": 0.47417744031739517, + "grad_norm": 2.0230678493180725, + "learning_rate": 5.659403063110508e-06, + "loss": 0.3218, + "step": 6932 + }, + { + "epoch": 0.4742458444490047, + "grad_norm": 1.7029027793868092, + "learning_rate": 5.6583049676874405e-06, + "loss": 0.2189, + "step": 6933 + }, + { + "epoch": 0.4743142485806143, + "grad_norm": 2.078324566487154, + "learning_rate": 5.657206839951588e-06, + "loss": 0.4973, + "step": 6934 + }, + { + "epoch": 0.4743826527122238, + "grad_norm": 2.437442460970451, + "learning_rate": 5.65610867995685e-06, + "loss": 0.5566, + "step": 6935 + }, + { + "epoch": 0.4744510568438334, + "grad_norm": 2.3285848816195505, + "learning_rate": 5.6550104877571285e-06, + "loss": 0.5065, + "step": 6936 + }, + { + "epoch": 0.4745194609754429, + "grad_norm": 1.5411925051167876, + "learning_rate": 5.6539122634063305e-06, + "loss": 0.2507, + "step": 6937 + }, + { + "epoch": 0.4745878651070525, + "grad_norm": 2.297897202232653, + "learning_rate": 5.652814006958361e-06, + "loss": 0.6503, + "step": 6938 + }, + { + "epoch": 0.474656269238662, + "grad_norm": 2.347854897826983, + "learning_rate": 5.651715718467127e-06, + "loss": 0.4338, + "step": 6939 + }, + { + "epoch": 0.4747246733702716, + "grad_norm": 2.1382936744203542, + "learning_rate": 5.650617397986538e-06, + "loss": 0.5345, + "step": 6940 + }, + { + "epoch": 0.4747930775018811, + "grad_norm": 2.264595205480016, + "learning_rate": 5.6495190455705066e-06, + "loss": 0.4156, + "step": 6941 + }, + { + "epoch": 0.4748614816334907, + "grad_norm": 2.3713154477039566, + "learning_rate": 5.648420661272943e-06, + "loss": 0.6495, + "step": 6942 + }, + { + "epoch": 0.4749298857651002, + "grad_norm": 2.2147282767091143, + "learning_rate": 5.647322245147762e-06, + "loss": 0.3238, + "step": 6943 + }, + { + "epoch": 0.4749982898967098, + "grad_norm": 2.453167234652493, + "learning_rate": 5.646223797248879e-06, + "loss": 0.4789, + "step": 6944 + }, + { + "epoch": 0.4750666940283193, + "grad_norm": 1.9317253140513335, + "learning_rate": 5.645125317630213e-06, + "loss": 0.5877, + "step": 6945 + }, + { + "epoch": 0.4751350981599289, + "grad_norm": 2.2568490206687097, + "learning_rate": 5.644026806345681e-06, + "loss": 0.2605, + "step": 6946 + }, + { + "epoch": 0.4752035022915384, + "grad_norm": 1.8351329562453396, + "learning_rate": 5.642928263449201e-06, + "loss": 0.2866, + "step": 6947 + }, + { + "epoch": 0.475271906423148, + "grad_norm": 2.5785215361847698, + "learning_rate": 5.641829688994699e-06, + "loss": 0.5434, + "step": 6948 + }, + { + "epoch": 0.4753403105547575, + "grad_norm": 2.242879023230441, + "learning_rate": 5.6407310830360966e-06, + "loss": 0.4741, + "step": 6949 + }, + { + "epoch": 0.4754087146863671, + "grad_norm": 1.5005219663990654, + "learning_rate": 5.639632445627318e-06, + "loss": 0.2825, + "step": 6950 + }, + { + "epoch": 0.4754771188179766, + "grad_norm": 1.9089180067002747, + "learning_rate": 5.63853377682229e-06, + "loss": 0.4466, + "step": 6951 + }, + { + "epoch": 0.47554552294958613, + "grad_norm": 2.4488093385327208, + "learning_rate": 5.637435076674942e-06, + "loss": 0.4994, + "step": 6952 + }, + { + "epoch": 0.4756139270811957, + "grad_norm": 1.9416063769712237, + "learning_rate": 5.636336345239202e-06, + "loss": 0.2458, + "step": 6953 + }, + { + "epoch": 0.47568233121280523, + "grad_norm": 2.2427228029430735, + "learning_rate": 5.6352375825689995e-06, + "loss": 0.5616, + "step": 6954 + }, + { + "epoch": 0.4757507353444148, + "grad_norm": 1.8915180233147875, + "learning_rate": 5.634138788718271e-06, + "loss": 0.3945, + "step": 6955 + }, + { + "epoch": 0.47581913947602433, + "grad_norm": 1.983879226418801, + "learning_rate": 5.633039963740948e-06, + "loss": 0.4693, + "step": 6956 + }, + { + "epoch": 0.4758875436076339, + "grad_norm": 2.1232681650399576, + "learning_rate": 5.631941107690967e-06, + "loss": 0.4247, + "step": 6957 + }, + { + "epoch": 0.47595594773924343, + "grad_norm": 1.9010278235327742, + "learning_rate": 5.630842220622263e-06, + "loss": 0.317, + "step": 6958 + }, + { + "epoch": 0.476024351870853, + "grad_norm": 1.9152530626462474, + "learning_rate": 5.62974330258878e-06, + "loss": 0.4298, + "step": 6959 + }, + { + "epoch": 0.47609275600246254, + "grad_norm": 2.574873405418641, + "learning_rate": 5.62864435364445e-06, + "loss": 0.4438, + "step": 6960 + }, + { + "epoch": 0.4761611601340721, + "grad_norm": 1.94125003301322, + "learning_rate": 5.627545373843223e-06, + "loss": 0.5886, + "step": 6961 + }, + { + "epoch": 0.47622956426568164, + "grad_norm": 1.9342875074157224, + "learning_rate": 5.626446363239038e-06, + "loss": 0.4549, + "step": 6962 + }, + { + "epoch": 0.4762979683972912, + "grad_norm": 1.9482621347271452, + "learning_rate": 5.62534732188584e-06, + "loss": 0.6379, + "step": 6963 + }, + { + "epoch": 0.47636637252890074, + "grad_norm": 2.5930464721138735, + "learning_rate": 5.624248249837575e-06, + "loss": 0.5695, + "step": 6964 + }, + { + "epoch": 0.4764347766605103, + "grad_norm": 2.1818732647327526, + "learning_rate": 5.623149147148193e-06, + "loss": 0.4209, + "step": 6965 + }, + { + "epoch": 0.47650318079211984, + "grad_norm": 1.6790293024942815, + "learning_rate": 5.6220500138716416e-06, + "loss": 0.2773, + "step": 6966 + }, + { + "epoch": 0.4765715849237294, + "grad_norm": 1.802931562986195, + "learning_rate": 5.62095085006187e-06, + "loss": 0.3374, + "step": 6967 + }, + { + "epoch": 0.47663998905533894, + "grad_norm": 1.805332242063052, + "learning_rate": 5.619851655772832e-06, + "loss": 0.3211, + "step": 6968 + }, + { + "epoch": 0.47670839318694846, + "grad_norm": 1.854282803237974, + "learning_rate": 5.618752431058484e-06, + "loss": 0.4571, + "step": 6969 + }, + { + "epoch": 0.47677679731855804, + "grad_norm": 2.059898122471929, + "learning_rate": 5.6176531759727755e-06, + "loss": 0.4467, + "step": 6970 + }, + { + "epoch": 0.47684520145016757, + "grad_norm": 1.2804608560659778, + "learning_rate": 5.6165538905696694e-06, + "loss": 0.1883, + "step": 6971 + }, + { + "epoch": 0.47691360558177714, + "grad_norm": 1.7916309668251897, + "learning_rate": 5.61545457490312e-06, + "loss": 0.3431, + "step": 6972 + }, + { + "epoch": 0.47698200971338667, + "grad_norm": 2.570900898982731, + "learning_rate": 5.614355229027088e-06, + "loss": 0.2044, + "step": 6973 + }, + { + "epoch": 0.47705041384499625, + "grad_norm": 1.8268242634564142, + "learning_rate": 5.613255852995535e-06, + "loss": 0.3082, + "step": 6974 + }, + { + "epoch": 0.47711881797660577, + "grad_norm": 2.259460197784031, + "learning_rate": 5.6121564468624235e-06, + "loss": 0.523, + "step": 6975 + }, + { + "epoch": 0.47718722210821535, + "grad_norm": 2.0309375119095563, + "learning_rate": 5.611057010681717e-06, + "loss": 0.5423, + "step": 6976 + }, + { + "epoch": 0.47725562623982487, + "grad_norm": 2.1604936208663905, + "learning_rate": 5.609957544507381e-06, + "loss": 0.1862, + "step": 6977 + }, + { + "epoch": 0.47732403037143445, + "grad_norm": 2.17310635905133, + "learning_rate": 5.608858048393384e-06, + "loss": 0.1972, + "step": 6978 + }, + { + "epoch": 0.47739243450304397, + "grad_norm": 2.228462058116425, + "learning_rate": 5.607758522393693e-06, + "loss": 0.5842, + "step": 6979 + }, + { + "epoch": 0.47746083863465355, + "grad_norm": 1.9208449922078177, + "learning_rate": 5.6066589665622804e-06, + "loss": 0.2767, + "step": 6980 + }, + { + "epoch": 0.4775292427662631, + "grad_norm": 1.6199781945719915, + "learning_rate": 5.605559380953116e-06, + "loss": 0.2312, + "step": 6981 + }, + { + "epoch": 0.47759764689787265, + "grad_norm": 2.2717220218161818, + "learning_rate": 5.604459765620171e-06, + "loss": 0.5636, + "step": 6982 + }, + { + "epoch": 0.4776660510294822, + "grad_norm": 2.4678355829213134, + "learning_rate": 5.603360120617424e-06, + "loss": 0.7642, + "step": 6983 + }, + { + "epoch": 0.47773445516109175, + "grad_norm": 1.7989148171872855, + "learning_rate": 5.6022604459988485e-06, + "loss": 0.2784, + "step": 6984 + }, + { + "epoch": 0.4778028592927013, + "grad_norm": 2.237779824111958, + "learning_rate": 5.601160741818422e-06, + "loss": 0.3178, + "step": 6985 + }, + { + "epoch": 0.47787126342431085, + "grad_norm": 1.839401722512025, + "learning_rate": 5.600061008130122e-06, + "loss": 0.4716, + "step": 6986 + }, + { + "epoch": 0.4779396675559204, + "grad_norm": 1.8190146555176492, + "learning_rate": 5.598961244987931e-06, + "loss": 0.3085, + "step": 6987 + }, + { + "epoch": 0.4780080716875299, + "grad_norm": 1.8114500416490409, + "learning_rate": 5.59786145244583e-06, + "loss": 0.276, + "step": 6988 + }, + { + "epoch": 0.4780764758191395, + "grad_norm": 1.1373223475179641, + "learning_rate": 5.596761630557801e-06, + "loss": 0.1452, + "step": 6989 + }, + { + "epoch": 0.478144879950749, + "grad_norm": 2.1296124538149015, + "learning_rate": 5.59566177937783e-06, + "loss": 0.3384, + "step": 6990 + }, + { + "epoch": 0.4782132840823586, + "grad_norm": 2.190426959556075, + "learning_rate": 5.594561898959903e-06, + "loss": 0.5635, + "step": 6991 + }, + { + "epoch": 0.4782816882139681, + "grad_norm": 1.8494519586655713, + "learning_rate": 5.593461989358005e-06, + "loss": 0.2169, + "step": 6992 + }, + { + "epoch": 0.4783500923455777, + "grad_norm": 2.56839444096636, + "learning_rate": 5.592362050626128e-06, + "loss": 0.8201, + "step": 6993 + }, + { + "epoch": 0.4784184964771872, + "grad_norm": 2.3749103215502925, + "learning_rate": 5.591262082818261e-06, + "loss": 0.514, + "step": 6994 + }, + { + "epoch": 0.4784869006087968, + "grad_norm": 2.0112824828036135, + "learning_rate": 5.590162085988397e-06, + "loss": 0.5077, + "step": 6995 + }, + { + "epoch": 0.4785553047404063, + "grad_norm": 1.6274695132186432, + "learning_rate": 5.589062060190527e-06, + "loss": 0.3174, + "step": 6996 + }, + { + "epoch": 0.4786237088720159, + "grad_norm": 2.190854443385166, + "learning_rate": 5.5879620054786454e-06, + "loss": 0.7344, + "step": 6997 + }, + { + "epoch": 0.4786921130036254, + "grad_norm": 1.9261848924266982, + "learning_rate": 5.58686192190675e-06, + "loss": 0.4618, + "step": 6998 + }, + { + "epoch": 0.478760517135235, + "grad_norm": 2.13557946623527, + "learning_rate": 5.585761809528839e-06, + "loss": 0.5529, + "step": 6999 + }, + { + "epoch": 0.4788289212668445, + "grad_norm": 8.113367644994689, + "learning_rate": 5.584661668398907e-06, + "loss": 0.2636, + "step": 7000 + }, + { + "epoch": 0.4788973253984541, + "grad_norm": 3.857292231059138, + "learning_rate": 5.58356149857096e-06, + "loss": 0.2569, + "step": 7001 + }, + { + "epoch": 0.4789657295300636, + "grad_norm": 1.8534889339860685, + "learning_rate": 5.5824613000989946e-06, + "loss": 0.3482, + "step": 7002 + }, + { + "epoch": 0.4790341336616732, + "grad_norm": 2.3435341799233207, + "learning_rate": 5.581361073037016e-06, + "loss": 0.6515, + "step": 7003 + }, + { + "epoch": 0.4791025377932827, + "grad_norm": 2.471737779939172, + "learning_rate": 5.580260817439029e-06, + "loss": 0.3692, + "step": 7004 + }, + { + "epoch": 0.4791709419248923, + "grad_norm": 1.9367686860362552, + "learning_rate": 5.57916053335904e-06, + "loss": 0.383, + "step": 7005 + }, + { + "epoch": 0.4792393460565018, + "grad_norm": 1.851768458053846, + "learning_rate": 5.578060220851055e-06, + "loss": 0.4879, + "step": 7006 + }, + { + "epoch": 0.47930775018811134, + "grad_norm": 2.046035463392153, + "learning_rate": 5.576959879969082e-06, + "loss": 0.5222, + "step": 7007 + }, + { + "epoch": 0.4793761543197209, + "grad_norm": 1.7896794750366083, + "learning_rate": 5.5758595107671325e-06, + "loss": 0.3324, + "step": 7008 + }, + { + "epoch": 0.47944455845133044, + "grad_norm": 1.8106354503507776, + "learning_rate": 5.5747591132992176e-06, + "loss": 0.3367, + "step": 7009 + }, + { + "epoch": 0.47951296258294, + "grad_norm": 1.7402571517190186, + "learning_rate": 5.573658687619349e-06, + "loss": 0.3402, + "step": 7010 + }, + { + "epoch": 0.47958136671454954, + "grad_norm": 2.047992081157455, + "learning_rate": 5.572558233781543e-06, + "loss": 0.4578, + "step": 7011 + }, + { + "epoch": 0.4796497708461591, + "grad_norm": 1.8770919097495593, + "learning_rate": 5.5714577518398125e-06, + "loss": 0.4158, + "step": 7012 + }, + { + "epoch": 0.47971817497776864, + "grad_norm": 1.5066335620076647, + "learning_rate": 5.5703572418481776e-06, + "loss": 0.2263, + "step": 7013 + }, + { + "epoch": 0.4797865791093782, + "grad_norm": 2.651083583662782, + "learning_rate": 5.5692567038606544e-06, + "loss": 0.3195, + "step": 7014 + }, + { + "epoch": 0.47985498324098774, + "grad_norm": 1.318872880182487, + "learning_rate": 5.568156137931264e-06, + "loss": 0.1813, + "step": 7015 + }, + { + "epoch": 0.4799233873725973, + "grad_norm": 1.9440703351921145, + "learning_rate": 5.567055544114026e-06, + "loss": 0.3561, + "step": 7016 + }, + { + "epoch": 0.47999179150420684, + "grad_norm": 2.312044659453414, + "learning_rate": 5.565954922462963e-06, + "loss": 0.4035, + "step": 7017 + }, + { + "epoch": 0.4800601956358164, + "grad_norm": 1.6274729215475865, + "learning_rate": 5.564854273032101e-06, + "loss": 0.2008, + "step": 7018 + }, + { + "epoch": 0.48012859976742595, + "grad_norm": 1.6444563225646154, + "learning_rate": 5.563753595875463e-06, + "loss": 0.3206, + "step": 7019 + }, + { + "epoch": 0.4801970038990355, + "grad_norm": 2.568726662021624, + "learning_rate": 5.562652891047076e-06, + "loss": 0.4405, + "step": 7020 + }, + { + "epoch": 0.48026540803064505, + "grad_norm": 1.9301442366400423, + "learning_rate": 5.561552158600968e-06, + "loss": 0.5337, + "step": 7021 + }, + { + "epoch": 0.4803338121622546, + "grad_norm": 2.0150232525754634, + "learning_rate": 5.560451398591168e-06, + "loss": 0.3587, + "step": 7022 + }, + { + "epoch": 0.48040221629386415, + "grad_norm": 1.9933658555643303, + "learning_rate": 5.559350611071707e-06, + "loss": 0.3938, + "step": 7023 + }, + { + "epoch": 0.48047062042547367, + "grad_norm": 2.0231133131338415, + "learning_rate": 5.558249796096618e-06, + "loss": 0.4421, + "step": 7024 + }, + { + "epoch": 0.48053902455708325, + "grad_norm": 7.9259098281320135, + "learning_rate": 5.557148953719932e-06, + "loss": 0.4313, + "step": 7025 + }, + { + "epoch": 0.4806074286886928, + "grad_norm": 1.9707585034940058, + "learning_rate": 5.556048083995685e-06, + "loss": 0.4734, + "step": 7026 + }, + { + "epoch": 0.48067583282030235, + "grad_norm": 1.8647871350059413, + "learning_rate": 5.554947186977913e-06, + "loss": 0.3139, + "step": 7027 + }, + { + "epoch": 0.4807442369519119, + "grad_norm": 1.8016628498803902, + "learning_rate": 5.5538462627206535e-06, + "loss": 0.2757, + "step": 7028 + }, + { + "epoch": 0.48081264108352145, + "grad_norm": 2.0337389552315304, + "learning_rate": 5.552745311277945e-06, + "loss": 0.238, + "step": 7029 + }, + { + "epoch": 0.480881045215131, + "grad_norm": 2.229309315142923, + "learning_rate": 5.551644332703828e-06, + "loss": 0.4468, + "step": 7030 + }, + { + "epoch": 0.48094944934674055, + "grad_norm": 2.049868833020718, + "learning_rate": 5.550543327052343e-06, + "loss": 0.6288, + "step": 7031 + }, + { + "epoch": 0.4810178534783501, + "grad_norm": 2.3577172853572312, + "learning_rate": 5.549442294377532e-06, + "loss": 0.6425, + "step": 7032 + }, + { + "epoch": 0.48108625760995966, + "grad_norm": 1.9046677291882954, + "learning_rate": 5.548341234733441e-06, + "loss": 0.3687, + "step": 7033 + }, + { + "epoch": 0.4811546617415692, + "grad_norm": 1.1613122979403072, + "learning_rate": 5.547240148174113e-06, + "loss": 0.181, + "step": 7034 + }, + { + "epoch": 0.48122306587317876, + "grad_norm": 1.9865978924243535, + "learning_rate": 5.5461390347535985e-06, + "loss": 0.226, + "step": 7035 + }, + { + "epoch": 0.4812914700047883, + "grad_norm": 2.593110476784389, + "learning_rate": 5.545037894525941e-06, + "loss": 0.4119, + "step": 7036 + }, + { + "epoch": 0.48135987413639786, + "grad_norm": 2.5294765924063274, + "learning_rate": 5.5439367275451905e-06, + "loss": 0.626, + "step": 7037 + }, + { + "epoch": 0.4814282782680074, + "grad_norm": 1.933066803000367, + "learning_rate": 5.5428355338654014e-06, + "loss": 0.5609, + "step": 7038 + }, + { + "epoch": 0.48149668239961696, + "grad_norm": 2.0925792153608236, + "learning_rate": 5.5417343135406206e-06, + "loss": 0.4179, + "step": 7039 + }, + { + "epoch": 0.4815650865312265, + "grad_norm": 2.5266771100624412, + "learning_rate": 5.540633066624906e-06, + "loss": 0.5821, + "step": 7040 + }, + { + "epoch": 0.48163349066283606, + "grad_norm": 2.135817998683959, + "learning_rate": 5.539531793172309e-06, + "loss": 0.5729, + "step": 7041 + }, + { + "epoch": 0.4817018947944456, + "grad_norm": 2.0212742424444734, + "learning_rate": 5.538430493236885e-06, + "loss": 0.5121, + "step": 7042 + }, + { + "epoch": 0.4817702989260551, + "grad_norm": 2.2918157066662546, + "learning_rate": 5.537329166872693e-06, + "loss": 0.3125, + "step": 7043 + }, + { + "epoch": 0.4818387030576647, + "grad_norm": 1.7091554479805837, + "learning_rate": 5.536227814133791e-06, + "loss": 0.2255, + "step": 7044 + }, + { + "epoch": 0.4819071071892742, + "grad_norm": 8.138179528888875, + "learning_rate": 5.53512643507424e-06, + "loss": 0.5109, + "step": 7045 + }, + { + "epoch": 0.4819755113208838, + "grad_norm": 1.414293191333199, + "learning_rate": 5.534025029748097e-06, + "loss": 0.1521, + "step": 7046 + }, + { + "epoch": 0.4820439154524933, + "grad_norm": 1.4424835820726492, + "learning_rate": 5.532923598209427e-06, + "loss": 0.149, + "step": 7047 + }, + { + "epoch": 0.4821123195841029, + "grad_norm": 2.4679446053478618, + "learning_rate": 5.531822140512294e-06, + "loss": 0.6544, + "step": 7048 + }, + { + "epoch": 0.4821807237157124, + "grad_norm": 2.0973969624278466, + "learning_rate": 5.530720656710762e-06, + "loss": 0.7497, + "step": 7049 + }, + { + "epoch": 0.482249127847322, + "grad_norm": 2.1387051653774964, + "learning_rate": 5.5296191468588975e-06, + "loss": 0.5211, + "step": 7050 + }, + { + "epoch": 0.4823175319789315, + "grad_norm": 2.175022770362355, + "learning_rate": 5.528517611010769e-06, + "loss": 0.3304, + "step": 7051 + }, + { + "epoch": 0.4823859361105411, + "grad_norm": 1.8766124666281156, + "learning_rate": 5.527416049220442e-06, + "loss": 0.3666, + "step": 7052 + }, + { + "epoch": 0.4824543402421506, + "grad_norm": 2.2674658695794054, + "learning_rate": 5.52631446154199e-06, + "loss": 0.2983, + "step": 7053 + }, + { + "epoch": 0.4825227443737602, + "grad_norm": 1.664862641863278, + "learning_rate": 5.525212848029482e-06, + "loss": 0.3346, + "step": 7054 + }, + { + "epoch": 0.4825911485053697, + "grad_norm": 2.1325445041730657, + "learning_rate": 5.524111208736991e-06, + "loss": 0.3971, + "step": 7055 + }, + { + "epoch": 0.4826595526369793, + "grad_norm": 1.875003949796967, + "learning_rate": 5.5230095437185905e-06, + "loss": 0.4325, + "step": 7056 + }, + { + "epoch": 0.4827279567685888, + "grad_norm": 2.1342961911163694, + "learning_rate": 5.521907853028356e-06, + "loss": 0.4278, + "step": 7057 + }, + { + "epoch": 0.4827963609001984, + "grad_norm": 1.8859567939890178, + "learning_rate": 5.520806136720365e-06, + "loss": 0.3785, + "step": 7058 + }, + { + "epoch": 0.4828647650318079, + "grad_norm": 3.1345052545624843, + "learning_rate": 5.519704394848693e-06, + "loss": 0.2018, + "step": 7059 + }, + { + "epoch": 0.4829331691634175, + "grad_norm": 2.0967009942616084, + "learning_rate": 5.51860262746742e-06, + "loss": 0.4232, + "step": 7060 + }, + { + "epoch": 0.483001573295027, + "grad_norm": 1.7172552891992874, + "learning_rate": 5.517500834630627e-06, + "loss": 0.1945, + "step": 7061 + }, + { + "epoch": 0.48306997742663654, + "grad_norm": 2.267054524455698, + "learning_rate": 5.516399016392392e-06, + "loss": 0.3378, + "step": 7062 + }, + { + "epoch": 0.4831383815582461, + "grad_norm": 2.1427607010812384, + "learning_rate": 5.5152971728068e-06, + "loss": 0.4416, + "step": 7063 + }, + { + "epoch": 0.48320678568985564, + "grad_norm": 2.339633693404998, + "learning_rate": 5.514195303927935e-06, + "loss": 0.5515, + "step": 7064 + }, + { + "epoch": 0.4832751898214652, + "grad_norm": 2.443604521799314, + "learning_rate": 5.513093409809881e-06, + "loss": 0.2965, + "step": 7065 + }, + { + "epoch": 0.48334359395307475, + "grad_norm": 1.6266699357880152, + "learning_rate": 5.511991490506725e-06, + "loss": 0.2092, + "step": 7066 + }, + { + "epoch": 0.4834119980846843, + "grad_norm": 2.3273601619760282, + "learning_rate": 5.5108895460725535e-06, + "loss": 0.6802, + "step": 7067 + }, + { + "epoch": 0.48348040221629385, + "grad_norm": 1.8856104918867875, + "learning_rate": 5.5097875765614575e-06, + "loss": 0.3464, + "step": 7068 + }, + { + "epoch": 0.4835488063479034, + "grad_norm": 2.082520214586045, + "learning_rate": 5.508685582027524e-06, + "loss": 0.5391, + "step": 7069 + }, + { + "epoch": 0.48361721047951295, + "grad_norm": 2.226428308960613, + "learning_rate": 5.507583562524846e-06, + "loss": 0.5715, + "step": 7070 + }, + { + "epoch": 0.4836856146111225, + "grad_norm": 1.9454446579189977, + "learning_rate": 5.506481518107517e-06, + "loss": 0.3102, + "step": 7071 + }, + { + "epoch": 0.48375401874273205, + "grad_norm": 2.1419766398130218, + "learning_rate": 5.505379448829629e-06, + "loss": 0.333, + "step": 7072 + }, + { + "epoch": 0.48382242287434163, + "grad_norm": 2.2448512281111546, + "learning_rate": 5.504277354745275e-06, + "loss": 0.3451, + "step": 7073 + }, + { + "epoch": 0.48389082700595115, + "grad_norm": 2.838455102049719, + "learning_rate": 5.503175235908556e-06, + "loss": 0.3918, + "step": 7074 + }, + { + "epoch": 0.48395923113756073, + "grad_norm": 1.7283078429774237, + "learning_rate": 5.502073092373564e-06, + "loss": 0.2951, + "step": 7075 + }, + { + "epoch": 0.48402763526917025, + "grad_norm": 2.6723901519734827, + "learning_rate": 5.500970924194402e-06, + "loss": 0.4806, + "step": 7076 + }, + { + "epoch": 0.48409603940077983, + "grad_norm": 1.6410125728761313, + "learning_rate": 5.499868731425167e-06, + "loss": 0.287, + "step": 7077 + }, + { + "epoch": 0.48416444353238935, + "grad_norm": 1.566959221664431, + "learning_rate": 5.498766514119963e-06, + "loss": 0.2224, + "step": 7078 + }, + { + "epoch": 0.48423284766399893, + "grad_norm": 1.7737708513648616, + "learning_rate": 5.497664272332888e-06, + "loss": 0.4292, + "step": 7079 + }, + { + "epoch": 0.48430125179560846, + "grad_norm": 2.1166263026577763, + "learning_rate": 5.496562006118048e-06, + "loss": 0.3915, + "step": 7080 + }, + { + "epoch": 0.484369655927218, + "grad_norm": 2.268527637965786, + "learning_rate": 5.495459715529545e-06, + "loss": 0.4483, + "step": 7081 + }, + { + "epoch": 0.48443806005882756, + "grad_norm": 1.9010557165454076, + "learning_rate": 5.494357400621488e-06, + "loss": 0.4872, + "step": 7082 + }, + { + "epoch": 0.4845064641904371, + "grad_norm": 2.3529702183806074, + "learning_rate": 5.493255061447984e-06, + "loss": 0.3998, + "step": 7083 + }, + { + "epoch": 0.48457486832204666, + "grad_norm": 2.1969829360525397, + "learning_rate": 5.492152698063138e-06, + "loss": 0.6019, + "step": 7084 + }, + { + "epoch": 0.4846432724536562, + "grad_norm": 1.8018915307685621, + "learning_rate": 5.491050310521061e-06, + "loss": 0.3351, + "step": 7085 + }, + { + "epoch": 0.48471167658526576, + "grad_norm": 3.869503583834209, + "learning_rate": 5.489947898875864e-06, + "loss": 0.1745, + "step": 7086 + }, + { + "epoch": 0.4847800807168753, + "grad_norm": 2.085862044066201, + "learning_rate": 5.488845463181658e-06, + "loss": 0.4755, + "step": 7087 + }, + { + "epoch": 0.48484848484848486, + "grad_norm": 3.3378578304958753, + "learning_rate": 5.4877430034925575e-06, + "loss": 0.4735, + "step": 7088 + }, + { + "epoch": 0.4849168889800944, + "grad_norm": 2.1706097574355114, + "learning_rate": 5.486640519862673e-06, + "loss": 0.2178, + "step": 7089 + }, + { + "epoch": 0.48498529311170396, + "grad_norm": 1.6781094790119184, + "learning_rate": 5.485538012346122e-06, + "loss": 0.182, + "step": 7090 + }, + { + "epoch": 0.4850536972433135, + "grad_norm": 1.8865812105391868, + "learning_rate": 5.484435480997022e-06, + "loss": 0.5768, + "step": 7091 + }, + { + "epoch": 0.48512210137492306, + "grad_norm": 1.2663712891466037, + "learning_rate": 5.4833329258694865e-06, + "loss": 0.1481, + "step": 7092 + }, + { + "epoch": 0.4851905055065326, + "grad_norm": 2.957728201747536, + "learning_rate": 5.482230347017639e-06, + "loss": 0.5194, + "step": 7093 + }, + { + "epoch": 0.48525890963814217, + "grad_norm": 1.9120753892132385, + "learning_rate": 5.481127744495599e-06, + "loss": 0.4539, + "step": 7094 + }, + { + "epoch": 0.4853273137697517, + "grad_norm": 1.457665798701054, + "learning_rate": 5.480025118357483e-06, + "loss": 0.2325, + "step": 7095 + }, + { + "epoch": 0.48539571790136127, + "grad_norm": 1.5735085780995108, + "learning_rate": 5.478922468657419e-06, + "loss": 0.269, + "step": 7096 + }, + { + "epoch": 0.4854641220329708, + "grad_norm": 2.450525096655828, + "learning_rate": 5.477819795449526e-06, + "loss": 0.4256, + "step": 7097 + }, + { + "epoch": 0.4855325261645803, + "grad_norm": 1.400907683052001, + "learning_rate": 5.47671709878793e-06, + "loss": 0.1335, + "step": 7098 + }, + { + "epoch": 0.4856009302961899, + "grad_norm": 2.10173818499154, + "learning_rate": 5.475614378726757e-06, + "loss": 0.3981, + "step": 7099 + }, + { + "epoch": 0.4856693344277994, + "grad_norm": 2.025929029041888, + "learning_rate": 5.474511635320133e-06, + "loss": 0.4224, + "step": 7100 + }, + { + "epoch": 0.485737738559409, + "grad_norm": 1.5627810225494694, + "learning_rate": 5.473408868622188e-06, + "loss": 0.1599, + "step": 7101 + }, + { + "epoch": 0.4858061426910185, + "grad_norm": 1.8855092296394673, + "learning_rate": 5.472306078687047e-06, + "loss": 0.4712, + "step": 7102 + }, + { + "epoch": 0.4858745468226281, + "grad_norm": 2.094984537525063, + "learning_rate": 5.471203265568845e-06, + "loss": 0.3066, + "step": 7103 + }, + { + "epoch": 0.4859429509542376, + "grad_norm": 2.490401569812346, + "learning_rate": 5.470100429321712e-06, + "loss": 0.326, + "step": 7104 + }, + { + "epoch": 0.4860113550858472, + "grad_norm": 3.2256344622092743, + "learning_rate": 5.468997569999778e-06, + "loss": 0.2242, + "step": 7105 + }, + { + "epoch": 0.4860797592174567, + "grad_norm": 2.108475359051964, + "learning_rate": 5.467894687657178e-06, + "loss": 0.3386, + "step": 7106 + }, + { + "epoch": 0.4861481633490663, + "grad_norm": 2.2024480002724065, + "learning_rate": 5.46679178234805e-06, + "loss": 0.2149, + "step": 7107 + }, + { + "epoch": 0.4862165674806758, + "grad_norm": 1.7018779817723866, + "learning_rate": 5.465688854126526e-06, + "loss": 0.3724, + "step": 7108 + }, + { + "epoch": 0.4862849716122854, + "grad_norm": 2.354868136959219, + "learning_rate": 5.464585903046744e-06, + "loss": 0.4972, + "step": 7109 + }, + { + "epoch": 0.4863533757438949, + "grad_norm": 2.2412740262196333, + "learning_rate": 5.463482929162842e-06, + "loss": 0.5015, + "step": 7110 + }, + { + "epoch": 0.4864217798755045, + "grad_norm": 1.7377377914239402, + "learning_rate": 5.46237993252896e-06, + "loss": 0.1889, + "step": 7111 + }, + { + "epoch": 0.486490184007114, + "grad_norm": 2.023402486225728, + "learning_rate": 5.461276913199238e-06, + "loss": 0.4074, + "step": 7112 + }, + { + "epoch": 0.4865585881387236, + "grad_norm": 2.6284745075821165, + "learning_rate": 5.4601738712278184e-06, + "loss": 0.4653, + "step": 7113 + }, + { + "epoch": 0.4866269922703331, + "grad_norm": 2.204440373521291, + "learning_rate": 5.459070806668843e-06, + "loss": 0.632, + "step": 7114 + }, + { + "epoch": 0.4866953964019427, + "grad_norm": 2.3061227887432656, + "learning_rate": 5.457967719576455e-06, + "loss": 0.5375, + "step": 7115 + }, + { + "epoch": 0.4867638005335522, + "grad_norm": 2.3392212410614732, + "learning_rate": 5.456864610004798e-06, + "loss": 0.463, + "step": 7116 + }, + { + "epoch": 0.48683220466516175, + "grad_norm": 2.5474552184016375, + "learning_rate": 5.455761478008023e-06, + "loss": 0.7365, + "step": 7117 + }, + { + "epoch": 0.48690060879677133, + "grad_norm": 2.064979291285572, + "learning_rate": 5.454658323640273e-06, + "loss": 0.3443, + "step": 7118 + }, + { + "epoch": 0.48696901292838085, + "grad_norm": 1.63027053984479, + "learning_rate": 5.453555146955696e-06, + "loss": 0.1773, + "step": 7119 + }, + { + "epoch": 0.48703741705999043, + "grad_norm": 1.9081847464263062, + "learning_rate": 5.452451948008443e-06, + "loss": 0.3377, + "step": 7120 + }, + { + "epoch": 0.48710582119159995, + "grad_norm": 2.609285764253385, + "learning_rate": 5.451348726852662e-06, + "loss": 0.5048, + "step": 7121 + }, + { + "epoch": 0.48717422532320953, + "grad_norm": 1.4389853694449866, + "learning_rate": 5.450245483542507e-06, + "loss": 0.1989, + "step": 7122 + }, + { + "epoch": 0.48724262945481905, + "grad_norm": 1.7515581572459251, + "learning_rate": 5.44914221813213e-06, + "loss": 0.3079, + "step": 7123 + }, + { + "epoch": 0.48731103358642863, + "grad_norm": 1.618977564940184, + "learning_rate": 5.448038930675684e-06, + "loss": 0.3425, + "step": 7124 + }, + { + "epoch": 0.48737943771803816, + "grad_norm": 1.5686245524261173, + "learning_rate": 5.446935621227323e-06, + "loss": 0.2176, + "step": 7125 + }, + { + "epoch": 0.48744784184964773, + "grad_norm": 1.8590313950580115, + "learning_rate": 5.445832289841203e-06, + "loss": 0.3195, + "step": 7126 + }, + { + "epoch": 0.48751624598125726, + "grad_norm": 1.6286994839444071, + "learning_rate": 5.444728936571483e-06, + "loss": 0.3807, + "step": 7127 + }, + { + "epoch": 0.48758465011286684, + "grad_norm": 2.04414451273631, + "learning_rate": 5.443625561472318e-06, + "loss": 0.4463, + "step": 7128 + }, + { + "epoch": 0.48765305424447636, + "grad_norm": 1.6597180674601557, + "learning_rate": 5.442522164597868e-06, + "loss": 0.2841, + "step": 7129 + }, + { + "epoch": 0.48772145837608594, + "grad_norm": 2.1193176476861186, + "learning_rate": 5.441418746002295e-06, + "loss": 0.6156, + "step": 7130 + }, + { + "epoch": 0.48778986250769546, + "grad_norm": 2.358205195816695, + "learning_rate": 5.440315305739756e-06, + "loss": 0.6092, + "step": 7131 + }, + { + "epoch": 0.48785826663930504, + "grad_norm": 1.864300007489988, + "learning_rate": 5.439211843864416e-06, + "loss": 0.3016, + "step": 7132 + }, + { + "epoch": 0.48792667077091456, + "grad_norm": 1.7840373249185397, + "learning_rate": 5.438108360430439e-06, + "loss": 0.4709, + "step": 7133 + }, + { + "epoch": 0.48799507490252414, + "grad_norm": 1.699844459998502, + "learning_rate": 5.437004855491988e-06, + "loss": 0.2079, + "step": 7134 + }, + { + "epoch": 0.48806347903413366, + "grad_norm": 2.236794488912316, + "learning_rate": 5.435901329103227e-06, + "loss": 0.6365, + "step": 7135 + }, + { + "epoch": 0.4881318831657432, + "grad_norm": 1.8248515165032895, + "learning_rate": 5.434797781318325e-06, + "loss": 0.2991, + "step": 7136 + }, + { + "epoch": 0.48820028729735276, + "grad_norm": 1.6274159976270208, + "learning_rate": 5.433694212191448e-06, + "loss": 0.2054, + "step": 7137 + }, + { + "epoch": 0.4882686914289623, + "grad_norm": 1.6202369309978206, + "learning_rate": 5.432590621776764e-06, + "loss": 0.3628, + "step": 7138 + }, + { + "epoch": 0.48833709556057187, + "grad_norm": 1.7792350306058764, + "learning_rate": 5.431487010128445e-06, + "loss": 0.2328, + "step": 7139 + }, + { + "epoch": 0.4884054996921814, + "grad_norm": 1.2167737233024436, + "learning_rate": 5.43038337730066e-06, + "loss": 0.1664, + "step": 7140 + }, + { + "epoch": 0.48847390382379097, + "grad_norm": 2.4079268421548443, + "learning_rate": 5.429279723347578e-06, + "loss": 0.3431, + "step": 7141 + }, + { + "epoch": 0.4885423079554005, + "grad_norm": 2.2741662263456286, + "learning_rate": 5.4281760483233755e-06, + "loss": 0.1726, + "step": 7142 + }, + { + "epoch": 0.48861071208701007, + "grad_norm": 1.7825436871713796, + "learning_rate": 5.427072352282225e-06, + "loss": 0.1346, + "step": 7143 + }, + { + "epoch": 0.4886791162186196, + "grad_norm": 1.7741089192657935, + "learning_rate": 5.425968635278303e-06, + "loss": 0.3433, + "step": 7144 + }, + { + "epoch": 0.48874752035022917, + "grad_norm": 2.0840196293688487, + "learning_rate": 5.424864897365781e-06, + "loss": 0.1757, + "step": 7145 + }, + { + "epoch": 0.4888159244818387, + "grad_norm": 1.7992365577402862, + "learning_rate": 5.423761138598839e-06, + "loss": 0.4111, + "step": 7146 + }, + { + "epoch": 0.48888432861344827, + "grad_norm": 2.236068490627875, + "learning_rate": 5.422657359031653e-06, + "loss": 0.3631, + "step": 7147 + }, + { + "epoch": 0.4889527327450578, + "grad_norm": 1.5588855087905507, + "learning_rate": 5.421553558718404e-06, + "loss": 0.2591, + "step": 7148 + }, + { + "epoch": 0.4890211368766674, + "grad_norm": 2.9682523812299153, + "learning_rate": 5.4204497377132705e-06, + "loss": 0.4517, + "step": 7149 + }, + { + "epoch": 0.4890895410082769, + "grad_norm": 2.0264572638288594, + "learning_rate": 5.419345896070434e-06, + "loss": 0.4496, + "step": 7150 + }, + { + "epoch": 0.4891579451398865, + "grad_norm": 2.646566844373131, + "learning_rate": 5.418242033844075e-06, + "loss": 0.5938, + "step": 7151 + }, + { + "epoch": 0.489226349271496, + "grad_norm": 2.095812358302529, + "learning_rate": 5.417138151088377e-06, + "loss": 0.2746, + "step": 7152 + }, + { + "epoch": 0.4892947534031055, + "grad_norm": 1.8242647422530993, + "learning_rate": 5.416034247857526e-06, + "loss": 0.3615, + "step": 7153 + }, + { + "epoch": 0.4893631575347151, + "grad_norm": 1.6306135026534472, + "learning_rate": 5.4149303242057036e-06, + "loss": 0.2812, + "step": 7154 + }, + { + "epoch": 0.4894315616663246, + "grad_norm": 2.8481176810958475, + "learning_rate": 5.413826380187097e-06, + "loss": 0.9065, + "step": 7155 + }, + { + "epoch": 0.4894999657979342, + "grad_norm": 1.932405094885856, + "learning_rate": 5.412722415855893e-06, + "loss": 0.4754, + "step": 7156 + }, + { + "epoch": 0.4895683699295437, + "grad_norm": 2.1403212853929645, + "learning_rate": 5.411618431266282e-06, + "loss": 0.5521, + "step": 7157 + }, + { + "epoch": 0.4896367740611533, + "grad_norm": 2.1856708781200003, + "learning_rate": 5.4105144264724484e-06, + "loss": 0.3628, + "step": 7158 + }, + { + "epoch": 0.4897051781927628, + "grad_norm": 2.04146872788717, + "learning_rate": 5.409410401528586e-06, + "loss": 0.3332, + "step": 7159 + }, + { + "epoch": 0.4897735823243724, + "grad_norm": 3.4895406075134647, + "learning_rate": 5.408306356488884e-06, + "loss": 0.6934, + "step": 7160 + }, + { + "epoch": 0.4898419864559819, + "grad_norm": 1.9508312031037525, + "learning_rate": 5.407202291407534e-06, + "loss": 0.4355, + "step": 7161 + }, + { + "epoch": 0.4899103905875915, + "grad_norm": 1.9364927888677372, + "learning_rate": 5.406098206338728e-06, + "loss": 0.3039, + "step": 7162 + }, + { + "epoch": 0.489978794719201, + "grad_norm": 2.1834311109105133, + "learning_rate": 5.404994101336662e-06, + "loss": 0.2681, + "step": 7163 + }, + { + "epoch": 0.4900471988508106, + "grad_norm": 1.6082656309866088, + "learning_rate": 5.403889976455531e-06, + "loss": 0.3318, + "step": 7164 + }, + { + "epoch": 0.49011560298242013, + "grad_norm": 2.32280049368533, + "learning_rate": 5.402785831749528e-06, + "loss": 0.4879, + "step": 7165 + }, + { + "epoch": 0.4901840071140297, + "grad_norm": 2.065808085235623, + "learning_rate": 5.401681667272851e-06, + "loss": 0.2697, + "step": 7166 + }, + { + "epoch": 0.49025241124563923, + "grad_norm": 2.15514702851867, + "learning_rate": 5.4005774830797e-06, + "loss": 0.336, + "step": 7167 + }, + { + "epoch": 0.4903208153772488, + "grad_norm": 2.113394531485554, + "learning_rate": 5.399473279224269e-06, + "loss": 0.5316, + "step": 7168 + }, + { + "epoch": 0.49038921950885833, + "grad_norm": 2.285597623275683, + "learning_rate": 5.3983690557607626e-06, + "loss": 0.5883, + "step": 7169 + }, + { + "epoch": 0.4904576236404679, + "grad_norm": 1.7918304661424864, + "learning_rate": 5.397264812743379e-06, + "loss": 0.3337, + "step": 7170 + }, + { + "epoch": 0.49052602777207743, + "grad_norm": 5.258846748056916, + "learning_rate": 5.396160550226319e-06, + "loss": 0.355, + "step": 7171 + }, + { + "epoch": 0.49059443190368696, + "grad_norm": 2.5663180982061773, + "learning_rate": 5.395056268263786e-06, + "loss": 0.2809, + "step": 7172 + }, + { + "epoch": 0.49066283603529653, + "grad_norm": 2.3120909084143957, + "learning_rate": 5.393951966909986e-06, + "loss": 0.6754, + "step": 7173 + }, + { + "epoch": 0.49073124016690606, + "grad_norm": 2.013163416370697, + "learning_rate": 5.392847646219118e-06, + "loss": 0.3373, + "step": 7174 + }, + { + "epoch": 0.49079964429851564, + "grad_norm": 2.724151550986977, + "learning_rate": 5.391743306245392e-06, + "loss": 0.3499, + "step": 7175 + }, + { + "epoch": 0.49086804843012516, + "grad_norm": 2.457558143292684, + "learning_rate": 5.390638947043013e-06, + "loss": 0.3378, + "step": 7176 + }, + { + "epoch": 0.49093645256173474, + "grad_norm": 2.4291809817449073, + "learning_rate": 5.3895345686661896e-06, + "loss": 0.18, + "step": 7177 + }, + { + "epoch": 0.49100485669334426, + "grad_norm": 1.9961627343606652, + "learning_rate": 5.388430171169126e-06, + "loss": 0.339, + "step": 7178 + }, + { + "epoch": 0.49107326082495384, + "grad_norm": 1.8836350148102023, + "learning_rate": 5.387325754606035e-06, + "loss": 0.2675, + "step": 7179 + }, + { + "epoch": 0.49114166495656336, + "grad_norm": 1.780972651589722, + "learning_rate": 5.386221319031125e-06, + "loss": 0.3499, + "step": 7180 + }, + { + "epoch": 0.49121006908817294, + "grad_norm": 2.2444467227898524, + "learning_rate": 5.385116864498607e-06, + "loss": 0.738, + "step": 7181 + }, + { + "epoch": 0.49127847321978246, + "grad_norm": 1.875836408345117, + "learning_rate": 5.384012391062695e-06, + "loss": 0.3932, + "step": 7182 + }, + { + "epoch": 0.49134687735139204, + "grad_norm": 2.2967343579265496, + "learning_rate": 5.382907898777601e-06, + "loss": 0.5123, + "step": 7183 + }, + { + "epoch": 0.49141528148300156, + "grad_norm": 1.8801566343165235, + "learning_rate": 5.381803387697536e-06, + "loss": 0.2906, + "step": 7184 + }, + { + "epoch": 0.49148368561461114, + "grad_norm": 1.926856787141095, + "learning_rate": 5.3806988578767185e-06, + "loss": 0.3506, + "step": 7185 + }, + { + "epoch": 0.49155208974622067, + "grad_norm": 1.6606984117129426, + "learning_rate": 5.379594309369363e-06, + "loss": 0.283, + "step": 7186 + }, + { + "epoch": 0.49162049387783024, + "grad_norm": 2.7261735097604474, + "learning_rate": 5.378489742229685e-06, + "loss": 0.2477, + "step": 7187 + }, + { + "epoch": 0.49168889800943977, + "grad_norm": 2.007649218380472, + "learning_rate": 5.377385156511903e-06, + "loss": 0.2134, + "step": 7188 + }, + { + "epoch": 0.49175730214104935, + "grad_norm": 2.0680778679271383, + "learning_rate": 5.376280552270235e-06, + "loss": 0.6101, + "step": 7189 + }, + { + "epoch": 0.49182570627265887, + "grad_norm": 2.0462210134609844, + "learning_rate": 5.3751759295589e-06, + "loss": 0.4688, + "step": 7190 + }, + { + "epoch": 0.4918941104042684, + "grad_norm": 1.8272825402898247, + "learning_rate": 5.374071288432119e-06, + "loss": 0.3099, + "step": 7191 + }, + { + "epoch": 0.49196251453587797, + "grad_norm": 1.9228514617563417, + "learning_rate": 5.372966628944114e-06, + "loss": 0.4544, + "step": 7192 + }, + { + "epoch": 0.4920309186674875, + "grad_norm": 2.4401849636975994, + "learning_rate": 5.371861951149106e-06, + "loss": 0.6637, + "step": 7193 + }, + { + "epoch": 0.49209932279909707, + "grad_norm": 1.8562180261073795, + "learning_rate": 5.370757255101315e-06, + "loss": 0.3209, + "step": 7194 + }, + { + "epoch": 0.4921677269307066, + "grad_norm": 1.6996027238593379, + "learning_rate": 5.3696525408549695e-06, + "loss": 0.2912, + "step": 7195 + }, + { + "epoch": 0.4922361310623162, + "grad_norm": 1.9676213132303793, + "learning_rate": 5.368547808464292e-06, + "loss": 0.3875, + "step": 7196 + }, + { + "epoch": 0.4923045351939257, + "grad_norm": 1.6338848854325065, + "learning_rate": 5.367443057983509e-06, + "loss": 0.3179, + "step": 7197 + }, + { + "epoch": 0.4923729393255353, + "grad_norm": 3.0474364814835617, + "learning_rate": 5.366338289466845e-06, + "loss": 0.2889, + "step": 7198 + }, + { + "epoch": 0.4924413434571448, + "grad_norm": 1.680678110989379, + "learning_rate": 5.36523350296853e-06, + "loss": 0.2179, + "step": 7199 + }, + { + "epoch": 0.4925097475887544, + "grad_norm": 2.02298367687566, + "learning_rate": 5.3641286985427895e-06, + "loss": 0.4876, + "step": 7200 + }, + { + "epoch": 0.4925781517203639, + "grad_norm": 2.303812235779653, + "learning_rate": 5.363023876243855e-06, + "loss": 0.3131, + "step": 7201 + }, + { + "epoch": 0.4926465558519735, + "grad_norm": 1.9296098917439903, + "learning_rate": 5.361919036125956e-06, + "loss": 0.3063, + "step": 7202 + }, + { + "epoch": 0.492714959983583, + "grad_norm": 2.0238991729864693, + "learning_rate": 5.360814178243323e-06, + "loss": 0.3976, + "step": 7203 + }, + { + "epoch": 0.4927833641151926, + "grad_norm": 2.5983684280583264, + "learning_rate": 5.3597093026501865e-06, + "loss": 0.5817, + "step": 7204 + }, + { + "epoch": 0.4928517682468021, + "grad_norm": 1.8679750609121768, + "learning_rate": 5.3586044094007805e-06, + "loss": 0.2943, + "step": 7205 + }, + { + "epoch": 0.4929201723784117, + "grad_norm": 2.2224009221384273, + "learning_rate": 5.35749949854934e-06, + "loss": 0.2648, + "step": 7206 + }, + { + "epoch": 0.4929885765100212, + "grad_norm": 1.8998102515299857, + "learning_rate": 5.356394570150097e-06, + "loss": 0.3697, + "step": 7207 + }, + { + "epoch": 0.4930569806416308, + "grad_norm": 1.6392946935218398, + "learning_rate": 5.355289624257287e-06, + "loss": 0.1888, + "step": 7208 + }, + { + "epoch": 0.4931253847732403, + "grad_norm": 1.6378833114639058, + "learning_rate": 5.354184660925148e-06, + "loss": 0.1792, + "step": 7209 + }, + { + "epoch": 0.49319378890484983, + "grad_norm": 1.3575069181299348, + "learning_rate": 5.353079680207913e-06, + "loss": 0.2278, + "step": 7210 + }, + { + "epoch": 0.4932621930364594, + "grad_norm": 2.223215032909506, + "learning_rate": 5.351974682159824e-06, + "loss": 0.4468, + "step": 7211 + }, + { + "epoch": 0.49333059716806893, + "grad_norm": 1.7577781461961501, + "learning_rate": 5.3508696668351175e-06, + "loss": 0.3429, + "step": 7212 + }, + { + "epoch": 0.4933990012996785, + "grad_norm": 2.156566717549716, + "learning_rate": 5.349764634288035e-06, + "loss": 0.3454, + "step": 7213 + }, + { + "epoch": 0.49346740543128803, + "grad_norm": 2.104731935090088, + "learning_rate": 5.348659584572813e-06, + "loss": 0.4129, + "step": 7214 + }, + { + "epoch": 0.4935358095628976, + "grad_norm": 1.744431389198262, + "learning_rate": 5.347554517743695e-06, + "loss": 0.3545, + "step": 7215 + }, + { + "epoch": 0.49360421369450713, + "grad_norm": 1.6077484410966028, + "learning_rate": 5.346449433854924e-06, + "loss": 0.2215, + "step": 7216 + }, + { + "epoch": 0.4936726178261167, + "grad_norm": 1.8688062490335302, + "learning_rate": 5.345344332960743e-06, + "loss": 0.1522, + "step": 7217 + }, + { + "epoch": 0.49374102195772623, + "grad_norm": 2.034613135971931, + "learning_rate": 5.344239215115392e-06, + "loss": 0.3671, + "step": 7218 + }, + { + "epoch": 0.4938094260893358, + "grad_norm": 1.9562598476527675, + "learning_rate": 5.34313408037312e-06, + "loss": 0.437, + "step": 7219 + }, + { + "epoch": 0.49387783022094534, + "grad_norm": 1.856376161044914, + "learning_rate": 5.342028928788169e-06, + "loss": 0.4502, + "step": 7220 + }, + { + "epoch": 0.4939462343525549, + "grad_norm": 6.839790313529394, + "learning_rate": 5.340923760414787e-06, + "loss": 0.2488, + "step": 7221 + }, + { + "epoch": 0.49401463848416444, + "grad_norm": 1.7401943501879955, + "learning_rate": 5.339818575307221e-06, + "loss": 0.3533, + "step": 7222 + }, + { + "epoch": 0.494083042615774, + "grad_norm": 2.583904366036664, + "learning_rate": 5.338713373519717e-06, + "loss": 0.6934, + "step": 7223 + }, + { + "epoch": 0.49415144674738354, + "grad_norm": 1.8961268361980694, + "learning_rate": 5.337608155106526e-06, + "loss": 0.229, + "step": 7224 + }, + { + "epoch": 0.4942198508789931, + "grad_norm": 2.395353608546944, + "learning_rate": 5.336502920121896e-06, + "loss": 0.5485, + "step": 7225 + }, + { + "epoch": 0.49428825501060264, + "grad_norm": 1.5841536331526884, + "learning_rate": 5.335397668620078e-06, + "loss": 0.2331, + "step": 7226 + }, + { + "epoch": 0.49435665914221216, + "grad_norm": 2.022284687491033, + "learning_rate": 5.334292400655322e-06, + "loss": 0.3974, + "step": 7227 + }, + { + "epoch": 0.49442506327382174, + "grad_norm": 1.7936395250586084, + "learning_rate": 5.333187116281883e-06, + "loss": 0.4162, + "step": 7228 + }, + { + "epoch": 0.49449346740543126, + "grad_norm": 2.185001035816637, + "learning_rate": 5.33208181555401e-06, + "loss": 0.4887, + "step": 7229 + }, + { + "epoch": 0.49456187153704084, + "grad_norm": 1.8400143323572058, + "learning_rate": 5.330976498525957e-06, + "loss": 0.3832, + "step": 7230 + }, + { + "epoch": 0.49463027566865037, + "grad_norm": 2.4098099161407234, + "learning_rate": 5.3298711652519795e-06, + "loss": 0.4298, + "step": 7231 + }, + { + "epoch": 0.49469867980025994, + "grad_norm": 1.7314610359523614, + "learning_rate": 5.328765815786333e-06, + "loss": 0.3269, + "step": 7232 + }, + { + "epoch": 0.49476708393186947, + "grad_norm": 2.260769264273259, + "learning_rate": 5.327660450183273e-06, + "loss": 0.4812, + "step": 7233 + }, + { + "epoch": 0.49483548806347905, + "grad_norm": 2.142526325166478, + "learning_rate": 5.326555068497054e-06, + "loss": 0.5795, + "step": 7234 + }, + { + "epoch": 0.49490389219508857, + "grad_norm": 2.4009596520330074, + "learning_rate": 5.325449670781937e-06, + "loss": 0.7888, + "step": 7235 + }, + { + "epoch": 0.49497229632669815, + "grad_norm": 2.3664485357114025, + "learning_rate": 5.324344257092179e-06, + "loss": 0.383, + "step": 7236 + }, + { + "epoch": 0.49504070045830767, + "grad_norm": 2.0629923334025477, + "learning_rate": 5.323238827482037e-06, + "loss": 0.3814, + "step": 7237 + }, + { + "epoch": 0.49510910458991725, + "grad_norm": 1.524927989830782, + "learning_rate": 5.322133382005774e-06, + "loss": 0.1667, + "step": 7238 + }, + { + "epoch": 0.49517750872152677, + "grad_norm": 1.7269826481402473, + "learning_rate": 5.321027920717649e-06, + "loss": 0.2575, + "step": 7239 + }, + { + "epoch": 0.49524591285313635, + "grad_norm": 2.1484616156871854, + "learning_rate": 5.319922443671922e-06, + "loss": 0.3591, + "step": 7240 + }, + { + "epoch": 0.4953143169847459, + "grad_norm": 2.4802120406309895, + "learning_rate": 5.318816950922856e-06, + "loss": 0.7318, + "step": 7241 + }, + { + "epoch": 0.49538272111635545, + "grad_norm": 2.408731172188152, + "learning_rate": 5.317711442524716e-06, + "loss": 0.2684, + "step": 7242 + }, + { + "epoch": 0.495451125247965, + "grad_norm": 2.0751256702181426, + "learning_rate": 5.316605918531765e-06, + "loss": 0.4298, + "step": 7243 + }, + { + "epoch": 0.49551952937957455, + "grad_norm": 2.2421705132349383, + "learning_rate": 5.315500378998265e-06, + "loss": 0.2709, + "step": 7244 + }, + { + "epoch": 0.4955879335111841, + "grad_norm": 2.623564435467827, + "learning_rate": 5.314394823978484e-06, + "loss": 0.6092, + "step": 7245 + }, + { + "epoch": 0.4956563376427936, + "grad_norm": 2.384170957961671, + "learning_rate": 5.313289253526685e-06, + "loss": 0.4983, + "step": 7246 + }, + { + "epoch": 0.4957247417744032, + "grad_norm": 2.170534244933525, + "learning_rate": 5.3121836676971385e-06, + "loss": 0.5028, + "step": 7247 + }, + { + "epoch": 0.4957931459060127, + "grad_norm": 2.1161500141297473, + "learning_rate": 5.31107806654411e-06, + "loss": 0.5464, + "step": 7248 + }, + { + "epoch": 0.4958615500376223, + "grad_norm": 2.0559900528220925, + "learning_rate": 5.3099724501218675e-06, + "loss": 0.5214, + "step": 7249 + }, + { + "epoch": 0.4959299541692318, + "grad_norm": 2.080387131926814, + "learning_rate": 5.308866818484679e-06, + "loss": 0.4202, + "step": 7250 + }, + { + "epoch": 0.4959983583008414, + "grad_norm": 2.2084235095710065, + "learning_rate": 5.307761171686817e-06, + "loss": 0.5937, + "step": 7251 + }, + { + "epoch": 0.4960667624324509, + "grad_norm": 1.9608599544861547, + "learning_rate": 5.30665550978255e-06, + "loss": 0.4827, + "step": 7252 + }, + { + "epoch": 0.4961351665640605, + "grad_norm": 2.1842970525835033, + "learning_rate": 5.305549832826151e-06, + "loss": 0.5062, + "step": 7253 + }, + { + "epoch": 0.49620357069567, + "grad_norm": 1.7777041310918942, + "learning_rate": 5.304444140871889e-06, + "loss": 0.4101, + "step": 7254 + }, + { + "epoch": 0.4962719748272796, + "grad_norm": 2.3177876269603006, + "learning_rate": 5.303338433974039e-06, + "loss": 0.5232, + "step": 7255 + }, + { + "epoch": 0.4963403789588891, + "grad_norm": 1.9993326668224112, + "learning_rate": 5.302232712186874e-06, + "loss": 0.3966, + "step": 7256 + }, + { + "epoch": 0.4964087830904987, + "grad_norm": 1.9023234936146818, + "learning_rate": 5.301126975564668e-06, + "loss": 0.2335, + "step": 7257 + }, + { + "epoch": 0.4964771872221082, + "grad_norm": 1.5253501452009295, + "learning_rate": 5.300021224161695e-06, + "loss": 0.2043, + "step": 7258 + }, + { + "epoch": 0.4965455913537178, + "grad_norm": 2.4176083420951326, + "learning_rate": 5.298915458032233e-06, + "loss": 0.6205, + "step": 7259 + }, + { + "epoch": 0.4966139954853273, + "grad_norm": 2.030769467367701, + "learning_rate": 5.297809677230556e-06, + "loss": 0.4248, + "step": 7260 + }, + { + "epoch": 0.4966823996169369, + "grad_norm": 2.0718562794180513, + "learning_rate": 5.296703881810941e-06, + "loss": 0.351, + "step": 7261 + }, + { + "epoch": 0.4967508037485464, + "grad_norm": 1.5812699570169173, + "learning_rate": 5.295598071827669e-06, + "loss": 0.2115, + "step": 7262 + }, + { + "epoch": 0.496819207880156, + "grad_norm": 1.8478840532259784, + "learning_rate": 5.294492247335016e-06, + "loss": 0.3722, + "step": 7263 + }, + { + "epoch": 0.4968876120117655, + "grad_norm": 2.8600268116107714, + "learning_rate": 5.29338640838726e-06, + "loss": 0.4896, + "step": 7264 + }, + { + "epoch": 0.49695601614337503, + "grad_norm": 1.6983092164844673, + "learning_rate": 5.292280555038682e-06, + "loss": 0.2282, + "step": 7265 + }, + { + "epoch": 0.4970244202749846, + "grad_norm": 2.4179003494768723, + "learning_rate": 5.291174687343565e-06, + "loss": 0.782, + "step": 7266 + }, + { + "epoch": 0.49709282440659414, + "grad_norm": 2.1471014966689363, + "learning_rate": 5.290068805356186e-06, + "loss": 0.5078, + "step": 7267 + }, + { + "epoch": 0.4971612285382037, + "grad_norm": 1.9524400768963344, + "learning_rate": 5.288962909130831e-06, + "loss": 0.4474, + "step": 7268 + }, + { + "epoch": 0.49722963266981324, + "grad_norm": 2.0026996839452873, + "learning_rate": 5.287856998721781e-06, + "loss": 0.3926, + "step": 7269 + }, + { + "epoch": 0.4972980368014228, + "grad_norm": 2.2456521470870645, + "learning_rate": 5.286751074183318e-06, + "loss": 0.3082, + "step": 7270 + }, + { + "epoch": 0.49736644093303234, + "grad_norm": 1.7798037388044716, + "learning_rate": 5.285645135569728e-06, + "loss": 0.1769, + "step": 7271 + }, + { + "epoch": 0.4974348450646419, + "grad_norm": 1.9984655864433027, + "learning_rate": 5.284539182935296e-06, + "loss": 0.5179, + "step": 7272 + }, + { + "epoch": 0.49750324919625144, + "grad_norm": 1.759672867408089, + "learning_rate": 5.2834332163343064e-06, + "loss": 0.1039, + "step": 7273 + }, + { + "epoch": 0.497571653327861, + "grad_norm": 2.121429005888761, + "learning_rate": 5.282327235821047e-06, + "loss": 0.3695, + "step": 7274 + }, + { + "epoch": 0.49764005745947054, + "grad_norm": 2.3960756763265416, + "learning_rate": 5.281221241449802e-06, + "loss": 0.5223, + "step": 7275 + }, + { + "epoch": 0.4977084615910801, + "grad_norm": 1.911534391461265, + "learning_rate": 5.2801152332748615e-06, + "loss": 0.2863, + "step": 7276 + }, + { + "epoch": 0.49777686572268964, + "grad_norm": 1.2694986779362698, + "learning_rate": 5.279009211350512e-06, + "loss": 0.1758, + "step": 7277 + }, + { + "epoch": 0.4978452698542992, + "grad_norm": 2.0398698246259834, + "learning_rate": 5.277903175731044e-06, + "loss": 0.4525, + "step": 7278 + }, + { + "epoch": 0.49791367398590874, + "grad_norm": 1.9075980696272348, + "learning_rate": 5.2767971264707445e-06, + "loss": 0.4309, + "step": 7279 + }, + { + "epoch": 0.4979820781175183, + "grad_norm": 2.136864073168389, + "learning_rate": 5.275691063623906e-06, + "loss": 0.4873, + "step": 7280 + }, + { + "epoch": 0.49805048224912785, + "grad_norm": 1.8736413165555308, + "learning_rate": 5.27458498724482e-06, + "loss": 0.3785, + "step": 7281 + }, + { + "epoch": 0.49811888638073737, + "grad_norm": 2.0035390759168674, + "learning_rate": 5.273478897387776e-06, + "loss": 0.5271, + "step": 7282 + }, + { + "epoch": 0.49818729051234695, + "grad_norm": 2.047357367687387, + "learning_rate": 5.2723727941070666e-06, + "loss": 0.5599, + "step": 7283 + }, + { + "epoch": 0.49825569464395647, + "grad_norm": 3.0094243276322485, + "learning_rate": 5.271266677456985e-06, + "loss": 0.5287, + "step": 7284 + }, + { + "epoch": 0.49832409877556605, + "grad_norm": 2.219802942401013, + "learning_rate": 5.270160547491827e-06, + "loss": 0.7146, + "step": 7285 + }, + { + "epoch": 0.49839250290717557, + "grad_norm": 2.309825920170866, + "learning_rate": 5.269054404265884e-06, + "loss": 0.4371, + "step": 7286 + }, + { + "epoch": 0.49846090703878515, + "grad_norm": 1.7995343742697893, + "learning_rate": 5.26794824783345e-06, + "loss": 0.2599, + "step": 7287 + }, + { + "epoch": 0.4985293111703947, + "grad_norm": 2.0935163901372804, + "learning_rate": 5.266842078248824e-06, + "loss": 0.4826, + "step": 7288 + }, + { + "epoch": 0.49859771530200425, + "grad_norm": 1.9667771612124523, + "learning_rate": 5.265735895566299e-06, + "loss": 0.3293, + "step": 7289 + }, + { + "epoch": 0.4986661194336138, + "grad_norm": 2.117203715920372, + "learning_rate": 5.264629699840173e-06, + "loss": 0.491, + "step": 7290 + }, + { + "epoch": 0.49873452356522335, + "grad_norm": 2.494812697850871, + "learning_rate": 5.2635234911247445e-06, + "loss": 0.6557, + "step": 7291 + }, + { + "epoch": 0.4988029276968329, + "grad_norm": 2.5860132165985585, + "learning_rate": 5.26241726947431e-06, + "loss": 0.3673, + "step": 7292 + }, + { + "epoch": 0.49887133182844245, + "grad_norm": 1.5331003992934935, + "learning_rate": 5.2613110349431686e-06, + "loss": 0.1506, + "step": 7293 + }, + { + "epoch": 0.498939735960052, + "grad_norm": 1.735507844275263, + "learning_rate": 5.260204787585619e-06, + "loss": 0.1783, + "step": 7294 + }, + { + "epoch": 0.49900814009166156, + "grad_norm": 8.515127534947238, + "learning_rate": 5.2590985274559625e-06, + "loss": 0.4535, + "step": 7295 + }, + { + "epoch": 0.4990765442232711, + "grad_norm": 1.7371581403348053, + "learning_rate": 5.2579922546085e-06, + "loss": 0.19, + "step": 7296 + }, + { + "epoch": 0.49914494835488066, + "grad_norm": 2.1495935693541375, + "learning_rate": 5.256885969097531e-06, + "loss": 0.296, + "step": 7297 + }, + { + "epoch": 0.4992133524864902, + "grad_norm": 2.150543041180792, + "learning_rate": 5.255779670977358e-06, + "loss": 0.6287, + "step": 7298 + }, + { + "epoch": 0.49928175661809976, + "grad_norm": 1.7301245082202437, + "learning_rate": 5.254673360302284e-06, + "loss": 0.2866, + "step": 7299 + }, + { + "epoch": 0.4993501607497093, + "grad_norm": 3.4721777535875877, + "learning_rate": 5.25356703712661e-06, + "loss": 0.5807, + "step": 7300 + }, + { + "epoch": 0.4994185648813188, + "grad_norm": 2.581865458280375, + "learning_rate": 5.252460701504643e-06, + "loss": 0.6036, + "step": 7301 + }, + { + "epoch": 0.4994869690129284, + "grad_norm": 2.2835186364228, + "learning_rate": 5.251354353490686e-06, + "loss": 0.6048, + "step": 7302 + }, + { + "epoch": 0.4995553731445379, + "grad_norm": 2.4088491051423104, + "learning_rate": 5.250247993139044e-06, + "loss": 0.782, + "step": 7303 + }, + { + "epoch": 0.4996237772761475, + "grad_norm": 2.900563593886202, + "learning_rate": 5.24914162050402e-06, + "loss": 0.332, + "step": 7304 + }, + { + "epoch": 0.499692181407757, + "grad_norm": 1.4139179951507406, + "learning_rate": 5.248035235639924e-06, + "loss": 0.2627, + "step": 7305 + }, + { + "epoch": 0.4997605855393666, + "grad_norm": 4.006634094673303, + "learning_rate": 5.24692883860106e-06, + "loss": 0.4424, + "step": 7306 + }, + { + "epoch": 0.4998289896709761, + "grad_norm": 2.1226674278619235, + "learning_rate": 5.245822429441736e-06, + "loss": 0.3766, + "step": 7307 + }, + { + "epoch": 0.4998973938025857, + "grad_norm": 2.4022325629578964, + "learning_rate": 5.244716008216261e-06, + "loss": 0.4571, + "step": 7308 + }, + { + "epoch": 0.4999657979341952, + "grad_norm": 2.0500883403212615, + "learning_rate": 5.243609574978942e-06, + "loss": 0.4533, + "step": 7309 + }, + { + "epoch": 0.5000342020658047, + "grad_norm": 1.6453026550266503, + "learning_rate": 5.242503129784087e-06, + "loss": 0.0968, + "step": 7310 + }, + { + "epoch": 0.5001026061974143, + "grad_norm": 1.5640857612869397, + "learning_rate": 5.2413966726860096e-06, + "loss": 0.3368, + "step": 7311 + }, + { + "epoch": 0.5001710103290239, + "grad_norm": 2.3521365545438773, + "learning_rate": 5.240290203739018e-06, + "loss": 0.6358, + "step": 7312 + }, + { + "epoch": 0.5002394144606335, + "grad_norm": 1.8665260955093554, + "learning_rate": 5.239183722997422e-06, + "loss": 0.3128, + "step": 7313 + }, + { + "epoch": 0.5003078185922429, + "grad_norm": 1.866035712702775, + "learning_rate": 5.2380772305155335e-06, + "loss": 0.5602, + "step": 7314 + }, + { + "epoch": 0.5003762227238525, + "grad_norm": 2.2089647314140954, + "learning_rate": 5.236970726347665e-06, + "loss": 0.47, + "step": 7315 + }, + { + "epoch": 0.5004446268554621, + "grad_norm": 2.35613333719363, + "learning_rate": 5.235864210548131e-06, + "loss": 0.3703, + "step": 7316 + }, + { + "epoch": 0.5005130309870717, + "grad_norm": 1.9959560503559743, + "learning_rate": 5.23475768317124e-06, + "loss": 0.3878, + "step": 7317 + }, + { + "epoch": 0.5005814351186811, + "grad_norm": 1.9608286717220695, + "learning_rate": 5.23365114427131e-06, + "loss": 0.439, + "step": 7318 + }, + { + "epoch": 0.5006498392502907, + "grad_norm": 1.5973673328418938, + "learning_rate": 5.232544593902652e-06, + "loss": 0.2945, + "step": 7319 + }, + { + "epoch": 0.5007182433819003, + "grad_norm": 3.518922124348948, + "learning_rate": 5.231438032119583e-06, + "loss": 0.3294, + "step": 7320 + }, + { + "epoch": 0.5007866475135098, + "grad_norm": 2.21282764961348, + "learning_rate": 5.230331458976417e-06, + "loss": 0.2746, + "step": 7321 + }, + { + "epoch": 0.5008550516451193, + "grad_norm": 3.4398643078905558, + "learning_rate": 5.229224874527473e-06, + "loss": 0.5344, + "step": 7322 + }, + { + "epoch": 0.5009234557767289, + "grad_norm": 1.8403390675261193, + "learning_rate": 5.228118278827064e-06, + "loss": 0.5253, + "step": 7323 + }, + { + "epoch": 0.5009918599083385, + "grad_norm": 1.4098796097613866, + "learning_rate": 5.227011671929506e-06, + "loss": 0.289, + "step": 7324 + }, + { + "epoch": 0.501060264039948, + "grad_norm": 2.0629169591729988, + "learning_rate": 5.225905053889122e-06, + "loss": 0.4175, + "step": 7325 + }, + { + "epoch": 0.5011286681715575, + "grad_norm": 1.879364845951256, + "learning_rate": 5.224798424760227e-06, + "loss": 0.2832, + "step": 7326 + }, + { + "epoch": 0.5011970723031671, + "grad_norm": 1.8308219619356454, + "learning_rate": 5.223691784597138e-06, + "loss": 0.4137, + "step": 7327 + }, + { + "epoch": 0.5012654764347767, + "grad_norm": 2.905794425861049, + "learning_rate": 5.222585133454177e-06, + "loss": 0.5915, + "step": 7328 + }, + { + "epoch": 0.5013338805663862, + "grad_norm": 1.545248218989121, + "learning_rate": 5.221478471385663e-06, + "loss": 0.2993, + "step": 7329 + }, + { + "epoch": 0.5014022846979957, + "grad_norm": 1.7623801248109152, + "learning_rate": 5.220371798445915e-06, + "loss": 0.3774, + "step": 7330 + }, + { + "epoch": 0.5014706888296053, + "grad_norm": 2.0320076776447706, + "learning_rate": 5.219265114689256e-06, + "loss": 0.4338, + "step": 7331 + }, + { + "epoch": 0.5015390929612149, + "grad_norm": 2.5834290591829143, + "learning_rate": 5.218158420170008e-06, + "loss": 0.353, + "step": 7332 + }, + { + "epoch": 0.5016074970928244, + "grad_norm": 1.7927170454216619, + "learning_rate": 5.217051714942488e-06, + "loss": 0.382, + "step": 7333 + }, + { + "epoch": 0.501675901224434, + "grad_norm": 2.0019929537649133, + "learning_rate": 5.2159449990610226e-06, + "loss": 0.5772, + "step": 7334 + }, + { + "epoch": 0.5017443053560435, + "grad_norm": 2.2064986945190994, + "learning_rate": 5.214838272579935e-06, + "loss": 0.6952, + "step": 7335 + }, + { + "epoch": 0.5018127094876531, + "grad_norm": 2.005273486395895, + "learning_rate": 5.213731535553546e-06, + "loss": 0.482, + "step": 7336 + }, + { + "epoch": 0.5018811136192626, + "grad_norm": 1.6870109591227034, + "learning_rate": 5.212624788036182e-06, + "loss": 0.2459, + "step": 7337 + }, + { + "epoch": 0.5019495177508722, + "grad_norm": 2.1005936271778562, + "learning_rate": 5.211518030082168e-06, + "loss": 0.5655, + "step": 7338 + }, + { + "epoch": 0.5020179218824817, + "grad_norm": 2.42039520831607, + "learning_rate": 5.2104112617458254e-06, + "loss": 0.7596, + "step": 7339 + }, + { + "epoch": 0.5020863260140912, + "grad_norm": 2.080168429281517, + "learning_rate": 5.209304483081482e-06, + "loss": 0.3884, + "step": 7340 + }, + { + "epoch": 0.5021547301457008, + "grad_norm": 2.5406542908879177, + "learning_rate": 5.208197694143466e-06, + "loss": 0.2777, + "step": 7341 + }, + { + "epoch": 0.5022231342773104, + "grad_norm": 1.9653719652750892, + "learning_rate": 5.207090894986101e-06, + "loss": 0.5011, + "step": 7342 + }, + { + "epoch": 0.5022915384089199, + "grad_norm": 2.060765158133226, + "learning_rate": 5.205984085663713e-06, + "loss": 0.4388, + "step": 7343 + }, + { + "epoch": 0.5023599425405294, + "grad_norm": 1.9203190669171724, + "learning_rate": 5.204877266230634e-06, + "loss": 0.3551, + "step": 7344 + }, + { + "epoch": 0.502428346672139, + "grad_norm": 1.9591662283530071, + "learning_rate": 5.203770436741188e-06, + "loss": 0.2238, + "step": 7345 + }, + { + "epoch": 0.5024967508037486, + "grad_norm": 2.449082110668276, + "learning_rate": 5.202663597249704e-06, + "loss": 0.7462, + "step": 7346 + }, + { + "epoch": 0.5025651549353581, + "grad_norm": 2.1042781499720333, + "learning_rate": 5.201556747810514e-06, + "loss": 0.3005, + "step": 7347 + }, + { + "epoch": 0.5026335590669676, + "grad_norm": 1.7664587280495547, + "learning_rate": 5.200449888477945e-06, + "loss": 0.3284, + "step": 7348 + }, + { + "epoch": 0.5027019631985772, + "grad_norm": 2.181970982823888, + "learning_rate": 5.1993430193063276e-06, + "loss": 0.4904, + "step": 7349 + }, + { + "epoch": 0.5027703673301868, + "grad_norm": 2.1109418948033887, + "learning_rate": 5.198236140349991e-06, + "loss": 0.4676, + "step": 7350 + }, + { + "epoch": 0.5028387714617963, + "grad_norm": 2.2108025122098693, + "learning_rate": 5.197129251663268e-06, + "loss": 0.6979, + "step": 7351 + }, + { + "epoch": 0.5029071755934058, + "grad_norm": 1.486465728445533, + "learning_rate": 5.196022353300491e-06, + "loss": 0.2446, + "step": 7352 + }, + { + "epoch": 0.5029755797250154, + "grad_norm": 1.7047382553474735, + "learning_rate": 5.194915445315988e-06, + "loss": 0.3953, + "step": 7353 + }, + { + "epoch": 0.503043983856625, + "grad_norm": 1.9830364167370824, + "learning_rate": 5.193808527764095e-06, + "loss": 0.1578, + "step": 7354 + }, + { + "epoch": 0.5031123879882345, + "grad_norm": 2.554095593397565, + "learning_rate": 5.192701600699143e-06, + "loss": 0.4949, + "step": 7355 + }, + { + "epoch": 0.503180792119844, + "grad_norm": 1.9646701527499089, + "learning_rate": 5.191594664175466e-06, + "loss": 0.2884, + "step": 7356 + }, + { + "epoch": 0.5032491962514536, + "grad_norm": 3.3971295086814277, + "learning_rate": 5.190487718247398e-06, + "loss": 0.3789, + "step": 7357 + }, + { + "epoch": 0.5033176003830632, + "grad_norm": 2.4199461854909714, + "learning_rate": 5.189380762969273e-06, + "loss": 0.4588, + "step": 7358 + }, + { + "epoch": 0.5033860045146726, + "grad_norm": 1.6577377295283733, + "learning_rate": 5.188273798395425e-06, + "loss": 0.3278, + "step": 7359 + }, + { + "epoch": 0.5034544086462822, + "grad_norm": 2.2024403042448637, + "learning_rate": 5.18716682458019e-06, + "loss": 0.3023, + "step": 7360 + }, + { + "epoch": 0.5035228127778918, + "grad_norm": 1.9449346282596018, + "learning_rate": 5.186059841577904e-06, + "loss": 0.4182, + "step": 7361 + }, + { + "epoch": 0.5035912169095014, + "grad_norm": 2.108973238796021, + "learning_rate": 5.184952849442902e-06, + "loss": 0.3334, + "step": 7362 + }, + { + "epoch": 0.5036596210411108, + "grad_norm": 1.9211969575155567, + "learning_rate": 5.183845848229521e-06, + "loss": 0.3477, + "step": 7363 + }, + { + "epoch": 0.5037280251727204, + "grad_norm": 1.355612579542062, + "learning_rate": 5.182738837992097e-06, + "loss": 0.2363, + "step": 7364 + }, + { + "epoch": 0.50379642930433, + "grad_norm": 1.5313879262347354, + "learning_rate": 5.1816318187849705e-06, + "loss": 0.2541, + "step": 7365 + }, + { + "epoch": 0.5038648334359396, + "grad_norm": 2.042198794624833, + "learning_rate": 5.180524790662476e-06, + "loss": 0.3607, + "step": 7366 + }, + { + "epoch": 0.503933237567549, + "grad_norm": 1.9668291727575389, + "learning_rate": 5.179417753678954e-06, + "loss": 0.3837, + "step": 7367 + }, + { + "epoch": 0.5040016416991586, + "grad_norm": 2.8387004434584484, + "learning_rate": 5.178310707888742e-06, + "loss": 0.4464, + "step": 7368 + }, + { + "epoch": 0.5040700458307682, + "grad_norm": 1.795883128861713, + "learning_rate": 5.177203653346178e-06, + "loss": 0.3741, + "step": 7369 + }, + { + "epoch": 0.5041384499623778, + "grad_norm": 1.8541511635900152, + "learning_rate": 5.176096590105603e-06, + "loss": 0.4982, + "step": 7370 + }, + { + "epoch": 0.5042068540939872, + "grad_norm": 2.051643023395992, + "learning_rate": 5.174989518221358e-06, + "loss": 0.4489, + "step": 7371 + }, + { + "epoch": 0.5042752582255968, + "grad_norm": 1.845375003612115, + "learning_rate": 5.173882437747784e-06, + "loss": 0.3614, + "step": 7372 + }, + { + "epoch": 0.5043436623572064, + "grad_norm": 2.2538838906063345, + "learning_rate": 5.172775348739217e-06, + "loss": 0.5447, + "step": 7373 + }, + { + "epoch": 0.504412066488816, + "grad_norm": 1.7549750378659306, + "learning_rate": 5.171668251250003e-06, + "loss": 0.1681, + "step": 7374 + }, + { + "epoch": 0.5044804706204254, + "grad_norm": 2.1340139728792997, + "learning_rate": 5.170561145334483e-06, + "loss": 0.3322, + "step": 7375 + }, + { + "epoch": 0.504548874752035, + "grad_norm": 1.6563982672274535, + "learning_rate": 5.169454031046998e-06, + "loss": 0.241, + "step": 7376 + }, + { + "epoch": 0.5046172788836446, + "grad_norm": 1.8373802927819327, + "learning_rate": 5.16834690844189e-06, + "loss": 0.3476, + "step": 7377 + }, + { + "epoch": 0.5046856830152541, + "grad_norm": 1.777743120702137, + "learning_rate": 5.167239777573503e-06, + "loss": 0.2753, + "step": 7378 + }, + { + "epoch": 0.5047540871468637, + "grad_norm": 2.2678874196690857, + "learning_rate": 5.1661326384961805e-06, + "loss": 0.5696, + "step": 7379 + }, + { + "epoch": 0.5048224912784732, + "grad_norm": 1.881993942872758, + "learning_rate": 5.165025491264266e-06, + "loss": 0.3697, + "step": 7380 + }, + { + "epoch": 0.5048908954100828, + "grad_norm": 2.3607224286299457, + "learning_rate": 5.163918335932105e-06, + "loss": 0.4538, + "step": 7381 + }, + { + "epoch": 0.5049592995416923, + "grad_norm": 1.9898480310832094, + "learning_rate": 5.162811172554038e-06, + "loss": 0.4034, + "step": 7382 + }, + { + "epoch": 0.5050277036733019, + "grad_norm": 2.619105146083865, + "learning_rate": 5.161704001184412e-06, + "loss": 0.4278, + "step": 7383 + }, + { + "epoch": 0.5050961078049114, + "grad_norm": 1.9074101083842934, + "learning_rate": 5.160596821877575e-06, + "loss": 0.4466, + "step": 7384 + }, + { + "epoch": 0.505164511936521, + "grad_norm": 2.1393209299893634, + "learning_rate": 5.15948963468787e-06, + "loss": 0.4797, + "step": 7385 + }, + { + "epoch": 0.5052329160681305, + "grad_norm": 1.928321327743178, + "learning_rate": 5.158382439669642e-06, + "loss": 0.2997, + "step": 7386 + }, + { + "epoch": 0.5053013201997401, + "grad_norm": 2.6608420340884305, + "learning_rate": 5.157275236877242e-06, + "loss": 0.3994, + "step": 7387 + }, + { + "epoch": 0.5053697243313496, + "grad_norm": 2.2721775535603985, + "learning_rate": 5.156168026365011e-06, + "loss": 0.7691, + "step": 7388 + }, + { + "epoch": 0.5054381284629592, + "grad_norm": 1.7802610831111885, + "learning_rate": 5.1550608081873e-06, + "loss": 0.2973, + "step": 7389 + }, + { + "epoch": 0.5055065325945687, + "grad_norm": 1.6298038013248206, + "learning_rate": 5.153953582398457e-06, + "loss": 0.3499, + "step": 7390 + }, + { + "epoch": 0.5055749367261783, + "grad_norm": 2.286422869154558, + "learning_rate": 5.152846349052827e-06, + "loss": 0.4125, + "step": 7391 + }, + { + "epoch": 0.5056433408577878, + "grad_norm": 1.6780416653540395, + "learning_rate": 5.151739108204761e-06, + "loss": 0.3853, + "step": 7392 + }, + { + "epoch": 0.5057117449893973, + "grad_norm": 2.2761996554107715, + "learning_rate": 5.150631859908606e-06, + "loss": 0.696, + "step": 7393 + }, + { + "epoch": 0.5057801491210069, + "grad_norm": 1.6201529794962208, + "learning_rate": 5.149524604218713e-06, + "loss": 0.2786, + "step": 7394 + }, + { + "epoch": 0.5058485532526165, + "grad_norm": 1.792102521363328, + "learning_rate": 5.14841734118943e-06, + "loss": 0.3068, + "step": 7395 + }, + { + "epoch": 0.505916957384226, + "grad_norm": 2.5699821886046563, + "learning_rate": 5.147310070875106e-06, + "loss": 0.4638, + "step": 7396 + }, + { + "epoch": 0.5059853615158355, + "grad_norm": 1.7067575846396388, + "learning_rate": 5.146202793330095e-06, + "loss": 0.3314, + "step": 7397 + }, + { + "epoch": 0.5060537656474451, + "grad_norm": 1.386856104535955, + "learning_rate": 5.1450955086087426e-06, + "loss": 0.2616, + "step": 7398 + }, + { + "epoch": 0.5061221697790547, + "grad_norm": 1.4108006333876302, + "learning_rate": 5.143988216765402e-06, + "loss": 0.3458, + "step": 7399 + }, + { + "epoch": 0.5061905739106642, + "grad_norm": 1.6002288336206973, + "learning_rate": 5.1428809178544255e-06, + "loss": 0.3218, + "step": 7400 + }, + { + "epoch": 0.5062589780422737, + "grad_norm": 1.6959845270089824, + "learning_rate": 5.141773611930164e-06, + "loss": 0.3045, + "step": 7401 + }, + { + "epoch": 0.5063273821738833, + "grad_norm": 1.869786855342558, + "learning_rate": 5.140666299046968e-06, + "loss": 0.3269, + "step": 7402 + }, + { + "epoch": 0.5063957863054929, + "grad_norm": 1.7879104403130526, + "learning_rate": 5.139558979259191e-06, + "loss": 0.4425, + "step": 7403 + }, + { + "epoch": 0.5064641904371024, + "grad_norm": 1.6985015144672666, + "learning_rate": 5.1384516526211856e-06, + "loss": 0.253, + "step": 7404 + }, + { + "epoch": 0.5065325945687119, + "grad_norm": 1.8734737979722726, + "learning_rate": 5.1373443191873055e-06, + "loss": 0.3806, + "step": 7405 + }, + { + "epoch": 0.5066009987003215, + "grad_norm": 1.9883557251357944, + "learning_rate": 5.1362369790119004e-06, + "loss": 0.5119, + "step": 7406 + }, + { + "epoch": 0.5066694028319311, + "grad_norm": 2.2210882451824148, + "learning_rate": 5.1351296321493295e-06, + "loss": 0.5469, + "step": 7407 + }, + { + "epoch": 0.5067378069635406, + "grad_norm": 2.2449435597931124, + "learning_rate": 5.134022278653941e-06, + "loss": 0.5871, + "step": 7408 + }, + { + "epoch": 0.5068062110951501, + "grad_norm": 1.7763987869211955, + "learning_rate": 5.132914918580093e-06, + "loss": 0.1485, + "step": 7409 + }, + { + "epoch": 0.5068746152267597, + "grad_norm": 1.7041518326916614, + "learning_rate": 5.131807551982139e-06, + "loss": 0.2391, + "step": 7410 + }, + { + "epoch": 0.5069430193583693, + "grad_norm": 1.9612350466093118, + "learning_rate": 5.130700178914435e-06, + "loss": 0.4124, + "step": 7411 + }, + { + "epoch": 0.5070114234899787, + "grad_norm": 2.1357643255369307, + "learning_rate": 5.1295927994313335e-06, + "loss": 0.4402, + "step": 7412 + }, + { + "epoch": 0.5070798276215883, + "grad_norm": 1.9225655338628875, + "learning_rate": 5.128485413587191e-06, + "loss": 0.3133, + "step": 7413 + }, + { + "epoch": 0.5071482317531979, + "grad_norm": 1.8097776694569077, + "learning_rate": 5.127378021436365e-06, + "loss": 0.2862, + "step": 7414 + }, + { + "epoch": 0.5072166358848075, + "grad_norm": 2.0904014770121195, + "learning_rate": 5.1262706230332115e-06, + "loss": 0.4601, + "step": 7415 + }, + { + "epoch": 0.5072850400164169, + "grad_norm": 1.8552182721173962, + "learning_rate": 5.125163218432084e-06, + "loss": 0.381, + "step": 7416 + }, + { + "epoch": 0.5073534441480265, + "grad_norm": 1.769504104810256, + "learning_rate": 5.124055807687344e-06, + "loss": 0.2531, + "step": 7417 + }, + { + "epoch": 0.5074218482796361, + "grad_norm": 1.810282745559647, + "learning_rate": 5.122948390853344e-06, + "loss": 0.218, + "step": 7418 + }, + { + "epoch": 0.5074902524112457, + "grad_norm": 2.1616152549373546, + "learning_rate": 5.121840967984443e-06, + "loss": 0.352, + "step": 7419 + }, + { + "epoch": 0.5075586565428551, + "grad_norm": 1.9848036415740145, + "learning_rate": 5.120733539135e-06, + "loss": 0.4605, + "step": 7420 + }, + { + "epoch": 0.5076270606744647, + "grad_norm": 1.7797852818367368, + "learning_rate": 5.119626104359372e-06, + "loss": 0.4348, + "step": 7421 + }, + { + "epoch": 0.5076954648060743, + "grad_norm": 1.2677889914492193, + "learning_rate": 5.118518663711916e-06, + "loss": 0.1273, + "step": 7422 + }, + { + "epoch": 0.5077638689376839, + "grad_norm": 1.9751316100645417, + "learning_rate": 5.117411217246991e-06, + "loss": 0.3794, + "step": 7423 + }, + { + "epoch": 0.5078322730692933, + "grad_norm": 1.731976302346, + "learning_rate": 5.116303765018959e-06, + "loss": 0.3682, + "step": 7424 + }, + { + "epoch": 0.5079006772009029, + "grad_norm": 1.792130292928941, + "learning_rate": 5.115196307082175e-06, + "loss": 0.3766, + "step": 7425 + }, + { + "epoch": 0.5079690813325125, + "grad_norm": 2.1893568196414397, + "learning_rate": 5.114088843491001e-06, + "loss": 0.2099, + "step": 7426 + }, + { + "epoch": 0.5080374854641221, + "grad_norm": 2.672715010411033, + "learning_rate": 5.112981374299795e-06, + "loss": 0.6264, + "step": 7427 + }, + { + "epoch": 0.5081058895957316, + "grad_norm": 2.024445393196121, + "learning_rate": 5.1118738995629165e-06, + "loss": 0.4698, + "step": 7428 + }, + { + "epoch": 0.5081742937273411, + "grad_norm": 2.1915834732607777, + "learning_rate": 5.110766419334728e-06, + "loss": 0.7129, + "step": 7429 + }, + { + "epoch": 0.5082426978589507, + "grad_norm": 1.9261681823756962, + "learning_rate": 5.109658933669589e-06, + "loss": 0.4279, + "step": 7430 + }, + { + "epoch": 0.5083111019905602, + "grad_norm": 1.9937506508302971, + "learning_rate": 5.10855144262186e-06, + "loss": 0.2751, + "step": 7431 + }, + { + "epoch": 0.5083795061221698, + "grad_norm": 2.1696206372164384, + "learning_rate": 5.1074439462459e-06, + "loss": 0.4688, + "step": 7432 + }, + { + "epoch": 0.5084479102537793, + "grad_norm": 2.229457973596998, + "learning_rate": 5.106336444596074e-06, + "loss": 0.426, + "step": 7433 + }, + { + "epoch": 0.5085163143853889, + "grad_norm": 1.864548382262807, + "learning_rate": 5.105228937726742e-06, + "loss": 0.3732, + "step": 7434 + }, + { + "epoch": 0.5085847185169984, + "grad_norm": 1.8655933294371672, + "learning_rate": 5.104121425692264e-06, + "loss": 0.2927, + "step": 7435 + }, + { + "epoch": 0.508653122648608, + "grad_norm": 2.0024649138002144, + "learning_rate": 5.1030139085470056e-06, + "loss": 0.3263, + "step": 7436 + }, + { + "epoch": 0.5087215267802175, + "grad_norm": 2.008317386661497, + "learning_rate": 5.101906386345328e-06, + "loss": 0.3363, + "step": 7437 + }, + { + "epoch": 0.5087899309118271, + "grad_norm": 2.5717473127017887, + "learning_rate": 5.100798859141591e-06, + "loss": 0.5131, + "step": 7438 + }, + { + "epoch": 0.5088583350434366, + "grad_norm": 2.8120877917532248, + "learning_rate": 5.099691326990158e-06, + "loss": 0.4016, + "step": 7439 + }, + { + "epoch": 0.5089267391750462, + "grad_norm": 2.670412432930667, + "learning_rate": 5.098583789945395e-06, + "loss": 0.4704, + "step": 7440 + }, + { + "epoch": 0.5089951433066557, + "grad_norm": 1.9402139713696012, + "learning_rate": 5.097476248061663e-06, + "loss": 0.2192, + "step": 7441 + }, + { + "epoch": 0.5090635474382653, + "grad_norm": 2.43308179194264, + "learning_rate": 5.096368701393326e-06, + "loss": 0.691, + "step": 7442 + }, + { + "epoch": 0.5091319515698748, + "grad_norm": 1.963833190684408, + "learning_rate": 5.095261149994748e-06, + "loss": 0.536, + "step": 7443 + }, + { + "epoch": 0.5092003557014844, + "grad_norm": 2.093954891171895, + "learning_rate": 5.094153593920294e-06, + "loss": 0.4593, + "step": 7444 + }, + { + "epoch": 0.5092687598330939, + "grad_norm": 2.270637603659685, + "learning_rate": 5.093046033224325e-06, + "loss": 0.5465, + "step": 7445 + }, + { + "epoch": 0.5093371639647035, + "grad_norm": 1.9179531803781076, + "learning_rate": 5.091938467961208e-06, + "loss": 0.1872, + "step": 7446 + }, + { + "epoch": 0.509405568096313, + "grad_norm": 3.081345127641342, + "learning_rate": 5.0908308981853085e-06, + "loss": 0.3313, + "step": 7447 + }, + { + "epoch": 0.5094739722279226, + "grad_norm": 3.6424050011287035, + "learning_rate": 5.089723323950988e-06, + "loss": 0.2739, + "step": 7448 + }, + { + "epoch": 0.5095423763595321, + "grad_norm": 2.176959696290403, + "learning_rate": 5.0886157453126125e-06, + "loss": 0.4682, + "step": 7449 + }, + { + "epoch": 0.5096107804911416, + "grad_norm": 2.1565258706573314, + "learning_rate": 5.087508162324551e-06, + "loss": 0.5941, + "step": 7450 + }, + { + "epoch": 0.5096791846227512, + "grad_norm": 1.9954618510105435, + "learning_rate": 5.086400575041166e-06, + "loss": 0.5341, + "step": 7451 + }, + { + "epoch": 0.5097475887543608, + "grad_norm": 3.80503292199829, + "learning_rate": 5.085292983516822e-06, + "loss": 0.3865, + "step": 7452 + }, + { + "epoch": 0.5098159928859703, + "grad_norm": 2.3444767905917154, + "learning_rate": 5.084185387805887e-06, + "loss": 0.7351, + "step": 7453 + }, + { + "epoch": 0.5098843970175798, + "grad_norm": 1.698688340987342, + "learning_rate": 5.083077787962727e-06, + "loss": 0.2746, + "step": 7454 + }, + { + "epoch": 0.5099528011491894, + "grad_norm": 3.1132920048430996, + "learning_rate": 5.081970184041707e-06, + "loss": 0.1274, + "step": 7455 + }, + { + "epoch": 0.510021205280799, + "grad_norm": 11.218239499008872, + "learning_rate": 5.080862576097197e-06, + "loss": 0.2302, + "step": 7456 + }, + { + "epoch": 0.5100896094124086, + "grad_norm": 1.7654203047293617, + "learning_rate": 5.0797549641835586e-06, + "loss": 0.3354, + "step": 7457 + }, + { + "epoch": 0.510158013544018, + "grad_norm": 1.785125623644796, + "learning_rate": 5.078647348355162e-06, + "loss": 0.3955, + "step": 7458 + }, + { + "epoch": 0.5102264176756276, + "grad_norm": 2.0237989912824212, + "learning_rate": 5.077539728666374e-06, + "loss": 0.3773, + "step": 7459 + }, + { + "epoch": 0.5102948218072372, + "grad_norm": 2.1008050681026913, + "learning_rate": 5.076432105171562e-06, + "loss": 0.3465, + "step": 7460 + }, + { + "epoch": 0.5103632259388468, + "grad_norm": 1.5919127065253895, + "learning_rate": 5.075324477925094e-06, + "loss": 0.1674, + "step": 7461 + }, + { + "epoch": 0.5104316300704562, + "grad_norm": 1.4638384611989748, + "learning_rate": 5.074216846981336e-06, + "loss": 0.287, + "step": 7462 + }, + { + "epoch": 0.5105000342020658, + "grad_norm": 2.2473384029142074, + "learning_rate": 5.073109212394656e-06, + "loss": 0.4621, + "step": 7463 + }, + { + "epoch": 0.5105684383336754, + "grad_norm": 1.7222737170881761, + "learning_rate": 5.072001574219424e-06, + "loss": 0.3129, + "step": 7464 + }, + { + "epoch": 0.510636842465285, + "grad_norm": 2.2736880895145712, + "learning_rate": 5.070893932510006e-06, + "loss": 0.5235, + "step": 7465 + }, + { + "epoch": 0.5107052465968944, + "grad_norm": 1.409892223928017, + "learning_rate": 5.069786287320772e-06, + "loss": 0.1704, + "step": 7466 + }, + { + "epoch": 0.510773650728504, + "grad_norm": 2.0863807382325867, + "learning_rate": 5.068678638706092e-06, + "loss": 0.3026, + "step": 7467 + }, + { + "epoch": 0.5108420548601136, + "grad_norm": 2.2866203312991704, + "learning_rate": 5.0675709867203295e-06, + "loss": 0.5328, + "step": 7468 + }, + { + "epoch": 0.510910458991723, + "grad_norm": 1.734605395848518, + "learning_rate": 5.066463331417858e-06, + "loss": 0.3495, + "step": 7469 + }, + { + "epoch": 0.5109788631233326, + "grad_norm": 2.34292324107188, + "learning_rate": 5.065355672853045e-06, + "loss": 0.3602, + "step": 7470 + }, + { + "epoch": 0.5110472672549422, + "grad_norm": 2.0903690854348, + "learning_rate": 5.064248011080262e-06, + "loss": 0.3941, + "step": 7471 + }, + { + "epoch": 0.5111156713865518, + "grad_norm": 2.3089025003743924, + "learning_rate": 5.063140346153873e-06, + "loss": 0.2767, + "step": 7472 + }, + { + "epoch": 0.5111840755181613, + "grad_norm": 2.2738048610601407, + "learning_rate": 5.062032678128253e-06, + "loss": 0.3408, + "step": 7473 + }, + { + "epoch": 0.5112524796497708, + "grad_norm": 1.8556137108644093, + "learning_rate": 5.060925007057768e-06, + "loss": 0.3759, + "step": 7474 + }, + { + "epoch": 0.5113208837813804, + "grad_norm": 1.871031164822911, + "learning_rate": 5.059817332996792e-06, + "loss": 0.4478, + "step": 7475 + }, + { + "epoch": 0.51138928791299, + "grad_norm": 2.1658336814091417, + "learning_rate": 5.058709655999691e-06, + "loss": 0.4387, + "step": 7476 + }, + { + "epoch": 0.5114576920445995, + "grad_norm": 2.0924585544305487, + "learning_rate": 5.057601976120837e-06, + "loss": 0.6772, + "step": 7477 + }, + { + "epoch": 0.511526096176209, + "grad_norm": 1.705666535564741, + "learning_rate": 5.056494293414599e-06, + "loss": 0.4093, + "step": 7478 + }, + { + "epoch": 0.5115945003078186, + "grad_norm": 2.045947017411562, + "learning_rate": 5.055386607935347e-06, + "loss": 0.374, + "step": 7479 + }, + { + "epoch": 0.5116629044394282, + "grad_norm": 2.065773230585365, + "learning_rate": 5.054278919737456e-06, + "loss": 0.3662, + "step": 7480 + }, + { + "epoch": 0.5117313085710377, + "grad_norm": 2.5564415069512116, + "learning_rate": 5.053171228875291e-06, + "loss": 0.1445, + "step": 7481 + }, + { + "epoch": 0.5117997127026472, + "grad_norm": 1.7417043841440205, + "learning_rate": 5.0520635354032265e-06, + "loss": 0.2681, + "step": 7482 + }, + { + "epoch": 0.5118681168342568, + "grad_norm": 2.143229157010183, + "learning_rate": 5.050955839375633e-06, + "loss": 0.2974, + "step": 7483 + }, + { + "epoch": 0.5119365209658664, + "grad_norm": 2.257428166238283, + "learning_rate": 5.0498481408468805e-06, + "loss": 0.536, + "step": 7484 + }, + { + "epoch": 0.5120049250974759, + "grad_norm": 1.6917206557902522, + "learning_rate": 5.048740439871339e-06, + "loss": 0.2541, + "step": 7485 + }, + { + "epoch": 0.5120733292290854, + "grad_norm": 2.328601113221814, + "learning_rate": 5.047632736503383e-06, + "loss": 0.5607, + "step": 7486 + }, + { + "epoch": 0.512141733360695, + "grad_norm": 1.8783952291692374, + "learning_rate": 5.046525030797381e-06, + "loss": 0.2738, + "step": 7487 + }, + { + "epoch": 0.5122101374923045, + "grad_norm": 1.5247272118496256, + "learning_rate": 5.045417322807706e-06, + "loss": 0.2142, + "step": 7488 + }, + { + "epoch": 0.5122785416239141, + "grad_norm": 1.9116138833252225, + "learning_rate": 5.04430961258873e-06, + "loss": 0.2655, + "step": 7489 + }, + { + "epoch": 0.5123469457555236, + "grad_norm": 1.7576821088113757, + "learning_rate": 5.043201900194825e-06, + "loss": 0.4053, + "step": 7490 + }, + { + "epoch": 0.5124153498871332, + "grad_norm": 1.9642858262959981, + "learning_rate": 5.0420941856803605e-06, + "loss": 0.4363, + "step": 7491 + }, + { + "epoch": 0.5124837540187427, + "grad_norm": 1.7470190681460642, + "learning_rate": 5.0409864690997115e-06, + "loss": 0.289, + "step": 7492 + }, + { + "epoch": 0.5125521581503523, + "grad_norm": 2.4302388482535964, + "learning_rate": 5.039878750507247e-06, + "loss": 0.5498, + "step": 7493 + }, + { + "epoch": 0.5126205622819618, + "grad_norm": 1.2468374448632265, + "learning_rate": 5.0387710299573435e-06, + "loss": 0.1364, + "step": 7494 + }, + { + "epoch": 0.5126889664135714, + "grad_norm": 2.048176246266195, + "learning_rate": 5.037663307504369e-06, + "loss": 0.4221, + "step": 7495 + }, + { + "epoch": 0.5127573705451809, + "grad_norm": 2.0364266675225404, + "learning_rate": 5.036555583202698e-06, + "loss": 0.5164, + "step": 7496 + }, + { + "epoch": 0.5128257746767905, + "grad_norm": 2.176855257094273, + "learning_rate": 5.035447857106701e-06, + "loss": 0.4837, + "step": 7497 + }, + { + "epoch": 0.5128941788084, + "grad_norm": 1.6844858341872588, + "learning_rate": 5.034340129270753e-06, + "loss": 0.2848, + "step": 7498 + }, + { + "epoch": 0.5129625829400096, + "grad_norm": 2.0283119447597056, + "learning_rate": 5.033232399749226e-06, + "loss": 0.4153, + "step": 7499 + }, + { + "epoch": 0.5130309870716191, + "grad_norm": 2.196787586345141, + "learning_rate": 5.032124668596492e-06, + "loss": 0.2993, + "step": 7500 + }, + { + "epoch": 0.5130993912032287, + "grad_norm": 2.410705757252175, + "learning_rate": 5.031016935866924e-06, + "loss": 0.6001, + "step": 7501 + }, + { + "epoch": 0.5131677953348383, + "grad_norm": 2.2017193902821295, + "learning_rate": 5.029909201614894e-06, + "loss": 0.2966, + "step": 7502 + }, + { + "epoch": 0.5132361994664478, + "grad_norm": 2.9016737543242646, + "learning_rate": 5.028801465894778e-06, + "loss": 0.4383, + "step": 7503 + }, + { + "epoch": 0.5133046035980573, + "grad_norm": 2.416801966107982, + "learning_rate": 5.027693728760946e-06, + "loss": 0.6138, + "step": 7504 + }, + { + "epoch": 0.5133730077296669, + "grad_norm": 2.3689136191956046, + "learning_rate": 5.026585990267771e-06, + "loss": 0.4831, + "step": 7505 + }, + { + "epoch": 0.5134414118612765, + "grad_norm": 3.049993161678465, + "learning_rate": 5.02547825046963e-06, + "loss": 0.5221, + "step": 7506 + }, + { + "epoch": 0.5135098159928859, + "grad_norm": 2.536012789912493, + "learning_rate": 5.024370509420892e-06, + "loss": 0.4837, + "step": 7507 + }, + { + "epoch": 0.5135782201244955, + "grad_norm": 2.193598135557885, + "learning_rate": 5.02326276717593e-06, + "loss": 0.2849, + "step": 7508 + }, + { + "epoch": 0.5136466242561051, + "grad_norm": 2.160392637925405, + "learning_rate": 5.022155023789121e-06, + "loss": 0.4264, + "step": 7509 + }, + { + "epoch": 0.5137150283877147, + "grad_norm": 1.8004658044176702, + "learning_rate": 5.021047279314837e-06, + "loss": 0.3646, + "step": 7510 + }, + { + "epoch": 0.5137834325193241, + "grad_norm": 1.6473305406957153, + "learning_rate": 5.019939533807449e-06, + "loss": 0.3603, + "step": 7511 + }, + { + "epoch": 0.5138518366509337, + "grad_norm": 2.2236300515755394, + "learning_rate": 5.018831787321333e-06, + "loss": 0.6022, + "step": 7512 + }, + { + "epoch": 0.5139202407825433, + "grad_norm": 1.9376784500112765, + "learning_rate": 5.017724039910863e-06, + "loss": 0.2247, + "step": 7513 + }, + { + "epoch": 0.5139886449141529, + "grad_norm": 1.6645525403291825, + "learning_rate": 5.0166162916304105e-06, + "loss": 0.2647, + "step": 7514 + }, + { + "epoch": 0.5140570490457623, + "grad_norm": 2.4489843448007695, + "learning_rate": 5.015508542534349e-06, + "loss": 0.5525, + "step": 7515 + }, + { + "epoch": 0.5141254531773719, + "grad_norm": 1.772793993383433, + "learning_rate": 5.0144007926770564e-06, + "loss": 0.3502, + "step": 7516 + }, + { + "epoch": 0.5141938573089815, + "grad_norm": 2.3782560715916143, + "learning_rate": 5.013293042112901e-06, + "loss": 0.3072, + "step": 7517 + }, + { + "epoch": 0.5142622614405911, + "grad_norm": 2.311810616243578, + "learning_rate": 5.01218529089626e-06, + "loss": 0.4413, + "step": 7518 + }, + { + "epoch": 0.5143306655722005, + "grad_norm": 2.319851550160227, + "learning_rate": 5.011077539081506e-06, + "loss": 0.3897, + "step": 7519 + }, + { + "epoch": 0.5143990697038101, + "grad_norm": 2.656614345919205, + "learning_rate": 5.009969786723013e-06, + "loss": 0.679, + "step": 7520 + }, + { + "epoch": 0.5144674738354197, + "grad_norm": 1.8366629891833284, + "learning_rate": 5.008862033875154e-06, + "loss": 0.1317, + "step": 7521 + }, + { + "epoch": 0.5145358779670292, + "grad_norm": 1.9541306843796655, + "learning_rate": 5.007754280592303e-06, + "loss": 0.3746, + "step": 7522 + }, + { + "epoch": 0.5146042820986387, + "grad_norm": 1.6556975572462114, + "learning_rate": 5.006646526928835e-06, + "loss": 0.2522, + "step": 7523 + }, + { + "epoch": 0.5146726862302483, + "grad_norm": 1.9679330160024666, + "learning_rate": 5.005538772939125e-06, + "loss": 0.2215, + "step": 7524 + }, + { + "epoch": 0.5147410903618579, + "grad_norm": 2.2337142294072674, + "learning_rate": 5.004431018677544e-06, + "loss": 0.5192, + "step": 7525 + }, + { + "epoch": 0.5148094944934674, + "grad_norm": 2.4252192081104598, + "learning_rate": 5.003323264198467e-06, + "loss": 0.7267, + "step": 7526 + }, + { + "epoch": 0.5148778986250769, + "grad_norm": 1.5048740146585258, + "learning_rate": 5.002215509556267e-06, + "loss": 0.1488, + "step": 7527 + }, + { + "epoch": 0.5149463027566865, + "grad_norm": 1.6250724042754443, + "learning_rate": 5.001107754805321e-06, + "loss": 0.3072, + "step": 7528 + }, + { + "epoch": 0.5150147068882961, + "grad_norm": 2.143701556133836, + "learning_rate": 5e-06, + "loss": 0.6196, + "step": 7529 + }, + { + "epoch": 0.5150831110199056, + "grad_norm": 3.8493682131619957, + "learning_rate": 4.998892245194681e-06, + "loss": 0.3341, + "step": 7530 + }, + { + "epoch": 0.5151515151515151, + "grad_norm": 1.9460484211901836, + "learning_rate": 4.9977844904437336e-06, + "loss": 0.362, + "step": 7531 + }, + { + "epoch": 0.5152199192831247, + "grad_norm": 2.7126793391188437, + "learning_rate": 4.996676735801535e-06, + "loss": 0.2917, + "step": 7532 + }, + { + "epoch": 0.5152883234147343, + "grad_norm": 2.231984400165727, + "learning_rate": 4.99556898132246e-06, + "loss": 0.3124, + "step": 7533 + }, + { + "epoch": 0.5153567275463438, + "grad_norm": 2.9969854955023116, + "learning_rate": 4.9944612270608765e-06, + "loss": 0.3033, + "step": 7534 + }, + { + "epoch": 0.5154251316779533, + "grad_norm": 1.855238886197318, + "learning_rate": 4.993353473071165e-06, + "loss": 0.2702, + "step": 7535 + }, + { + "epoch": 0.5154935358095629, + "grad_norm": 2.0095788629798386, + "learning_rate": 4.992245719407698e-06, + "loss": 0.2976, + "step": 7536 + }, + { + "epoch": 0.5155619399411725, + "grad_norm": 3.3221962106888525, + "learning_rate": 4.991137966124849e-06, + "loss": 0.3579, + "step": 7537 + }, + { + "epoch": 0.515630344072782, + "grad_norm": 1.978413465296947, + "learning_rate": 4.990030213276989e-06, + "loss": 0.4211, + "step": 7538 + }, + { + "epoch": 0.5156987482043915, + "grad_norm": 1.8006705210769938, + "learning_rate": 4.988922460918496e-06, + "loss": 0.3501, + "step": 7539 + }, + { + "epoch": 0.5157671523360011, + "grad_norm": 2.5467785600528923, + "learning_rate": 4.987814709103742e-06, + "loss": 0.6847, + "step": 7540 + }, + { + "epoch": 0.5158355564676106, + "grad_norm": 6.3161896601847936, + "learning_rate": 4.9867069578871e-06, + "loss": 0.2047, + "step": 7541 + }, + { + "epoch": 0.5159039605992202, + "grad_norm": 1.6312453492781598, + "learning_rate": 4.985599207322946e-06, + "loss": 0.2407, + "step": 7542 + }, + { + "epoch": 0.5159723647308297, + "grad_norm": 1.8102903626752957, + "learning_rate": 4.984491457465652e-06, + "loss": 0.3231, + "step": 7543 + }, + { + "epoch": 0.5160407688624393, + "grad_norm": 1.746371458358401, + "learning_rate": 4.98338370836959e-06, + "loss": 0.1707, + "step": 7544 + }, + { + "epoch": 0.5161091729940488, + "grad_norm": 1.7141799597888496, + "learning_rate": 4.982275960089138e-06, + "loss": 0.2087, + "step": 7545 + }, + { + "epoch": 0.5161775771256584, + "grad_norm": 1.7826950042882825, + "learning_rate": 4.9811682126786675e-06, + "loss": 0.3782, + "step": 7546 + }, + { + "epoch": 0.516245981257268, + "grad_norm": 2.2275839152622257, + "learning_rate": 4.980060466192552e-06, + "loss": 0.679, + "step": 7547 + }, + { + "epoch": 0.5163143853888775, + "grad_norm": 1.7667354492939407, + "learning_rate": 4.978952720685165e-06, + "loss": 0.3241, + "step": 7548 + }, + { + "epoch": 0.516382789520487, + "grad_norm": 2.056726187640598, + "learning_rate": 4.9778449762108795e-06, + "loss": 0.2738, + "step": 7549 + }, + { + "epoch": 0.5164511936520966, + "grad_norm": 1.9961141447156339, + "learning_rate": 4.976737232824071e-06, + "loss": 0.2581, + "step": 7550 + }, + { + "epoch": 0.5165195977837062, + "grad_norm": 2.1854134485665315, + "learning_rate": 4.97562949057911e-06, + "loss": 0.687, + "step": 7551 + }, + { + "epoch": 0.5165880019153157, + "grad_norm": 2.0343288050151687, + "learning_rate": 4.974521749530373e-06, + "loss": 0.4959, + "step": 7552 + }, + { + "epoch": 0.5166564060469252, + "grad_norm": 1.7361210805818625, + "learning_rate": 4.97341400973223e-06, + "loss": 0.4074, + "step": 7553 + }, + { + "epoch": 0.5167248101785348, + "grad_norm": 2.0373446144928944, + "learning_rate": 4.972306271239055e-06, + "loss": 0.408, + "step": 7554 + }, + { + "epoch": 0.5167932143101444, + "grad_norm": 1.9250650254017478, + "learning_rate": 4.971198534105222e-06, + "loss": 0.4365, + "step": 7555 + }, + { + "epoch": 0.5168616184417539, + "grad_norm": 2.379967664130315, + "learning_rate": 4.970090798385106e-06, + "loss": 0.4049, + "step": 7556 + }, + { + "epoch": 0.5169300225733634, + "grad_norm": 2.08369527454288, + "learning_rate": 4.968983064133078e-06, + "loss": 0.727, + "step": 7557 + }, + { + "epoch": 0.516998426704973, + "grad_norm": 2.474646407695567, + "learning_rate": 4.9678753314035094e-06, + "loss": 0.2734, + "step": 7558 + }, + { + "epoch": 0.5170668308365826, + "grad_norm": 1.844706957800244, + "learning_rate": 4.966767600250776e-06, + "loss": 0.3451, + "step": 7559 + }, + { + "epoch": 0.517135234968192, + "grad_norm": 1.7939016537241537, + "learning_rate": 4.965659870729249e-06, + "loss": 0.2572, + "step": 7560 + }, + { + "epoch": 0.5172036390998016, + "grad_norm": 1.8714876457158705, + "learning_rate": 4.9645521428933005e-06, + "loss": 0.2418, + "step": 7561 + }, + { + "epoch": 0.5172720432314112, + "grad_norm": 1.92339807341365, + "learning_rate": 4.963444416797305e-06, + "loss": 0.3508, + "step": 7562 + }, + { + "epoch": 0.5173404473630208, + "grad_norm": 2.9723849078741233, + "learning_rate": 4.962336692495634e-06, + "loss": 0.463, + "step": 7563 + }, + { + "epoch": 0.5174088514946302, + "grad_norm": 1.764645368006559, + "learning_rate": 4.961228970042658e-06, + "loss": 0.3892, + "step": 7564 + }, + { + "epoch": 0.5174772556262398, + "grad_norm": 1.8327156813692733, + "learning_rate": 4.960121249492753e-06, + "loss": 0.434, + "step": 7565 + }, + { + "epoch": 0.5175456597578494, + "grad_norm": 1.6668140515089904, + "learning_rate": 4.959013530900289e-06, + "loss": 0.2679, + "step": 7566 + }, + { + "epoch": 0.517614063889459, + "grad_norm": 2.6319276924629063, + "learning_rate": 4.95790581431964e-06, + "loss": 0.3788, + "step": 7567 + }, + { + "epoch": 0.5176824680210684, + "grad_norm": 1.9895784159960739, + "learning_rate": 4.9567980998051765e-06, + "loss": 0.5385, + "step": 7568 + }, + { + "epoch": 0.517750872152678, + "grad_norm": 2.3692296035800773, + "learning_rate": 4.955690387411271e-06, + "loss": 0.2645, + "step": 7569 + }, + { + "epoch": 0.5178192762842876, + "grad_norm": 1.6157808982148585, + "learning_rate": 4.954582677192296e-06, + "loss": 0.2433, + "step": 7570 + }, + { + "epoch": 0.5178876804158972, + "grad_norm": 2.20881405957152, + "learning_rate": 4.95347496920262e-06, + "loss": 0.4217, + "step": 7571 + }, + { + "epoch": 0.5179560845475066, + "grad_norm": 2.4272621168069315, + "learning_rate": 4.952367263496619e-06, + "loss": 0.508, + "step": 7572 + }, + { + "epoch": 0.5180244886791162, + "grad_norm": 2.1034294365782302, + "learning_rate": 4.951259560128662e-06, + "loss": 0.4625, + "step": 7573 + }, + { + "epoch": 0.5180928928107258, + "grad_norm": 2.3032683771709324, + "learning_rate": 4.950151859153121e-06, + "loss": 0.503, + "step": 7574 + }, + { + "epoch": 0.5181612969423354, + "grad_norm": 1.9849007690943543, + "learning_rate": 4.949044160624368e-06, + "loss": 0.3414, + "step": 7575 + }, + { + "epoch": 0.5182297010739448, + "grad_norm": 1.8076147927505894, + "learning_rate": 4.947936464596774e-06, + "loss": 0.2647, + "step": 7576 + }, + { + "epoch": 0.5182981052055544, + "grad_norm": 2.0838363902382464, + "learning_rate": 4.94682877112471e-06, + "loss": 0.4376, + "step": 7577 + }, + { + "epoch": 0.518366509337164, + "grad_norm": 1.731099790811694, + "learning_rate": 4.945721080262546e-06, + "loss": 0.2426, + "step": 7578 + }, + { + "epoch": 0.5184349134687735, + "grad_norm": 2.35537606305262, + "learning_rate": 4.9446133920646535e-06, + "loss": 0.7079, + "step": 7579 + }, + { + "epoch": 0.518503317600383, + "grad_norm": 1.857521672180171, + "learning_rate": 4.943505706585403e-06, + "loss": 0.297, + "step": 7580 + }, + { + "epoch": 0.5185717217319926, + "grad_norm": 2.289643998220257, + "learning_rate": 4.942398023879166e-06, + "loss": 0.6424, + "step": 7581 + }, + { + "epoch": 0.5186401258636022, + "grad_norm": 1.933048533627131, + "learning_rate": 4.941290344000312e-06, + "loss": 0.3924, + "step": 7582 + }, + { + "epoch": 0.5187085299952117, + "grad_norm": 2.5255226608183743, + "learning_rate": 4.940182667003211e-06, + "loss": 0.5373, + "step": 7583 + }, + { + "epoch": 0.5187769341268212, + "grad_norm": 1.9050128001811801, + "learning_rate": 4.9390749929422324e-06, + "loss": 0.3422, + "step": 7584 + }, + { + "epoch": 0.5188453382584308, + "grad_norm": 1.2036243120968597, + "learning_rate": 4.937967321871748e-06, + "loss": 0.1416, + "step": 7585 + }, + { + "epoch": 0.5189137423900404, + "grad_norm": 1.7501139816562359, + "learning_rate": 4.9368596538461285e-06, + "loss": 0.4783, + "step": 7586 + }, + { + "epoch": 0.5189821465216499, + "grad_norm": 1.799603174098307, + "learning_rate": 4.93575198891974e-06, + "loss": 0.487, + "step": 7587 + }, + { + "epoch": 0.5190505506532594, + "grad_norm": 3.174634588044527, + "learning_rate": 4.934644327146956e-06, + "loss": 0.3534, + "step": 7588 + }, + { + "epoch": 0.519118954784869, + "grad_norm": 1.5601333504735984, + "learning_rate": 4.933536668582144e-06, + "loss": 0.1993, + "step": 7589 + }, + { + "epoch": 0.5191873589164786, + "grad_norm": 2.190475840939038, + "learning_rate": 4.932429013279671e-06, + "loss": 0.4483, + "step": 7590 + }, + { + "epoch": 0.5192557630480881, + "grad_norm": 1.5505685588354339, + "learning_rate": 4.931321361293911e-06, + "loss": 0.2913, + "step": 7591 + }, + { + "epoch": 0.5193241671796976, + "grad_norm": 2.1924807180450654, + "learning_rate": 4.93021371267923e-06, + "loss": 0.6026, + "step": 7592 + }, + { + "epoch": 0.5193925713113072, + "grad_norm": 1.5673973867148787, + "learning_rate": 4.929106067489995e-06, + "loss": 0.2301, + "step": 7593 + }, + { + "epoch": 0.5194609754429168, + "grad_norm": 2.128158837104032, + "learning_rate": 4.927998425780576e-06, + "loss": 0.1919, + "step": 7594 + }, + { + "epoch": 0.5195293795745263, + "grad_norm": 2.0436316849645126, + "learning_rate": 4.926890787605345e-06, + "loss": 0.4321, + "step": 7595 + }, + { + "epoch": 0.5195977837061359, + "grad_norm": 1.6659430413126266, + "learning_rate": 4.925783153018666e-06, + "loss": 0.3637, + "step": 7596 + }, + { + "epoch": 0.5196661878377454, + "grad_norm": 1.5205194108994733, + "learning_rate": 4.924675522074907e-06, + "loss": 0.2873, + "step": 7597 + }, + { + "epoch": 0.5197345919693549, + "grad_norm": 1.9326972578597006, + "learning_rate": 4.923567894828438e-06, + "loss": 0.472, + "step": 7598 + }, + { + "epoch": 0.5198029961009645, + "grad_norm": 2.278196276686523, + "learning_rate": 4.922460271333627e-06, + "loss": 0.5901, + "step": 7599 + }, + { + "epoch": 0.519871400232574, + "grad_norm": 2.3872996152549124, + "learning_rate": 4.921352651644839e-06, + "loss": 0.3573, + "step": 7600 + }, + { + "epoch": 0.5199398043641836, + "grad_norm": 1.5424407973203818, + "learning_rate": 4.920245035816443e-06, + "loss": 0.2328, + "step": 7601 + }, + { + "epoch": 0.5200082084957931, + "grad_norm": 1.9620177838263702, + "learning_rate": 4.919137423902807e-06, + "loss": 0.4237, + "step": 7602 + }, + { + "epoch": 0.5200766126274027, + "grad_norm": 1.9843661841249252, + "learning_rate": 4.9180298159582946e-06, + "loss": 0.386, + "step": 7603 + }, + { + "epoch": 0.5201450167590123, + "grad_norm": 2.249335137788056, + "learning_rate": 4.916922212037274e-06, + "loss": 0.6977, + "step": 7604 + }, + { + "epoch": 0.5202134208906218, + "grad_norm": 2.297643088673055, + "learning_rate": 4.915814612194114e-06, + "loss": 0.3374, + "step": 7605 + }, + { + "epoch": 0.5202818250222313, + "grad_norm": 1.9701018990023387, + "learning_rate": 4.9147070164831795e-06, + "loss": 0.3675, + "step": 7606 + }, + { + "epoch": 0.5203502291538409, + "grad_norm": 2.437855358385186, + "learning_rate": 4.913599424958836e-06, + "loss": 0.3917, + "step": 7607 + }, + { + "epoch": 0.5204186332854505, + "grad_norm": 2.07713465289235, + "learning_rate": 4.91249183767545e-06, + "loss": 0.3258, + "step": 7608 + }, + { + "epoch": 0.52048703741706, + "grad_norm": 1.8416484558562978, + "learning_rate": 4.911384254687388e-06, + "loss": 0.435, + "step": 7609 + }, + { + "epoch": 0.5205554415486695, + "grad_norm": 1.7246089257037742, + "learning_rate": 4.910276676049014e-06, + "loss": 0.3153, + "step": 7610 + }, + { + "epoch": 0.5206238456802791, + "grad_norm": 2.020461364477712, + "learning_rate": 4.909169101814695e-06, + "loss": 0.4324, + "step": 7611 + }, + { + "epoch": 0.5206922498118887, + "grad_norm": 2.715052995919743, + "learning_rate": 4.908061532038794e-06, + "loss": 0.4635, + "step": 7612 + }, + { + "epoch": 0.5207606539434982, + "grad_norm": 1.9203442160437016, + "learning_rate": 4.906953966775677e-06, + "loss": 0.3274, + "step": 7613 + }, + { + "epoch": 0.5208290580751077, + "grad_norm": 1.5486471736692868, + "learning_rate": 4.905846406079707e-06, + "loss": 0.2622, + "step": 7614 + }, + { + "epoch": 0.5208974622067173, + "grad_norm": 1.8785621740420415, + "learning_rate": 4.904738850005252e-06, + "loss": 0.3321, + "step": 7615 + }, + { + "epoch": 0.5209658663383269, + "grad_norm": 2.5319455274403007, + "learning_rate": 4.903631298606674e-06, + "loss": 0.1964, + "step": 7616 + }, + { + "epoch": 0.5210342704699363, + "grad_norm": 1.97120297220431, + "learning_rate": 4.9025237519383375e-06, + "loss": 0.3069, + "step": 7617 + }, + { + "epoch": 0.5211026746015459, + "grad_norm": 1.852277191266299, + "learning_rate": 4.901416210054606e-06, + "loss": 0.2169, + "step": 7618 + }, + { + "epoch": 0.5211710787331555, + "grad_norm": 2.320998979737544, + "learning_rate": 4.900308673009843e-06, + "loss": 0.5729, + "step": 7619 + }, + { + "epoch": 0.5212394828647651, + "grad_norm": 2.1156532457570703, + "learning_rate": 4.899201140858412e-06, + "loss": 0.1917, + "step": 7620 + }, + { + "epoch": 0.5213078869963745, + "grad_norm": 2.267582734831889, + "learning_rate": 4.898093613654675e-06, + "loss": 0.6513, + "step": 7621 + }, + { + "epoch": 0.5213762911279841, + "grad_norm": 2.2038333410179494, + "learning_rate": 4.896986091452996e-06, + "loss": 0.3462, + "step": 7622 + }, + { + "epoch": 0.5214446952595937, + "grad_norm": 1.6323450094594083, + "learning_rate": 4.895878574307737e-06, + "loss": 0.2455, + "step": 7623 + }, + { + "epoch": 0.5215130993912033, + "grad_norm": 2.091770677731951, + "learning_rate": 4.894771062273258e-06, + "loss": 0.2382, + "step": 7624 + }, + { + "epoch": 0.5215815035228127, + "grad_norm": 2.3874379774096295, + "learning_rate": 4.893663555403927e-06, + "loss": 0.5696, + "step": 7625 + }, + { + "epoch": 0.5216499076544223, + "grad_norm": 2.3639534301709206, + "learning_rate": 4.892556053754101e-06, + "loss": 0.3642, + "step": 7626 + }, + { + "epoch": 0.5217183117860319, + "grad_norm": 2.2108039074224113, + "learning_rate": 4.891448557378142e-06, + "loss": 0.2976, + "step": 7627 + }, + { + "epoch": 0.5217867159176415, + "grad_norm": 1.6145881181051256, + "learning_rate": 4.890341066330413e-06, + "loss": 0.3143, + "step": 7628 + }, + { + "epoch": 0.5218551200492509, + "grad_norm": 2.432071374941451, + "learning_rate": 4.889233580665274e-06, + "loss": 0.5741, + "step": 7629 + }, + { + "epoch": 0.5219235241808605, + "grad_norm": 2.044775643353119, + "learning_rate": 4.888126100437085e-06, + "loss": 0.1929, + "step": 7630 + }, + { + "epoch": 0.5219919283124701, + "grad_norm": 1.5434221012129719, + "learning_rate": 4.8870186257002075e-06, + "loss": 0.1942, + "step": 7631 + }, + { + "epoch": 0.5220603324440797, + "grad_norm": 1.8230461338026176, + "learning_rate": 4.885911156509002e-06, + "loss": 0.3326, + "step": 7632 + }, + { + "epoch": 0.5221287365756891, + "grad_norm": 2.0113622256892683, + "learning_rate": 4.884803692917826e-06, + "loss": 0.3417, + "step": 7633 + }, + { + "epoch": 0.5221971407072987, + "grad_norm": 2.0944626506312902, + "learning_rate": 4.883696234981042e-06, + "loss": 0.4601, + "step": 7634 + }, + { + "epoch": 0.5222655448389083, + "grad_norm": 2.266669742819391, + "learning_rate": 4.882588782753009e-06, + "loss": 0.4288, + "step": 7635 + }, + { + "epoch": 0.5223339489705178, + "grad_norm": 1.8059208987971898, + "learning_rate": 4.881481336288086e-06, + "loss": 0.298, + "step": 7636 + }, + { + "epoch": 0.5224023531021273, + "grad_norm": 2.0565428533979278, + "learning_rate": 4.88037389564063e-06, + "loss": 0.3553, + "step": 7637 + }, + { + "epoch": 0.5224707572337369, + "grad_norm": 1.7749563762515772, + "learning_rate": 4.879266460865002e-06, + "loss": 0.358, + "step": 7638 + }, + { + "epoch": 0.5225391613653465, + "grad_norm": 2.593719689065702, + "learning_rate": 4.878159032015559e-06, + "loss": 0.6825, + "step": 7639 + }, + { + "epoch": 0.522607565496956, + "grad_norm": 1.862920618245743, + "learning_rate": 4.877051609146658e-06, + "loss": 0.4326, + "step": 7640 + }, + { + "epoch": 0.5226759696285656, + "grad_norm": 1.9591915785495468, + "learning_rate": 4.875944192312659e-06, + "loss": 0.5898, + "step": 7641 + }, + { + "epoch": 0.5227443737601751, + "grad_norm": 2.6275590502693915, + "learning_rate": 4.8748367815679185e-06, + "loss": 0.5241, + "step": 7642 + }, + { + "epoch": 0.5228127778917847, + "grad_norm": 0.8759161312730372, + "learning_rate": 4.87372937696679e-06, + "loss": 0.0755, + "step": 7643 + }, + { + "epoch": 0.5228811820233942, + "grad_norm": 4.7182905592417415, + "learning_rate": 4.872621978563635e-06, + "loss": 0.2959, + "step": 7644 + }, + { + "epoch": 0.5229495861550038, + "grad_norm": 1.541070557595474, + "learning_rate": 4.871514586412809e-06, + "loss": 0.2978, + "step": 7645 + }, + { + "epoch": 0.5230179902866133, + "grad_norm": 1.4010719971995809, + "learning_rate": 4.870407200568669e-06, + "loss": 0.2671, + "step": 7646 + }, + { + "epoch": 0.5230863944182229, + "grad_norm": 1.873047234989645, + "learning_rate": 4.869299821085567e-06, + "loss": 0.3903, + "step": 7647 + }, + { + "epoch": 0.5231547985498324, + "grad_norm": 1.8642866593012948, + "learning_rate": 4.868192448017862e-06, + "loss": 0.5243, + "step": 7648 + }, + { + "epoch": 0.523223202681442, + "grad_norm": 1.7599484838727313, + "learning_rate": 4.867085081419909e-06, + "loss": 0.3365, + "step": 7649 + }, + { + "epoch": 0.5232916068130515, + "grad_norm": 2.176030348484619, + "learning_rate": 4.86597772134606e-06, + "loss": 0.4957, + "step": 7650 + }, + { + "epoch": 0.523360010944661, + "grad_norm": 2.189589908761176, + "learning_rate": 4.864870367850674e-06, + "loss": 0.4925, + "step": 7651 + }, + { + "epoch": 0.5234284150762706, + "grad_norm": 2.1519647115664493, + "learning_rate": 4.863763020988101e-06, + "loss": 0.3393, + "step": 7652 + }, + { + "epoch": 0.5234968192078802, + "grad_norm": 1.6050199800321712, + "learning_rate": 4.862655680812696e-06, + "loss": 0.2466, + "step": 7653 + }, + { + "epoch": 0.5235652233394897, + "grad_norm": 7.4567573061422, + "learning_rate": 4.861548347378814e-06, + "loss": 0.3585, + "step": 7654 + }, + { + "epoch": 0.5236336274710992, + "grad_norm": 2.2591960818992844, + "learning_rate": 4.860441020740809e-06, + "loss": 0.585, + "step": 7655 + }, + { + "epoch": 0.5237020316027088, + "grad_norm": 1.75102046394522, + "learning_rate": 4.859333700953032e-06, + "loss": 0.2412, + "step": 7656 + }, + { + "epoch": 0.5237704357343184, + "grad_norm": 2.7547569530124782, + "learning_rate": 4.858226388069837e-06, + "loss": 0.4972, + "step": 7657 + }, + { + "epoch": 0.5238388398659279, + "grad_norm": 2.197572827374658, + "learning_rate": 4.857119082145575e-06, + "loss": 0.5845, + "step": 7658 + }, + { + "epoch": 0.5239072439975374, + "grad_norm": 1.6343786494868946, + "learning_rate": 4.856011783234599e-06, + "loss": 0.3729, + "step": 7659 + }, + { + "epoch": 0.523975648129147, + "grad_norm": 2.5081760582668737, + "learning_rate": 4.854904491391258e-06, + "loss": 0.8181, + "step": 7660 + }, + { + "epoch": 0.5240440522607566, + "grad_norm": 1.9260138164244545, + "learning_rate": 4.853797206669908e-06, + "loss": 0.4631, + "step": 7661 + }, + { + "epoch": 0.5241124563923661, + "grad_norm": 1.9696048477687935, + "learning_rate": 4.852689929124896e-06, + "loss": 0.5918, + "step": 7662 + }, + { + "epoch": 0.5241808605239756, + "grad_norm": 2.5450709219532994, + "learning_rate": 4.8515826588105706e-06, + "loss": 0.4743, + "step": 7663 + }, + { + "epoch": 0.5242492646555852, + "grad_norm": 2.1722255122064174, + "learning_rate": 4.850475395781287e-06, + "loss": 0.2853, + "step": 7664 + }, + { + "epoch": 0.5243176687871948, + "grad_norm": 1.9381617369702135, + "learning_rate": 4.849368140091395e-06, + "loss": 0.5479, + "step": 7665 + }, + { + "epoch": 0.5243860729188043, + "grad_norm": 2.2351969260894946, + "learning_rate": 4.848260891795241e-06, + "loss": 0.5231, + "step": 7666 + }, + { + "epoch": 0.5244544770504138, + "grad_norm": 2.0853267661415145, + "learning_rate": 4.8471536509471736e-06, + "loss": 0.3316, + "step": 7667 + }, + { + "epoch": 0.5245228811820234, + "grad_norm": 1.9400285134297033, + "learning_rate": 4.846046417601545e-06, + "loss": 0.5161, + "step": 7668 + }, + { + "epoch": 0.524591285313633, + "grad_norm": 1.4427995718900828, + "learning_rate": 4.844939191812702e-06, + "loss": 0.284, + "step": 7669 + }, + { + "epoch": 0.5246596894452424, + "grad_norm": 1.7625114535236885, + "learning_rate": 4.84383197363499e-06, + "loss": 0.3202, + "step": 7670 + }, + { + "epoch": 0.524728093576852, + "grad_norm": 2.1881004761030094, + "learning_rate": 4.842724763122762e-06, + "loss": 0.164, + "step": 7671 + }, + { + "epoch": 0.5247964977084616, + "grad_norm": 1.6213600790179672, + "learning_rate": 4.841617560330359e-06, + "loss": 0.1715, + "step": 7672 + }, + { + "epoch": 0.5248649018400712, + "grad_norm": 2.130150884368167, + "learning_rate": 4.8405103653121315e-06, + "loss": 0.5642, + "step": 7673 + }, + { + "epoch": 0.5249333059716806, + "grad_norm": 2.412341717483768, + "learning_rate": 4.839403178122425e-06, + "loss": 0.4184, + "step": 7674 + }, + { + "epoch": 0.5250017101032902, + "grad_norm": 1.8247407252197119, + "learning_rate": 4.838295998815588e-06, + "loss": 0.4054, + "step": 7675 + }, + { + "epoch": 0.5250701142348998, + "grad_norm": 1.9118291943772208, + "learning_rate": 4.837188827445964e-06, + "loss": 0.3756, + "step": 7676 + }, + { + "epoch": 0.5251385183665094, + "grad_norm": 1.8419351131466073, + "learning_rate": 4.836081664067898e-06, + "loss": 0.3313, + "step": 7677 + }, + { + "epoch": 0.5252069224981188, + "grad_norm": 2.205137551033176, + "learning_rate": 4.834974508735735e-06, + "loss": 0.4726, + "step": 7678 + }, + { + "epoch": 0.5252753266297284, + "grad_norm": 2.0069960258840838, + "learning_rate": 4.83386736150382e-06, + "loss": 0.5043, + "step": 7679 + }, + { + "epoch": 0.525343730761338, + "grad_norm": 1.771095005972223, + "learning_rate": 4.832760222426498e-06, + "loss": 0.2592, + "step": 7680 + }, + { + "epoch": 0.5254121348929476, + "grad_norm": 5.159267594474703, + "learning_rate": 4.831653091558113e-06, + "loss": 0.5261, + "step": 7681 + }, + { + "epoch": 0.525480539024557, + "grad_norm": 2.4429973056540617, + "learning_rate": 4.830545968953004e-06, + "loss": 0.5527, + "step": 7682 + }, + { + "epoch": 0.5255489431561666, + "grad_norm": 1.883222262323448, + "learning_rate": 4.8294388546655184e-06, + "loss": 0.5572, + "step": 7683 + }, + { + "epoch": 0.5256173472877762, + "grad_norm": 2.644516796934676, + "learning_rate": 4.828331748749997e-06, + "loss": 0.456, + "step": 7684 + }, + { + "epoch": 0.5256857514193858, + "grad_norm": 2.0626335173074364, + "learning_rate": 4.827224651260784e-06, + "loss": 0.6465, + "step": 7685 + }, + { + "epoch": 0.5257541555509953, + "grad_norm": 2.319156210912679, + "learning_rate": 4.826117562252219e-06, + "loss": 0.5868, + "step": 7686 + }, + { + "epoch": 0.5258225596826048, + "grad_norm": 1.8373926321113117, + "learning_rate": 4.825010481778643e-06, + "loss": 0.3342, + "step": 7687 + }, + { + "epoch": 0.5258909638142144, + "grad_norm": 2.347795879001041, + "learning_rate": 4.823903409894398e-06, + "loss": 0.2942, + "step": 7688 + }, + { + "epoch": 0.5259593679458239, + "grad_norm": 2.303398353679778, + "learning_rate": 4.822796346653824e-06, + "loss": 0.3979, + "step": 7689 + }, + { + "epoch": 0.5260277720774335, + "grad_norm": 1.854368481921912, + "learning_rate": 4.821689292111262e-06, + "loss": 0.4312, + "step": 7690 + }, + { + "epoch": 0.526096176209043, + "grad_norm": 2.8974066852341442, + "learning_rate": 4.8205822463210494e-06, + "loss": 0.282, + "step": 7691 + }, + { + "epoch": 0.5261645803406526, + "grad_norm": 1.6717201589023447, + "learning_rate": 4.819475209337526e-06, + "loss": 0.216, + "step": 7692 + }, + { + "epoch": 0.5262329844722621, + "grad_norm": 1.9478727176794626, + "learning_rate": 4.818368181215031e-06, + "loss": 0.3161, + "step": 7693 + }, + { + "epoch": 0.5263013886038717, + "grad_norm": 2.034927743395158, + "learning_rate": 4.817261162007903e-06, + "loss": 0.4877, + "step": 7694 + }, + { + "epoch": 0.5263697927354812, + "grad_norm": 1.9933225502351903, + "learning_rate": 4.816154151770481e-06, + "loss": 0.3088, + "step": 7695 + }, + { + "epoch": 0.5264381968670908, + "grad_norm": 1.7939378159916157, + "learning_rate": 4.815047150557099e-06, + "loss": 0.2817, + "step": 7696 + }, + { + "epoch": 0.5265066009987003, + "grad_norm": 1.80329814291157, + "learning_rate": 4.813940158422098e-06, + "loss": 0.4158, + "step": 7697 + }, + { + "epoch": 0.5265750051303099, + "grad_norm": 2.5696038483813703, + "learning_rate": 4.812833175419812e-06, + "loss": 0.6365, + "step": 7698 + }, + { + "epoch": 0.5266434092619194, + "grad_norm": 2.0887255306283086, + "learning_rate": 4.811726201604576e-06, + "loss": 0.3822, + "step": 7699 + }, + { + "epoch": 0.526711813393529, + "grad_norm": 1.9530832400625968, + "learning_rate": 4.810619237030729e-06, + "loss": 0.4804, + "step": 7700 + }, + { + "epoch": 0.5267802175251385, + "grad_norm": 1.9051316529685873, + "learning_rate": 4.809512281752604e-06, + "loss": 0.5541, + "step": 7701 + }, + { + "epoch": 0.5268486216567481, + "grad_norm": 1.4045215315572208, + "learning_rate": 4.808405335824536e-06, + "loss": 0.2575, + "step": 7702 + }, + { + "epoch": 0.5269170257883576, + "grad_norm": 2.002040001449565, + "learning_rate": 4.807298399300857e-06, + "loss": 0.5573, + "step": 7703 + }, + { + "epoch": 0.5269854299199672, + "grad_norm": 2.9204227285875093, + "learning_rate": 4.806191472235906e-06, + "loss": 0.3712, + "step": 7704 + }, + { + "epoch": 0.5270538340515767, + "grad_norm": 2.417323493214302, + "learning_rate": 4.805084554684013e-06, + "loss": 0.2308, + "step": 7705 + }, + { + "epoch": 0.5271222381831863, + "grad_norm": 1.569507844645029, + "learning_rate": 4.80397764669951e-06, + "loss": 0.2486, + "step": 7706 + }, + { + "epoch": 0.5271906423147958, + "grad_norm": 2.1008123881400405, + "learning_rate": 4.802870748336732e-06, + "loss": 0.5525, + "step": 7707 + }, + { + "epoch": 0.5272590464464053, + "grad_norm": 1.5778656312047417, + "learning_rate": 4.80176385965001e-06, + "loss": 0.3414, + "step": 7708 + }, + { + "epoch": 0.5273274505780149, + "grad_norm": 1.7060149809961074, + "learning_rate": 4.800656980693674e-06, + "loss": 0.2917, + "step": 7709 + }, + { + "epoch": 0.5273958547096245, + "grad_norm": 2.329214225299422, + "learning_rate": 4.799550111522057e-06, + "loss": 0.3751, + "step": 7710 + }, + { + "epoch": 0.527464258841234, + "grad_norm": 2.7120669208177883, + "learning_rate": 4.798443252189488e-06, + "loss": 0.4456, + "step": 7711 + }, + { + "epoch": 0.5275326629728435, + "grad_norm": 2.0597724905755084, + "learning_rate": 4.797336402750297e-06, + "loss": 0.4636, + "step": 7712 + }, + { + "epoch": 0.5276010671044531, + "grad_norm": 2.229614294570601, + "learning_rate": 4.796229563258812e-06, + "loss": 0.3229, + "step": 7713 + }, + { + "epoch": 0.5276694712360627, + "grad_norm": 2.076658717553425, + "learning_rate": 4.795122733769367e-06, + "loss": 0.1967, + "step": 7714 + }, + { + "epoch": 0.5277378753676722, + "grad_norm": 2.1187917958415996, + "learning_rate": 4.7940159143362874e-06, + "loss": 0.4294, + "step": 7715 + }, + { + "epoch": 0.5278062794992817, + "grad_norm": 1.7373071834552711, + "learning_rate": 4.7929091050139004e-06, + "loss": 0.2437, + "step": 7716 + }, + { + "epoch": 0.5278746836308913, + "grad_norm": 1.7757716299888144, + "learning_rate": 4.791802305856536e-06, + "loss": 0.4105, + "step": 7717 + }, + { + "epoch": 0.5279430877625009, + "grad_norm": 1.388369193267494, + "learning_rate": 4.790695516918519e-06, + "loss": 0.2497, + "step": 7718 + }, + { + "epoch": 0.5280114918941105, + "grad_norm": 1.3641186481403278, + "learning_rate": 4.789588738254176e-06, + "loss": 0.1642, + "step": 7719 + }, + { + "epoch": 0.5280798960257199, + "grad_norm": 2.0089200041603394, + "learning_rate": 4.788481969917835e-06, + "loss": 0.3979, + "step": 7720 + }, + { + "epoch": 0.5281483001573295, + "grad_norm": 2.2020866368454546, + "learning_rate": 4.78737521196382e-06, + "loss": 0.5448, + "step": 7721 + }, + { + "epoch": 0.5282167042889391, + "grad_norm": 1.8437680227400965, + "learning_rate": 4.786268464446455e-06, + "loss": 0.3538, + "step": 7722 + }, + { + "epoch": 0.5282851084205487, + "grad_norm": 1.4735549794972844, + "learning_rate": 4.785161727420066e-06, + "loss": 0.2282, + "step": 7723 + }, + { + "epoch": 0.5283535125521581, + "grad_norm": 1.7040275535798748, + "learning_rate": 4.784055000938978e-06, + "loss": 0.308, + "step": 7724 + }, + { + "epoch": 0.5284219166837677, + "grad_norm": 1.9042724373056366, + "learning_rate": 4.782948285057514e-06, + "loss": 0.2605, + "step": 7725 + }, + { + "epoch": 0.5284903208153773, + "grad_norm": 1.7371962343515492, + "learning_rate": 4.781841579829995e-06, + "loss": 0.2528, + "step": 7726 + }, + { + "epoch": 0.5285587249469867, + "grad_norm": 2.5664300293307045, + "learning_rate": 4.780734885310745e-06, + "loss": 0.4174, + "step": 7727 + }, + { + "epoch": 0.5286271290785963, + "grad_norm": 2.005971584134495, + "learning_rate": 4.779628201554086e-06, + "loss": 0.3424, + "step": 7728 + }, + { + "epoch": 0.5286955332102059, + "grad_norm": 2.879582898811444, + "learning_rate": 4.778521528614339e-06, + "loss": 0.3028, + "step": 7729 + }, + { + "epoch": 0.5287639373418155, + "grad_norm": 1.7629133052706725, + "learning_rate": 4.777414866545824e-06, + "loss": 0.3516, + "step": 7730 + }, + { + "epoch": 0.528832341473425, + "grad_norm": 1.7570065579305552, + "learning_rate": 4.776308215402864e-06, + "loss": 0.4349, + "step": 7731 + }, + { + "epoch": 0.5289007456050345, + "grad_norm": 2.483946619348267, + "learning_rate": 4.775201575239775e-06, + "loss": 0.7153, + "step": 7732 + }, + { + "epoch": 0.5289691497366441, + "grad_norm": 2.3354131563031517, + "learning_rate": 4.774094946110878e-06, + "loss": 0.6423, + "step": 7733 + }, + { + "epoch": 0.5290375538682537, + "grad_norm": 2.7722101853970864, + "learning_rate": 4.772988328070493e-06, + "loss": 0.2204, + "step": 7734 + }, + { + "epoch": 0.5291059579998632, + "grad_norm": 1.814996525068871, + "learning_rate": 4.771881721172939e-06, + "loss": 0.2602, + "step": 7735 + }, + { + "epoch": 0.5291743621314727, + "grad_norm": 1.5378899198089138, + "learning_rate": 4.770775125472529e-06, + "loss": 0.2839, + "step": 7736 + }, + { + "epoch": 0.5292427662630823, + "grad_norm": 1.6468885215352578, + "learning_rate": 4.7696685410235835e-06, + "loss": 0.2189, + "step": 7737 + }, + { + "epoch": 0.5293111703946919, + "grad_norm": 1.8489079237947839, + "learning_rate": 4.768561967880419e-06, + "loss": 0.3079, + "step": 7738 + }, + { + "epoch": 0.5293795745263014, + "grad_norm": 1.7399721258427328, + "learning_rate": 4.76745540609735e-06, + "loss": 0.2986, + "step": 7739 + }, + { + "epoch": 0.5294479786579109, + "grad_norm": 2.1186429612795314, + "learning_rate": 4.766348855728693e-06, + "loss": 0.3193, + "step": 7740 + }, + { + "epoch": 0.5295163827895205, + "grad_norm": 1.6677062232258124, + "learning_rate": 4.765242316828763e-06, + "loss": 0.1564, + "step": 7741 + }, + { + "epoch": 0.5295847869211301, + "grad_norm": 1.8227560644701597, + "learning_rate": 4.764135789451871e-06, + "loss": 0.3077, + "step": 7742 + }, + { + "epoch": 0.5296531910527396, + "grad_norm": 2.230329933456481, + "learning_rate": 4.7630292736523345e-06, + "loss": 0.5839, + "step": 7743 + }, + { + "epoch": 0.5297215951843491, + "grad_norm": 2.4372106435395673, + "learning_rate": 4.761922769484467e-06, + "loss": 0.3134, + "step": 7744 + }, + { + "epoch": 0.5297899993159587, + "grad_norm": 2.1004275479936942, + "learning_rate": 4.76081627700258e-06, + "loss": 0.3485, + "step": 7745 + }, + { + "epoch": 0.5298584034475682, + "grad_norm": 1.8833691854320713, + "learning_rate": 4.759709796260983e-06, + "loss": 0.3986, + "step": 7746 + }, + { + "epoch": 0.5299268075791778, + "grad_norm": 1.8921262361041489, + "learning_rate": 4.758603327313991e-06, + "loss": 0.2903, + "step": 7747 + }, + { + "epoch": 0.5299952117107873, + "grad_norm": 2.704075332023939, + "learning_rate": 4.757496870215914e-06, + "loss": 0.2242, + "step": 7748 + }, + { + "epoch": 0.5300636158423969, + "grad_norm": 1.8927837767932196, + "learning_rate": 4.75639042502106e-06, + "loss": 0.2402, + "step": 7749 + }, + { + "epoch": 0.5301320199740064, + "grad_norm": 1.6691219265401784, + "learning_rate": 4.755283991783742e-06, + "loss": 0.202, + "step": 7750 + }, + { + "epoch": 0.530200424105616, + "grad_norm": 1.7872837106688673, + "learning_rate": 4.754177570558267e-06, + "loss": 0.36, + "step": 7751 + }, + { + "epoch": 0.5302688282372255, + "grad_norm": 1.6391525586554136, + "learning_rate": 4.753071161398941e-06, + "loss": 0.2168, + "step": 7752 + }, + { + "epoch": 0.5303372323688351, + "grad_norm": 2.1190216945201024, + "learning_rate": 4.751964764360077e-06, + "loss": 0.4523, + "step": 7753 + }, + { + "epoch": 0.5304056365004446, + "grad_norm": 1.3558294629814687, + "learning_rate": 4.750858379495981e-06, + "loss": 0.1069, + "step": 7754 + }, + { + "epoch": 0.5304740406320542, + "grad_norm": 2.0475871543477053, + "learning_rate": 4.749752006860958e-06, + "loss": 0.3145, + "step": 7755 + }, + { + "epoch": 0.5305424447636637, + "grad_norm": 1.8021258590803622, + "learning_rate": 4.7486456465093145e-06, + "loss": 0.285, + "step": 7756 + }, + { + "epoch": 0.5306108488952733, + "grad_norm": 2.9563666217858304, + "learning_rate": 4.7475392984953575e-06, + "loss": 0.2518, + "step": 7757 + }, + { + "epoch": 0.5306792530268828, + "grad_norm": 2.258878123122853, + "learning_rate": 4.746432962873392e-06, + "loss": 0.4608, + "step": 7758 + }, + { + "epoch": 0.5307476571584924, + "grad_norm": 1.9568092083173918, + "learning_rate": 4.745326639697718e-06, + "loss": 0.4778, + "step": 7759 + }, + { + "epoch": 0.530816061290102, + "grad_norm": 1.9704258613132588, + "learning_rate": 4.744220329022644e-06, + "loss": 0.3787, + "step": 7760 + }, + { + "epoch": 0.5308844654217115, + "grad_norm": 1.8917482326129287, + "learning_rate": 4.743114030902472e-06, + "loss": 0.3871, + "step": 7761 + }, + { + "epoch": 0.530952869553321, + "grad_norm": 2.12190082292757, + "learning_rate": 4.742007745391501e-06, + "loss": 0.4937, + "step": 7762 + }, + { + "epoch": 0.5310212736849306, + "grad_norm": 1.741894740152622, + "learning_rate": 4.7409014725440375e-06, + "loss": 0.2371, + "step": 7763 + }, + { + "epoch": 0.5310896778165402, + "grad_norm": 1.5756558184426135, + "learning_rate": 4.739795212414382e-06, + "loss": 0.279, + "step": 7764 + }, + { + "epoch": 0.5311580819481496, + "grad_norm": 2.1608534062173237, + "learning_rate": 4.738688965056833e-06, + "loss": 0.5575, + "step": 7765 + }, + { + "epoch": 0.5312264860797592, + "grad_norm": 1.7297975431067778, + "learning_rate": 4.7375827305256915e-06, + "loss": 0.1367, + "step": 7766 + }, + { + "epoch": 0.5312948902113688, + "grad_norm": 2.394379814798308, + "learning_rate": 4.736476508875257e-06, + "loss": 0.4915, + "step": 7767 + }, + { + "epoch": 0.5313632943429784, + "grad_norm": 2.0791902178294985, + "learning_rate": 4.735370300159829e-06, + "loss": 0.308, + "step": 7768 + }, + { + "epoch": 0.5314316984745878, + "grad_norm": 1.8192245441106305, + "learning_rate": 4.734264104433702e-06, + "loss": 0.4306, + "step": 7769 + }, + { + "epoch": 0.5315001026061974, + "grad_norm": 1.8891666793809219, + "learning_rate": 4.7331579217511785e-06, + "loss": 0.2575, + "step": 7770 + }, + { + "epoch": 0.531568506737807, + "grad_norm": 2.0661166714721912, + "learning_rate": 4.732051752166551e-06, + "loss": 0.3256, + "step": 7771 + }, + { + "epoch": 0.5316369108694166, + "grad_norm": 1.481916474647247, + "learning_rate": 4.730945595734118e-06, + "loss": 0.1921, + "step": 7772 + }, + { + "epoch": 0.531705315001026, + "grad_norm": 1.5358463245739495, + "learning_rate": 4.729839452508174e-06, + "loss": 0.2728, + "step": 7773 + }, + { + "epoch": 0.5317737191326356, + "grad_norm": 3.2203390310083932, + "learning_rate": 4.728733322543015e-06, + "loss": 0.2919, + "step": 7774 + }, + { + "epoch": 0.5318421232642452, + "grad_norm": 1.4472568346958676, + "learning_rate": 4.727627205892934e-06, + "loss": 0.2331, + "step": 7775 + }, + { + "epoch": 0.5319105273958548, + "grad_norm": 2.2579550582600505, + "learning_rate": 4.7265211026122245e-06, + "loss": 0.5915, + "step": 7776 + }, + { + "epoch": 0.5319789315274642, + "grad_norm": 2.851888455090205, + "learning_rate": 4.725415012755182e-06, + "loss": 0.2314, + "step": 7777 + }, + { + "epoch": 0.5320473356590738, + "grad_norm": 1.3194729037171473, + "learning_rate": 4.7243089363760945e-06, + "loss": 0.2167, + "step": 7778 + }, + { + "epoch": 0.5321157397906834, + "grad_norm": 2.075082451836795, + "learning_rate": 4.723202873529256e-06, + "loss": 0.3474, + "step": 7779 + }, + { + "epoch": 0.5321841439222929, + "grad_norm": 2.391496325008114, + "learning_rate": 4.722096824268959e-06, + "loss": 0.2996, + "step": 7780 + }, + { + "epoch": 0.5322525480539024, + "grad_norm": 2.2419395509056286, + "learning_rate": 4.72099078864949e-06, + "loss": 0.5261, + "step": 7781 + }, + { + "epoch": 0.532320952185512, + "grad_norm": 1.6487633758172988, + "learning_rate": 4.71988476672514e-06, + "loss": 0.1125, + "step": 7782 + }, + { + "epoch": 0.5323893563171216, + "grad_norm": 2.012711659549699, + "learning_rate": 4.7187787585501986e-06, + "loss": 0.3769, + "step": 7783 + }, + { + "epoch": 0.532457760448731, + "grad_norm": 2.1555110867785263, + "learning_rate": 4.717672764178955e-06, + "loss": 0.3653, + "step": 7784 + }, + { + "epoch": 0.5325261645803406, + "grad_norm": 3.0021909124427064, + "learning_rate": 4.716566783665695e-06, + "loss": 0.5763, + "step": 7785 + }, + { + "epoch": 0.5325945687119502, + "grad_norm": 2.124597932139836, + "learning_rate": 4.715460817064705e-06, + "loss": 0.6412, + "step": 7786 + }, + { + "epoch": 0.5326629728435598, + "grad_norm": 2.052118326911828, + "learning_rate": 4.714354864430273e-06, + "loss": 0.4867, + "step": 7787 + }, + { + "epoch": 0.5327313769751693, + "grad_norm": 2.1635682130255454, + "learning_rate": 4.713248925816683e-06, + "loss": 0.4594, + "step": 7788 + }, + { + "epoch": 0.5327997811067788, + "grad_norm": 1.8968654607190458, + "learning_rate": 4.712143001278221e-06, + "loss": 0.2549, + "step": 7789 + }, + { + "epoch": 0.5328681852383884, + "grad_norm": 2.355343500606213, + "learning_rate": 4.711037090869172e-06, + "loss": 0.4492, + "step": 7790 + }, + { + "epoch": 0.532936589369998, + "grad_norm": 2.1919445899758867, + "learning_rate": 4.709931194643815e-06, + "loss": 0.2541, + "step": 7791 + }, + { + "epoch": 0.5330049935016075, + "grad_norm": 1.540302099334807, + "learning_rate": 4.708825312656437e-06, + "loss": 0.299, + "step": 7792 + }, + { + "epoch": 0.533073397633217, + "grad_norm": 1.5460166596772518, + "learning_rate": 4.7077194449613185e-06, + "loss": 0.2537, + "step": 7793 + }, + { + "epoch": 0.5331418017648266, + "grad_norm": 1.8191107110399602, + "learning_rate": 4.706613591612742e-06, + "loss": 0.2166, + "step": 7794 + }, + { + "epoch": 0.5332102058964362, + "grad_norm": 2.5063936254201113, + "learning_rate": 4.7055077526649855e-06, + "loss": 0.3239, + "step": 7795 + }, + { + "epoch": 0.5332786100280457, + "grad_norm": 2.15059539623777, + "learning_rate": 4.704401928172332e-06, + "loss": 0.4308, + "step": 7796 + }, + { + "epoch": 0.5333470141596552, + "grad_norm": 2.5393324840497, + "learning_rate": 4.7032961181890595e-06, + "loss": 0.4629, + "step": 7797 + }, + { + "epoch": 0.5334154182912648, + "grad_norm": 1.8877138313533632, + "learning_rate": 4.7021903227694455e-06, + "loss": 0.4416, + "step": 7798 + }, + { + "epoch": 0.5334838224228743, + "grad_norm": 2.683199624517326, + "learning_rate": 4.701084541967769e-06, + "loss": 0.4814, + "step": 7799 + }, + { + "epoch": 0.5335522265544839, + "grad_norm": 1.8804621172963516, + "learning_rate": 4.699978775838307e-06, + "loss": 0.3782, + "step": 7800 + }, + { + "epoch": 0.5336206306860934, + "grad_norm": 2.2012911056003546, + "learning_rate": 4.698873024435334e-06, + "loss": 0.5913, + "step": 7801 + }, + { + "epoch": 0.533689034817703, + "grad_norm": 1.9140482376501475, + "learning_rate": 4.697767287813127e-06, + "loss": 0.309, + "step": 7802 + }, + { + "epoch": 0.5337574389493125, + "grad_norm": 2.2703128164912063, + "learning_rate": 4.6966615660259615e-06, + "loss": 0.3558, + "step": 7803 + }, + { + "epoch": 0.5338258430809221, + "grad_norm": 1.8881691122473865, + "learning_rate": 4.6955558591281125e-06, + "loss": 0.4582, + "step": 7804 + }, + { + "epoch": 0.5338942472125316, + "grad_norm": 1.8000050855935628, + "learning_rate": 4.694450167173851e-06, + "loss": 0.4512, + "step": 7805 + }, + { + "epoch": 0.5339626513441412, + "grad_norm": 2.155610653173363, + "learning_rate": 4.693344490217451e-06, + "loss": 0.4924, + "step": 7806 + }, + { + "epoch": 0.5340310554757507, + "grad_norm": 1.9994030076355065, + "learning_rate": 4.692238828313185e-06, + "loss": 0.275, + "step": 7807 + }, + { + "epoch": 0.5340994596073603, + "grad_norm": 2.4812671127497925, + "learning_rate": 4.691133181515322e-06, + "loss": 0.611, + "step": 7808 + }, + { + "epoch": 0.5341678637389699, + "grad_norm": 1.959518124100154, + "learning_rate": 4.690027549878135e-06, + "loss": 0.3652, + "step": 7809 + }, + { + "epoch": 0.5342362678705794, + "grad_norm": 2.2214700855689693, + "learning_rate": 4.6889219334558926e-06, + "loss": 0.5097, + "step": 7810 + }, + { + "epoch": 0.5343046720021889, + "grad_norm": 2.058665728532169, + "learning_rate": 4.687816332302863e-06, + "loss": 0.4002, + "step": 7811 + }, + { + "epoch": 0.5343730761337985, + "grad_norm": 2.4742924126710117, + "learning_rate": 4.686710746473314e-06, + "loss": 0.7073, + "step": 7812 + }, + { + "epoch": 0.534441480265408, + "grad_norm": 1.7380185432369457, + "learning_rate": 4.685605176021517e-06, + "loss": 0.2323, + "step": 7813 + }, + { + "epoch": 0.5345098843970176, + "grad_norm": 2.2983402975911424, + "learning_rate": 4.684499621001736e-06, + "loss": 0.3055, + "step": 7814 + }, + { + "epoch": 0.5345782885286271, + "grad_norm": 2.0592960484353, + "learning_rate": 4.683394081468236e-06, + "loss": 0.3004, + "step": 7815 + }, + { + "epoch": 0.5346466926602367, + "grad_norm": 2.5034763603922845, + "learning_rate": 4.6822885574752845e-06, + "loss": 0.6566, + "step": 7816 + }, + { + "epoch": 0.5347150967918463, + "grad_norm": 1.8671054263412161, + "learning_rate": 4.681183049077145e-06, + "loss": 0.3031, + "step": 7817 + }, + { + "epoch": 0.5347835009234557, + "grad_norm": 2.105160833996635, + "learning_rate": 4.68007755632808e-06, + "loss": 0.3754, + "step": 7818 + }, + { + "epoch": 0.5348519050550653, + "grad_norm": 1.839874380711176, + "learning_rate": 4.678972079282354e-06, + "loss": 0.2546, + "step": 7819 + }, + { + "epoch": 0.5349203091866749, + "grad_norm": 1.8794870886505075, + "learning_rate": 4.677866617994229e-06, + "loss": 0.3889, + "step": 7820 + }, + { + "epoch": 0.5349887133182845, + "grad_norm": 2.0069054992945716, + "learning_rate": 4.676761172517965e-06, + "loss": 0.3223, + "step": 7821 + }, + { + "epoch": 0.5350571174498939, + "grad_norm": 2.420916681501761, + "learning_rate": 4.675655742907822e-06, + "loss": 0.1618, + "step": 7822 + }, + { + "epoch": 0.5351255215815035, + "grad_norm": 2.2217683007310827, + "learning_rate": 4.674550329218063e-06, + "loss": 0.3254, + "step": 7823 + }, + { + "epoch": 0.5351939257131131, + "grad_norm": 2.993432480692078, + "learning_rate": 4.673444931502947e-06, + "loss": 0.4624, + "step": 7824 + }, + { + "epoch": 0.5352623298447227, + "grad_norm": 2.130011392426358, + "learning_rate": 4.672339549816728e-06, + "loss": 0.1878, + "step": 7825 + }, + { + "epoch": 0.5353307339763321, + "grad_norm": 2.081392687697591, + "learning_rate": 4.671234184213668e-06, + "loss": 0.6085, + "step": 7826 + }, + { + "epoch": 0.5353991381079417, + "grad_norm": 1.9198326563551493, + "learning_rate": 4.670128834748021e-06, + "loss": 0.3962, + "step": 7827 + }, + { + "epoch": 0.5354675422395513, + "grad_norm": 2.5876485229104094, + "learning_rate": 4.669023501474045e-06, + "loss": 0.6631, + "step": 7828 + }, + { + "epoch": 0.5355359463711609, + "grad_norm": 2.204933148222715, + "learning_rate": 4.667918184445992e-06, + "loss": 0.6098, + "step": 7829 + }, + { + "epoch": 0.5356043505027703, + "grad_norm": 2.3998130836069835, + "learning_rate": 4.66681288371812e-06, + "loss": 0.5034, + "step": 7830 + }, + { + "epoch": 0.5356727546343799, + "grad_norm": 1.8744763119834735, + "learning_rate": 4.665707599344679e-06, + "loss": 0.414, + "step": 7831 + }, + { + "epoch": 0.5357411587659895, + "grad_norm": 2.3138299352500526, + "learning_rate": 4.6646023313799225e-06, + "loss": 0.6334, + "step": 7832 + }, + { + "epoch": 0.5358095628975991, + "grad_norm": 2.0239172996278856, + "learning_rate": 4.663497079878105e-06, + "loss": 0.3291, + "step": 7833 + }, + { + "epoch": 0.5358779670292085, + "grad_norm": 2.1739779143981646, + "learning_rate": 4.662391844893476e-06, + "loss": 0.4482, + "step": 7834 + }, + { + "epoch": 0.5359463711608181, + "grad_norm": 2.096047407811681, + "learning_rate": 4.661286626480284e-06, + "loss": 0.4617, + "step": 7835 + }, + { + "epoch": 0.5360147752924277, + "grad_norm": 1.9100084150089012, + "learning_rate": 4.660181424692781e-06, + "loss": 0.4726, + "step": 7836 + }, + { + "epoch": 0.5360831794240372, + "grad_norm": 2.1768720210950625, + "learning_rate": 4.659076239585215e-06, + "loss": 0.52, + "step": 7837 + }, + { + "epoch": 0.5361515835556467, + "grad_norm": 1.9838928553049513, + "learning_rate": 4.657971071211833e-06, + "loss": 0.3935, + "step": 7838 + }, + { + "epoch": 0.5362199876872563, + "grad_norm": 1.9220674852571955, + "learning_rate": 4.656865919626883e-06, + "loss": 0.5946, + "step": 7839 + }, + { + "epoch": 0.5362883918188659, + "grad_norm": 2.031179415870284, + "learning_rate": 4.65576078488461e-06, + "loss": 0.4241, + "step": 7840 + }, + { + "epoch": 0.5363567959504754, + "grad_norm": 1.971167109965489, + "learning_rate": 4.654655667039258e-06, + "loss": 0.2726, + "step": 7841 + }, + { + "epoch": 0.5364252000820849, + "grad_norm": 1.958679858240846, + "learning_rate": 4.653550566145075e-06, + "loss": 0.1947, + "step": 7842 + }, + { + "epoch": 0.5364936042136945, + "grad_norm": 1.9284202646373283, + "learning_rate": 4.652445482256305e-06, + "loss": 0.4309, + "step": 7843 + }, + { + "epoch": 0.5365620083453041, + "grad_norm": 5.4546760495583895, + "learning_rate": 4.651340415427188e-06, + "loss": 0.486, + "step": 7844 + }, + { + "epoch": 0.5366304124769136, + "grad_norm": 1.2431990520410816, + "learning_rate": 4.650235365711966e-06, + "loss": 0.151, + "step": 7845 + }, + { + "epoch": 0.5366988166085231, + "grad_norm": 1.6874590334511421, + "learning_rate": 4.649130333164883e-06, + "loss": 0.3383, + "step": 7846 + }, + { + "epoch": 0.5367672207401327, + "grad_norm": 1.8744806444593836, + "learning_rate": 4.6480253178401775e-06, + "loss": 0.4967, + "step": 7847 + }, + { + "epoch": 0.5368356248717423, + "grad_norm": 2.2787968104163894, + "learning_rate": 4.646920319792088e-06, + "loss": 0.4203, + "step": 7848 + }, + { + "epoch": 0.5369040290033518, + "grad_norm": 2.429587806790658, + "learning_rate": 4.645815339074855e-06, + "loss": 0.6703, + "step": 7849 + }, + { + "epoch": 0.5369724331349613, + "grad_norm": 2.2615939288705054, + "learning_rate": 4.6447103757427156e-06, + "loss": 0.2894, + "step": 7850 + }, + { + "epoch": 0.5370408372665709, + "grad_norm": 1.8755367226118704, + "learning_rate": 4.6436054298499035e-06, + "loss": 0.2379, + "step": 7851 + }, + { + "epoch": 0.5371092413981805, + "grad_norm": 1.715909101243912, + "learning_rate": 4.6425005014506605e-06, + "loss": 0.3662, + "step": 7852 + }, + { + "epoch": 0.53717764552979, + "grad_norm": 1.5557178840019485, + "learning_rate": 4.6413955905992195e-06, + "loss": 0.2079, + "step": 7853 + }, + { + "epoch": 0.5372460496613995, + "grad_norm": 1.8078185988894548, + "learning_rate": 4.640290697349815e-06, + "loss": 0.3653, + "step": 7854 + }, + { + "epoch": 0.5373144537930091, + "grad_norm": 2.0530878156549135, + "learning_rate": 4.639185821756679e-06, + "loss": 0.2048, + "step": 7855 + }, + { + "epoch": 0.5373828579246186, + "grad_norm": 2.208625298818868, + "learning_rate": 4.638080963874046e-06, + "loss": 0.3844, + "step": 7856 + }, + { + "epoch": 0.5374512620562282, + "grad_norm": 1.9999305910494036, + "learning_rate": 4.636976123756147e-06, + "loss": 0.2843, + "step": 7857 + }, + { + "epoch": 0.5375196661878378, + "grad_norm": 2.203502994603268, + "learning_rate": 4.635871301457212e-06, + "loss": 0.5084, + "step": 7858 + }, + { + "epoch": 0.5375880703194473, + "grad_norm": 1.4195193752617226, + "learning_rate": 4.634766497031472e-06, + "loss": 0.1741, + "step": 7859 + }, + { + "epoch": 0.5376564744510568, + "grad_norm": 1.8812127420151443, + "learning_rate": 4.633661710533158e-06, + "loss": 0.3021, + "step": 7860 + }, + { + "epoch": 0.5377248785826664, + "grad_norm": 1.8836651687657742, + "learning_rate": 4.632556942016493e-06, + "loss": 0.4415, + "step": 7861 + }, + { + "epoch": 0.537793282714276, + "grad_norm": 2.1343247707929867, + "learning_rate": 4.631452191535709e-06, + "loss": 0.5508, + "step": 7862 + }, + { + "epoch": 0.5378616868458855, + "grad_norm": 1.8082716059660375, + "learning_rate": 4.630347459145031e-06, + "loss": 0.4086, + "step": 7863 + }, + { + "epoch": 0.537930090977495, + "grad_norm": 2.011682995228472, + "learning_rate": 4.6292427448986856e-06, + "loss": 0.4756, + "step": 7864 + }, + { + "epoch": 0.5379984951091046, + "grad_norm": 1.6199743036561904, + "learning_rate": 4.628138048850896e-06, + "loss": 0.3391, + "step": 7865 + }, + { + "epoch": 0.5380668992407142, + "grad_norm": 1.5081329252301185, + "learning_rate": 4.627033371055887e-06, + "loss": 0.2226, + "step": 7866 + }, + { + "epoch": 0.5381353033723237, + "grad_norm": 1.5752085657384034, + "learning_rate": 4.625928711567882e-06, + "loss": 0.2476, + "step": 7867 + }, + { + "epoch": 0.5382037075039332, + "grad_norm": 1.366997367717173, + "learning_rate": 4.6248240704411005e-06, + "loss": 0.191, + "step": 7868 + }, + { + "epoch": 0.5382721116355428, + "grad_norm": 2.3594529221153406, + "learning_rate": 4.623719447729767e-06, + "loss": 0.5732, + "step": 7869 + }, + { + "epoch": 0.5383405157671524, + "grad_norm": 1.2425547362845295, + "learning_rate": 4.6226148434880995e-06, + "loss": 0.1974, + "step": 7870 + }, + { + "epoch": 0.5384089198987619, + "grad_norm": 1.904285685188478, + "learning_rate": 4.6215102577703154e-06, + "loss": 0.4041, + "step": 7871 + }, + { + "epoch": 0.5384773240303714, + "grad_norm": 1.7532488009145333, + "learning_rate": 4.620405690630638e-06, + "loss": 0.2006, + "step": 7872 + }, + { + "epoch": 0.538545728161981, + "grad_norm": 2.021125216067707, + "learning_rate": 4.619301142123282e-06, + "loss": 0.4472, + "step": 7873 + }, + { + "epoch": 0.5386141322935906, + "grad_norm": 1.6312937312275042, + "learning_rate": 4.618196612302465e-06, + "loss": 0.2222, + "step": 7874 + }, + { + "epoch": 0.5386825364252, + "grad_norm": 2.697037678229069, + "learning_rate": 4.617092101222401e-06, + "loss": 0.6528, + "step": 7875 + }, + { + "epoch": 0.5387509405568096, + "grad_norm": 2.007422323945057, + "learning_rate": 4.615987608937307e-06, + "loss": 0.4564, + "step": 7876 + }, + { + "epoch": 0.5388193446884192, + "grad_norm": 2.3225316696553624, + "learning_rate": 4.614883135501394e-06, + "loss": 0.6075, + "step": 7877 + }, + { + "epoch": 0.5388877488200288, + "grad_norm": 1.6109104008064932, + "learning_rate": 4.6137786809688765e-06, + "loss": 0.1611, + "step": 7878 + }, + { + "epoch": 0.5389561529516382, + "grad_norm": 1.9177169906494012, + "learning_rate": 4.612674245393967e-06, + "loss": 0.2641, + "step": 7879 + }, + { + "epoch": 0.5390245570832478, + "grad_norm": 2.1325056113765184, + "learning_rate": 4.611569828830876e-06, + "loss": 0.3118, + "step": 7880 + }, + { + "epoch": 0.5390929612148574, + "grad_norm": 4.514682631844804, + "learning_rate": 4.610465431333813e-06, + "loss": 0.5411, + "step": 7881 + }, + { + "epoch": 0.539161365346467, + "grad_norm": 2.1618827625255506, + "learning_rate": 4.609361052956987e-06, + "loss": 0.4992, + "step": 7882 + }, + { + "epoch": 0.5392297694780764, + "grad_norm": 2.0344861734135584, + "learning_rate": 4.608256693754608e-06, + "loss": 0.328, + "step": 7883 + }, + { + "epoch": 0.539298173609686, + "grad_norm": 1.8397030624734494, + "learning_rate": 4.607152353780883e-06, + "loss": 0.3091, + "step": 7884 + }, + { + "epoch": 0.5393665777412956, + "grad_norm": 2.5993453765193744, + "learning_rate": 4.6060480330900156e-06, + "loss": 0.3137, + "step": 7885 + }, + { + "epoch": 0.5394349818729052, + "grad_norm": 2.4515239569100093, + "learning_rate": 4.604943731736215e-06, + "loss": 0.4873, + "step": 7886 + }, + { + "epoch": 0.5395033860045146, + "grad_norm": 2.2969855197604527, + "learning_rate": 4.603839449773682e-06, + "loss": 0.3646, + "step": 7887 + }, + { + "epoch": 0.5395717901361242, + "grad_norm": 2.1512523140803412, + "learning_rate": 4.602735187256624e-06, + "loss": 0.5462, + "step": 7888 + }, + { + "epoch": 0.5396401942677338, + "grad_norm": 2.958958419837047, + "learning_rate": 4.60163094423924e-06, + "loss": 0.6183, + "step": 7889 + }, + { + "epoch": 0.5397085983993434, + "grad_norm": 1.7855260641236792, + "learning_rate": 4.600526720775732e-06, + "loss": 0.2206, + "step": 7890 + }, + { + "epoch": 0.5397770025309528, + "grad_norm": 1.589507721202105, + "learning_rate": 4.5994225169203025e-06, + "loss": 0.2281, + "step": 7891 + }, + { + "epoch": 0.5398454066625624, + "grad_norm": 2.16386961163391, + "learning_rate": 4.5983183327271495e-06, + "loss": 0.4313, + "step": 7892 + }, + { + "epoch": 0.539913810794172, + "grad_norm": 3.032534061878992, + "learning_rate": 4.597214168250474e-06, + "loss": 0.3897, + "step": 7893 + }, + { + "epoch": 0.5399822149257815, + "grad_norm": 2.1602897257896365, + "learning_rate": 4.596110023544471e-06, + "loss": 0.3563, + "step": 7894 + }, + { + "epoch": 0.540050619057391, + "grad_norm": 2.129625541894056, + "learning_rate": 4.595005898663339e-06, + "loss": 0.4143, + "step": 7895 + }, + { + "epoch": 0.5401190231890006, + "grad_norm": 1.288104175036514, + "learning_rate": 4.593901793661274e-06, + "loss": 0.1736, + "step": 7896 + }, + { + "epoch": 0.5401874273206102, + "grad_norm": 1.7209884026035238, + "learning_rate": 4.592797708592468e-06, + "loss": 0.3666, + "step": 7897 + }, + { + "epoch": 0.5402558314522197, + "grad_norm": 2.165251749386382, + "learning_rate": 4.591693643511118e-06, + "loss": 0.3878, + "step": 7898 + }, + { + "epoch": 0.5403242355838292, + "grad_norm": 1.6147974754166352, + "learning_rate": 4.590589598471416e-06, + "loss": 0.2524, + "step": 7899 + }, + { + "epoch": 0.5403926397154388, + "grad_norm": 1.975854912859511, + "learning_rate": 4.589485573527553e-06, + "loss": 0.407, + "step": 7900 + }, + { + "epoch": 0.5404610438470484, + "grad_norm": 1.9276844586341984, + "learning_rate": 4.588381568733719e-06, + "loss": 0.2954, + "step": 7901 + }, + { + "epoch": 0.5405294479786579, + "grad_norm": 1.9424787701915809, + "learning_rate": 4.587277584144107e-06, + "loss": 0.2596, + "step": 7902 + }, + { + "epoch": 0.5405978521102675, + "grad_norm": 1.8919653712522655, + "learning_rate": 4.586173619812904e-06, + "loss": 0.4121, + "step": 7903 + }, + { + "epoch": 0.540666256241877, + "grad_norm": 2.341123275662305, + "learning_rate": 4.585069675794297e-06, + "loss": 0.3908, + "step": 7904 + }, + { + "epoch": 0.5407346603734866, + "grad_norm": 1.5014612312583575, + "learning_rate": 4.583965752142476e-06, + "loss": 0.2824, + "step": 7905 + }, + { + "epoch": 0.5408030645050961, + "grad_norm": 1.6260892078684077, + "learning_rate": 4.582861848911625e-06, + "loss": 0.2792, + "step": 7906 + }, + { + "epoch": 0.5408714686367057, + "grad_norm": 2.613740592695152, + "learning_rate": 4.581757966155926e-06, + "loss": 0.3629, + "step": 7907 + }, + { + "epoch": 0.5409398727683152, + "grad_norm": 1.9605153833912832, + "learning_rate": 4.580654103929568e-06, + "loss": 0.2835, + "step": 7908 + }, + { + "epoch": 0.5410082768999247, + "grad_norm": 2.5488923397233005, + "learning_rate": 4.579550262286731e-06, + "loss": 0.3464, + "step": 7909 + }, + { + "epoch": 0.5410766810315343, + "grad_norm": 1.8696913869384781, + "learning_rate": 4.578446441281597e-06, + "loss": 0.3045, + "step": 7910 + }, + { + "epoch": 0.5411450851631439, + "grad_norm": 2.1528820119425873, + "learning_rate": 4.5773426409683466e-06, + "loss": 0.4414, + "step": 7911 + }, + { + "epoch": 0.5412134892947534, + "grad_norm": 2.454096601718757, + "learning_rate": 4.576238861401162e-06, + "loss": 0.468, + "step": 7912 + }, + { + "epoch": 0.5412818934263629, + "grad_norm": 2.048255025678457, + "learning_rate": 4.57513510263422e-06, + "loss": 0.4416, + "step": 7913 + }, + { + "epoch": 0.5413502975579725, + "grad_norm": 2.489628934376094, + "learning_rate": 4.574031364721699e-06, + "loss": 0.2927, + "step": 7914 + }, + { + "epoch": 0.5414187016895821, + "grad_norm": 3.26289316380723, + "learning_rate": 4.572927647717776e-06, + "loss": 0.4376, + "step": 7915 + }, + { + "epoch": 0.5414871058211916, + "grad_norm": 1.6869170718548807, + "learning_rate": 4.571823951676625e-06, + "loss": 0.3226, + "step": 7916 + }, + { + "epoch": 0.5415555099528011, + "grad_norm": 2.32441872129173, + "learning_rate": 4.5707202766524225e-06, + "loss": 0.5613, + "step": 7917 + }, + { + "epoch": 0.5416239140844107, + "grad_norm": 4.2866797577350075, + "learning_rate": 4.569616622699343e-06, + "loss": 0.1696, + "step": 7918 + }, + { + "epoch": 0.5416923182160203, + "grad_norm": 2.5455275288443695, + "learning_rate": 4.568512989871557e-06, + "loss": 0.6508, + "step": 7919 + }, + { + "epoch": 0.5417607223476298, + "grad_norm": 2.488863327477198, + "learning_rate": 4.567409378223237e-06, + "loss": 0.5612, + "step": 7920 + }, + { + "epoch": 0.5418291264792393, + "grad_norm": 1.419910666986908, + "learning_rate": 4.566305787808552e-06, + "loss": 0.1836, + "step": 7921 + }, + { + "epoch": 0.5418975306108489, + "grad_norm": 22.27772481245603, + "learning_rate": 4.565202218681676e-06, + "loss": 0.3204, + "step": 7922 + }, + { + "epoch": 0.5419659347424585, + "grad_norm": 3.326435327316362, + "learning_rate": 4.564098670896774e-06, + "loss": 0.6189, + "step": 7923 + }, + { + "epoch": 0.542034338874068, + "grad_norm": 2.609670782171582, + "learning_rate": 4.5629951445080135e-06, + "loss": 0.4462, + "step": 7924 + }, + { + "epoch": 0.5421027430056775, + "grad_norm": 2.7011817275479264, + "learning_rate": 4.561891639569562e-06, + "loss": 0.5253, + "step": 7925 + }, + { + "epoch": 0.5421711471372871, + "grad_norm": 2.405306370982474, + "learning_rate": 4.560788156135586e-06, + "loss": 0.1994, + "step": 7926 + }, + { + "epoch": 0.5422395512688967, + "grad_norm": 1.5006130723983662, + "learning_rate": 4.559684694260245e-06, + "loss": 0.1927, + "step": 7927 + }, + { + "epoch": 0.5423079554005061, + "grad_norm": 2.3355780168605, + "learning_rate": 4.558581253997708e-06, + "loss": 0.7455, + "step": 7928 + }, + { + "epoch": 0.5423763595321157, + "grad_norm": 2.0524943435535246, + "learning_rate": 4.557477835402134e-06, + "loss": 0.3872, + "step": 7929 + }, + { + "epoch": 0.5424447636637253, + "grad_norm": 2.0780133740841014, + "learning_rate": 4.5563744385276845e-06, + "loss": 0.3402, + "step": 7930 + }, + { + "epoch": 0.5425131677953349, + "grad_norm": 2.0422170835364652, + "learning_rate": 4.555271063428518e-06, + "loss": 0.4688, + "step": 7931 + }, + { + "epoch": 0.5425815719269443, + "grad_norm": 2.0467008669507543, + "learning_rate": 4.554167710158798e-06, + "loss": 0.4039, + "step": 7932 + }, + { + "epoch": 0.5426499760585539, + "grad_norm": 2.457820998902853, + "learning_rate": 4.5530643787726785e-06, + "loss": 0.1966, + "step": 7933 + }, + { + "epoch": 0.5427183801901635, + "grad_norm": 2.3129650498927345, + "learning_rate": 4.551961069324318e-06, + "loss": 0.4555, + "step": 7934 + }, + { + "epoch": 0.5427867843217731, + "grad_norm": 1.8368058403458, + "learning_rate": 4.550857781867872e-06, + "loss": 0.3514, + "step": 7935 + }, + { + "epoch": 0.5428551884533825, + "grad_norm": 2.5198550348933346, + "learning_rate": 4.549754516457495e-06, + "loss": 0.599, + "step": 7936 + }, + { + "epoch": 0.5429235925849921, + "grad_norm": 2.308748604733263, + "learning_rate": 4.54865127314734e-06, + "loss": 0.5004, + "step": 7937 + }, + { + "epoch": 0.5429919967166017, + "grad_norm": 2.103626255395438, + "learning_rate": 4.54754805199156e-06, + "loss": 0.4872, + "step": 7938 + }, + { + "epoch": 0.5430604008482113, + "grad_norm": 2.8969236569397396, + "learning_rate": 4.546444853044308e-06, + "loss": 0.4176, + "step": 7939 + }, + { + "epoch": 0.5431288049798207, + "grad_norm": 2.4573425496034713, + "learning_rate": 4.5453416763597284e-06, + "loss": 0.7931, + "step": 7940 + }, + { + "epoch": 0.5431972091114303, + "grad_norm": 2.16421627365939, + "learning_rate": 4.544238521991978e-06, + "loss": 0.287, + "step": 7941 + }, + { + "epoch": 0.5432656132430399, + "grad_norm": 2.365092682000251, + "learning_rate": 4.543135389995202e-06, + "loss": 0.5047, + "step": 7942 + }, + { + "epoch": 0.5433340173746495, + "grad_norm": 1.5281789110967443, + "learning_rate": 4.542032280423547e-06, + "loss": 0.2611, + "step": 7943 + }, + { + "epoch": 0.543402421506259, + "grad_norm": 2.3493291318312273, + "learning_rate": 4.540929193331159e-06, + "loss": 0.5992, + "step": 7944 + }, + { + "epoch": 0.5434708256378685, + "grad_norm": 1.9739387309091672, + "learning_rate": 4.539826128772183e-06, + "loss": 0.4324, + "step": 7945 + }, + { + "epoch": 0.5435392297694781, + "grad_norm": 3.3073974257052416, + "learning_rate": 4.538723086800763e-06, + "loss": 0.5864, + "step": 7946 + }, + { + "epoch": 0.5436076339010876, + "grad_norm": 2.304820987908915, + "learning_rate": 4.537620067471041e-06, + "loss": 0.3449, + "step": 7947 + }, + { + "epoch": 0.5436760380326972, + "grad_norm": 2.0534821082172376, + "learning_rate": 4.53651707083716e-06, + "loss": 0.2239, + "step": 7948 + }, + { + "epoch": 0.5437444421643067, + "grad_norm": 2.902850429004987, + "learning_rate": 4.535414096953259e-06, + "loss": 0.2968, + "step": 7949 + }, + { + "epoch": 0.5438128462959163, + "grad_norm": 2.03089477660824, + "learning_rate": 4.534311145873476e-06, + "loss": 0.2491, + "step": 7950 + }, + { + "epoch": 0.5438812504275258, + "grad_norm": 2.080611003978771, + "learning_rate": 4.533208217651951e-06, + "loss": 0.2082, + "step": 7951 + }, + { + "epoch": 0.5439496545591354, + "grad_norm": 1.9592319761822046, + "learning_rate": 4.532105312342821e-06, + "loss": 0.3927, + "step": 7952 + }, + { + "epoch": 0.5440180586907449, + "grad_norm": 2.1314803144457315, + "learning_rate": 4.531002430000224e-06, + "loss": 0.4415, + "step": 7953 + }, + { + "epoch": 0.5440864628223545, + "grad_norm": 2.080432481983223, + "learning_rate": 4.529899570678289e-06, + "loss": 0.3459, + "step": 7954 + }, + { + "epoch": 0.544154866953964, + "grad_norm": 3.7609127761795227, + "learning_rate": 4.528796734431156e-06, + "loss": 0.2304, + "step": 7955 + }, + { + "epoch": 0.5442232710855736, + "grad_norm": 2.110635289118265, + "learning_rate": 4.527693921312954e-06, + "loss": 0.3605, + "step": 7956 + }, + { + "epoch": 0.5442916752171831, + "grad_norm": 1.9407257637486994, + "learning_rate": 4.526591131377814e-06, + "loss": 0.4657, + "step": 7957 + }, + { + "epoch": 0.5443600793487927, + "grad_norm": 2.1961244730936014, + "learning_rate": 4.525488364679869e-06, + "loss": 0.3019, + "step": 7958 + }, + { + "epoch": 0.5444284834804022, + "grad_norm": 1.9850078947681373, + "learning_rate": 4.5243856212732466e-06, + "loss": 0.5415, + "step": 7959 + }, + { + "epoch": 0.5444968876120118, + "grad_norm": 2.1339365823062946, + "learning_rate": 4.523282901212071e-06, + "loss": 0.4704, + "step": 7960 + }, + { + "epoch": 0.5445652917436213, + "grad_norm": 1.777407529755522, + "learning_rate": 4.522180204550475e-06, + "loss": 0.5064, + "step": 7961 + }, + { + "epoch": 0.5446336958752309, + "grad_norm": 2.6552833481183136, + "learning_rate": 4.521077531342582e-06, + "loss": 0.4809, + "step": 7962 + }, + { + "epoch": 0.5447021000068404, + "grad_norm": 2.598045870594776, + "learning_rate": 4.5199748816425174e-06, + "loss": 0.1752, + "step": 7963 + }, + { + "epoch": 0.54477050413845, + "grad_norm": 2.703177743738951, + "learning_rate": 4.518872255504402e-06, + "loss": 0.3371, + "step": 7964 + }, + { + "epoch": 0.5448389082700595, + "grad_norm": 2.960746537545146, + "learning_rate": 4.5177696529823615e-06, + "loss": 0.2666, + "step": 7965 + }, + { + "epoch": 0.544907312401669, + "grad_norm": 2.3732673919321354, + "learning_rate": 4.516667074130514e-06, + "loss": 0.5654, + "step": 7966 + }, + { + "epoch": 0.5449757165332786, + "grad_norm": 1.985855159695769, + "learning_rate": 4.515564519002981e-06, + "loss": 0.1612, + "step": 7967 + }, + { + "epoch": 0.5450441206648882, + "grad_norm": 2.525777915411983, + "learning_rate": 4.514461987653881e-06, + "loss": 0.3898, + "step": 7968 + }, + { + "epoch": 0.5451125247964977, + "grad_norm": 1.8088563312378803, + "learning_rate": 4.5133594801373305e-06, + "loss": 0.2268, + "step": 7969 + }, + { + "epoch": 0.5451809289281072, + "grad_norm": 1.866902932225325, + "learning_rate": 4.512256996507445e-06, + "loss": 0.2638, + "step": 7970 + }, + { + "epoch": 0.5452493330597168, + "grad_norm": 1.9858601899484711, + "learning_rate": 4.511154536818343e-06, + "loss": 0.4496, + "step": 7971 + }, + { + "epoch": 0.5453177371913264, + "grad_norm": 1.6527179605403248, + "learning_rate": 4.510052101124137e-06, + "loss": 0.2953, + "step": 7972 + }, + { + "epoch": 0.545386141322936, + "grad_norm": 2.2797178126536357, + "learning_rate": 4.508949689478941e-06, + "loss": 0.5989, + "step": 7973 + }, + { + "epoch": 0.5454545454545454, + "grad_norm": 2.8171792047667283, + "learning_rate": 4.507847301936863e-06, + "loss": 0.5459, + "step": 7974 + }, + { + "epoch": 0.545522949586155, + "grad_norm": 2.230769315867903, + "learning_rate": 4.506744938552018e-06, + "loss": 0.3972, + "step": 7975 + }, + { + "epoch": 0.5455913537177646, + "grad_norm": 2.6555297828836446, + "learning_rate": 4.505642599378513e-06, + "loss": 0.2495, + "step": 7976 + }, + { + "epoch": 0.5456597578493741, + "grad_norm": 2.5295838889096944, + "learning_rate": 4.5045402844704555e-06, + "loss": 0.4108, + "step": 7977 + }, + { + "epoch": 0.5457281619809836, + "grad_norm": 1.9721854375116101, + "learning_rate": 4.503437993881955e-06, + "loss": 0.4769, + "step": 7978 + }, + { + "epoch": 0.5457965661125932, + "grad_norm": 1.1047926543824405, + "learning_rate": 4.502335727667114e-06, + "loss": 0.1258, + "step": 7979 + }, + { + "epoch": 0.5458649702442028, + "grad_norm": 7.813743965294277, + "learning_rate": 4.501233485880039e-06, + "loss": 0.2415, + "step": 7980 + }, + { + "epoch": 0.5459333743758124, + "grad_norm": 1.6257023302287397, + "learning_rate": 4.500131268574832e-06, + "loss": 0.3699, + "step": 7981 + }, + { + "epoch": 0.5460017785074218, + "grad_norm": 1.978446484757579, + "learning_rate": 4.499029075805599e-06, + "loss": 0.4077, + "step": 7982 + }, + { + "epoch": 0.5460701826390314, + "grad_norm": 2.3261448516685332, + "learning_rate": 4.497926907626437e-06, + "loss": 0.3797, + "step": 7983 + }, + { + "epoch": 0.546138586770641, + "grad_norm": 1.5477375409702037, + "learning_rate": 4.496824764091446e-06, + "loss": 0.2085, + "step": 7984 + }, + { + "epoch": 0.5462069909022504, + "grad_norm": 1.852735598469442, + "learning_rate": 4.4957226452547256e-06, + "loss": 0.5001, + "step": 7985 + }, + { + "epoch": 0.54627539503386, + "grad_norm": 2.660998643765109, + "learning_rate": 4.494620551170374e-06, + "loss": 0.4455, + "step": 7986 + }, + { + "epoch": 0.5463437991654696, + "grad_norm": 1.9707663594923914, + "learning_rate": 4.493518481892485e-06, + "loss": 0.3267, + "step": 7987 + }, + { + "epoch": 0.5464122032970792, + "grad_norm": 2.036058112587789, + "learning_rate": 4.492416437475156e-06, + "loss": 0.3865, + "step": 7988 + }, + { + "epoch": 0.5464806074286886, + "grad_norm": 2.2972278907763086, + "learning_rate": 4.4913144179724775e-06, + "loss": 0.4704, + "step": 7989 + }, + { + "epoch": 0.5465490115602982, + "grad_norm": 1.9807261162439396, + "learning_rate": 4.490212423438544e-06, + "loss": 0.4009, + "step": 7990 + }, + { + "epoch": 0.5466174156919078, + "grad_norm": 1.867844517883193, + "learning_rate": 4.489110453927447e-06, + "loss": 0.4496, + "step": 7991 + }, + { + "epoch": 0.5466858198235174, + "grad_norm": 1.7361531679039672, + "learning_rate": 4.488008509493277e-06, + "loss": 0.3395, + "step": 7992 + }, + { + "epoch": 0.5467542239551269, + "grad_norm": 1.8174308143274198, + "learning_rate": 4.48690659019012e-06, + "loss": 0.3809, + "step": 7993 + }, + { + "epoch": 0.5468226280867364, + "grad_norm": 1.9736685733112966, + "learning_rate": 4.485804696072067e-06, + "loss": 0.3511, + "step": 7994 + }, + { + "epoch": 0.546891032218346, + "grad_norm": 2.3493158437735584, + "learning_rate": 4.484702827193202e-06, + "loss": 0.5359, + "step": 7995 + }, + { + "epoch": 0.5469594363499556, + "grad_norm": 1.7469422329160955, + "learning_rate": 4.48360098360761e-06, + "loss": 0.33, + "step": 7996 + }, + { + "epoch": 0.547027840481565, + "grad_norm": 2.718897333208892, + "learning_rate": 4.482499165369376e-06, + "loss": 0.5614, + "step": 7997 + }, + { + "epoch": 0.5470962446131746, + "grad_norm": 2.3119621070122993, + "learning_rate": 4.481397372532582e-06, + "loss": 0.332, + "step": 7998 + }, + { + "epoch": 0.5471646487447842, + "grad_norm": 1.948606926374352, + "learning_rate": 4.480295605151308e-06, + "loss": 0.4764, + "step": 7999 + }, + { + "epoch": 0.5472330528763938, + "grad_norm": 2.319621378148816, + "learning_rate": 4.479193863279635e-06, + "loss": 0.6821, + "step": 8000 + }, + { + "epoch": 0.5473014570080033, + "grad_norm": 2.4324917387650533, + "learning_rate": 4.478092146971645e-06, + "loss": 0.5614, + "step": 8001 + }, + { + "epoch": 0.5473698611396128, + "grad_norm": 1.7775768043393454, + "learning_rate": 4.476990456281411e-06, + "loss": 0.3602, + "step": 8002 + }, + { + "epoch": 0.5474382652712224, + "grad_norm": 1.7635879454386665, + "learning_rate": 4.47588879126301e-06, + "loss": 0.3359, + "step": 8003 + }, + { + "epoch": 0.5475066694028319, + "grad_norm": 1.9040399466098743, + "learning_rate": 4.47478715197052e-06, + "loss": 0.4762, + "step": 8004 + }, + { + "epoch": 0.5475750735344415, + "grad_norm": 1.9924941705764443, + "learning_rate": 4.473685538458013e-06, + "loss": 0.4168, + "step": 8005 + }, + { + "epoch": 0.547643477666051, + "grad_norm": 2.416772124165423, + "learning_rate": 4.472583950779559e-06, + "loss": 0.6814, + "step": 8006 + }, + { + "epoch": 0.5477118817976606, + "grad_norm": 2.0802379134865068, + "learning_rate": 4.471482388989233e-06, + "loss": 0.5699, + "step": 8007 + }, + { + "epoch": 0.5477802859292701, + "grad_norm": 2.6344758051798176, + "learning_rate": 4.470380853141104e-06, + "loss": 0.617, + "step": 8008 + }, + { + "epoch": 0.5478486900608797, + "grad_norm": 1.9105685321560766, + "learning_rate": 4.469279343289239e-06, + "loss": 0.2759, + "step": 8009 + }, + { + "epoch": 0.5479170941924892, + "grad_norm": 1.9081821577155977, + "learning_rate": 4.468177859487706e-06, + "loss": 0.3114, + "step": 8010 + }, + { + "epoch": 0.5479854983240988, + "grad_norm": 2.067209817660637, + "learning_rate": 4.467076401790573e-06, + "loss": 0.2676, + "step": 8011 + }, + { + "epoch": 0.5480539024557083, + "grad_norm": 1.9882470823098914, + "learning_rate": 4.465974970251905e-06, + "loss": 0.5467, + "step": 8012 + }, + { + "epoch": 0.5481223065873179, + "grad_norm": 1.9425508283595427, + "learning_rate": 4.464873564925762e-06, + "loss": 0.3842, + "step": 8013 + }, + { + "epoch": 0.5481907107189274, + "grad_norm": 2.2856009678233074, + "learning_rate": 4.4637721858662095e-06, + "loss": 0.3595, + "step": 8014 + }, + { + "epoch": 0.548259114850537, + "grad_norm": 2.1706804757260625, + "learning_rate": 4.462670833127308e-06, + "loss": 0.3529, + "step": 8015 + }, + { + "epoch": 0.5483275189821465, + "grad_norm": 2.2628998052778604, + "learning_rate": 4.461569506763116e-06, + "loss": 0.4191, + "step": 8016 + }, + { + "epoch": 0.5483959231137561, + "grad_norm": 2.0318433921887693, + "learning_rate": 4.4604682068276935e-06, + "loss": 0.3851, + "step": 8017 + }, + { + "epoch": 0.5484643272453656, + "grad_norm": 2.153941528950132, + "learning_rate": 4.459366933375097e-06, + "loss": 0.3189, + "step": 8018 + }, + { + "epoch": 0.5485327313769752, + "grad_norm": 2.0245618531620555, + "learning_rate": 4.45826568645938e-06, + "loss": 0.3261, + "step": 8019 + }, + { + "epoch": 0.5486011355085847, + "grad_norm": 1.8055472658549852, + "learning_rate": 4.457164466134599e-06, + "loss": 0.4018, + "step": 8020 + }, + { + "epoch": 0.5486695396401943, + "grad_norm": 1.905640019864012, + "learning_rate": 4.4560632724548094e-06, + "loss": 0.3806, + "step": 8021 + }, + { + "epoch": 0.5487379437718038, + "grad_norm": 2.2998551932132565, + "learning_rate": 4.454962105474061e-06, + "loss": 0.4808, + "step": 8022 + }, + { + "epoch": 0.5488063479034133, + "grad_norm": 1.9504095159494768, + "learning_rate": 4.453860965246403e-06, + "loss": 0.3713, + "step": 8023 + }, + { + "epoch": 0.5488747520350229, + "grad_norm": 2.169879621065156, + "learning_rate": 4.452759851825888e-06, + "loss": 0.3961, + "step": 8024 + }, + { + "epoch": 0.5489431561666325, + "grad_norm": 1.4111414608805866, + "learning_rate": 4.451658765266561e-06, + "loss": 0.2468, + "step": 8025 + }, + { + "epoch": 0.549011560298242, + "grad_norm": 2.288521784170949, + "learning_rate": 4.450557705622469e-06, + "loss": 0.5805, + "step": 8026 + }, + { + "epoch": 0.5490799644298515, + "grad_norm": 2.8666906731916253, + "learning_rate": 4.4494566729476595e-06, + "loss": 0.3573, + "step": 8027 + }, + { + "epoch": 0.5491483685614611, + "grad_norm": 2.213696190919437, + "learning_rate": 4.448355667296175e-06, + "loss": 0.4564, + "step": 8028 + }, + { + "epoch": 0.5492167726930707, + "grad_norm": 2.4468219108230045, + "learning_rate": 4.447254688722057e-06, + "loss": 0.7064, + "step": 8029 + }, + { + "epoch": 0.5492851768246803, + "grad_norm": 2.4205988505867, + "learning_rate": 4.4461537372793465e-06, + "loss": 0.7155, + "step": 8030 + }, + { + "epoch": 0.5493535809562897, + "grad_norm": 2.6095428869637685, + "learning_rate": 4.445052813022088e-06, + "loss": 0.7902, + "step": 8031 + }, + { + "epoch": 0.5494219850878993, + "grad_norm": 2.2190776233831944, + "learning_rate": 4.443951916004316e-06, + "loss": 0.5862, + "step": 8032 + }, + { + "epoch": 0.5494903892195089, + "grad_norm": 2.951298466712223, + "learning_rate": 4.44285104628007e-06, + "loss": 0.5646, + "step": 8033 + }, + { + "epoch": 0.5495587933511185, + "grad_norm": 2.144101455922822, + "learning_rate": 4.441750203903385e-06, + "loss": 0.2459, + "step": 8034 + }, + { + "epoch": 0.5496271974827279, + "grad_norm": 1.560962211138353, + "learning_rate": 4.440649388928295e-06, + "loss": 0.1999, + "step": 8035 + }, + { + "epoch": 0.5496956016143375, + "grad_norm": 2.7400369416699832, + "learning_rate": 4.439548601408834e-06, + "loss": 0.4363, + "step": 8036 + }, + { + "epoch": 0.5497640057459471, + "grad_norm": 2.6859861812207675, + "learning_rate": 4.4384478413990355e-06, + "loss": 0.4638, + "step": 8037 + }, + { + "epoch": 0.5498324098775565, + "grad_norm": 1.7388694111033103, + "learning_rate": 4.437347108952927e-06, + "loss": 0.3147, + "step": 8038 + }, + { + "epoch": 0.5499008140091661, + "grad_norm": 1.8366768788798333, + "learning_rate": 4.436246404124539e-06, + "loss": 0.2584, + "step": 8039 + }, + { + "epoch": 0.5499692181407757, + "grad_norm": 1.9715589511855722, + "learning_rate": 4.4351457269679e-06, + "loss": 0.4526, + "step": 8040 + }, + { + "epoch": 0.5500376222723853, + "grad_norm": 1.6057134891907359, + "learning_rate": 4.434045077537038e-06, + "loss": 0.2265, + "step": 8041 + }, + { + "epoch": 0.5501060264039948, + "grad_norm": 2.507229888835358, + "learning_rate": 4.432944455885976e-06, + "loss": 0.4709, + "step": 8042 + }, + { + "epoch": 0.5501744305356043, + "grad_norm": 2.0554003476136185, + "learning_rate": 4.431843862068737e-06, + "loss": 0.4814, + "step": 8043 + }, + { + "epoch": 0.5502428346672139, + "grad_norm": 1.748292434749339, + "learning_rate": 4.430743296139347e-06, + "loss": 0.3243, + "step": 8044 + }, + { + "epoch": 0.5503112387988235, + "grad_norm": 3.8809992203956183, + "learning_rate": 4.429642758151824e-06, + "loss": 0.3354, + "step": 8045 + }, + { + "epoch": 0.550379642930433, + "grad_norm": 1.8901308966714272, + "learning_rate": 4.428542248160188e-06, + "loss": 0.4343, + "step": 8046 + }, + { + "epoch": 0.5504480470620425, + "grad_norm": 2.3893415583715725, + "learning_rate": 4.427441766218459e-06, + "loss": 0.3213, + "step": 8047 + }, + { + "epoch": 0.5505164511936521, + "grad_norm": 1.8400082403301672, + "learning_rate": 4.4263413123806534e-06, + "loss": 0.3398, + "step": 8048 + }, + { + "epoch": 0.5505848553252617, + "grad_norm": 2.617335164473533, + "learning_rate": 4.425240886700783e-06, + "loss": 0.702, + "step": 8049 + }, + { + "epoch": 0.5506532594568712, + "grad_norm": 1.8962014377161849, + "learning_rate": 4.4241404892328675e-06, + "loss": 0.3313, + "step": 8050 + }, + { + "epoch": 0.5507216635884807, + "grad_norm": 1.594894433048233, + "learning_rate": 4.423040120030919e-06, + "loss": 0.2939, + "step": 8051 + }, + { + "epoch": 0.5507900677200903, + "grad_norm": 1.445642704230624, + "learning_rate": 4.421939779148947e-06, + "loss": 0.216, + "step": 8052 + }, + { + "epoch": 0.5508584718516999, + "grad_norm": 2.104534663858151, + "learning_rate": 4.420839466640961e-06, + "loss": 0.5278, + "step": 8053 + }, + { + "epoch": 0.5509268759833094, + "grad_norm": 2.383756713147605, + "learning_rate": 4.4197391825609716e-06, + "loss": 0.5788, + "step": 8054 + }, + { + "epoch": 0.5509952801149189, + "grad_norm": 3.097958120715615, + "learning_rate": 4.418638926962985e-06, + "loss": 0.3302, + "step": 8055 + }, + { + "epoch": 0.5510636842465285, + "grad_norm": 2.0907851977475334, + "learning_rate": 4.417538699901007e-06, + "loss": 0.394, + "step": 8056 + }, + { + "epoch": 0.551132088378138, + "grad_norm": 2.1878152926606584, + "learning_rate": 4.416438501429042e-06, + "loss": 0.4111, + "step": 8057 + }, + { + "epoch": 0.5512004925097476, + "grad_norm": 1.78263861086492, + "learning_rate": 4.4153383316010945e-06, + "loss": 0.2731, + "step": 8058 + }, + { + "epoch": 0.5512688966413571, + "grad_norm": 2.0762676744437956, + "learning_rate": 4.414238190471163e-06, + "loss": 0.495, + "step": 8059 + }, + { + "epoch": 0.5513373007729667, + "grad_norm": 2.501103682082128, + "learning_rate": 4.4131380780932494e-06, + "loss": 0.6222, + "step": 8060 + }, + { + "epoch": 0.5514057049045762, + "grad_norm": 2.5948993595173806, + "learning_rate": 4.412037994521355e-06, + "loss": 0.492, + "step": 8061 + }, + { + "epoch": 0.5514741090361858, + "grad_norm": 1.932750695365255, + "learning_rate": 4.410937939809476e-06, + "loss": 0.2247, + "step": 8062 + }, + { + "epoch": 0.5515425131677953, + "grad_norm": 1.7942942704071263, + "learning_rate": 4.409837914011604e-06, + "loss": 0.3958, + "step": 8063 + }, + { + "epoch": 0.5516109172994049, + "grad_norm": 2.019405165523443, + "learning_rate": 4.40873791718174e-06, + "loss": 0.4707, + "step": 8064 + }, + { + "epoch": 0.5516793214310144, + "grad_norm": 2.1754106038115086, + "learning_rate": 4.407637949373874e-06, + "loss": 0.4867, + "step": 8065 + }, + { + "epoch": 0.551747725562624, + "grad_norm": 2.153674225041536, + "learning_rate": 4.406538010641996e-06, + "loss": 0.6001, + "step": 8066 + }, + { + "epoch": 0.5518161296942335, + "grad_norm": 1.8217482041784203, + "learning_rate": 4.4054381010401e-06, + "loss": 0.2529, + "step": 8067 + }, + { + "epoch": 0.5518845338258431, + "grad_norm": 2.203624966212372, + "learning_rate": 4.404338220622173e-06, + "loss": 0.5247, + "step": 8068 + }, + { + "epoch": 0.5519529379574526, + "grad_norm": 2.3947937091908837, + "learning_rate": 4.4032383694422e-06, + "loss": 0.213, + "step": 8069 + }, + { + "epoch": 0.5520213420890622, + "grad_norm": 2.1726865898683263, + "learning_rate": 4.402138547554171e-06, + "loss": 0.3513, + "step": 8070 + }, + { + "epoch": 0.5520897462206718, + "grad_norm": 1.9096373377825482, + "learning_rate": 4.40103875501207e-06, + "loss": 0.376, + "step": 8071 + }, + { + "epoch": 0.5521581503522813, + "grad_norm": 2.2393647545599573, + "learning_rate": 4.39993899186988e-06, + "loss": 0.1986, + "step": 8072 + }, + { + "epoch": 0.5522265544838908, + "grad_norm": 2.509437106113356, + "learning_rate": 4.398839258181579e-06, + "loss": 0.3908, + "step": 8073 + }, + { + "epoch": 0.5522949586155004, + "grad_norm": 1.2082605367414592, + "learning_rate": 4.397739554001153e-06, + "loss": 0.1558, + "step": 8074 + }, + { + "epoch": 0.55236336274711, + "grad_norm": 2.2314347147950784, + "learning_rate": 4.3966398793825776e-06, + "loss": 0.3861, + "step": 8075 + }, + { + "epoch": 0.5524317668787194, + "grad_norm": 2.0365351583586775, + "learning_rate": 4.39554023437983e-06, + "loss": 0.4982, + "step": 8076 + }, + { + "epoch": 0.552500171010329, + "grad_norm": 2.193781590438053, + "learning_rate": 4.3944406190468865e-06, + "loss": 0.5795, + "step": 8077 + }, + { + "epoch": 0.5525685751419386, + "grad_norm": 1.7908824966855799, + "learning_rate": 4.39334103343772e-06, + "loss": 0.1951, + "step": 8078 + }, + { + "epoch": 0.5526369792735482, + "grad_norm": 2.0889233995804095, + "learning_rate": 4.3922414776063075e-06, + "loss": 0.4698, + "step": 8079 + }, + { + "epoch": 0.5527053834051576, + "grad_norm": 2.7470875185886348, + "learning_rate": 4.391141951606617e-06, + "loss": 0.8101, + "step": 8080 + }, + { + "epoch": 0.5527737875367672, + "grad_norm": 1.9176129246077838, + "learning_rate": 4.39004245549262e-06, + "loss": 0.5452, + "step": 8081 + }, + { + "epoch": 0.5528421916683768, + "grad_norm": 2.1779836606068725, + "learning_rate": 4.388942989318286e-06, + "loss": 0.4884, + "step": 8082 + }, + { + "epoch": 0.5529105957999864, + "grad_norm": 2.0069906096243506, + "learning_rate": 4.387843553137578e-06, + "loss": 0.3962, + "step": 8083 + }, + { + "epoch": 0.5529789999315958, + "grad_norm": 1.753890889191852, + "learning_rate": 4.386744147004467e-06, + "loss": 0.2964, + "step": 8084 + }, + { + "epoch": 0.5530474040632054, + "grad_norm": 1.7829852059818172, + "learning_rate": 4.3856447709729135e-06, + "loss": 0.2685, + "step": 8085 + }, + { + "epoch": 0.553115808194815, + "grad_norm": 2.460287226185305, + "learning_rate": 4.384545425096882e-06, + "loss": 0.825, + "step": 8086 + }, + { + "epoch": 0.5531842123264246, + "grad_norm": 2.094062611030971, + "learning_rate": 4.383446109430333e-06, + "loss": 0.4497, + "step": 8087 + }, + { + "epoch": 0.553252616458034, + "grad_norm": 2.0186192133555583, + "learning_rate": 4.382346824027225e-06, + "loss": 0.4897, + "step": 8088 + }, + { + "epoch": 0.5533210205896436, + "grad_norm": 2.598623442749426, + "learning_rate": 4.381247568941518e-06, + "loss": 0.4535, + "step": 8089 + }, + { + "epoch": 0.5533894247212532, + "grad_norm": 2.428747222101701, + "learning_rate": 4.380148344227168e-06, + "loss": 0.4691, + "step": 8090 + }, + { + "epoch": 0.5534578288528628, + "grad_norm": 1.560702625279734, + "learning_rate": 4.3790491499381316e-06, + "loss": 0.2968, + "step": 8091 + }, + { + "epoch": 0.5535262329844722, + "grad_norm": 2.1683258174992734, + "learning_rate": 4.37794998612836e-06, + "loss": 0.3562, + "step": 8092 + }, + { + "epoch": 0.5535946371160818, + "grad_norm": 2.0179836534045044, + "learning_rate": 4.376850852851808e-06, + "loss": 0.4169, + "step": 8093 + }, + { + "epoch": 0.5536630412476914, + "grad_norm": 2.228131795286969, + "learning_rate": 4.375751750162426e-06, + "loss": 0.6883, + "step": 8094 + }, + { + "epoch": 0.5537314453793009, + "grad_norm": 1.582588248451411, + "learning_rate": 4.374652678114162e-06, + "loss": 0.3329, + "step": 8095 + }, + { + "epoch": 0.5537998495109104, + "grad_norm": 2.1529115353753174, + "learning_rate": 4.373553636760964e-06, + "loss": 0.5221, + "step": 8096 + }, + { + "epoch": 0.55386825364252, + "grad_norm": 2.327078984321098, + "learning_rate": 4.3724546261567795e-06, + "loss": 0.5726, + "step": 8097 + }, + { + "epoch": 0.5539366577741296, + "grad_norm": 2.1226599445107284, + "learning_rate": 4.371355646355551e-06, + "loss": 0.4762, + "step": 8098 + }, + { + "epoch": 0.5540050619057391, + "grad_norm": 2.170246941144106, + "learning_rate": 4.370256697411221e-06, + "loss": 0.6029, + "step": 8099 + }, + { + "epoch": 0.5540734660373486, + "grad_norm": 1.4280946697044412, + "learning_rate": 4.369157779377738e-06, + "loss": 0.1716, + "step": 8100 + }, + { + "epoch": 0.5541418701689582, + "grad_norm": 2.4410097700522813, + "learning_rate": 4.368058892309035e-06, + "loss": 0.2789, + "step": 8101 + }, + { + "epoch": 0.5542102743005678, + "grad_norm": 2.0622010231066032, + "learning_rate": 4.3669600362590525e-06, + "loss": 0.5388, + "step": 8102 + }, + { + "epoch": 0.5542786784321773, + "grad_norm": 2.3869465720769987, + "learning_rate": 4.36586121128173e-06, + "loss": 0.3404, + "step": 8103 + }, + { + "epoch": 0.5543470825637868, + "grad_norm": 1.7722178731248455, + "learning_rate": 4.364762417431001e-06, + "loss": 0.2151, + "step": 8104 + }, + { + "epoch": 0.5544154866953964, + "grad_norm": 1.729321338350415, + "learning_rate": 4.3636636547608e-06, + "loss": 0.136, + "step": 8105 + }, + { + "epoch": 0.554483890827006, + "grad_norm": 1.2956265104573008, + "learning_rate": 4.36256492332506e-06, + "loss": 0.1298, + "step": 8106 + }, + { + "epoch": 0.5545522949586155, + "grad_norm": 2.6172866176226663, + "learning_rate": 4.361466223177711e-06, + "loss": 0.5298, + "step": 8107 + }, + { + "epoch": 0.554620699090225, + "grad_norm": 2.1110923416644596, + "learning_rate": 4.360367554372684e-06, + "loss": 0.6167, + "step": 8108 + }, + { + "epoch": 0.5546891032218346, + "grad_norm": 1.884265671274014, + "learning_rate": 4.359268916963903e-06, + "loss": 0.1788, + "step": 8109 + }, + { + "epoch": 0.5547575073534442, + "grad_norm": 1.9728514470885044, + "learning_rate": 4.358170311005302e-06, + "loss": 0.4645, + "step": 8110 + }, + { + "epoch": 0.5548259114850537, + "grad_norm": 2.263705222910269, + "learning_rate": 4.357071736550799e-06, + "loss": 0.6319, + "step": 8111 + }, + { + "epoch": 0.5548943156166632, + "grad_norm": 2.1916109829224153, + "learning_rate": 4.355973193654321e-06, + "loss": 0.5752, + "step": 8112 + }, + { + "epoch": 0.5549627197482728, + "grad_norm": 1.7365824595730956, + "learning_rate": 4.354874682369789e-06, + "loss": 0.3356, + "step": 8113 + }, + { + "epoch": 0.5550311238798823, + "grad_norm": 1.6655700373930136, + "learning_rate": 4.353776202751123e-06, + "loss": 0.3159, + "step": 8114 + }, + { + "epoch": 0.5550995280114919, + "grad_norm": 2.505637164235056, + "learning_rate": 4.35267775485224e-06, + "loss": 0.4616, + "step": 8115 + }, + { + "epoch": 0.5551679321431015, + "grad_norm": 1.6789155472559691, + "learning_rate": 4.351579338727059e-06, + "loss": 0.3058, + "step": 8116 + }, + { + "epoch": 0.555236336274711, + "grad_norm": 1.905950483649132, + "learning_rate": 4.350480954429497e-06, + "loss": 0.4209, + "step": 8117 + }, + { + "epoch": 0.5553047404063205, + "grad_norm": 1.5671842122447552, + "learning_rate": 4.349382602013464e-06, + "loss": 0.2968, + "step": 8118 + }, + { + "epoch": 0.5553731445379301, + "grad_norm": 2.2511737000701086, + "learning_rate": 4.348284281532874e-06, + "loss": 0.3703, + "step": 8119 + }, + { + "epoch": 0.5554415486695397, + "grad_norm": 2.673461443082074, + "learning_rate": 4.34718599304164e-06, + "loss": 0.3363, + "step": 8120 + }, + { + "epoch": 0.5555099528011492, + "grad_norm": 1.4536180070076912, + "learning_rate": 4.346087736593671e-06, + "loss": 0.0713, + "step": 8121 + }, + { + "epoch": 0.5555783569327587, + "grad_norm": 1.71085463719483, + "learning_rate": 4.344989512242872e-06, + "loss": 0.1591, + "step": 8122 + }, + { + "epoch": 0.5556467610643683, + "grad_norm": 1.9231666047248344, + "learning_rate": 4.343891320043152e-06, + "loss": 0.3694, + "step": 8123 + }, + { + "epoch": 0.5557151651959779, + "grad_norm": 2.2418246125098733, + "learning_rate": 4.342793160048414e-06, + "loss": 0.4063, + "step": 8124 + }, + { + "epoch": 0.5557835693275874, + "grad_norm": 1.6778178210928127, + "learning_rate": 4.34169503231256e-06, + "loss": 0.2561, + "step": 8125 + }, + { + "epoch": 0.5558519734591969, + "grad_norm": 1.7986190831827193, + "learning_rate": 4.340596936889495e-06, + "loss": 0.2233, + "step": 8126 + }, + { + "epoch": 0.5559203775908065, + "grad_norm": 2.749894123856929, + "learning_rate": 4.3394988738331155e-06, + "loss": 0.4415, + "step": 8127 + }, + { + "epoch": 0.5559887817224161, + "grad_norm": 1.8665167509559204, + "learning_rate": 4.33840084319732e-06, + "loss": 0.2937, + "step": 8128 + }, + { + "epoch": 0.5560571858540256, + "grad_norm": 1.5124375534368155, + "learning_rate": 4.337302845036005e-06, + "loss": 0.1794, + "step": 8129 + }, + { + "epoch": 0.5561255899856351, + "grad_norm": 2.0134379845843164, + "learning_rate": 4.336204879403069e-06, + "loss": 0.5153, + "step": 8130 + }, + { + "epoch": 0.5561939941172447, + "grad_norm": 2.0825012635551183, + "learning_rate": 4.335106946352402e-06, + "loss": 0.3416, + "step": 8131 + }, + { + "epoch": 0.5562623982488543, + "grad_norm": 1.949472827444149, + "learning_rate": 4.334009045937897e-06, + "loss": 0.3981, + "step": 8132 + }, + { + "epoch": 0.5563308023804637, + "grad_norm": 1.8072657251191042, + "learning_rate": 4.332911178213444e-06, + "loss": 0.2897, + "step": 8133 + }, + { + "epoch": 0.5563992065120733, + "grad_norm": 2.031799638184287, + "learning_rate": 4.331813343232932e-06, + "loss": 0.3818, + "step": 8134 + }, + { + "epoch": 0.5564676106436829, + "grad_norm": 2.319001861605035, + "learning_rate": 4.330715541050246e-06, + "loss": 0.3874, + "step": 8135 + }, + { + "epoch": 0.5565360147752925, + "grad_norm": 2.705838170332639, + "learning_rate": 4.329617771719275e-06, + "loss": 0.4356, + "step": 8136 + }, + { + "epoch": 0.5566044189069019, + "grad_norm": 1.8848129068877582, + "learning_rate": 4.328520035293902e-06, + "loss": 0.3269, + "step": 8137 + }, + { + "epoch": 0.5566728230385115, + "grad_norm": 1.8669895242735297, + "learning_rate": 4.327422331828005e-06, + "loss": 0.5511, + "step": 8138 + }, + { + "epoch": 0.5567412271701211, + "grad_norm": 2.40581352436705, + "learning_rate": 4.3263246613754685e-06, + "loss": 0.249, + "step": 8139 + }, + { + "epoch": 0.5568096313017307, + "grad_norm": 2.0760590961360754, + "learning_rate": 4.325227023990173e-06, + "loss": 0.2662, + "step": 8140 + }, + { + "epoch": 0.5568780354333401, + "grad_norm": 2.291366933959412, + "learning_rate": 4.324129419725993e-06, + "loss": 0.2846, + "step": 8141 + }, + { + "epoch": 0.5569464395649497, + "grad_norm": 1.618496889960772, + "learning_rate": 4.323031848636803e-06, + "loss": 0.2565, + "step": 8142 + }, + { + "epoch": 0.5570148436965593, + "grad_norm": 1.2048405362060253, + "learning_rate": 4.321934310776481e-06, + "loss": 0.2141, + "step": 8143 + }, + { + "epoch": 0.5570832478281689, + "grad_norm": 2.1783174358613318, + "learning_rate": 4.320836806198897e-06, + "loss": 0.3486, + "step": 8144 + }, + { + "epoch": 0.5571516519597783, + "grad_norm": 2.5838498450311467, + "learning_rate": 4.319739334957921e-06, + "loss": 0.4907, + "step": 8145 + }, + { + "epoch": 0.5572200560913879, + "grad_norm": 1.969525456977989, + "learning_rate": 4.318641897107425e-06, + "loss": 0.3781, + "step": 8146 + }, + { + "epoch": 0.5572884602229975, + "grad_norm": 2.0593033351208576, + "learning_rate": 4.3175444927012744e-06, + "loss": 0.3582, + "step": 8147 + }, + { + "epoch": 0.5573568643546071, + "grad_norm": 1.9703518088380252, + "learning_rate": 4.316447121793334e-06, + "loss": 0.4227, + "step": 8148 + }, + { + "epoch": 0.5574252684862165, + "grad_norm": 2.19775862524835, + "learning_rate": 4.315349784437471e-06, + "loss": 0.5112, + "step": 8149 + }, + { + "epoch": 0.5574936726178261, + "grad_norm": 2.100733895014577, + "learning_rate": 4.314252480687547e-06, + "loss": 0.2441, + "step": 8150 + }, + { + "epoch": 0.5575620767494357, + "grad_norm": 2.9461043935020546, + "learning_rate": 4.313155210597422e-06, + "loss": 0.2219, + "step": 8151 + }, + { + "epoch": 0.5576304808810452, + "grad_norm": 1.4534194478767741, + "learning_rate": 4.312057974220956e-06, + "loss": 0.2093, + "step": 8152 + }, + { + "epoch": 0.5576988850126547, + "grad_norm": 1.4520628723712492, + "learning_rate": 4.3109607716120076e-06, + "loss": 0.2197, + "step": 8153 + }, + { + "epoch": 0.5577672891442643, + "grad_norm": 2.014963747572821, + "learning_rate": 4.3098636028244315e-06, + "loss": 0.5652, + "step": 8154 + }, + { + "epoch": 0.5578356932758739, + "grad_norm": 1.645770362713689, + "learning_rate": 4.308766467912082e-06, + "loss": 0.1833, + "step": 8155 + }, + { + "epoch": 0.5579040974074834, + "grad_norm": 1.2154279221208204, + "learning_rate": 4.307669366928813e-06, + "loss": 0.2123, + "step": 8156 + }, + { + "epoch": 0.557972501539093, + "grad_norm": 1.974990777706392, + "learning_rate": 4.306572299928475e-06, + "loss": 0.3418, + "step": 8157 + }, + { + "epoch": 0.5580409056707025, + "grad_norm": 2.0020364064827745, + "learning_rate": 4.305475266964914e-06, + "loss": 0.5993, + "step": 8158 + }, + { + "epoch": 0.5581093098023121, + "grad_norm": 2.554039123336694, + "learning_rate": 4.304378268091982e-06, + "loss": 0.3915, + "step": 8159 + }, + { + "epoch": 0.5581777139339216, + "grad_norm": 2.5813541072698034, + "learning_rate": 4.303281303363526e-06, + "loss": 0.655, + "step": 8160 + }, + { + "epoch": 0.5582461180655311, + "grad_norm": 2.520364336780266, + "learning_rate": 4.302184372833387e-06, + "loss": 0.6435, + "step": 8161 + }, + { + "epoch": 0.5583145221971407, + "grad_norm": 2.098832384989448, + "learning_rate": 4.301087476555408e-06, + "loss": 0.2916, + "step": 8162 + }, + { + "epoch": 0.5583829263287503, + "grad_norm": 2.077544052262916, + "learning_rate": 4.299990614583432e-06, + "loss": 0.2337, + "step": 8163 + }, + { + "epoch": 0.5584513304603598, + "grad_norm": 2.1049966088129484, + "learning_rate": 4.298893786971296e-06, + "loss": 0.5442, + "step": 8164 + }, + { + "epoch": 0.5585197345919694, + "grad_norm": 1.8964732333797898, + "learning_rate": 4.297796993772837e-06, + "loss": 0.4181, + "step": 8165 + }, + { + "epoch": 0.5585881387235789, + "grad_norm": 2.277985811102253, + "learning_rate": 4.296700235041895e-06, + "loss": 0.2916, + "step": 8166 + }, + { + "epoch": 0.5586565428551884, + "grad_norm": 2.400271826365822, + "learning_rate": 4.295603510832302e-06, + "loss": 0.2545, + "step": 8167 + }, + { + "epoch": 0.558724946986798, + "grad_norm": 2.3386667058150694, + "learning_rate": 4.294506821197887e-06, + "loss": 0.6305, + "step": 8168 + }, + { + "epoch": 0.5587933511184076, + "grad_norm": 2.2694459922127157, + "learning_rate": 4.293410166192485e-06, + "loss": 0.4857, + "step": 8169 + }, + { + "epoch": 0.5588617552500171, + "grad_norm": 1.534142705616106, + "learning_rate": 4.292313545869925e-06, + "loss": 0.1995, + "step": 8170 + }, + { + "epoch": 0.5589301593816266, + "grad_norm": 2.18447336212833, + "learning_rate": 4.291216960284034e-06, + "loss": 0.4169, + "step": 8171 + }, + { + "epoch": 0.5589985635132362, + "grad_norm": 1.4656040583621175, + "learning_rate": 4.290120409488636e-06, + "loss": 0.1873, + "step": 8172 + }, + { + "epoch": 0.5590669676448458, + "grad_norm": 1.9864407547758236, + "learning_rate": 4.2890238935375585e-06, + "loss": 0.472, + "step": 8173 + }, + { + "epoch": 0.5591353717764553, + "grad_norm": 2.4452666988288176, + "learning_rate": 4.2879274124846206e-06, + "loss": 0.5064, + "step": 8174 + }, + { + "epoch": 0.5592037759080648, + "grad_norm": 2.4563100268767735, + "learning_rate": 4.2868309663836425e-06, + "loss": 0.4958, + "step": 8175 + }, + { + "epoch": 0.5592721800396744, + "grad_norm": 1.640182717062421, + "learning_rate": 4.285734555288446e-06, + "loss": 0.2497, + "step": 8176 + }, + { + "epoch": 0.559340584171284, + "grad_norm": 2.8321919047070856, + "learning_rate": 4.284638179252845e-06, + "loss": 0.6522, + "step": 8177 + }, + { + "epoch": 0.5594089883028935, + "grad_norm": 1.493675348978249, + "learning_rate": 4.283541838330657e-06, + "loss": 0.1669, + "step": 8178 + }, + { + "epoch": 0.559477392434503, + "grad_norm": 2.1700185622023453, + "learning_rate": 4.2824455325756955e-06, + "loss": 0.4868, + "step": 8179 + }, + { + "epoch": 0.5595457965661126, + "grad_norm": 2.2566035772963495, + "learning_rate": 4.281349262041773e-06, + "loss": 0.5629, + "step": 8180 + }, + { + "epoch": 0.5596142006977222, + "grad_norm": 2.336983360477768, + "learning_rate": 4.2802530267827e-06, + "loss": 0.394, + "step": 8181 + }, + { + "epoch": 0.5596826048293317, + "grad_norm": 1.763352045100841, + "learning_rate": 4.279156826852282e-06, + "loss": 0.3417, + "step": 8182 + }, + { + "epoch": 0.5597510089609412, + "grad_norm": 2.0044985633829726, + "learning_rate": 4.278060662304329e-06, + "loss": 0.4706, + "step": 8183 + }, + { + "epoch": 0.5598194130925508, + "grad_norm": 3.145125433659837, + "learning_rate": 4.276964533192644e-06, + "loss": 0.2101, + "step": 8184 + }, + { + "epoch": 0.5598878172241604, + "grad_norm": 2.5218302329248585, + "learning_rate": 4.275868439571033e-06, + "loss": 0.4755, + "step": 8185 + }, + { + "epoch": 0.5599562213557698, + "grad_norm": 1.700001968003985, + "learning_rate": 4.274772381493295e-06, + "loss": 0.1961, + "step": 8186 + }, + { + "epoch": 0.5600246254873794, + "grad_norm": 2.0779648984458445, + "learning_rate": 4.273676359013229e-06, + "loss": 0.2066, + "step": 8187 + }, + { + "epoch": 0.560093029618989, + "grad_norm": 2.2801635911774074, + "learning_rate": 4.272580372184636e-06, + "loss": 0.2685, + "step": 8188 + }, + { + "epoch": 0.5601614337505986, + "grad_norm": 2.10423833816259, + "learning_rate": 4.27148442106131e-06, + "loss": 0.4039, + "step": 8189 + }, + { + "epoch": 0.560229837882208, + "grad_norm": 1.984048238268311, + "learning_rate": 4.270388505697048e-06, + "loss": 0.211, + "step": 8190 + }, + { + "epoch": 0.5602982420138176, + "grad_norm": 2.2358910279206277, + "learning_rate": 4.2692926261456415e-06, + "loss": 0.4572, + "step": 8191 + }, + { + "epoch": 0.5603666461454272, + "grad_norm": 2.4080632548927188, + "learning_rate": 4.26819678246088e-06, + "loss": 0.73, + "step": 8192 + }, + { + "epoch": 0.5604350502770368, + "grad_norm": 1.978982393988814, + "learning_rate": 4.2671009746965556e-06, + "loss": 0.3319, + "step": 8193 + }, + { + "epoch": 0.5605034544086462, + "grad_norm": 1.9976688970912133, + "learning_rate": 4.2660052029064535e-06, + "loss": 0.543, + "step": 8194 + }, + { + "epoch": 0.5605718585402558, + "grad_norm": 2.2942098478827706, + "learning_rate": 4.264909467144362e-06, + "loss": 0.2148, + "step": 8195 + }, + { + "epoch": 0.5606402626718654, + "grad_norm": 1.9844412905216013, + "learning_rate": 4.2638137674640625e-06, + "loss": 0.3127, + "step": 8196 + }, + { + "epoch": 0.560708666803475, + "grad_norm": 1.4733962974572243, + "learning_rate": 4.262718103919337e-06, + "loss": 0.1474, + "step": 8197 + }, + { + "epoch": 0.5607770709350844, + "grad_norm": 2.289258004650599, + "learning_rate": 4.261622476563969e-06, + "loss": 0.4929, + "step": 8198 + }, + { + "epoch": 0.560845475066694, + "grad_norm": 2.471670113812146, + "learning_rate": 4.260526885451734e-06, + "loss": 0.4436, + "step": 8199 + }, + { + "epoch": 0.5609138791983036, + "grad_norm": 1.9758940233511941, + "learning_rate": 4.2594313306364125e-06, + "loss": 0.4893, + "step": 8200 + }, + { + "epoch": 0.5609822833299132, + "grad_norm": 1.989948034979821, + "learning_rate": 4.258335812171775e-06, + "loss": 0.1987, + "step": 8201 + }, + { + "epoch": 0.5610506874615226, + "grad_norm": 2.2147964257191646, + "learning_rate": 4.257240330111599e-06, + "loss": 0.4208, + "step": 8202 + }, + { + "epoch": 0.5611190915931322, + "grad_norm": 2.715811972552144, + "learning_rate": 4.256144884509655e-06, + "loss": 0.4401, + "step": 8203 + }, + { + "epoch": 0.5611874957247418, + "grad_norm": 2.072816723342336, + "learning_rate": 4.25504947541971e-06, + "loss": 0.2096, + "step": 8204 + }, + { + "epoch": 0.5612558998563513, + "grad_norm": 2.0030930934632303, + "learning_rate": 4.253954102895536e-06, + "loss": 0.3046, + "step": 8205 + }, + { + "epoch": 0.5613243039879608, + "grad_norm": 1.9622695509645809, + "learning_rate": 4.252858766990897e-06, + "loss": 0.4219, + "step": 8206 + }, + { + "epoch": 0.5613927081195704, + "grad_norm": 2.0490804431415315, + "learning_rate": 4.251763467759557e-06, + "loss": 0.363, + "step": 8207 + }, + { + "epoch": 0.56146111225118, + "grad_norm": 3.899854612697181, + "learning_rate": 4.2506682052552775e-06, + "loss": 0.3637, + "step": 8208 + }, + { + "epoch": 0.5615295163827895, + "grad_norm": 1.2411685699267134, + "learning_rate": 4.249572979531823e-06, + "loss": 0.205, + "step": 8209 + }, + { + "epoch": 0.561597920514399, + "grad_norm": 2.3752765180835764, + "learning_rate": 4.248477790642951e-06, + "loss": 0.6909, + "step": 8210 + }, + { + "epoch": 0.5616663246460086, + "grad_norm": 2.292227558651863, + "learning_rate": 4.247382638642416e-06, + "loss": 0.4609, + "step": 8211 + }, + { + "epoch": 0.5617347287776182, + "grad_norm": 1.8801521009312072, + "learning_rate": 4.246287523583978e-06, + "loss": 0.1499, + "step": 8212 + }, + { + "epoch": 0.5618031329092277, + "grad_norm": 2.1585606792345184, + "learning_rate": 4.245192445521387e-06, + "loss": 0.4621, + "step": 8213 + }, + { + "epoch": 0.5618715370408373, + "grad_norm": 2.207621483662352, + "learning_rate": 4.2440974045083935e-06, + "loss": 0.5609, + "step": 8214 + }, + { + "epoch": 0.5619399411724468, + "grad_norm": 1.9378069895909125, + "learning_rate": 4.2430024005987524e-06, + "loss": 0.4465, + "step": 8215 + }, + { + "epoch": 0.5620083453040564, + "grad_norm": 1.7337839562900714, + "learning_rate": 4.241907433846208e-06, + "loss": 0.318, + "step": 8216 + }, + { + "epoch": 0.5620767494356659, + "grad_norm": 2.777879429917562, + "learning_rate": 4.240812504304508e-06, + "loss": 0.6093, + "step": 8217 + }, + { + "epoch": 0.5621451535672755, + "grad_norm": 2.3631864859757465, + "learning_rate": 4.239717612027394e-06, + "loss": 0.3912, + "step": 8218 + }, + { + "epoch": 0.562213557698885, + "grad_norm": 2.3877599914153222, + "learning_rate": 4.238622757068614e-06, + "loss": 0.5806, + "step": 8219 + }, + { + "epoch": 0.5622819618304946, + "grad_norm": 2.270802665351707, + "learning_rate": 4.2375279394819045e-06, + "loss": 0.6254, + "step": 8220 + }, + { + "epoch": 0.5623503659621041, + "grad_norm": 2.2872579316220496, + "learning_rate": 4.2364331593210055e-06, + "loss": 0.5289, + "step": 8221 + }, + { + "epoch": 0.5624187700937137, + "grad_norm": 2.10682125919258, + "learning_rate": 4.235338416639655e-06, + "loss": 0.4285, + "step": 8222 + }, + { + "epoch": 0.5624871742253232, + "grad_norm": 1.8395087874685572, + "learning_rate": 4.234243711491588e-06, + "loss": 0.4812, + "step": 8223 + }, + { + "epoch": 0.5625555783569327, + "grad_norm": 1.258913704299779, + "learning_rate": 4.2331490439305355e-06, + "loss": 0.2196, + "step": 8224 + }, + { + "epoch": 0.5626239824885423, + "grad_norm": 1.7822931982673322, + "learning_rate": 4.232054414010233e-06, + "loss": 0.3553, + "step": 8225 + }, + { + "epoch": 0.5626923866201519, + "grad_norm": 1.4638735673700503, + "learning_rate": 4.230959821784408e-06, + "loss": 0.2201, + "step": 8226 + }, + { + "epoch": 0.5627607907517614, + "grad_norm": 2.117919570087327, + "learning_rate": 4.229865267306787e-06, + "loss": 0.3835, + "step": 8227 + }, + { + "epoch": 0.5628291948833709, + "grad_norm": 1.7115379742948915, + "learning_rate": 4.228770750631097e-06, + "loss": 0.2494, + "step": 8228 + }, + { + "epoch": 0.5628975990149805, + "grad_norm": 2.6930859034779666, + "learning_rate": 4.227676271811065e-06, + "loss": 0.5034, + "step": 8229 + }, + { + "epoch": 0.5629660031465901, + "grad_norm": 2.104453679062152, + "learning_rate": 4.22658183090041e-06, + "loss": 0.4816, + "step": 8230 + }, + { + "epoch": 0.5630344072781996, + "grad_norm": 1.8871647246480856, + "learning_rate": 4.225487427952854e-06, + "loss": 0.3744, + "step": 8231 + }, + { + "epoch": 0.5631028114098091, + "grad_norm": 2.967826769978155, + "learning_rate": 4.224393063022115e-06, + "loss": 0.4111, + "step": 8232 + }, + { + "epoch": 0.5631712155414187, + "grad_norm": 1.6808894691725362, + "learning_rate": 4.223298736161911e-06, + "loss": 0.3474, + "step": 8233 + }, + { + "epoch": 0.5632396196730283, + "grad_norm": 2.1518082411196864, + "learning_rate": 4.222204447425954e-06, + "loss": 0.4018, + "step": 8234 + }, + { + "epoch": 0.5633080238046378, + "grad_norm": 1.847832322552226, + "learning_rate": 4.22111019686796e-06, + "loss": 0.3869, + "step": 8235 + }, + { + "epoch": 0.5633764279362473, + "grad_norm": 1.9716271350464252, + "learning_rate": 4.220015984541638e-06, + "loss": 0.4435, + "step": 8236 + }, + { + "epoch": 0.5634448320678569, + "grad_norm": 2.4648102940920156, + "learning_rate": 4.218921810500698e-06, + "loss": 0.4557, + "step": 8237 + }, + { + "epoch": 0.5635132361994665, + "grad_norm": 2.316908629106213, + "learning_rate": 4.217827674798845e-06, + "loss": 0.5139, + "step": 8238 + }, + { + "epoch": 0.563581640331076, + "grad_norm": 1.811301381775668, + "learning_rate": 4.21673357748979e-06, + "loss": 0.337, + "step": 8239 + }, + { + "epoch": 0.5636500444626855, + "grad_norm": 1.9349036201604455, + "learning_rate": 4.2156395186272325e-06, + "loss": 0.425, + "step": 8240 + }, + { + "epoch": 0.5637184485942951, + "grad_norm": 2.738259466413847, + "learning_rate": 4.2145454982648745e-06, + "loss": 0.2138, + "step": 8241 + }, + { + "epoch": 0.5637868527259047, + "grad_norm": 1.5082549202103857, + "learning_rate": 4.213451516456418e-06, + "loss": 0.1773, + "step": 8242 + }, + { + "epoch": 0.5638552568575141, + "grad_norm": 1.9685786490576869, + "learning_rate": 4.212357573255558e-06, + "loss": 0.4479, + "step": 8243 + }, + { + "epoch": 0.5639236609891237, + "grad_norm": 1.6811398902629024, + "learning_rate": 4.211263668715992e-06, + "loss": 0.3332, + "step": 8244 + }, + { + "epoch": 0.5639920651207333, + "grad_norm": 1.8214936987736225, + "learning_rate": 4.210169802891414e-06, + "loss": 0.2314, + "step": 8245 + }, + { + "epoch": 0.5640604692523429, + "grad_norm": 2.6170240436582053, + "learning_rate": 4.209075975835517e-06, + "loss": 0.7385, + "step": 8246 + }, + { + "epoch": 0.5641288733839523, + "grad_norm": 2.3274156270576016, + "learning_rate": 4.207982187601988e-06, + "loss": 0.5874, + "step": 8247 + }, + { + "epoch": 0.5641972775155619, + "grad_norm": 2.3263630360065153, + "learning_rate": 4.206888438244518e-06, + "loss": 0.3384, + "step": 8248 + }, + { + "epoch": 0.5642656816471715, + "grad_norm": 2.606346462943701, + "learning_rate": 4.205794727816795e-06, + "loss": 0.5959, + "step": 8249 + }, + { + "epoch": 0.5643340857787811, + "grad_norm": 1.910843906777457, + "learning_rate": 4.204701056372502e-06, + "loss": 0.4981, + "step": 8250 + }, + { + "epoch": 0.5644024899103905, + "grad_norm": 2.45275107042536, + "learning_rate": 4.203607423965321e-06, + "loss": 0.3045, + "step": 8251 + }, + { + "epoch": 0.5644708940420001, + "grad_norm": 1.8120648091833427, + "learning_rate": 4.202513830648934e-06, + "loss": 0.4289, + "step": 8252 + }, + { + "epoch": 0.5645392981736097, + "grad_norm": 2.7492635998071684, + "learning_rate": 4.20142027647702e-06, + "loss": 0.7277, + "step": 8253 + }, + { + "epoch": 0.5646077023052193, + "grad_norm": 2.308053897534116, + "learning_rate": 4.200326761503253e-06, + "loss": 0.4443, + "step": 8254 + }, + { + "epoch": 0.5646761064368288, + "grad_norm": 1.7513673838716024, + "learning_rate": 4.1992332857813116e-06, + "loss": 0.3068, + "step": 8255 + }, + { + "epoch": 0.5647445105684383, + "grad_norm": 1.3757659474766861, + "learning_rate": 4.198139849364868e-06, + "loss": 0.197, + "step": 8256 + }, + { + "epoch": 0.5648129147000479, + "grad_norm": 1.6585328501060639, + "learning_rate": 4.197046452307589e-06, + "loss": 0.1981, + "step": 8257 + }, + { + "epoch": 0.5648813188316575, + "grad_norm": 1.9278774923362765, + "learning_rate": 4.195953094663149e-06, + "loss": 0.1798, + "step": 8258 + }, + { + "epoch": 0.564949722963267, + "grad_norm": 2.7573494779153656, + "learning_rate": 4.194859776485216e-06, + "loss": 0.3988, + "step": 8259 + }, + { + "epoch": 0.5650181270948765, + "grad_norm": 1.766351942621998, + "learning_rate": 4.193766497827452e-06, + "loss": 0.4146, + "step": 8260 + }, + { + "epoch": 0.5650865312264861, + "grad_norm": 2.139086828518401, + "learning_rate": 4.192673258743521e-06, + "loss": 0.5337, + "step": 8261 + }, + { + "epoch": 0.5651549353580956, + "grad_norm": 1.645600553440375, + "learning_rate": 4.191580059287087e-06, + "loss": 0.256, + "step": 8262 + }, + { + "epoch": 0.5652233394897052, + "grad_norm": 2.1152394577345714, + "learning_rate": 4.190486899511806e-06, + "loss": 0.3679, + "step": 8263 + }, + { + "epoch": 0.5652917436213147, + "grad_norm": 2.0231325368944124, + "learning_rate": 4.189393779471337e-06, + "loss": 0.2855, + "step": 8264 + }, + { + "epoch": 0.5653601477529243, + "grad_norm": 2.043836012030683, + "learning_rate": 4.188300699219336e-06, + "loss": 0.3525, + "step": 8265 + }, + { + "epoch": 0.5654285518845338, + "grad_norm": 2.6155017092233748, + "learning_rate": 4.187207658809458e-06, + "loss": 0.3425, + "step": 8266 + }, + { + "epoch": 0.5654969560161434, + "grad_norm": 2.264799980695961, + "learning_rate": 4.186114658295349e-06, + "loss": 0.2708, + "step": 8267 + }, + { + "epoch": 0.5655653601477529, + "grad_norm": 1.7998539156709052, + "learning_rate": 4.185021697730664e-06, + "loss": 0.3516, + "step": 8268 + }, + { + "epoch": 0.5656337642793625, + "grad_norm": 2.2894792730676583, + "learning_rate": 4.183928777169052e-06, + "loss": 0.5546, + "step": 8269 + }, + { + "epoch": 0.565702168410972, + "grad_norm": 2.277998933064259, + "learning_rate": 4.182835896664155e-06, + "loss": 0.5154, + "step": 8270 + }, + { + "epoch": 0.5657705725425816, + "grad_norm": 2.359905672708841, + "learning_rate": 4.181743056269618e-06, + "loss": 0.6368, + "step": 8271 + }, + { + "epoch": 0.5658389766741911, + "grad_norm": 2.096258326327962, + "learning_rate": 4.180650256039083e-06, + "loss": 0.3398, + "step": 8272 + }, + { + "epoch": 0.5659073808058007, + "grad_norm": 1.5643746693306366, + "learning_rate": 4.179557496026192e-06, + "loss": 0.2677, + "step": 8273 + }, + { + "epoch": 0.5659757849374102, + "grad_norm": 1.6171794098730026, + "learning_rate": 4.1784647762845785e-06, + "loss": 0.3046, + "step": 8274 + }, + { + "epoch": 0.5660441890690198, + "grad_norm": 1.792917940670639, + "learning_rate": 4.177372096867882e-06, + "loss": 0.3853, + "step": 8275 + }, + { + "epoch": 0.5661125932006293, + "grad_norm": 1.9321204535487648, + "learning_rate": 4.176279457829735e-06, + "loss": 0.311, + "step": 8276 + }, + { + "epoch": 0.5661809973322389, + "grad_norm": 1.5798604553579942, + "learning_rate": 4.175186859223769e-06, + "loss": 0.3763, + "step": 8277 + }, + { + "epoch": 0.5662494014638484, + "grad_norm": 2.025857637982865, + "learning_rate": 4.174094301103616e-06, + "loss": 0.4748, + "step": 8278 + }, + { + "epoch": 0.566317805595458, + "grad_norm": 2.6077162158578027, + "learning_rate": 4.173001783522903e-06, + "loss": 0.3979, + "step": 8279 + }, + { + "epoch": 0.5663862097270675, + "grad_norm": 2.511782637950855, + "learning_rate": 4.1719093065352565e-06, + "loss": 0.5862, + "step": 8280 + }, + { + "epoch": 0.566454613858677, + "grad_norm": 2.083937325344222, + "learning_rate": 4.1708168701942996e-06, + "loss": 0.4897, + "step": 8281 + }, + { + "epoch": 0.5665230179902866, + "grad_norm": 1.9927831544670038, + "learning_rate": 4.169724474553654e-06, + "loss": 0.225, + "step": 8282 + }, + { + "epoch": 0.5665914221218962, + "grad_norm": 1.8631448035993623, + "learning_rate": 4.168632119666943e-06, + "loss": 0.3845, + "step": 8283 + }, + { + "epoch": 0.5666598262535057, + "grad_norm": 2.9723423254449512, + "learning_rate": 4.16753980558778e-06, + "loss": 0.4746, + "step": 8284 + }, + { + "epoch": 0.5667282303851152, + "grad_norm": 1.6124249603494811, + "learning_rate": 4.166447532369784e-06, + "loss": 0.2561, + "step": 8285 + }, + { + "epoch": 0.5667966345167248, + "grad_norm": 2.9998177135241093, + "learning_rate": 4.1653553000665675e-06, + "loss": 0.2919, + "step": 8286 + }, + { + "epoch": 0.5668650386483344, + "grad_norm": 1.8857559132483908, + "learning_rate": 4.164263108731743e-06, + "loss": 0.3733, + "step": 8287 + }, + { + "epoch": 0.566933442779944, + "grad_norm": 1.8215580186839468, + "learning_rate": 4.1631709584189214e-06, + "loss": 0.2702, + "step": 8288 + }, + { + "epoch": 0.5670018469115534, + "grad_norm": 2.0469469174279697, + "learning_rate": 4.162078849181711e-06, + "loss": 0.4103, + "step": 8289 + }, + { + "epoch": 0.567070251043163, + "grad_norm": 2.083068682232353, + "learning_rate": 4.160986781073717e-06, + "loss": 0.4714, + "step": 8290 + }, + { + "epoch": 0.5671386551747726, + "grad_norm": 2.479670833298475, + "learning_rate": 4.1598947541485425e-06, + "loss": 0.2935, + "step": 8291 + }, + { + "epoch": 0.5672070593063822, + "grad_norm": 2.0254660273231053, + "learning_rate": 4.158802768459791e-06, + "loss": 0.2851, + "step": 8292 + }, + { + "epoch": 0.5672754634379916, + "grad_norm": 1.5681950826328925, + "learning_rate": 4.157710824061061e-06, + "loss": 0.3019, + "step": 8293 + }, + { + "epoch": 0.5673438675696012, + "grad_norm": 2.06817773844529, + "learning_rate": 4.156618921005952e-06, + "loss": 0.5162, + "step": 8294 + }, + { + "epoch": 0.5674122717012108, + "grad_norm": 2.27061657060591, + "learning_rate": 4.155527059348059e-06, + "loss": 0.4532, + "step": 8295 + }, + { + "epoch": 0.5674806758328202, + "grad_norm": 2.5240837249959163, + "learning_rate": 4.154435239140974e-06, + "loss": 0.3545, + "step": 8296 + }, + { + "epoch": 0.5675490799644298, + "grad_norm": 2.0867772787177348, + "learning_rate": 4.1533434604382915e-06, + "loss": 0.4271, + "step": 8297 + }, + { + "epoch": 0.5676174840960394, + "grad_norm": 1.9736206152462137, + "learning_rate": 4.152251723293601e-06, + "loss": 0.4709, + "step": 8298 + }, + { + "epoch": 0.567685888227649, + "grad_norm": 2.624595741523844, + "learning_rate": 4.15116002776049e-06, + "loss": 0.9253, + "step": 8299 + }, + { + "epoch": 0.5677542923592585, + "grad_norm": 1.7883403686643988, + "learning_rate": 4.150068373892542e-06, + "loss": 0.2392, + "step": 8300 + }, + { + "epoch": 0.567822696490868, + "grad_norm": 1.2456979332208924, + "learning_rate": 4.1489767617433446e-06, + "loss": 0.1194, + "step": 8301 + }, + { + "epoch": 0.5678911006224776, + "grad_norm": 1.8313884766563007, + "learning_rate": 4.147885191366477e-06, + "loss": 0.2402, + "step": 8302 + }, + { + "epoch": 0.5679595047540872, + "grad_norm": 2.660795986312371, + "learning_rate": 4.146793662815518e-06, + "loss": 0.2602, + "step": 8303 + }, + { + "epoch": 0.5680279088856967, + "grad_norm": 2.6201639587117875, + "learning_rate": 4.1457021761440475e-06, + "loss": 0.4159, + "step": 8304 + }, + { + "epoch": 0.5680963130173062, + "grad_norm": 1.8430776825034723, + "learning_rate": 4.144610731405639e-06, + "loss": 0.337, + "step": 8305 + }, + { + "epoch": 0.5681647171489158, + "grad_norm": 1.8394326562688406, + "learning_rate": 4.143519328653866e-06, + "loss": 0.3853, + "step": 8306 + }, + { + "epoch": 0.5682331212805254, + "grad_norm": 2.107525350100325, + "learning_rate": 4.142427967942299e-06, + "loss": 0.3955, + "step": 8307 + }, + { + "epoch": 0.5683015254121349, + "grad_norm": 2.3926348371985595, + "learning_rate": 4.14133664932451e-06, + "loss": 0.6849, + "step": 8308 + }, + { + "epoch": 0.5683699295437444, + "grad_norm": 2.0605570788379426, + "learning_rate": 4.1402453728540655e-06, + "loss": 0.4473, + "step": 8309 + }, + { + "epoch": 0.568438333675354, + "grad_norm": 1.9373150629565103, + "learning_rate": 4.139154138584528e-06, + "loss": 0.4341, + "step": 8310 + }, + { + "epoch": 0.5685067378069636, + "grad_norm": 1.7681777485004202, + "learning_rate": 4.138062946569465e-06, + "loss": 0.3328, + "step": 8311 + }, + { + "epoch": 0.5685751419385731, + "grad_norm": 2.3605049029830436, + "learning_rate": 4.136971796862433e-06, + "loss": 0.4806, + "step": 8312 + }, + { + "epoch": 0.5686435460701826, + "grad_norm": 2.1038160330959554, + "learning_rate": 4.135880689516992e-06, + "loss": 0.46, + "step": 8313 + }, + { + "epoch": 0.5687119502017922, + "grad_norm": 1.8574654687364636, + "learning_rate": 4.134789624586701e-06, + "loss": 0.3341, + "step": 8314 + }, + { + "epoch": 0.5687803543334017, + "grad_norm": 2.2100871981813093, + "learning_rate": 4.133698602125113e-06, + "loss": 0.4723, + "step": 8315 + }, + { + "epoch": 0.5688487584650113, + "grad_norm": 2.6050166319720693, + "learning_rate": 4.1326076221857805e-06, + "loss": 0.5285, + "step": 8316 + }, + { + "epoch": 0.5689171625966208, + "grad_norm": 2.628036740666029, + "learning_rate": 4.131516684822252e-06, + "loss": 0.5089, + "step": 8317 + }, + { + "epoch": 0.5689855667282304, + "grad_norm": 2.0014608047641893, + "learning_rate": 4.130425790088082e-06, + "loss": 0.4386, + "step": 8318 + }, + { + "epoch": 0.5690539708598399, + "grad_norm": 1.7008294725568962, + "learning_rate": 4.129334938036813e-06, + "loss": 0.328, + "step": 8319 + }, + { + "epoch": 0.5691223749914495, + "grad_norm": 1.9463417402106746, + "learning_rate": 4.1282441287219886e-06, + "loss": 0.3112, + "step": 8320 + }, + { + "epoch": 0.569190779123059, + "grad_norm": 2.2145208564072445, + "learning_rate": 4.127153362197153e-06, + "loss": 0.3932, + "step": 8321 + }, + { + "epoch": 0.5692591832546686, + "grad_norm": 5.57270472708821, + "learning_rate": 4.126062638515846e-06, + "loss": 0.3784, + "step": 8322 + }, + { + "epoch": 0.5693275873862781, + "grad_norm": 1.882014901045449, + "learning_rate": 4.124971957731603e-06, + "loss": 0.442, + "step": 8323 + }, + { + "epoch": 0.5693959915178877, + "grad_norm": 1.8136282566206168, + "learning_rate": 4.1238813198979635e-06, + "loss": 0.2395, + "step": 8324 + }, + { + "epoch": 0.5694643956494972, + "grad_norm": 2.761911145405584, + "learning_rate": 4.122790725068459e-06, + "loss": 0.2429, + "step": 8325 + }, + { + "epoch": 0.5695327997811068, + "grad_norm": 2.0584201558588284, + "learning_rate": 4.12170017329662e-06, + "loss": 0.4053, + "step": 8326 + }, + { + "epoch": 0.5696012039127163, + "grad_norm": 2.2754277827949445, + "learning_rate": 4.120609664635978e-06, + "loss": 0.1648, + "step": 8327 + }, + { + "epoch": 0.5696696080443259, + "grad_norm": 2.8061507241460304, + "learning_rate": 4.119519199140062e-06, + "loss": 0.4756, + "step": 8328 + }, + { + "epoch": 0.5697380121759354, + "grad_norm": 2.1550104545131683, + "learning_rate": 4.118428776862395e-06, + "loss": 0.4546, + "step": 8329 + }, + { + "epoch": 0.569806416307545, + "grad_norm": 2.144664708138273, + "learning_rate": 4.117338397856499e-06, + "loss": 0.5423, + "step": 8330 + }, + { + "epoch": 0.5698748204391545, + "grad_norm": 2.9156449447064148, + "learning_rate": 4.1162480621759e-06, + "loss": 0.4667, + "step": 8331 + }, + { + "epoch": 0.5699432245707641, + "grad_norm": 2.9732032002013673, + "learning_rate": 4.115157769874112e-06, + "loss": 0.3893, + "step": 8332 + }, + { + "epoch": 0.5700116287023737, + "grad_norm": 2.098640154314934, + "learning_rate": 4.114067521004652e-06, + "loss": 0.4753, + "step": 8333 + }, + { + "epoch": 0.5700800328339831, + "grad_norm": 1.90424929828388, + "learning_rate": 4.112977315621039e-06, + "loss": 0.4118, + "step": 8334 + }, + { + "epoch": 0.5701484369655927, + "grad_norm": 1.7867126566487348, + "learning_rate": 4.1118871537767816e-06, + "loss": 0.381, + "step": 8335 + }, + { + "epoch": 0.5702168410972023, + "grad_norm": 2.1835093200411313, + "learning_rate": 4.110797035525391e-06, + "loss": 0.4406, + "step": 8336 + }, + { + "epoch": 0.5702852452288119, + "grad_norm": 1.6220314163606258, + "learning_rate": 4.109706960920375e-06, + "loss": 0.301, + "step": 8337 + }, + { + "epoch": 0.5703536493604213, + "grad_norm": 2.2725736477797374, + "learning_rate": 4.108616930015241e-06, + "loss": 0.5787, + "step": 8338 + }, + { + "epoch": 0.5704220534920309, + "grad_norm": 2.237339094439403, + "learning_rate": 4.107526942863493e-06, + "loss": 0.4466, + "step": 8339 + }, + { + "epoch": 0.5704904576236405, + "grad_norm": 1.9783064871964107, + "learning_rate": 4.106436999518633e-06, + "loss": 0.4342, + "step": 8340 + }, + { + "epoch": 0.5705588617552501, + "grad_norm": 1.3168366970761536, + "learning_rate": 4.10534710003416e-06, + "loss": 0.2461, + "step": 8341 + }, + { + "epoch": 0.5706272658868595, + "grad_norm": 2.2586239607224474, + "learning_rate": 4.104257244463571e-06, + "loss": 0.2043, + "step": 8342 + }, + { + "epoch": 0.5706956700184691, + "grad_norm": 1.649094504912737, + "learning_rate": 4.103167432860361e-06, + "loss": 0.3369, + "step": 8343 + }, + { + "epoch": 0.5707640741500787, + "grad_norm": 2.0151608989975776, + "learning_rate": 4.102077665278026e-06, + "loss": 0.3691, + "step": 8344 + }, + { + "epoch": 0.5708324782816883, + "grad_norm": 1.703515406142555, + "learning_rate": 4.100987941770054e-06, + "loss": 0.4172, + "step": 8345 + }, + { + "epoch": 0.5709008824132977, + "grad_norm": 2.1435912179115144, + "learning_rate": 4.099898262389934e-06, + "loss": 0.4795, + "step": 8346 + }, + { + "epoch": 0.5709692865449073, + "grad_norm": 1.8277484065750558, + "learning_rate": 4.098808627191155e-06, + "loss": 0.3498, + "step": 8347 + }, + { + "epoch": 0.5710376906765169, + "grad_norm": 2.233831132707928, + "learning_rate": 4.097719036227201e-06, + "loss": 0.6474, + "step": 8348 + }, + { + "epoch": 0.5711060948081265, + "grad_norm": 1.726530303482271, + "learning_rate": 4.0966294895515536e-06, + "loss": 0.2961, + "step": 8349 + }, + { + "epoch": 0.5711744989397359, + "grad_norm": 1.3043255018238515, + "learning_rate": 4.095539987217693e-06, + "loss": 0.1181, + "step": 8350 + }, + { + "epoch": 0.5712429030713455, + "grad_norm": 1.1389693612610352, + "learning_rate": 4.094450529279097e-06, + "loss": 0.11, + "step": 8351 + }, + { + "epoch": 0.5713113072029551, + "grad_norm": 1.648328438536387, + "learning_rate": 4.0933611157892426e-06, + "loss": 0.3411, + "step": 8352 + }, + { + "epoch": 0.5713797113345646, + "grad_norm": 3.0988747507999785, + "learning_rate": 4.092271746801602e-06, + "loss": 0.5914, + "step": 8353 + }, + { + "epoch": 0.5714481154661741, + "grad_norm": 1.606866430028276, + "learning_rate": 4.0911824223696484e-06, + "loss": 0.2465, + "step": 8354 + }, + { + "epoch": 0.5715165195977837, + "grad_norm": 2.0738738109035, + "learning_rate": 4.09009314254685e-06, + "loss": 0.5167, + "step": 8355 + }, + { + "epoch": 0.5715849237293933, + "grad_norm": 3.0596436005650456, + "learning_rate": 4.089003907386672e-06, + "loss": 0.6516, + "step": 8356 + }, + { + "epoch": 0.5716533278610028, + "grad_norm": 1.372171483772787, + "learning_rate": 4.087914716942582e-06, + "loss": 0.1596, + "step": 8357 + }, + { + "epoch": 0.5717217319926123, + "grad_norm": 1.612090188354312, + "learning_rate": 4.0868255712680445e-06, + "loss": 0.2572, + "step": 8358 + }, + { + "epoch": 0.5717901361242219, + "grad_norm": 2.0956600927821465, + "learning_rate": 4.085736470416517e-06, + "loss": 0.2291, + "step": 8359 + }, + { + "epoch": 0.5718585402558315, + "grad_norm": 2.063590231209559, + "learning_rate": 4.084647414441457e-06, + "loss": 0.3806, + "step": 8360 + }, + { + "epoch": 0.571926944387441, + "grad_norm": 1.4655163180971782, + "learning_rate": 4.083558403396323e-06, + "loss": 0.14, + "step": 8361 + }, + { + "epoch": 0.5719953485190505, + "grad_norm": 1.8965387974953634, + "learning_rate": 4.082469437334569e-06, + "loss": 0.3917, + "step": 8362 + }, + { + "epoch": 0.5720637526506601, + "grad_norm": 2.0694313099347963, + "learning_rate": 4.0813805163096435e-06, + "loss": 0.3235, + "step": 8363 + }, + { + "epoch": 0.5721321567822697, + "grad_norm": 2.109351995130766, + "learning_rate": 4.080291640374999e-06, + "loss": 0.1342, + "step": 8364 + }, + { + "epoch": 0.5722005609138792, + "grad_norm": 1.9891941021027528, + "learning_rate": 4.079202809584083e-06, + "loss": 0.5286, + "step": 8365 + }, + { + "epoch": 0.5722689650454887, + "grad_norm": 2.2321074590600993, + "learning_rate": 4.078114023990338e-06, + "loss": 0.3464, + "step": 8366 + }, + { + "epoch": 0.5723373691770983, + "grad_norm": 1.8845696777317378, + "learning_rate": 4.077025283647208e-06, + "loss": 0.3498, + "step": 8367 + }, + { + "epoch": 0.5724057733087079, + "grad_norm": 2.2180057136919538, + "learning_rate": 4.075936588608136e-06, + "loss": 0.582, + "step": 8368 + }, + { + "epoch": 0.5724741774403174, + "grad_norm": 1.4753604744437983, + "learning_rate": 4.074847938926558e-06, + "loss": 0.2074, + "step": 8369 + }, + { + "epoch": 0.572542581571927, + "grad_norm": 1.7615778670047584, + "learning_rate": 4.073759334655911e-06, + "loss": 0.3332, + "step": 8370 + }, + { + "epoch": 0.5726109857035365, + "grad_norm": 2.283737250669875, + "learning_rate": 4.072670775849628e-06, + "loss": 0.6205, + "step": 8371 + }, + { + "epoch": 0.572679389835146, + "grad_norm": 1.7402384465744734, + "learning_rate": 4.071582262561142e-06, + "loss": 0.2024, + "step": 8372 + }, + { + "epoch": 0.5727477939667556, + "grad_norm": 1.867921457525601, + "learning_rate": 4.070493794843881e-06, + "loss": 0.3451, + "step": 8373 + }, + { + "epoch": 0.5728161980983651, + "grad_norm": 1.8758566488915276, + "learning_rate": 4.069405372751274e-06, + "loss": 0.3492, + "step": 8374 + }, + { + "epoch": 0.5728846022299747, + "grad_norm": 2.4442499003125984, + "learning_rate": 4.0683169963367455e-06, + "loss": 0.3737, + "step": 8375 + }, + { + "epoch": 0.5729530063615842, + "grad_norm": 2.87240406679626, + "learning_rate": 4.067228665653715e-06, + "loss": 0.2084, + "step": 8376 + }, + { + "epoch": 0.5730214104931938, + "grad_norm": 1.7974498782843062, + "learning_rate": 4.066140380755607e-06, + "loss": 0.3457, + "step": 8377 + }, + { + "epoch": 0.5730898146248034, + "grad_norm": 1.620957182681427, + "learning_rate": 4.06505214169584e-06, + "loss": 0.2131, + "step": 8378 + }, + { + "epoch": 0.5731582187564129, + "grad_norm": 2.523017432027304, + "learning_rate": 4.063963948527829e-06, + "loss": 0.5858, + "step": 8379 + }, + { + "epoch": 0.5732266228880224, + "grad_norm": 1.8409453250923975, + "learning_rate": 4.062875801304986e-06, + "loss": 0.2876, + "step": 8380 + }, + { + "epoch": 0.573295027019632, + "grad_norm": 2.33494698693014, + "learning_rate": 4.061787700080725e-06, + "loss": 0.518, + "step": 8381 + }, + { + "epoch": 0.5733634311512416, + "grad_norm": 1.8094496751110623, + "learning_rate": 4.060699644908455e-06, + "loss": 0.4241, + "step": 8382 + }, + { + "epoch": 0.5734318352828511, + "grad_norm": 1.795286940516374, + "learning_rate": 4.059611635841581e-06, + "loss": 0.2111, + "step": 8383 + }, + { + "epoch": 0.5735002394144606, + "grad_norm": 1.9062574730398312, + "learning_rate": 4.05852367293351e-06, + "loss": 0.4241, + "step": 8384 + }, + { + "epoch": 0.5735686435460702, + "grad_norm": 2.176938422109092, + "learning_rate": 4.057435756237641e-06, + "loss": 0.5658, + "step": 8385 + }, + { + "epoch": 0.5736370476776798, + "grad_norm": 2.127467276341687, + "learning_rate": 4.056347885807378e-06, + "loss": 0.5816, + "step": 8386 + }, + { + "epoch": 0.5737054518092893, + "grad_norm": 1.8418899100016979, + "learning_rate": 4.055260061696118e-06, + "loss": 0.3151, + "step": 8387 + }, + { + "epoch": 0.5737738559408988, + "grad_norm": 1.9063925767953336, + "learning_rate": 4.054172283957256e-06, + "loss": 0.4984, + "step": 8388 + }, + { + "epoch": 0.5738422600725084, + "grad_norm": 2.5979993780001744, + "learning_rate": 4.053084552644186e-06, + "loss": 0.4138, + "step": 8389 + }, + { + "epoch": 0.573910664204118, + "grad_norm": 2.1196804079918237, + "learning_rate": 4.051996867810299e-06, + "loss": 0.3485, + "step": 8390 + }, + { + "epoch": 0.5739790683357274, + "grad_norm": 3.0856010543329893, + "learning_rate": 4.050909229508984e-06, + "loss": 0.5188, + "step": 8391 + }, + { + "epoch": 0.574047472467337, + "grad_norm": 2.3873232124444397, + "learning_rate": 4.049821637793625e-06, + "loss": 0.4469, + "step": 8392 + }, + { + "epoch": 0.5741158765989466, + "grad_norm": 2.6334447073371012, + "learning_rate": 4.048734092717611e-06, + "loss": 0.6275, + "step": 8393 + }, + { + "epoch": 0.5741842807305562, + "grad_norm": 1.6217728674525864, + "learning_rate": 4.04764659433432e-06, + "loss": 0.2909, + "step": 8394 + }, + { + "epoch": 0.5742526848621656, + "grad_norm": 2.1071357426742634, + "learning_rate": 4.046559142697133e-06, + "loss": 0.3366, + "step": 8395 + }, + { + "epoch": 0.5743210889937752, + "grad_norm": 1.7824823441573991, + "learning_rate": 4.045471737859427e-06, + "loss": 0.3451, + "step": 8396 + }, + { + "epoch": 0.5743894931253848, + "grad_norm": 2.2974664289971023, + "learning_rate": 4.0443843798745776e-06, + "loss": 0.4627, + "step": 8397 + }, + { + "epoch": 0.5744578972569944, + "grad_norm": 1.7767541683009107, + "learning_rate": 4.043297068795958e-06, + "loss": 0.3882, + "step": 8398 + }, + { + "epoch": 0.5745263013886038, + "grad_norm": 2.1583224785714674, + "learning_rate": 4.042209804676937e-06, + "loss": 0.5943, + "step": 8399 + }, + { + "epoch": 0.5745947055202134, + "grad_norm": 2.5661782638103756, + "learning_rate": 4.041122587570886e-06, + "loss": 0.4655, + "step": 8400 + }, + { + "epoch": 0.574663109651823, + "grad_norm": 2.0108373099787196, + "learning_rate": 4.040035417531168e-06, + "loss": 0.4189, + "step": 8401 + }, + { + "epoch": 0.5747315137834326, + "grad_norm": 2.0386018196110562, + "learning_rate": 4.038948294611147e-06, + "loss": 0.4743, + "step": 8402 + }, + { + "epoch": 0.574799917915042, + "grad_norm": 1.6152226448786324, + "learning_rate": 4.037861218864185e-06, + "loss": 0.3069, + "step": 8403 + }, + { + "epoch": 0.5748683220466516, + "grad_norm": 2.0636214761899754, + "learning_rate": 4.03677419034364e-06, + "loss": 0.5936, + "step": 8404 + }, + { + "epoch": 0.5749367261782612, + "grad_norm": 2.6164739461611632, + "learning_rate": 4.035687209102869e-06, + "loss": 0.4474, + "step": 8405 + }, + { + "epoch": 0.5750051303098708, + "grad_norm": 3.7043264307204478, + "learning_rate": 4.034600275195225e-06, + "loss": 0.4987, + "step": 8406 + }, + { + "epoch": 0.5750735344414802, + "grad_norm": 1.5765568618749213, + "learning_rate": 4.033513388674063e-06, + "loss": 0.2836, + "step": 8407 + }, + { + "epoch": 0.5751419385730898, + "grad_norm": 2.2421858651291475, + "learning_rate": 4.032426549592732e-06, + "loss": 0.5644, + "step": 8408 + }, + { + "epoch": 0.5752103427046994, + "grad_norm": 1.564695653201811, + "learning_rate": 4.031339758004577e-06, + "loss": 0.199, + "step": 8409 + }, + { + "epoch": 0.5752787468363089, + "grad_norm": 2.186045524158478, + "learning_rate": 4.030253013962944e-06, + "loss": 0.6664, + "step": 8410 + }, + { + "epoch": 0.5753471509679184, + "grad_norm": 2.4582616699948088, + "learning_rate": 4.029166317521177e-06, + "loss": 0.3485, + "step": 8411 + }, + { + "epoch": 0.575415555099528, + "grad_norm": 2.4422967370143955, + "learning_rate": 4.028079668732615e-06, + "loss": 0.5774, + "step": 8412 + }, + { + "epoch": 0.5754839592311376, + "grad_norm": 1.806405178914785, + "learning_rate": 4.026993067650596e-06, + "loss": 0.2845, + "step": 8413 + }, + { + "epoch": 0.5755523633627471, + "grad_norm": 2.1391898655782073, + "learning_rate": 4.025906514328456e-06, + "loss": 0.3522, + "step": 8414 + }, + { + "epoch": 0.5756207674943566, + "grad_norm": 1.8983506076848446, + "learning_rate": 4.024820008819528e-06, + "loss": 0.5231, + "step": 8415 + }, + { + "epoch": 0.5756891716259662, + "grad_norm": 2.306496624566004, + "learning_rate": 4.023733551177141e-06, + "loss": 0.5965, + "step": 8416 + }, + { + "epoch": 0.5757575757575758, + "grad_norm": 2.0991491265742184, + "learning_rate": 4.022647141454628e-06, + "loss": 0.6603, + "step": 8417 + }, + { + "epoch": 0.5758259798891853, + "grad_norm": 1.8400449865480875, + "learning_rate": 4.021560779705313e-06, + "loss": 0.3851, + "step": 8418 + }, + { + "epoch": 0.5758943840207948, + "grad_norm": 1.8919625595102643, + "learning_rate": 4.020474465982519e-06, + "loss": 0.3303, + "step": 8419 + }, + { + "epoch": 0.5759627881524044, + "grad_norm": 2.428987230415355, + "learning_rate": 4.019388200339569e-06, + "loss": 0.3298, + "step": 8420 + }, + { + "epoch": 0.576031192284014, + "grad_norm": 2.0242476406413448, + "learning_rate": 4.018301982829781e-06, + "loss": 0.2626, + "step": 8421 + }, + { + "epoch": 0.5760995964156235, + "grad_norm": 3.582699911924669, + "learning_rate": 4.017215813506472e-06, + "loss": 0.5152, + "step": 8422 + }, + { + "epoch": 0.576168000547233, + "grad_norm": 2.1415530998149475, + "learning_rate": 4.0161296924229575e-06, + "loss": 0.2586, + "step": 8423 + }, + { + "epoch": 0.5762364046788426, + "grad_norm": 2.536332902150346, + "learning_rate": 4.0150436196325486e-06, + "loss": 0.5105, + "step": 8424 + }, + { + "epoch": 0.5763048088104521, + "grad_norm": 1.7266917245068738, + "learning_rate": 4.013957595188554e-06, + "loss": 0.5089, + "step": 8425 + }, + { + "epoch": 0.5763732129420617, + "grad_norm": 2.4985356870422932, + "learning_rate": 4.012871619144282e-06, + "loss": 0.5548, + "step": 8426 + }, + { + "epoch": 0.5764416170736713, + "grad_norm": 2.086223075548966, + "learning_rate": 4.011785691553039e-06, + "loss": 0.2902, + "step": 8427 + }, + { + "epoch": 0.5765100212052808, + "grad_norm": 1.7144290796359392, + "learning_rate": 4.0106998124681265e-06, + "loss": 0.3762, + "step": 8428 + }, + { + "epoch": 0.5765784253368903, + "grad_norm": 2.16066408310142, + "learning_rate": 4.009613981942843e-06, + "loss": 0.668, + "step": 8429 + }, + { + "epoch": 0.5766468294684999, + "grad_norm": 1.8993793945286768, + "learning_rate": 4.0085282000304894e-06, + "loss": 0.5833, + "step": 8430 + }, + { + "epoch": 0.5767152336001095, + "grad_norm": 1.7828307541189372, + "learning_rate": 4.007442466784359e-06, + "loss": 0.3676, + "step": 8431 + }, + { + "epoch": 0.576783637731719, + "grad_norm": 1.881991444820363, + "learning_rate": 4.006356782257746e-06, + "loss": 0.3021, + "step": 8432 + }, + { + "epoch": 0.5768520418633285, + "grad_norm": 1.844089650534592, + "learning_rate": 4.0052711465039394e-06, + "loss": 0.2265, + "step": 8433 + }, + { + "epoch": 0.5769204459949381, + "grad_norm": 1.9345579190268836, + "learning_rate": 4.004185559576229e-06, + "loss": 0.2825, + "step": 8434 + }, + { + "epoch": 0.5769888501265477, + "grad_norm": 1.784706267690861, + "learning_rate": 4.0031000215278995e-06, + "loss": 0.3231, + "step": 8435 + }, + { + "epoch": 0.5770572542581572, + "grad_norm": 2.101776098011037, + "learning_rate": 4.002014532412233e-06, + "loss": 0.3063, + "step": 8436 + }, + { + "epoch": 0.5771256583897667, + "grad_norm": 1.522899865380883, + "learning_rate": 4.000929092282515e-06, + "loss": 0.3184, + "step": 8437 + }, + { + "epoch": 0.5771940625213763, + "grad_norm": 1.7787456935889414, + "learning_rate": 3.999843701192022e-06, + "loss": 0.3096, + "step": 8438 + }, + { + "epoch": 0.5772624666529859, + "grad_norm": 2.0060728901083777, + "learning_rate": 3.998758359194028e-06, + "loss": 0.3233, + "step": 8439 + }, + { + "epoch": 0.5773308707845954, + "grad_norm": 2.456346789484141, + "learning_rate": 3.997673066341811e-06, + "loss": 0.3375, + "step": 8440 + }, + { + "epoch": 0.5773992749162049, + "grad_norm": 2.0255523808334925, + "learning_rate": 3.996587822688639e-06, + "loss": 0.356, + "step": 8441 + }, + { + "epoch": 0.5774676790478145, + "grad_norm": 1.8542869177053323, + "learning_rate": 3.995502628287781e-06, + "loss": 0.4826, + "step": 8442 + }, + { + "epoch": 0.5775360831794241, + "grad_norm": 2.3667772019345112, + "learning_rate": 3.994417483192506e-06, + "loss": 0.3012, + "step": 8443 + }, + { + "epoch": 0.5776044873110335, + "grad_norm": 1.6838615739928682, + "learning_rate": 3.993332387456077e-06, + "loss": 0.3445, + "step": 8444 + }, + { + "epoch": 0.5776728914426431, + "grad_norm": 3.090785571440936, + "learning_rate": 3.992247341131754e-06, + "loss": 0.5389, + "step": 8445 + }, + { + "epoch": 0.5777412955742527, + "grad_norm": 1.6600954044636038, + "learning_rate": 3.991162344272798e-06, + "loss": 0.377, + "step": 8446 + }, + { + "epoch": 0.5778096997058623, + "grad_norm": 1.9926949397060894, + "learning_rate": 3.990077396932467e-06, + "loss": 0.291, + "step": 8447 + }, + { + "epoch": 0.5778781038374717, + "grad_norm": 3.523015929586458, + "learning_rate": 3.988992499164015e-06, + "loss": 0.6252, + "step": 8448 + }, + { + "epoch": 0.5779465079690813, + "grad_norm": 1.9655297145581867, + "learning_rate": 3.987907651020691e-06, + "loss": 0.3069, + "step": 8449 + }, + { + "epoch": 0.5780149121006909, + "grad_norm": 1.8687401863784572, + "learning_rate": 3.9868228525557485e-06, + "loss": 0.4593, + "step": 8450 + }, + { + "epoch": 0.5780833162323005, + "grad_norm": 1.9908996458441182, + "learning_rate": 3.985738103822433e-06, + "loss": 0.3234, + "step": 8451 + }, + { + "epoch": 0.5781517203639099, + "grad_norm": 1.7797252294344053, + "learning_rate": 3.984653404873988e-06, + "loss": 0.3712, + "step": 8452 + }, + { + "epoch": 0.5782201244955195, + "grad_norm": 2.0273122693155283, + "learning_rate": 3.983568755763658e-06, + "loss": 0.1227, + "step": 8453 + }, + { + "epoch": 0.5782885286271291, + "grad_norm": 1.9434121896285916, + "learning_rate": 3.982484156544682e-06, + "loss": 0.2078, + "step": 8454 + }, + { + "epoch": 0.5783569327587387, + "grad_norm": 2.950829163501432, + "learning_rate": 3.981399607270295e-06, + "loss": 0.5502, + "step": 8455 + }, + { + "epoch": 0.5784253368903481, + "grad_norm": 2.4068591040515024, + "learning_rate": 3.9803151079937345e-06, + "loss": 0.3431, + "step": 8456 + }, + { + "epoch": 0.5784937410219577, + "grad_norm": 2.0273131917660714, + "learning_rate": 3.979230658768234e-06, + "loss": 0.5419, + "step": 8457 + }, + { + "epoch": 0.5785621451535673, + "grad_norm": 1.8261340785067552, + "learning_rate": 3.9781462596470225e-06, + "loss": 0.2941, + "step": 8458 + }, + { + "epoch": 0.5786305492851769, + "grad_norm": 2.2335948649213853, + "learning_rate": 3.977061910683325e-06, + "loss": 0.5228, + "step": 8459 + }, + { + "epoch": 0.5786989534167863, + "grad_norm": 2.3035288727062233, + "learning_rate": 3.975977611930371e-06, + "loss": 0.7463, + "step": 8460 + }, + { + "epoch": 0.5787673575483959, + "grad_norm": 2.4241125414857714, + "learning_rate": 3.974893363441379e-06, + "loss": 0.4607, + "step": 8461 + }, + { + "epoch": 0.5788357616800055, + "grad_norm": 2.833258527347904, + "learning_rate": 3.973809165269571e-06, + "loss": 0.6971, + "step": 8462 + }, + { + "epoch": 0.578904165811615, + "grad_norm": 1.5624235468279972, + "learning_rate": 3.972725017468166e-06, + "loss": 0.3479, + "step": 8463 + }, + { + "epoch": 0.5789725699432245, + "grad_norm": 2.122571588427754, + "learning_rate": 3.971640920090377e-06, + "loss": 0.1718, + "step": 8464 + }, + { + "epoch": 0.5790409740748341, + "grad_norm": 2.146185516572907, + "learning_rate": 3.9705568731894155e-06, + "loss": 0.3971, + "step": 8465 + }, + { + "epoch": 0.5791093782064437, + "grad_norm": 1.963701587374052, + "learning_rate": 3.969472876818496e-06, + "loss": 0.3116, + "step": 8466 + }, + { + "epoch": 0.5791777823380532, + "grad_norm": 2.239792524962151, + "learning_rate": 3.968388931030823e-06, + "loss": 0.3642, + "step": 8467 + }, + { + "epoch": 0.5792461864696627, + "grad_norm": 3.0420189331250507, + "learning_rate": 3.967305035879605e-06, + "loss": 0.4205, + "step": 8468 + }, + { + "epoch": 0.5793145906012723, + "grad_norm": 2.779137462727516, + "learning_rate": 3.96622119141804e-06, + "loss": 0.532, + "step": 8469 + }, + { + "epoch": 0.5793829947328819, + "grad_norm": 2.3014804056291935, + "learning_rate": 3.965137397699333e-06, + "loss": 0.2531, + "step": 8470 + }, + { + "epoch": 0.5794513988644914, + "grad_norm": 2.1783314113151206, + "learning_rate": 3.96405365477668e-06, + "loss": 0.2584, + "step": 8471 + }, + { + "epoch": 0.579519802996101, + "grad_norm": 1.6763044554230522, + "learning_rate": 3.962969962703274e-06, + "loss": 0.1972, + "step": 8472 + }, + { + "epoch": 0.5795882071277105, + "grad_norm": 2.126946385340526, + "learning_rate": 3.961886321532312e-06, + "loss": 0.3566, + "step": 8473 + }, + { + "epoch": 0.5796566112593201, + "grad_norm": 2.4457979543056134, + "learning_rate": 3.960802731316982e-06, + "loss": 0.7298, + "step": 8474 + }, + { + "epoch": 0.5797250153909296, + "grad_norm": 2.6961265274571278, + "learning_rate": 3.95971919211047e-06, + "loss": 0.559, + "step": 8475 + }, + { + "epoch": 0.5797934195225392, + "grad_norm": 2.283491388924669, + "learning_rate": 3.958635703965964e-06, + "loss": 0.3066, + "step": 8476 + }, + { + "epoch": 0.5798618236541487, + "grad_norm": 1.9728828375580363, + "learning_rate": 3.9575522669366485e-06, + "loss": 0.5259, + "step": 8477 + }, + { + "epoch": 0.5799302277857583, + "grad_norm": 1.7867458786800128, + "learning_rate": 3.956468881075701e-06, + "loss": 0.3854, + "step": 8478 + }, + { + "epoch": 0.5799986319173678, + "grad_norm": 2.2711679659731696, + "learning_rate": 3.955385546436299e-06, + "loss": 0.5498, + "step": 8479 + }, + { + "epoch": 0.5800670360489774, + "grad_norm": 1.7254481711279255, + "learning_rate": 3.954302263071621e-06, + "loss": 0.3223, + "step": 8480 + }, + { + "epoch": 0.5801354401805869, + "grad_norm": 2.049181344973984, + "learning_rate": 3.9532190310348366e-06, + "loss": 0.5318, + "step": 8481 + }, + { + "epoch": 0.5802038443121964, + "grad_norm": 1.76700893110394, + "learning_rate": 3.952135850379115e-06, + "loss": 0.3235, + "step": 8482 + }, + { + "epoch": 0.580272248443806, + "grad_norm": 2.0180309337762985, + "learning_rate": 3.951052721157629e-06, + "loss": 0.414, + "step": 8483 + }, + { + "epoch": 0.5803406525754156, + "grad_norm": 2.56302632764912, + "learning_rate": 3.949969643423538e-06, + "loss": 0.2694, + "step": 8484 + }, + { + "epoch": 0.5804090567070251, + "grad_norm": 2.0672992586704257, + "learning_rate": 3.948886617230008e-06, + "loss": 0.331, + "step": 8485 + }, + { + "epoch": 0.5804774608386346, + "grad_norm": 1.736264917448265, + "learning_rate": 3.9478036426301984e-06, + "loss": 0.2495, + "step": 8486 + }, + { + "epoch": 0.5805458649702442, + "grad_norm": 1.7664230914454, + "learning_rate": 3.946720719677268e-06, + "loss": 0.2599, + "step": 8487 + }, + { + "epoch": 0.5806142691018538, + "grad_norm": 1.8902407287188154, + "learning_rate": 3.945637848424372e-06, + "loss": 0.2937, + "step": 8488 + }, + { + "epoch": 0.5806826732334633, + "grad_norm": 2.5072939149386624, + "learning_rate": 3.9445550289246595e-06, + "loss": 0.6221, + "step": 8489 + }, + { + "epoch": 0.5807510773650728, + "grad_norm": 2.4355657006265, + "learning_rate": 3.9434722612312846e-06, + "loss": 0.4678, + "step": 8490 + }, + { + "epoch": 0.5808194814966824, + "grad_norm": 1.7581213616185356, + "learning_rate": 3.942389545397392e-06, + "loss": 0.382, + "step": 8491 + }, + { + "epoch": 0.580887885628292, + "grad_norm": 2.226506342932599, + "learning_rate": 3.941306881476129e-06, + "loss": 0.2607, + "step": 8492 + }, + { + "epoch": 0.5809562897599015, + "grad_norm": 2.1111374732791988, + "learning_rate": 3.940224269520636e-06, + "loss": 0.4276, + "step": 8493 + }, + { + "epoch": 0.581024693891511, + "grad_norm": 2.6037796727960645, + "learning_rate": 3.939141709584053e-06, + "loss": 0.2632, + "step": 8494 + }, + { + "epoch": 0.5810930980231206, + "grad_norm": 2.4103293488105386, + "learning_rate": 3.9380592017195165e-06, + "loss": 0.3579, + "step": 8495 + }, + { + "epoch": 0.5811615021547302, + "grad_norm": 2.1095215958411875, + "learning_rate": 3.936976745980163e-06, + "loss": 0.4681, + "step": 8496 + }, + { + "epoch": 0.5812299062863397, + "grad_norm": 2.3672851252804676, + "learning_rate": 3.935894342419126e-06, + "loss": 0.2522, + "step": 8497 + }, + { + "epoch": 0.5812983104179492, + "grad_norm": 2.113342058643095, + "learning_rate": 3.934811991089531e-06, + "loss": 0.4572, + "step": 8498 + }, + { + "epoch": 0.5813667145495588, + "grad_norm": 2.2361685683455175, + "learning_rate": 3.93372969204451e-06, + "loss": 0.3738, + "step": 8499 + }, + { + "epoch": 0.5814351186811684, + "grad_norm": 2.139691358072249, + "learning_rate": 3.932647445337183e-06, + "loss": 0.4775, + "step": 8500 + }, + { + "epoch": 0.5815035228127778, + "grad_norm": 2.7069469785577867, + "learning_rate": 3.931565251020673e-06, + "loss": 0.5745, + "step": 8501 + }, + { + "epoch": 0.5815719269443874, + "grad_norm": 1.937128681318196, + "learning_rate": 3.930483109148101e-06, + "loss": 0.4267, + "step": 8502 + }, + { + "epoch": 0.581640331075997, + "grad_norm": 2.439553697771455, + "learning_rate": 3.9294010197725826e-06, + "loss": 0.3085, + "step": 8503 + }, + { + "epoch": 0.5817087352076066, + "grad_norm": 4.2769527587556, + "learning_rate": 3.928318982947232e-06, + "loss": 0.2058, + "step": 8504 + }, + { + "epoch": 0.581777139339216, + "grad_norm": 1.4020686982294637, + "learning_rate": 3.927236998725158e-06, + "loss": 0.2155, + "step": 8505 + }, + { + "epoch": 0.5818455434708256, + "grad_norm": 1.7977213317681737, + "learning_rate": 3.926155067159475e-06, + "loss": 0.378, + "step": 8506 + }, + { + "epoch": 0.5819139476024352, + "grad_norm": 2.159448338655767, + "learning_rate": 3.925073188303287e-06, + "loss": 0.3927, + "step": 8507 + }, + { + "epoch": 0.5819823517340448, + "grad_norm": 1.4567040313767639, + "learning_rate": 3.923991362209697e-06, + "loss": 0.1174, + "step": 8508 + }, + { + "epoch": 0.5820507558656542, + "grad_norm": 1.1796900783914848, + "learning_rate": 3.922909588931808e-06, + "loss": 0.0966, + "step": 8509 + }, + { + "epoch": 0.5821191599972638, + "grad_norm": 2.1155933135461615, + "learning_rate": 3.921827868522718e-06, + "loss": 0.5172, + "step": 8510 + }, + { + "epoch": 0.5821875641288734, + "grad_norm": 1.9311293647643204, + "learning_rate": 3.92074620103552e-06, + "loss": 0.275, + "step": 8511 + }, + { + "epoch": 0.582255968260483, + "grad_norm": 2.1007871580200423, + "learning_rate": 3.919664586523312e-06, + "loss": 0.2852, + "step": 8512 + }, + { + "epoch": 0.5823243723920924, + "grad_norm": 4.444340738768761, + "learning_rate": 3.9185830250391835e-06, + "loss": 0.2658, + "step": 8513 + }, + { + "epoch": 0.582392776523702, + "grad_norm": 2.4588800579362466, + "learning_rate": 3.917501516636221e-06, + "loss": 0.678, + "step": 8514 + }, + { + "epoch": 0.5824611806553116, + "grad_norm": 2.6220725174935446, + "learning_rate": 3.91642006136751e-06, + "loss": 0.5286, + "step": 8515 + }, + { + "epoch": 0.5825295847869212, + "grad_norm": 2.9274777178400933, + "learning_rate": 3.915338659286137e-06, + "loss": 0.2905, + "step": 8516 + }, + { + "epoch": 0.5825979889185307, + "grad_norm": 1.6305919039559778, + "learning_rate": 3.9142573104451805e-06, + "loss": 0.1972, + "step": 8517 + }, + { + "epoch": 0.5826663930501402, + "grad_norm": 1.8325535613269022, + "learning_rate": 3.913176014897717e-06, + "loss": 0.4034, + "step": 8518 + }, + { + "epoch": 0.5827347971817498, + "grad_norm": 1.735059439866695, + "learning_rate": 3.912094772696825e-06, + "loss": 0.2478, + "step": 8519 + }, + { + "epoch": 0.5828032013133593, + "grad_norm": 1.6584747145958947, + "learning_rate": 3.911013583895573e-06, + "loss": 0.2867, + "step": 8520 + }, + { + "epoch": 0.5828716054449689, + "grad_norm": 2.9613587374225165, + "learning_rate": 3.909932448547034e-06, + "loss": 0.356, + "step": 8521 + }, + { + "epoch": 0.5829400095765784, + "grad_norm": 1.0172544942202493, + "learning_rate": 3.908851366704274e-06, + "loss": 0.0666, + "step": 8522 + }, + { + "epoch": 0.583008413708188, + "grad_norm": 2.1572315221188108, + "learning_rate": 3.907770338420359e-06, + "loss": 0.5485, + "step": 8523 + }, + { + "epoch": 0.5830768178397975, + "grad_norm": 4.2774923220397785, + "learning_rate": 3.906689363748348e-06, + "loss": 0.238, + "step": 8524 + }, + { + "epoch": 0.5831452219714071, + "grad_norm": 1.8836124904970801, + "learning_rate": 3.905608442741302e-06, + "loss": 0.2987, + "step": 8525 + }, + { + "epoch": 0.5832136261030166, + "grad_norm": 2.1728834582784504, + "learning_rate": 3.90452757545228e-06, + "loss": 0.3564, + "step": 8526 + }, + { + "epoch": 0.5832820302346262, + "grad_norm": 2.0377216315107742, + "learning_rate": 3.903446761934336e-06, + "loss": 0.4648, + "step": 8527 + }, + { + "epoch": 0.5833504343662357, + "grad_norm": 2.546598951907636, + "learning_rate": 3.902366002240518e-06, + "loss": 0.4222, + "step": 8528 + }, + { + "epoch": 0.5834188384978453, + "grad_norm": 2.2009380862922976, + "learning_rate": 3.9012852964238776e-06, + "loss": 0.4227, + "step": 8529 + }, + { + "epoch": 0.5834872426294548, + "grad_norm": 1.8558432391859703, + "learning_rate": 3.900204644537462e-06, + "loss": 0.2555, + "step": 8530 + }, + { + "epoch": 0.5835556467610644, + "grad_norm": 2.41433147055431, + "learning_rate": 3.8991240466343115e-06, + "loss": 0.3369, + "step": 8531 + }, + { + "epoch": 0.5836240508926739, + "grad_norm": 2.610748706066641, + "learning_rate": 3.89804350276747e-06, + "loss": 0.3975, + "step": 8532 + }, + { + "epoch": 0.5836924550242835, + "grad_norm": 2.289478836995619, + "learning_rate": 3.896963012989974e-06, + "loss": 0.3926, + "step": 8533 + }, + { + "epoch": 0.583760859155893, + "grad_norm": 2.3444928581452804, + "learning_rate": 3.8958825773548594e-06, + "loss": 0.3322, + "step": 8534 + }, + { + "epoch": 0.5838292632875026, + "grad_norm": 2.0095863447681537, + "learning_rate": 3.894802195915158e-06, + "loss": 0.2415, + "step": 8535 + }, + { + "epoch": 0.5838976674191121, + "grad_norm": 2.2172046403981795, + "learning_rate": 3.893721868723905e-06, + "loss": 0.4347, + "step": 8536 + }, + { + "epoch": 0.5839660715507217, + "grad_norm": 2.0509333308914406, + "learning_rate": 3.892641595834124e-06, + "loss": 0.4727, + "step": 8537 + }, + { + "epoch": 0.5840344756823312, + "grad_norm": 1.8477942695440925, + "learning_rate": 3.89156137729884e-06, + "loss": 0.1762, + "step": 8538 + }, + { + "epoch": 0.5841028798139407, + "grad_norm": 2.0529346313939643, + "learning_rate": 3.8904812131710776e-06, + "loss": 0.2869, + "step": 8539 + }, + { + "epoch": 0.5841712839455503, + "grad_norm": 1.790641820538992, + "learning_rate": 3.889401103503854e-06, + "loss": 0.3312, + "step": 8540 + }, + { + "epoch": 0.5842396880771599, + "grad_norm": 1.8537286851021493, + "learning_rate": 3.888321048350187e-06, + "loss": 0.2275, + "step": 8541 + }, + { + "epoch": 0.5843080922087694, + "grad_norm": 2.1846948826026287, + "learning_rate": 3.887241047763092e-06, + "loss": 0.4906, + "step": 8542 + }, + { + "epoch": 0.5843764963403789, + "grad_norm": 1.8592900088464086, + "learning_rate": 3.88616110179558e-06, + "loss": 0.373, + "step": 8543 + }, + { + "epoch": 0.5844449004719885, + "grad_norm": 2.15350359009124, + "learning_rate": 3.885081210500658e-06, + "loss": 0.5024, + "step": 8544 + }, + { + "epoch": 0.5845133046035981, + "grad_norm": 1.8025473418083355, + "learning_rate": 3.884001373931335e-06, + "loss": 0.4054, + "step": 8545 + }, + { + "epoch": 0.5845817087352077, + "grad_norm": 2.4237381997590934, + "learning_rate": 3.882921592140614e-06, + "loss": 0.6888, + "step": 8546 + }, + { + "epoch": 0.5846501128668171, + "grad_norm": 1.8620570457728676, + "learning_rate": 3.881841865181497e-06, + "loss": 0.4134, + "step": 8547 + }, + { + "epoch": 0.5847185169984267, + "grad_norm": 1.6333836225134044, + "learning_rate": 3.880762193106978e-06, + "loss": 0.1584, + "step": 8548 + }, + { + "epoch": 0.5847869211300363, + "grad_norm": 2.2601078206269096, + "learning_rate": 3.879682575970057e-06, + "loss": 0.6171, + "step": 8549 + }, + { + "epoch": 0.5848553252616459, + "grad_norm": 1.96200834724661, + "learning_rate": 3.878603013823726e-06, + "loss": 0.1012, + "step": 8550 + }, + { + "epoch": 0.5849237293932553, + "grad_norm": 1.9270126978498407, + "learning_rate": 3.8775235067209725e-06, + "loss": 0.4072, + "step": 8551 + }, + { + "epoch": 0.5849921335248649, + "grad_norm": 2.8731079250672242, + "learning_rate": 3.876444054714787e-06, + "loss": 0.4731, + "step": 8552 + }, + { + "epoch": 0.5850605376564745, + "grad_norm": 1.6069359375094185, + "learning_rate": 3.875364657858153e-06, + "loss": 0.1284, + "step": 8553 + }, + { + "epoch": 0.585128941788084, + "grad_norm": 2.4370207376529134, + "learning_rate": 3.874285316204049e-06, + "loss": 0.5814, + "step": 8554 + }, + { + "epoch": 0.5851973459196935, + "grad_norm": 2.399349424479757, + "learning_rate": 3.87320602980546e-06, + "loss": 0.8012, + "step": 8555 + }, + { + "epoch": 0.5852657500513031, + "grad_norm": 1.6049576848533516, + "learning_rate": 3.8721267987153614e-06, + "loss": 0.2423, + "step": 8556 + }, + { + "epoch": 0.5853341541829127, + "grad_norm": 2.507167198368074, + "learning_rate": 3.871047622986726e-06, + "loss": 0.696, + "step": 8557 + }, + { + "epoch": 0.5854025583145221, + "grad_norm": 2.402541361799071, + "learning_rate": 3.8699685026725235e-06, + "loss": 0.1831, + "step": 8558 + }, + { + "epoch": 0.5854709624461317, + "grad_norm": 2.250162532684517, + "learning_rate": 3.868889437825724e-06, + "loss": 0.4521, + "step": 8559 + }, + { + "epoch": 0.5855393665777413, + "grad_norm": 2.851725032825184, + "learning_rate": 3.867810428499294e-06, + "loss": 0.4222, + "step": 8560 + }, + { + "epoch": 0.5856077707093509, + "grad_norm": 2.8889221433769414, + "learning_rate": 3.866731474746194e-06, + "loss": 0.3782, + "step": 8561 + }, + { + "epoch": 0.5856761748409604, + "grad_norm": 2.113709750071423, + "learning_rate": 3.865652576619388e-06, + "loss": 0.3585, + "step": 8562 + }, + { + "epoch": 0.5857445789725699, + "grad_norm": 1.9127390665343385, + "learning_rate": 3.864573734171829e-06, + "loss": 0.1066, + "step": 8563 + }, + { + "epoch": 0.5858129831041795, + "grad_norm": 2.076710248101276, + "learning_rate": 3.863494947456473e-06, + "loss": 0.1233, + "step": 8564 + }, + { + "epoch": 0.5858813872357891, + "grad_norm": 1.7717378681004876, + "learning_rate": 3.862416216526274e-06, + "loss": 0.4263, + "step": 8565 + }, + { + "epoch": 0.5859497913673986, + "grad_norm": 3.112826755714306, + "learning_rate": 3.861337541434183e-06, + "loss": 0.5016, + "step": 8566 + }, + { + "epoch": 0.5860181954990081, + "grad_norm": 2.2403313566732077, + "learning_rate": 3.860258922233142e-06, + "loss": 0.3229, + "step": 8567 + }, + { + "epoch": 0.5860865996306177, + "grad_norm": 2.240306883006877, + "learning_rate": 3.8591803589760966e-06, + "loss": 0.4951, + "step": 8568 + }, + { + "epoch": 0.5861550037622273, + "grad_norm": 2.0807831764770857, + "learning_rate": 3.85810185171599e-06, + "loss": 0.2902, + "step": 8569 + }, + { + "epoch": 0.5862234078938368, + "grad_norm": 2.342274493840712, + "learning_rate": 3.857023400505757e-06, + "loss": 0.7136, + "step": 8570 + }, + { + "epoch": 0.5862918120254463, + "grad_norm": 1.3240070145608815, + "learning_rate": 3.855945005398335e-06, + "loss": 0.185, + "step": 8571 + }, + { + "epoch": 0.5863602161570559, + "grad_norm": 1.6952935582522293, + "learning_rate": 3.8548666664466575e-06, + "loss": 0.2154, + "step": 8572 + }, + { + "epoch": 0.5864286202886654, + "grad_norm": 2.08860290267924, + "learning_rate": 3.853788383703654e-06, + "loss": 0.3827, + "step": 8573 + }, + { + "epoch": 0.586497024420275, + "grad_norm": 2.215997748502042, + "learning_rate": 3.8527101572222495e-06, + "loss": 0.417, + "step": 8574 + }, + { + "epoch": 0.5865654285518845, + "grad_norm": 1.856263349947463, + "learning_rate": 3.851631987055371e-06, + "loss": 0.3069, + "step": 8575 + }, + { + "epoch": 0.5866338326834941, + "grad_norm": 2.3325208472682792, + "learning_rate": 3.850553873255942e-06, + "loss": 0.1948, + "step": 8576 + }, + { + "epoch": 0.5867022368151036, + "grad_norm": 2.644264117544133, + "learning_rate": 3.849475815876878e-06, + "loss": 0.6228, + "step": 8577 + }, + { + "epoch": 0.5867706409467132, + "grad_norm": 1.46429354697967, + "learning_rate": 3.848397814971098e-06, + "loss": 0.2822, + "step": 8578 + }, + { + "epoch": 0.5868390450783227, + "grad_norm": 1.6338717935049272, + "learning_rate": 3.8473198705915135e-06, + "loss": 0.2385, + "step": 8579 + }, + { + "epoch": 0.5869074492099323, + "grad_norm": 2.0652513936602253, + "learning_rate": 3.846241982791037e-06, + "loss": 0.3152, + "step": 8580 + }, + { + "epoch": 0.5869758533415418, + "grad_norm": 4.97918956410824, + "learning_rate": 3.845164151622574e-06, + "loss": 0.4857, + "step": 8581 + }, + { + "epoch": 0.5870442574731514, + "grad_norm": 1.9035303969723028, + "learning_rate": 3.844086377139032e-06, + "loss": 0.4123, + "step": 8582 + }, + { + "epoch": 0.5871126616047609, + "grad_norm": 1.7344901888631175, + "learning_rate": 3.843008659393312e-06, + "loss": 0.3441, + "step": 8583 + }, + { + "epoch": 0.5871810657363705, + "grad_norm": 2.4042158446304, + "learning_rate": 3.841930998438313e-06, + "loss": 0.3318, + "step": 8584 + }, + { + "epoch": 0.58724946986798, + "grad_norm": 2.083123823995914, + "learning_rate": 3.840853394326934e-06, + "loss": 0.4479, + "step": 8585 + }, + { + "epoch": 0.5873178739995896, + "grad_norm": 2.5920662585435363, + "learning_rate": 3.839775847112068e-06, + "loss": 0.511, + "step": 8586 + }, + { + "epoch": 0.5873862781311991, + "grad_norm": 2.424350802051757, + "learning_rate": 3.838698356846607e-06, + "loss": 0.5675, + "step": 8587 + }, + { + "epoch": 0.5874546822628087, + "grad_norm": 1.7222653202740992, + "learning_rate": 3.837620923583436e-06, + "loss": 0.3101, + "step": 8588 + }, + { + "epoch": 0.5875230863944182, + "grad_norm": 2.601889317693267, + "learning_rate": 3.836543547375445e-06, + "loss": 0.6132, + "step": 8589 + }, + { + "epoch": 0.5875914905260278, + "grad_norm": 2.3151723653548224, + "learning_rate": 3.835466228275514e-06, + "loss": 0.4066, + "step": 8590 + }, + { + "epoch": 0.5876598946576373, + "grad_norm": 1.701751194274145, + "learning_rate": 3.834388966336525e-06, + "loss": 0.2197, + "step": 8591 + }, + { + "epoch": 0.5877282987892468, + "grad_norm": 1.801683245514077, + "learning_rate": 3.8333117616113545e-06, + "loss": 0.2775, + "step": 8592 + }, + { + "epoch": 0.5877967029208564, + "grad_norm": 2.1154954983337193, + "learning_rate": 3.832234614152876e-06, + "loss": 0.3754, + "step": 8593 + }, + { + "epoch": 0.587865107052466, + "grad_norm": 1.882549599935104, + "learning_rate": 3.831157524013961e-06, + "loss": 0.2328, + "step": 8594 + }, + { + "epoch": 0.5879335111840756, + "grad_norm": 1.8330356715691207, + "learning_rate": 3.830080491247479e-06, + "loss": 0.2999, + "step": 8595 + }, + { + "epoch": 0.588001915315685, + "grad_norm": 2.824316669418066, + "learning_rate": 3.829003515906297e-06, + "loss": 0.6062, + "step": 8596 + }, + { + "epoch": 0.5880703194472946, + "grad_norm": 2.080432335151342, + "learning_rate": 3.827926598043276e-06, + "loss": 0.4664, + "step": 8597 + }, + { + "epoch": 0.5881387235789042, + "grad_norm": 2.0719210727453885, + "learning_rate": 3.826849737711279e-06, + "loss": 0.4547, + "step": 8598 + }, + { + "epoch": 0.5882071277105138, + "grad_norm": 1.8404705615900634, + "learning_rate": 3.825772934963161e-06, + "loss": 0.2871, + "step": 8599 + }, + { + "epoch": 0.5882755318421232, + "grad_norm": 2.0884183425515546, + "learning_rate": 3.824696189851778e-06, + "loss": 0.3062, + "step": 8600 + }, + { + "epoch": 0.5883439359737328, + "grad_norm": 1.7456102491193286, + "learning_rate": 3.823619502429982e-06, + "loss": 0.3782, + "step": 8601 + }, + { + "epoch": 0.5884123401053424, + "grad_norm": 1.4106297741438598, + "learning_rate": 3.822542872750621e-06, + "loss": 0.1536, + "step": 8602 + }, + { + "epoch": 0.588480744236952, + "grad_norm": 2.6575203045906393, + "learning_rate": 3.821466300866541e-06, + "loss": 0.3305, + "step": 8603 + }, + { + "epoch": 0.5885491483685614, + "grad_norm": 3.3553408863721104, + "learning_rate": 3.820389786830586e-06, + "loss": 0.4608, + "step": 8604 + }, + { + "epoch": 0.588617552500171, + "grad_norm": 2.635769962985612, + "learning_rate": 3.819313330695596e-06, + "loss": 0.723, + "step": 8605 + }, + { + "epoch": 0.5886859566317806, + "grad_norm": 1.9615020965019943, + "learning_rate": 3.818236932514411e-06, + "loss": 0.3117, + "step": 8606 + }, + { + "epoch": 0.5887543607633902, + "grad_norm": 1.9566530939711284, + "learning_rate": 3.8171605923398625e-06, + "loss": 0.4395, + "step": 8607 + }, + { + "epoch": 0.5888227648949996, + "grad_norm": 2.3682714567707426, + "learning_rate": 3.816084310224785e-06, + "loss": 0.4473, + "step": 8608 + }, + { + "epoch": 0.5888911690266092, + "grad_norm": 2.3722465075924344, + "learning_rate": 3.815008086222007e-06, + "loss": 0.4189, + "step": 8609 + }, + { + "epoch": 0.5889595731582188, + "grad_norm": 1.9824132947313255, + "learning_rate": 3.8139319203843533e-06, + "loss": 0.2932, + "step": 8610 + }, + { + "epoch": 0.5890279772898283, + "grad_norm": 1.702908169653875, + "learning_rate": 3.81285581276465e-06, + "loss": 0.2042, + "step": 8611 + }, + { + "epoch": 0.5890963814214378, + "grad_norm": 1.5987373602160933, + "learning_rate": 3.811779763415715e-06, + "loss": 0.2551, + "step": 8612 + }, + { + "epoch": 0.5891647855530474, + "grad_norm": 1.6495299990230903, + "learning_rate": 3.810703772390366e-06, + "loss": 0.2568, + "step": 8613 + }, + { + "epoch": 0.589233189684657, + "grad_norm": 2.8471861086211763, + "learning_rate": 3.8096278397414187e-06, + "loss": 0.3562, + "step": 8614 + }, + { + "epoch": 0.5893015938162665, + "grad_norm": 2.2024226658631494, + "learning_rate": 3.8085519655216876e-06, + "loss": 0.4975, + "step": 8615 + }, + { + "epoch": 0.589369997947876, + "grad_norm": 1.6150017816501272, + "learning_rate": 3.8074761497839787e-06, + "loss": 0.3065, + "step": 8616 + }, + { + "epoch": 0.5894384020794856, + "grad_norm": 2.1167784080959913, + "learning_rate": 3.806400392581098e-06, + "loss": 0.4035, + "step": 8617 + }, + { + "epoch": 0.5895068062110952, + "grad_norm": 2.4129856108832444, + "learning_rate": 3.8053246939658504e-06, + "loss": 0.3967, + "step": 8618 + }, + { + "epoch": 0.5895752103427047, + "grad_norm": 2.2460384229397126, + "learning_rate": 3.804249053991037e-06, + "loss": 0.3443, + "step": 8619 + }, + { + "epoch": 0.5896436144743142, + "grad_norm": 1.916934863189041, + "learning_rate": 3.8031734727094525e-06, + "loss": 0.1416, + "step": 8620 + }, + { + "epoch": 0.5897120186059238, + "grad_norm": 1.7994046731776552, + "learning_rate": 3.8020979501738946e-06, + "loss": 0.2035, + "step": 8621 + }, + { + "epoch": 0.5897804227375334, + "grad_norm": 2.4128760845184063, + "learning_rate": 3.8010224864371543e-06, + "loss": 0.3611, + "step": 8622 + }, + { + "epoch": 0.5898488268691429, + "grad_norm": 1.4122841807287676, + "learning_rate": 3.799947081552019e-06, + "loss": 0.1428, + "step": 8623 + }, + { + "epoch": 0.5899172310007524, + "grad_norm": 1.8587568842459603, + "learning_rate": 3.7988717355712744e-06, + "loss": 0.2565, + "step": 8624 + }, + { + "epoch": 0.589985635132362, + "grad_norm": 1.92635450671301, + "learning_rate": 3.7977964485477077e-06, + "loss": 0.2646, + "step": 8625 + }, + { + "epoch": 0.5900540392639716, + "grad_norm": 2.5628854153543394, + "learning_rate": 3.7967212205340965e-06, + "loss": 0.3259, + "step": 8626 + }, + { + "epoch": 0.5901224433955811, + "grad_norm": 2.345607098115444, + "learning_rate": 3.7956460515832173e-06, + "loss": 0.4438, + "step": 8627 + }, + { + "epoch": 0.5901908475271906, + "grad_norm": 1.524099653568251, + "learning_rate": 3.7945709417478467e-06, + "loss": 0.3064, + "step": 8628 + }, + { + "epoch": 0.5902592516588002, + "grad_norm": 2.632144299845218, + "learning_rate": 3.793495891080755e-06, + "loss": 0.2582, + "step": 8629 + }, + { + "epoch": 0.5903276557904097, + "grad_norm": 2.0848939692799986, + "learning_rate": 3.7924208996347106e-06, + "loss": 0.3429, + "step": 8630 + }, + { + "epoch": 0.5903960599220193, + "grad_norm": 1.593076933124057, + "learning_rate": 3.791345967462481e-06, + "loss": 0.2257, + "step": 8631 + }, + { + "epoch": 0.5904644640536288, + "grad_norm": 1.3727837442017077, + "learning_rate": 3.7902710946168275e-06, + "loss": 0.2169, + "step": 8632 + }, + { + "epoch": 0.5905328681852384, + "grad_norm": 2.2649664086046717, + "learning_rate": 3.789196281150509e-06, + "loss": 0.6259, + "step": 8633 + }, + { + "epoch": 0.5906012723168479, + "grad_norm": 1.6550970135408423, + "learning_rate": 3.788121527116283e-06, + "loss": 0.2121, + "step": 8634 + }, + { + "epoch": 0.5906696764484575, + "grad_norm": 2.0194477123049386, + "learning_rate": 3.787046832566907e-06, + "loss": 0.4015, + "step": 8635 + }, + { + "epoch": 0.590738080580067, + "grad_norm": 2.097946612189516, + "learning_rate": 3.7859721975551294e-06, + "loss": 0.2431, + "step": 8636 + }, + { + "epoch": 0.5908064847116766, + "grad_norm": 2.6948296764088346, + "learning_rate": 3.7848976221336982e-06, + "loss": 0.2296, + "step": 8637 + }, + { + "epoch": 0.5908748888432861, + "grad_norm": 1.2441532547577685, + "learning_rate": 3.78382310635536e-06, + "loss": 0.182, + "step": 8638 + }, + { + "epoch": 0.5909432929748957, + "grad_norm": 1.7703874045246304, + "learning_rate": 3.782748650272857e-06, + "loss": 0.3295, + "step": 8639 + }, + { + "epoch": 0.5910116971065053, + "grad_norm": 2.2689454443741828, + "learning_rate": 3.7816742539389274e-06, + "loss": 0.5887, + "step": 8640 + }, + { + "epoch": 0.5910801012381148, + "grad_norm": 1.894912807149586, + "learning_rate": 3.780599917406309e-06, + "loss": 0.1359, + "step": 8641 + }, + { + "epoch": 0.5911485053697243, + "grad_norm": 2.1393596222681723, + "learning_rate": 3.7795256407277354e-06, + "loss": 0.4389, + "step": 8642 + }, + { + "epoch": 0.5912169095013339, + "grad_norm": 2.302755460893349, + "learning_rate": 3.778451423955936e-06, + "loss": 0.4272, + "step": 8643 + }, + { + "epoch": 0.5912853136329435, + "grad_norm": 2.4459275639094464, + "learning_rate": 3.777377267143638e-06, + "loss": 0.6474, + "step": 8644 + }, + { + "epoch": 0.591353717764553, + "grad_norm": 1.7144076416281993, + "learning_rate": 3.7763031703435706e-06, + "loss": 0.1697, + "step": 8645 + }, + { + "epoch": 0.5914221218961625, + "grad_norm": 3.3063155328208818, + "learning_rate": 3.775229133608452e-06, + "loss": 0.4791, + "step": 8646 + }, + { + "epoch": 0.5914905260277721, + "grad_norm": 2.4231701617967953, + "learning_rate": 3.7741551569910016e-06, + "loss": 0.3299, + "step": 8647 + }, + { + "epoch": 0.5915589301593817, + "grad_norm": 2.4753378418370584, + "learning_rate": 3.773081240543936e-06, + "loss": 0.499, + "step": 8648 + }, + { + "epoch": 0.5916273342909911, + "grad_norm": 1.8731483137648313, + "learning_rate": 3.7720073843199685e-06, + "loss": 0.4031, + "step": 8649 + }, + { + "epoch": 0.5916957384226007, + "grad_norm": 1.7075226816176123, + "learning_rate": 3.770933588371807e-06, + "loss": 0.2778, + "step": 8650 + }, + { + "epoch": 0.5917641425542103, + "grad_norm": 1.8127291057746666, + "learning_rate": 3.769859852752161e-06, + "loss": 0.3179, + "step": 8651 + }, + { + "epoch": 0.5918325466858199, + "grad_norm": 2.0206674851040467, + "learning_rate": 3.768786177513734e-06, + "loss": 0.575, + "step": 8652 + }, + { + "epoch": 0.5919009508174293, + "grad_norm": 1.9199562806998134, + "learning_rate": 3.767712562709225e-06, + "loss": 0.5439, + "step": 8653 + }, + { + "epoch": 0.5919693549490389, + "grad_norm": 2.447408451256972, + "learning_rate": 3.7666390083913344e-06, + "loss": 0.4303, + "step": 8654 + }, + { + "epoch": 0.5920377590806485, + "grad_norm": 2.474010139056764, + "learning_rate": 3.765565514612759e-06, + "loss": 0.4663, + "step": 8655 + }, + { + "epoch": 0.5921061632122581, + "grad_norm": 1.404498731724963, + "learning_rate": 3.7644920814261885e-06, + "loss": 0.2693, + "step": 8656 + }, + { + "epoch": 0.5921745673438675, + "grad_norm": 2.1296736427802285, + "learning_rate": 3.763418708884312e-06, + "loss": 0.5023, + "step": 8657 + }, + { + "epoch": 0.5922429714754771, + "grad_norm": 1.8915997742982682, + "learning_rate": 3.7623453970398175e-06, + "loss": 0.2269, + "step": 8658 + }, + { + "epoch": 0.5923113756070867, + "grad_norm": 1.946197094158402, + "learning_rate": 3.7612721459453883e-06, + "loss": 0.4095, + "step": 8659 + }, + { + "epoch": 0.5923797797386963, + "grad_norm": 2.0249816054815613, + "learning_rate": 3.760198955653702e-06, + "loss": 0.3514, + "step": 8660 + }, + { + "epoch": 0.5924481838703057, + "grad_norm": 1.5806018305672087, + "learning_rate": 3.7591258262174397e-06, + "loss": 0.2492, + "step": 8661 + }, + { + "epoch": 0.5925165880019153, + "grad_norm": 2.51916814745566, + "learning_rate": 3.758052757689273e-06, + "loss": 0.653, + "step": 8662 + }, + { + "epoch": 0.5925849921335249, + "grad_norm": 1.8904242881920867, + "learning_rate": 3.756979750121873e-06, + "loss": 0.4593, + "step": 8663 + }, + { + "epoch": 0.5926533962651345, + "grad_norm": 2.2042944455712234, + "learning_rate": 3.7559068035679103e-06, + "loss": 0.4873, + "step": 8664 + }, + { + "epoch": 0.5927218003967439, + "grad_norm": 2.143719397888077, + "learning_rate": 3.7548339180800497e-06, + "loss": 0.2279, + "step": 8665 + }, + { + "epoch": 0.5927902045283535, + "grad_norm": 1.991451763352814, + "learning_rate": 3.7537610937109535e-06, + "loss": 0.4794, + "step": 8666 + }, + { + "epoch": 0.5928586086599631, + "grad_norm": 1.7823172015826956, + "learning_rate": 3.7526883305132807e-06, + "loss": 0.3492, + "step": 8667 + }, + { + "epoch": 0.5929270127915726, + "grad_norm": 2.3980135154609705, + "learning_rate": 3.751615628539688e-06, + "loss": 0.4205, + "step": 8668 + }, + { + "epoch": 0.5929954169231821, + "grad_norm": 2.9384271294900577, + "learning_rate": 3.750542987842829e-06, + "loss": 0.4761, + "step": 8669 + }, + { + "epoch": 0.5930638210547917, + "grad_norm": 2.1386350584998177, + "learning_rate": 3.7494704084753524e-06, + "loss": 0.528, + "step": 8670 + }, + { + "epoch": 0.5931322251864013, + "grad_norm": 2.411833181283114, + "learning_rate": 3.7483978904899083e-06, + "loss": 0.3567, + "step": 8671 + }, + { + "epoch": 0.5932006293180108, + "grad_norm": 2.1846707371704466, + "learning_rate": 3.7473254339391397e-06, + "loss": 0.4047, + "step": 8672 + }, + { + "epoch": 0.5932690334496203, + "grad_norm": 1.5849333764839453, + "learning_rate": 3.7462530388756852e-06, + "loss": 0.2928, + "step": 8673 + }, + { + "epoch": 0.5933374375812299, + "grad_norm": 1.906184091757501, + "learning_rate": 3.7451807053521873e-06, + "loss": 0.3233, + "step": 8674 + }, + { + "epoch": 0.5934058417128395, + "grad_norm": 1.7488405865840242, + "learning_rate": 3.7441084334212806e-06, + "loss": 0.3224, + "step": 8675 + }, + { + "epoch": 0.593474245844449, + "grad_norm": 1.8584133193702614, + "learning_rate": 3.743036223135597e-06, + "loss": 0.3454, + "step": 8676 + }, + { + "epoch": 0.5935426499760585, + "grad_norm": 1.7184639606009868, + "learning_rate": 3.741964074547764e-06, + "loss": 0.2591, + "step": 8677 + }, + { + "epoch": 0.5936110541076681, + "grad_norm": 2.584683526245575, + "learning_rate": 3.740891987710411e-06, + "loss": 0.1308, + "step": 8678 + }, + { + "epoch": 0.5936794582392777, + "grad_norm": 2.2218128407578486, + "learning_rate": 3.739819962676159e-06, + "loss": 0.3397, + "step": 8679 + }, + { + "epoch": 0.5937478623708872, + "grad_norm": 1.7129394846160155, + "learning_rate": 3.7387479994976272e-06, + "loss": 0.2903, + "step": 8680 + }, + { + "epoch": 0.5938162665024967, + "grad_norm": 2.1815112151557767, + "learning_rate": 3.7376760982274353e-06, + "loss": 0.5374, + "step": 8681 + }, + { + "epoch": 0.5938846706341063, + "grad_norm": 1.393765861908499, + "learning_rate": 3.736604258918195e-06, + "loss": 0.1132, + "step": 8682 + }, + { + "epoch": 0.5939530747657158, + "grad_norm": 1.970762284055367, + "learning_rate": 3.7355324816225193e-06, + "loss": 0.4658, + "step": 8683 + }, + { + "epoch": 0.5940214788973254, + "grad_norm": 2.054723804573774, + "learning_rate": 3.734460766393014e-06, + "loss": 0.4463, + "step": 8684 + }, + { + "epoch": 0.594089883028935, + "grad_norm": 2.044907472372116, + "learning_rate": 3.7333891132822876e-06, + "loss": 0.5308, + "step": 8685 + }, + { + "epoch": 0.5941582871605445, + "grad_norm": 2.1493024853417553, + "learning_rate": 3.7323175223429396e-06, + "loss": 0.6513, + "step": 8686 + }, + { + "epoch": 0.594226691292154, + "grad_norm": 1.6419430524476533, + "learning_rate": 3.731245993627569e-06, + "loss": 0.3476, + "step": 8687 + }, + { + "epoch": 0.5942950954237636, + "grad_norm": 2.1987089037022973, + "learning_rate": 3.7301745271887718e-06, + "loss": 0.4335, + "step": 8688 + }, + { + "epoch": 0.5943634995553732, + "grad_norm": 1.989189000688178, + "learning_rate": 3.7291031230791413e-06, + "loss": 0.4908, + "step": 8689 + }, + { + "epoch": 0.5944319036869827, + "grad_norm": 2.573935989572106, + "learning_rate": 3.7280317813512655e-06, + "loss": 0.5243, + "step": 8690 + }, + { + "epoch": 0.5945003078185922, + "grad_norm": 1.9710061754268324, + "learning_rate": 3.7269605020577326e-06, + "loss": 0.2448, + "step": 8691 + }, + { + "epoch": 0.5945687119502018, + "grad_norm": 1.8433337752045358, + "learning_rate": 3.725889285251125e-06, + "loss": 0.4218, + "step": 8692 + }, + { + "epoch": 0.5946371160818114, + "grad_norm": 1.931939856916259, + "learning_rate": 3.7248181309840242e-06, + "loss": 0.5677, + "step": 8693 + }, + { + "epoch": 0.5947055202134209, + "grad_norm": 2.136377443325705, + "learning_rate": 3.723747039309007e-06, + "loss": 0.31, + "step": 8694 + }, + { + "epoch": 0.5947739243450304, + "grad_norm": 3.01651082839426, + "learning_rate": 3.7226760102786497e-06, + "loss": 0.3067, + "step": 8695 + }, + { + "epoch": 0.59484232847664, + "grad_norm": 1.6406569614021946, + "learning_rate": 3.721605043945521e-06, + "loss": 0.1779, + "step": 8696 + }, + { + "epoch": 0.5949107326082496, + "grad_norm": 1.4819293630121664, + "learning_rate": 3.7205341403621896e-06, + "loss": 0.2566, + "step": 8697 + }, + { + "epoch": 0.5949791367398591, + "grad_norm": 1.3549463326218047, + "learning_rate": 3.719463299581222e-06, + "loss": 0.147, + "step": 8698 + }, + { + "epoch": 0.5950475408714686, + "grad_norm": 1.4218116709471025, + "learning_rate": 3.7183925216551784e-06, + "loss": 0.1746, + "step": 8699 + }, + { + "epoch": 0.5951159450030782, + "grad_norm": 2.214225736642956, + "learning_rate": 3.7173218066366195e-06, + "loss": 0.5726, + "step": 8700 + }, + { + "epoch": 0.5951843491346878, + "grad_norm": 1.5413978885785407, + "learning_rate": 3.7162511545781004e-06, + "loss": 0.1782, + "step": 8701 + }, + { + "epoch": 0.5952527532662972, + "grad_norm": 1.9299783912125965, + "learning_rate": 3.7151805655321727e-06, + "loss": 0.4743, + "step": 8702 + }, + { + "epoch": 0.5953211573979068, + "grad_norm": 2.451377388898971, + "learning_rate": 3.7141100395513874e-06, + "loss": 0.4716, + "step": 8703 + }, + { + "epoch": 0.5953895615295164, + "grad_norm": 2.041832474784457, + "learning_rate": 3.7130395766882908e-06, + "loss": 0.5204, + "step": 8704 + }, + { + "epoch": 0.595457965661126, + "grad_norm": 2.345960998336165, + "learning_rate": 3.711969176995427e-06, + "loss": 0.419, + "step": 8705 + }, + { + "epoch": 0.5955263697927354, + "grad_norm": 2.978379193292491, + "learning_rate": 3.7108988405253353e-06, + "loss": 0.1344, + "step": 8706 + }, + { + "epoch": 0.595594773924345, + "grad_norm": 1.8733669441115095, + "learning_rate": 3.7098285673305546e-06, + "loss": 0.3371, + "step": 8707 + }, + { + "epoch": 0.5956631780559546, + "grad_norm": 2.152224144494168, + "learning_rate": 3.7087583574636175e-06, + "loss": 0.407, + "step": 8708 + }, + { + "epoch": 0.5957315821875642, + "grad_norm": 2.4536081676524764, + "learning_rate": 3.707688210977055e-06, + "loss": 0.5136, + "step": 8709 + }, + { + "epoch": 0.5957999863191736, + "grad_norm": 2.1204966834068673, + "learning_rate": 3.7066181279233964e-06, + "loss": 0.4354, + "step": 8710 + }, + { + "epoch": 0.5958683904507832, + "grad_norm": 2.427965283633502, + "learning_rate": 3.705548108355166e-06, + "loss": 0.6643, + "step": 8711 + }, + { + "epoch": 0.5959367945823928, + "grad_norm": 2.3884113852822506, + "learning_rate": 3.704478152324884e-06, + "loss": 0.4441, + "step": 8712 + }, + { + "epoch": 0.5960051987140024, + "grad_norm": 2.1314907449798315, + "learning_rate": 3.7034082598850698e-06, + "loss": 0.2784, + "step": 8713 + }, + { + "epoch": 0.5960736028456118, + "grad_norm": 2.155885603424289, + "learning_rate": 3.7023384310882404e-06, + "loss": 0.5728, + "step": 8714 + }, + { + "epoch": 0.5961420069772214, + "grad_norm": 2.7679258382530345, + "learning_rate": 3.701268665986908e-06, + "loss": 0.4169, + "step": 8715 + }, + { + "epoch": 0.596210411108831, + "grad_norm": 1.5428709892496497, + "learning_rate": 3.7001989646335796e-06, + "loss": 0.3312, + "step": 8716 + }, + { + "epoch": 0.5962788152404406, + "grad_norm": 1.9851306843077836, + "learning_rate": 3.699129327080764e-06, + "loss": 0.4674, + "step": 8717 + }, + { + "epoch": 0.59634721937205, + "grad_norm": 2.118573639430513, + "learning_rate": 3.6980597533809626e-06, + "loss": 0.4206, + "step": 8718 + }, + { + "epoch": 0.5964156235036596, + "grad_norm": 8.678800769356874, + "learning_rate": 3.6969902435866743e-06, + "loss": 0.6354, + "step": 8719 + }, + { + "epoch": 0.5964840276352692, + "grad_norm": 1.9740225035839867, + "learning_rate": 3.6959207977503988e-06, + "loss": 0.4486, + "step": 8720 + }, + { + "epoch": 0.5965524317668787, + "grad_norm": 1.9398436292447787, + "learning_rate": 3.6948514159246275e-06, + "loss": 0.4242, + "step": 8721 + }, + { + "epoch": 0.5966208358984882, + "grad_norm": 2.0901525322794865, + "learning_rate": 3.6937820981618495e-06, + "loss": 0.3723, + "step": 8722 + }, + { + "epoch": 0.5966892400300978, + "grad_norm": 2.3996732522894466, + "learning_rate": 3.6927128445145533e-06, + "loss": 0.5681, + "step": 8723 + }, + { + "epoch": 0.5967576441617074, + "grad_norm": 2.079899152479664, + "learning_rate": 3.691643655035225e-06, + "loss": 0.5477, + "step": 8724 + }, + { + "epoch": 0.5968260482933169, + "grad_norm": 2.7364739345826536, + "learning_rate": 3.690574529776345e-06, + "loss": 0.381, + "step": 8725 + }, + { + "epoch": 0.5968944524249264, + "grad_norm": 1.800683339533376, + "learning_rate": 3.6895054687903892e-06, + "loss": 0.3727, + "step": 8726 + }, + { + "epoch": 0.596962856556536, + "grad_norm": 1.4543826905536144, + "learning_rate": 3.688436472129834e-06, + "loss": 0.145, + "step": 8727 + }, + { + "epoch": 0.5970312606881456, + "grad_norm": 1.9744044498470354, + "learning_rate": 3.6873675398471504e-06, + "loss": 0.4331, + "step": 8728 + }, + { + "epoch": 0.5970996648197551, + "grad_norm": 2.1487011210223095, + "learning_rate": 3.686298671994805e-06, + "loss": 0.3369, + "step": 8729 + }, + { + "epoch": 0.5971680689513647, + "grad_norm": 2.33256769959063, + "learning_rate": 3.6852298686252664e-06, + "loss": 0.4044, + "step": 8730 + }, + { + "epoch": 0.5972364730829742, + "grad_norm": 2.4295664079425836, + "learning_rate": 3.6841611297909942e-06, + "loss": 0.7924, + "step": 8731 + }, + { + "epoch": 0.5973048772145838, + "grad_norm": 1.4736299558982184, + "learning_rate": 3.683092455544447e-06, + "loss": 0.1908, + "step": 8732 + }, + { + "epoch": 0.5973732813461933, + "grad_norm": 1.7827106435031983, + "learning_rate": 3.682023845938081e-06, + "loss": 0.2555, + "step": 8733 + }, + { + "epoch": 0.5974416854778029, + "grad_norm": 1.7716237174247287, + "learning_rate": 3.68095530102435e-06, + "loss": 0.2314, + "step": 8734 + }, + { + "epoch": 0.5975100896094124, + "grad_norm": 1.743800015740592, + "learning_rate": 3.6798868208557025e-06, + "loss": 0.2004, + "step": 8735 + }, + { + "epoch": 0.597578493741022, + "grad_norm": 2.2904436937425836, + "learning_rate": 3.678818405484584e-06, + "loss": 0.2275, + "step": 8736 + }, + { + "epoch": 0.5976468978726315, + "grad_norm": 2.013881557732608, + "learning_rate": 3.6777500549634395e-06, + "loss": 0.3663, + "step": 8737 + }, + { + "epoch": 0.5977153020042411, + "grad_norm": 2.675010235936198, + "learning_rate": 3.676681769344707e-06, + "loss": 0.4979, + "step": 8738 + }, + { + "epoch": 0.5977837061358506, + "grad_norm": 1.9094693873874045, + "learning_rate": 3.6756135486808227e-06, + "loss": 0.5317, + "step": 8739 + }, + { + "epoch": 0.5978521102674601, + "grad_norm": 1.1868386999688894, + "learning_rate": 3.6745453930242213e-06, + "loss": 0.1376, + "step": 8740 + }, + { + "epoch": 0.5979205143990697, + "grad_norm": 1.841099243887531, + "learning_rate": 3.673477302427333e-06, + "loss": 0.3212, + "step": 8741 + }, + { + "epoch": 0.5979889185306793, + "grad_norm": 1.8209378105766905, + "learning_rate": 3.6724092769425833e-06, + "loss": 0.2763, + "step": 8742 + }, + { + "epoch": 0.5980573226622888, + "grad_norm": 2.0355366298291067, + "learning_rate": 3.6713413166223967e-06, + "loss": 0.2586, + "step": 8743 + }, + { + "epoch": 0.5981257267938983, + "grad_norm": 2.223074198279933, + "learning_rate": 3.670273421519195e-06, + "loss": 0.4045, + "step": 8744 + }, + { + "epoch": 0.5981941309255079, + "grad_norm": 1.313581651078552, + "learning_rate": 3.669205591685396e-06, + "loss": 0.2588, + "step": 8745 + }, + { + "epoch": 0.5982625350571175, + "grad_norm": 2.2807154486406898, + "learning_rate": 3.6681378271734115e-06, + "loss": 0.3123, + "step": 8746 + }, + { + "epoch": 0.598330939188727, + "grad_norm": 2.22274346339743, + "learning_rate": 3.667070128035655e-06, + "loss": 0.4788, + "step": 8747 + }, + { + "epoch": 0.5983993433203365, + "grad_norm": 2.642536876325522, + "learning_rate": 3.666002494324533e-06, + "loss": 0.5359, + "step": 8748 + }, + { + "epoch": 0.5984677474519461, + "grad_norm": 1.942827468232725, + "learning_rate": 3.6649349260924504e-06, + "loss": 0.3606, + "step": 8749 + }, + { + "epoch": 0.5985361515835557, + "grad_norm": 1.8953476788557344, + "learning_rate": 3.6638674233918085e-06, + "loss": 0.4555, + "step": 8750 + }, + { + "epoch": 0.5986045557151652, + "grad_norm": 3.2330554998048786, + "learning_rate": 3.6627999862750062e-06, + "loss": 0.6561, + "step": 8751 + }, + { + "epoch": 0.5986729598467747, + "grad_norm": 1.9825795078088622, + "learning_rate": 3.6617326147944355e-06, + "loss": 0.409, + "step": 8752 + }, + { + "epoch": 0.5987413639783843, + "grad_norm": 2.45028493572115, + "learning_rate": 3.6606653090024918e-06, + "loss": 0.5623, + "step": 8753 + }, + { + "epoch": 0.5988097681099939, + "grad_norm": 2.6930790590027134, + "learning_rate": 3.659598068951563e-06, + "loss": 0.6213, + "step": 8754 + }, + { + "epoch": 0.5988781722416034, + "grad_norm": 2.143650833093899, + "learning_rate": 3.658530894694034e-06, + "loss": 0.3109, + "step": 8755 + }, + { + "epoch": 0.5989465763732129, + "grad_norm": 2.0998905579114204, + "learning_rate": 3.6574637862822854e-06, + "loss": 0.2825, + "step": 8756 + }, + { + "epoch": 0.5990149805048225, + "grad_norm": 2.1597626725920245, + "learning_rate": 3.656396743768698e-06, + "loss": 0.3623, + "step": 8757 + }, + { + "epoch": 0.5990833846364321, + "grad_norm": 2.0809940417012873, + "learning_rate": 3.655329767205647e-06, + "loss": 0.4386, + "step": 8758 + }, + { + "epoch": 0.5991517887680415, + "grad_norm": 2.0712677065599516, + "learning_rate": 3.654262856645503e-06, + "loss": 0.1811, + "step": 8759 + }, + { + "epoch": 0.5992201928996511, + "grad_norm": 2.373456961907962, + "learning_rate": 3.6531960121406383e-06, + "loss": 0.3979, + "step": 8760 + }, + { + "epoch": 0.5992885970312607, + "grad_norm": 2.1820656166962378, + "learning_rate": 3.652129233743416e-06, + "loss": 0.3673, + "step": 8761 + }, + { + "epoch": 0.5993570011628703, + "grad_norm": 2.0879989895653748, + "learning_rate": 3.651062521506198e-06, + "loss": 0.5693, + "step": 8762 + }, + { + "epoch": 0.5994254052944797, + "grad_norm": 2.163732806544582, + "learning_rate": 3.649995875481347e-06, + "loss": 0.5354, + "step": 8763 + }, + { + "epoch": 0.5994938094260893, + "grad_norm": 2.4719136465357447, + "learning_rate": 3.6489292957212176e-06, + "loss": 0.3656, + "step": 8764 + }, + { + "epoch": 0.5995622135576989, + "grad_norm": 1.9626917534286388, + "learning_rate": 3.6478627822781632e-06, + "loss": 0.4598, + "step": 8765 + }, + { + "epoch": 0.5996306176893085, + "grad_norm": 1.3284225439110509, + "learning_rate": 3.6467963352045317e-06, + "loss": 0.1605, + "step": 8766 + }, + { + "epoch": 0.5996990218209179, + "grad_norm": 2.3364392510387586, + "learning_rate": 3.6457299545526713e-06, + "loss": 0.6381, + "step": 8767 + }, + { + "epoch": 0.5997674259525275, + "grad_norm": 1.4490331237855882, + "learning_rate": 3.6446636403749246e-06, + "loss": 0.133, + "step": 8768 + }, + { + "epoch": 0.5998358300841371, + "grad_norm": 1.8380766925411054, + "learning_rate": 3.6435973927236313e-06, + "loss": 0.3097, + "step": 8769 + }, + { + "epoch": 0.5999042342157467, + "grad_norm": 2.038686867531071, + "learning_rate": 3.642531211651128e-06, + "loss": 0.3718, + "step": 8770 + }, + { + "epoch": 0.5999726383473561, + "grad_norm": 1.7231924904829246, + "learning_rate": 3.6414650972097475e-06, + "loss": 0.277, + "step": 8771 + }, + { + "epoch": 0.6000410424789657, + "grad_norm": 2.0631954075593373, + "learning_rate": 3.6403990494518194e-06, + "loss": 0.2984, + "step": 8772 + }, + { + "epoch": 0.6001094466105753, + "grad_norm": 3.3461220919317016, + "learning_rate": 3.639333068429671e-06, + "loss": 0.2858, + "step": 8773 + }, + { + "epoch": 0.6001778507421849, + "grad_norm": 1.5383284459349849, + "learning_rate": 3.6382671541956286e-06, + "loss": 0.2088, + "step": 8774 + }, + { + "epoch": 0.6002462548737943, + "grad_norm": 2.3332993719206616, + "learning_rate": 3.6372013068020097e-06, + "loss": 0.3737, + "step": 8775 + }, + { + "epoch": 0.6003146590054039, + "grad_norm": 1.7636863226314645, + "learning_rate": 3.63613552630113e-06, + "loss": 0.1844, + "step": 8776 + }, + { + "epoch": 0.6003830631370135, + "grad_norm": 2.1977011422767, + "learning_rate": 3.6350698127453067e-06, + "loss": 0.4771, + "step": 8777 + }, + { + "epoch": 0.600451467268623, + "grad_norm": 2.0919135281103722, + "learning_rate": 3.6340041661868473e-06, + "loss": 0.3091, + "step": 8778 + }, + { + "epoch": 0.6005198714002326, + "grad_norm": 1.996214726732993, + "learning_rate": 3.6329385866780587e-06, + "loss": 0.439, + "step": 8779 + }, + { + "epoch": 0.6005882755318421, + "grad_norm": 2.1096518158299795, + "learning_rate": 3.6318730742712473e-06, + "loss": 0.4924, + "step": 8780 + }, + { + "epoch": 0.6006566796634517, + "grad_norm": 1.7503001049415383, + "learning_rate": 3.6308076290187123e-06, + "loss": 0.3365, + "step": 8781 + }, + { + "epoch": 0.6007250837950612, + "grad_norm": 1.682332480282936, + "learning_rate": 3.6297422509727478e-06, + "loss": 0.3489, + "step": 8782 + }, + { + "epoch": 0.6007934879266708, + "grad_norm": 2.070911212936782, + "learning_rate": 3.6286769401856513e-06, + "loss": 0.4248, + "step": 8783 + }, + { + "epoch": 0.6008618920582803, + "grad_norm": 2.0030492247533442, + "learning_rate": 3.627611696709714e-06, + "loss": 0.2531, + "step": 8784 + }, + { + "epoch": 0.6009302961898899, + "grad_norm": 1.8315558355535928, + "learning_rate": 3.6265465205972227e-06, + "loss": 0.3086, + "step": 8785 + }, + { + "epoch": 0.6009987003214994, + "grad_norm": 1.6640490455262618, + "learning_rate": 3.625481411900459e-06, + "loss": 0.2921, + "step": 8786 + }, + { + "epoch": 0.601067104453109, + "grad_norm": 1.4688087563226562, + "learning_rate": 3.624416370671705e-06, + "loss": 0.2696, + "step": 8787 + }, + { + "epoch": 0.6011355085847185, + "grad_norm": 2.388931303993859, + "learning_rate": 3.6233513969632393e-06, + "loss": 0.3857, + "step": 8788 + }, + { + "epoch": 0.6012039127163281, + "grad_norm": 1.5301530957449974, + "learning_rate": 3.6222864908273334e-06, + "loss": 0.1816, + "step": 8789 + }, + { + "epoch": 0.6012723168479376, + "grad_norm": 1.9552985342440594, + "learning_rate": 3.621221652316261e-06, + "loss": 0.1622, + "step": 8790 + }, + { + "epoch": 0.6013407209795472, + "grad_norm": 2.008748233511907, + "learning_rate": 3.6201568814822864e-06, + "loss": 0.1498, + "step": 8791 + }, + { + "epoch": 0.6014091251111567, + "grad_norm": 4.16786320989781, + "learning_rate": 3.6190921783776755e-06, + "loss": 0.332, + "step": 8792 + }, + { + "epoch": 0.6014775292427663, + "grad_norm": 2.084653709342481, + "learning_rate": 3.6180275430546887e-06, + "loss": 0.3805, + "step": 8793 + }, + { + "epoch": 0.6015459333743758, + "grad_norm": 2.253303883682701, + "learning_rate": 3.6169629755655845e-06, + "loss": 0.6292, + "step": 8794 + }, + { + "epoch": 0.6016143375059854, + "grad_norm": 1.3742045220163874, + "learning_rate": 3.6158984759626157e-06, + "loss": 0.2355, + "step": 8795 + }, + { + "epoch": 0.6016827416375949, + "grad_norm": 1.6719775034774782, + "learning_rate": 3.6148340442980336e-06, + "loss": 0.2744, + "step": 8796 + }, + { + "epoch": 0.6017511457692044, + "grad_norm": 1.045350181288167, + "learning_rate": 3.613769680624085e-06, + "loss": 0.1069, + "step": 8797 + }, + { + "epoch": 0.601819549900814, + "grad_norm": 1.7731905609827283, + "learning_rate": 3.612705384993015e-06, + "loss": 0.2733, + "step": 8798 + }, + { + "epoch": 0.6018879540324236, + "grad_norm": 2.7299951188948595, + "learning_rate": 3.611641157457064e-06, + "loss": 0.4519, + "step": 8799 + }, + { + "epoch": 0.6019563581640331, + "grad_norm": 1.8098752731875256, + "learning_rate": 3.610576998068469e-06, + "loss": 0.3851, + "step": 8800 + }, + { + "epoch": 0.6020247622956426, + "grad_norm": 2.341227865567646, + "learning_rate": 3.6095129068794645e-06, + "loss": 0.2802, + "step": 8801 + }, + { + "epoch": 0.6020931664272522, + "grad_norm": 1.9418618878332259, + "learning_rate": 3.6084488839422806e-06, + "loss": 0.561, + "step": 8802 + }, + { + "epoch": 0.6021615705588618, + "grad_norm": 2.5858548661135665, + "learning_rate": 3.6073849293091444e-06, + "loss": 0.324, + "step": 8803 + }, + { + "epoch": 0.6022299746904713, + "grad_norm": 2.4665927636399427, + "learning_rate": 3.6063210430322824e-06, + "loss": 0.571, + "step": 8804 + }, + { + "epoch": 0.6022983788220808, + "grad_norm": 2.223239936013126, + "learning_rate": 3.605257225163913e-06, + "loss": 0.2008, + "step": 8805 + }, + { + "epoch": 0.6023667829536904, + "grad_norm": 2.351077793391948, + "learning_rate": 3.6041934757562546e-06, + "loss": 0.3752, + "step": 8806 + }, + { + "epoch": 0.6024351870853, + "grad_norm": 2.3392924326360274, + "learning_rate": 3.603129794861521e-06, + "loss": 0.5039, + "step": 8807 + }, + { + "epoch": 0.6025035912169096, + "grad_norm": 2.180593702152361, + "learning_rate": 3.6020661825319216e-06, + "loss": 0.5017, + "step": 8808 + }, + { + "epoch": 0.602571995348519, + "grad_norm": 1.7314933258129594, + "learning_rate": 3.601002638819665e-06, + "loss": 0.3095, + "step": 8809 + }, + { + "epoch": 0.6026403994801286, + "grad_norm": 3.0152147636660414, + "learning_rate": 3.5999391637769553e-06, + "loss": 0.4224, + "step": 8810 + }, + { + "epoch": 0.6027088036117382, + "grad_norm": 2.3450551722264565, + "learning_rate": 3.598875757455991e-06, + "loss": 0.3626, + "step": 8811 + }, + { + "epoch": 0.6027772077433476, + "grad_norm": 1.9781116965752001, + "learning_rate": 3.597812419908969e-06, + "loss": 0.1407, + "step": 8812 + }, + { + "epoch": 0.6028456118749572, + "grad_norm": 1.5097718380581826, + "learning_rate": 3.596749151188087e-06, + "loss": 0.2099, + "step": 8813 + }, + { + "epoch": 0.6029140160065668, + "grad_norm": 2.232808437570623, + "learning_rate": 3.595685951345533e-06, + "loss": 0.4476, + "step": 8814 + }, + { + "epoch": 0.6029824201381764, + "grad_norm": 2.3788244925545174, + "learning_rate": 3.594622820433492e-06, + "loss": 0.5893, + "step": 8815 + }, + { + "epoch": 0.6030508242697858, + "grad_norm": 2.2224498813876115, + "learning_rate": 3.5935597585041516e-06, + "loss": 0.3945, + "step": 8816 + }, + { + "epoch": 0.6031192284013954, + "grad_norm": 2.339453102579338, + "learning_rate": 3.592496765609689e-06, + "loss": 0.6856, + "step": 8817 + }, + { + "epoch": 0.603187632533005, + "grad_norm": 6.029970557139367, + "learning_rate": 3.5914338418022815e-06, + "loss": 0.3564, + "step": 8818 + }, + { + "epoch": 0.6032560366646146, + "grad_norm": 3.5811165044113147, + "learning_rate": 3.5903709871341034e-06, + "loss": 0.7713, + "step": 8819 + }, + { + "epoch": 0.603324440796224, + "grad_norm": 2.257291317880906, + "learning_rate": 3.5893082016573248e-06, + "loss": 0.2685, + "step": 8820 + }, + { + "epoch": 0.6033928449278336, + "grad_norm": 3.1887877705279957, + "learning_rate": 3.588245485424111e-06, + "loss": 0.2949, + "step": 8821 + }, + { + "epoch": 0.6034612490594432, + "grad_norm": 2.4020295224499617, + "learning_rate": 3.587182838486624e-06, + "loss": 0.4479, + "step": 8822 + }, + { + "epoch": 0.6035296531910528, + "grad_norm": 2.5695095953256493, + "learning_rate": 3.586120260897029e-06, + "loss": 0.5586, + "step": 8823 + }, + { + "epoch": 0.6035980573226623, + "grad_norm": 1.75844942209864, + "learning_rate": 3.585057752707478e-06, + "loss": 0.3956, + "step": 8824 + }, + { + "epoch": 0.6036664614542718, + "grad_norm": 2.884108662940063, + "learning_rate": 3.583995313970125e-06, + "loss": 0.3021, + "step": 8825 + }, + { + "epoch": 0.6037348655858814, + "grad_norm": 2.9461278014641725, + "learning_rate": 3.582932944737121e-06, + "loss": 0.7839, + "step": 8826 + }, + { + "epoch": 0.603803269717491, + "grad_norm": 2.1564855481458873, + "learning_rate": 3.58187064506061e-06, + "loss": 0.3198, + "step": 8827 + }, + { + "epoch": 0.6038716738491005, + "grad_norm": 1.5409126730833387, + "learning_rate": 3.5808084149927356e-06, + "loss": 0.1846, + "step": 8828 + }, + { + "epoch": 0.60394007798071, + "grad_norm": 1.7240665830142399, + "learning_rate": 3.5797462545856382e-06, + "loss": 0.1876, + "step": 8829 + }, + { + "epoch": 0.6040084821123196, + "grad_norm": 1.8851948778010439, + "learning_rate": 3.5786841638914526e-06, + "loss": 0.3256, + "step": 8830 + }, + { + "epoch": 0.6040768862439291, + "grad_norm": 1.8630185051970047, + "learning_rate": 3.5776221429623115e-06, + "loss": 0.2444, + "step": 8831 + }, + { + "epoch": 0.6041452903755387, + "grad_norm": 2.6101443361514547, + "learning_rate": 3.576560191850342e-06, + "loss": 0.3711, + "step": 8832 + }, + { + "epoch": 0.6042136945071482, + "grad_norm": 1.7606941127719948, + "learning_rate": 3.5754983106076757e-06, + "loss": 0.2871, + "step": 8833 + }, + { + "epoch": 0.6042820986387578, + "grad_norm": 1.7683849326971317, + "learning_rate": 3.57443649928643e-06, + "loss": 0.1475, + "step": 8834 + }, + { + "epoch": 0.6043505027703673, + "grad_norm": 2.0790386163545813, + "learning_rate": 3.573374757938724e-06, + "loss": 0.2315, + "step": 8835 + }, + { + "epoch": 0.6044189069019769, + "grad_norm": 1.880701670467239, + "learning_rate": 3.572313086616675e-06, + "loss": 0.4337, + "step": 8836 + }, + { + "epoch": 0.6044873110335864, + "grad_norm": 1.8522513512549696, + "learning_rate": 3.5712514853723943e-06, + "loss": 0.297, + "step": 8837 + }, + { + "epoch": 0.604555715165196, + "grad_norm": 2.4619831043747498, + "learning_rate": 3.5701899542579888e-06, + "loss": 0.3789, + "step": 8838 + }, + { + "epoch": 0.6046241192968055, + "grad_norm": 1.885680918238993, + "learning_rate": 3.5691284933255653e-06, + "loss": 0.3367, + "step": 8839 + }, + { + "epoch": 0.6046925234284151, + "grad_norm": 2.12669574171578, + "learning_rate": 3.5680671026272263e-06, + "loss": 0.2784, + "step": 8840 + }, + { + "epoch": 0.6047609275600246, + "grad_norm": 1.742075463186499, + "learning_rate": 3.5670057822150662e-06, + "loss": 0.2384, + "step": 8841 + }, + { + "epoch": 0.6048293316916342, + "grad_norm": 1.6220844872192361, + "learning_rate": 3.5659445321411815e-06, + "loss": 0.3312, + "step": 8842 + }, + { + "epoch": 0.6048977358232437, + "grad_norm": 2.2864074607235376, + "learning_rate": 3.5648833524576664e-06, + "loss": 0.6295, + "step": 8843 + }, + { + "epoch": 0.6049661399548533, + "grad_norm": 2.0408292913481234, + "learning_rate": 3.563822243216607e-06, + "loss": 0.2603, + "step": 8844 + }, + { + "epoch": 0.6050345440864628, + "grad_norm": 1.990059010994146, + "learning_rate": 3.5627612044700854e-06, + "loss": 0.4787, + "step": 8845 + }, + { + "epoch": 0.6051029482180724, + "grad_norm": 1.9965643318718365, + "learning_rate": 3.561700236270186e-06, + "loss": 0.3624, + "step": 8846 + }, + { + "epoch": 0.6051713523496819, + "grad_norm": 1.9334845078560152, + "learning_rate": 3.5606393386689837e-06, + "loss": 0.3823, + "step": 8847 + }, + { + "epoch": 0.6052397564812915, + "grad_norm": 3.059333633828982, + "learning_rate": 3.5595785117185523e-06, + "loss": 0.6427, + "step": 8848 + }, + { + "epoch": 0.605308160612901, + "grad_norm": 1.7103832427076229, + "learning_rate": 3.5585177554709648e-06, + "loss": 0.1837, + "step": 8849 + }, + { + "epoch": 0.6053765647445105, + "grad_norm": 2.4412485056460915, + "learning_rate": 3.557457069978286e-06, + "loss": 0.2807, + "step": 8850 + }, + { + "epoch": 0.6054449688761201, + "grad_norm": 2.2780426680309094, + "learning_rate": 3.556396455292579e-06, + "loss": 0.2085, + "step": 8851 + }, + { + "epoch": 0.6055133730077297, + "grad_norm": 1.7365593429078172, + "learning_rate": 3.5553359114659054e-06, + "loss": 0.397, + "step": 8852 + }, + { + "epoch": 0.6055817771393393, + "grad_norm": 1.9271537269588512, + "learning_rate": 3.5542754385503224e-06, + "loss": 0.3367, + "step": 8853 + }, + { + "epoch": 0.6056501812709487, + "grad_norm": 2.2843637380048394, + "learning_rate": 3.5532150365978825e-06, + "loss": 0.3894, + "step": 8854 + }, + { + "epoch": 0.6057185854025583, + "grad_norm": 2.0290239681132087, + "learning_rate": 3.5521547056606344e-06, + "loss": 0.3291, + "step": 8855 + }, + { + "epoch": 0.6057869895341679, + "grad_norm": 1.9334447244571158, + "learning_rate": 3.5510944457906255e-06, + "loss": 0.1616, + "step": 8856 + }, + { + "epoch": 0.6058553936657775, + "grad_norm": 2.707038538098674, + "learning_rate": 3.550034257039898e-06, + "loss": 0.4409, + "step": 8857 + }, + { + "epoch": 0.6059237977973869, + "grad_norm": 1.6950133934377996, + "learning_rate": 3.54897413946049e-06, + "loss": 0.3448, + "step": 8858 + }, + { + "epoch": 0.6059922019289965, + "grad_norm": 2.2642822825006985, + "learning_rate": 3.5479140931044393e-06, + "loss": 0.3385, + "step": 8859 + }, + { + "epoch": 0.6060606060606061, + "grad_norm": 2.1297901873275733, + "learning_rate": 3.546854118023777e-06, + "loss": 0.5529, + "step": 8860 + }, + { + "epoch": 0.6061290101922157, + "grad_norm": 2.3719371571393952, + "learning_rate": 3.545794214270529e-06, + "loss": 0.4892, + "step": 8861 + }, + { + "epoch": 0.6061974143238251, + "grad_norm": 1.8451772919685072, + "learning_rate": 3.544734381896725e-06, + "loss": 0.2813, + "step": 8862 + }, + { + "epoch": 0.6062658184554347, + "grad_norm": 1.9876216926146553, + "learning_rate": 3.543674620954386e-06, + "loss": 0.509, + "step": 8863 + }, + { + "epoch": 0.6063342225870443, + "grad_norm": 2.6158699476975555, + "learning_rate": 3.542614931495528e-06, + "loss": 0.6629, + "step": 8864 + }, + { + "epoch": 0.6064026267186539, + "grad_norm": 2.4825504668850193, + "learning_rate": 3.541555313572167e-06, + "loss": 0.6012, + "step": 8865 + }, + { + "epoch": 0.6064710308502633, + "grad_norm": 2.1320658064879683, + "learning_rate": 3.540495767236315e-06, + "loss": 0.6406, + "step": 8866 + }, + { + "epoch": 0.6065394349818729, + "grad_norm": 1.6766056487191685, + "learning_rate": 3.5394362925399783e-06, + "loss": 0.2685, + "step": 8867 + }, + { + "epoch": 0.6066078391134825, + "grad_norm": 2.8829756796535815, + "learning_rate": 3.53837688953516e-06, + "loss": 0.2716, + "step": 8868 + }, + { + "epoch": 0.606676243245092, + "grad_norm": 1.4081545542304674, + "learning_rate": 3.537317558273864e-06, + "loss": 0.2222, + "step": 8869 + }, + { + "epoch": 0.6067446473767015, + "grad_norm": 1.7511636237299593, + "learning_rate": 3.536258298808084e-06, + "loss": 0.3493, + "step": 8870 + }, + { + "epoch": 0.6068130515083111, + "grad_norm": 1.7664389506455107, + "learning_rate": 3.5351991111898145e-06, + "loss": 0.3597, + "step": 8871 + }, + { + "epoch": 0.6068814556399207, + "grad_norm": 2.044099130324149, + "learning_rate": 3.5341399954710466e-06, + "loss": 0.4624, + "step": 8872 + }, + { + "epoch": 0.6069498597715302, + "grad_norm": 1.1911246232664594, + "learning_rate": 3.5330809517037668e-06, + "loss": 0.117, + "step": 8873 + }, + { + "epoch": 0.6070182639031397, + "grad_norm": 2.005114676199783, + "learning_rate": 3.532021979939958e-06, + "loss": 0.4168, + "step": 8874 + }, + { + "epoch": 0.6070866680347493, + "grad_norm": 1.466231103679395, + "learning_rate": 3.5309630802315975e-06, + "loss": 0.1221, + "step": 8875 + }, + { + "epoch": 0.6071550721663589, + "grad_norm": 2.282619979437587, + "learning_rate": 3.5299042526306648e-06, + "loss": 0.2862, + "step": 8876 + }, + { + "epoch": 0.6072234762979684, + "grad_norm": 1.829397040196236, + "learning_rate": 3.5288454971891308e-06, + "loss": 0.2828, + "step": 8877 + }, + { + "epoch": 0.6072918804295779, + "grad_norm": 2.071958150532841, + "learning_rate": 3.5277868139589623e-06, + "loss": 0.3416, + "step": 8878 + }, + { + "epoch": 0.6073602845611875, + "grad_norm": 1.6998793709956328, + "learning_rate": 3.526728202992127e-06, + "loss": 0.1454, + "step": 8879 + }, + { + "epoch": 0.6074286886927971, + "grad_norm": 1.7913721243547975, + "learning_rate": 3.5256696643405856e-06, + "loss": 0.4687, + "step": 8880 + }, + { + "epoch": 0.6074970928244066, + "grad_norm": 3.1253661227329608, + "learning_rate": 3.5246111980562945e-06, + "loss": 0.7696, + "step": 8881 + }, + { + "epoch": 0.6075654969560161, + "grad_norm": 1.66136979571727, + "learning_rate": 3.5235528041912125e-06, + "loss": 0.2511, + "step": 8882 + }, + { + "epoch": 0.6076339010876257, + "grad_norm": 2.521089246596318, + "learning_rate": 3.522494482797289e-06, + "loss": 0.5098, + "step": 8883 + }, + { + "epoch": 0.6077023052192353, + "grad_norm": 2.479828197643676, + "learning_rate": 3.5214362339264706e-06, + "loss": 0.4758, + "step": 8884 + }, + { + "epoch": 0.6077707093508448, + "grad_norm": 1.43751109160415, + "learning_rate": 3.5203780576307013e-06, + "loss": 0.1401, + "step": 8885 + }, + { + "epoch": 0.6078391134824543, + "grad_norm": 2.0631860184566095, + "learning_rate": 3.5193199539619226e-06, + "loss": 0.303, + "step": 8886 + }, + { + "epoch": 0.6079075176140639, + "grad_norm": 1.5819782860449687, + "learning_rate": 3.5182619229720715e-06, + "loss": 0.197, + "step": 8887 + }, + { + "epoch": 0.6079759217456734, + "grad_norm": 2.0379553379681226, + "learning_rate": 3.5172039647130784e-06, + "loss": 0.4901, + "step": 8888 + }, + { + "epoch": 0.608044325877283, + "grad_norm": 1.9738314423419037, + "learning_rate": 3.5161460792368766e-06, + "loss": 0.4878, + "step": 8889 + }, + { + "epoch": 0.6081127300088925, + "grad_norm": 1.7200363893873438, + "learning_rate": 3.51508826659539e-06, + "loss": 0.2587, + "step": 8890 + }, + { + "epoch": 0.6081811341405021, + "grad_norm": 1.9479671504028735, + "learning_rate": 3.514030526840542e-06, + "loss": 0.3842, + "step": 8891 + }, + { + "epoch": 0.6082495382721116, + "grad_norm": 1.6984883854392836, + "learning_rate": 3.51297286002425e-06, + "loss": 0.3187, + "step": 8892 + }, + { + "epoch": 0.6083179424037212, + "grad_norm": 2.169820486149899, + "learning_rate": 3.511915266198433e-06, + "loss": 0.3351, + "step": 8893 + }, + { + "epoch": 0.6083863465353307, + "grad_norm": 1.7420592667125345, + "learning_rate": 3.5108577454150005e-06, + "loss": 0.2859, + "step": 8894 + }, + { + "epoch": 0.6084547506669403, + "grad_norm": 2.62360543150349, + "learning_rate": 3.5098002977258593e-06, + "loss": 0.4951, + "step": 8895 + }, + { + "epoch": 0.6085231547985498, + "grad_norm": 2.5879719732202635, + "learning_rate": 3.508742923182917e-06, + "loss": 0.1715, + "step": 8896 + }, + { + "epoch": 0.6085915589301594, + "grad_norm": 2.372869163106911, + "learning_rate": 3.5076856218380723e-06, + "loss": 0.3263, + "step": 8897 + }, + { + "epoch": 0.608659963061769, + "grad_norm": 1.9652520432395488, + "learning_rate": 3.5066283937432244e-06, + "loss": 0.3533, + "step": 8898 + }, + { + "epoch": 0.6087283671933785, + "grad_norm": 2.309142381569529, + "learning_rate": 3.505571238950267e-06, + "loss": 0.5089, + "step": 8899 + }, + { + "epoch": 0.608796771324988, + "grad_norm": 2.5811063578727422, + "learning_rate": 3.5045141575110875e-06, + "loss": 0.2692, + "step": 8900 + }, + { + "epoch": 0.6088651754565976, + "grad_norm": 2.023221232665418, + "learning_rate": 3.5034571494775753e-06, + "loss": 0.3945, + "step": 8901 + }, + { + "epoch": 0.6089335795882072, + "grad_norm": 1.8243921634497924, + "learning_rate": 3.5024002149016135e-06, + "loss": 0.2549, + "step": 8902 + }, + { + "epoch": 0.6090019837198167, + "grad_norm": 2.922378425272086, + "learning_rate": 3.5013433538350816e-06, + "loss": 0.4506, + "step": 8903 + }, + { + "epoch": 0.6090703878514262, + "grad_norm": 2.4269539276028396, + "learning_rate": 3.5002865663298536e-06, + "loss": 0.4031, + "step": 8904 + }, + { + "epoch": 0.6091387919830358, + "grad_norm": 2.15826448724432, + "learning_rate": 3.4992298524378044e-06, + "loss": 0.6622, + "step": 8905 + }, + { + "epoch": 0.6092071961146454, + "grad_norm": 2.34929588934626, + "learning_rate": 3.4981732122108003e-06, + "loss": 0.3633, + "step": 8906 + }, + { + "epoch": 0.6092756002462548, + "grad_norm": 1.9961870399368746, + "learning_rate": 3.4971166457007072e-06, + "loss": 0.4637, + "step": 8907 + }, + { + "epoch": 0.6093440043778644, + "grad_norm": 2.8640729946645536, + "learning_rate": 3.4960601529593874e-06, + "loss": 0.3101, + "step": 8908 + }, + { + "epoch": 0.609412408509474, + "grad_norm": 1.2302760836216555, + "learning_rate": 3.4950037340386973e-06, + "loss": 0.1219, + "step": 8909 + }, + { + "epoch": 0.6094808126410836, + "grad_norm": 3.0113683118178645, + "learning_rate": 3.4939473889904908e-06, + "loss": 0.2494, + "step": 8910 + }, + { + "epoch": 0.609549216772693, + "grad_norm": 2.082807950689523, + "learning_rate": 3.4928911178666175e-06, + "loss": 0.6243, + "step": 8911 + }, + { + "epoch": 0.6096176209043026, + "grad_norm": 2.711190786264685, + "learning_rate": 3.4918349207189283e-06, + "loss": 0.3193, + "step": 8912 + }, + { + "epoch": 0.6096860250359122, + "grad_norm": 1.8942361188855459, + "learning_rate": 3.4907787975992645e-06, + "loss": 0.3997, + "step": 8913 + }, + { + "epoch": 0.6097544291675218, + "grad_norm": 1.8144148051606073, + "learning_rate": 3.4897227485594637e-06, + "loss": 0.3094, + "step": 8914 + }, + { + "epoch": 0.6098228332991312, + "grad_norm": 1.704338529446287, + "learning_rate": 3.4886667736513647e-06, + "loss": 0.2576, + "step": 8915 + }, + { + "epoch": 0.6098912374307408, + "grad_norm": 1.9597621754218086, + "learning_rate": 3.4876108729267984e-06, + "loss": 0.4487, + "step": 8916 + }, + { + "epoch": 0.6099596415623504, + "grad_norm": 2.484049175897132, + "learning_rate": 3.486555046437593e-06, + "loss": 0.5079, + "step": 8917 + }, + { + "epoch": 0.61002804569396, + "grad_norm": 4.556538933487063, + "learning_rate": 3.485499294235575e-06, + "loss": 0.3273, + "step": 8918 + }, + { + "epoch": 0.6100964498255694, + "grad_norm": 1.8277640271619722, + "learning_rate": 3.4844436163725648e-06, + "loss": 0.3324, + "step": 8919 + }, + { + "epoch": 0.610164853957179, + "grad_norm": 2.226469185195519, + "learning_rate": 3.48338801290038e-06, + "loss": 0.3056, + "step": 8920 + }, + { + "epoch": 0.6102332580887886, + "grad_norm": 2.2578056559739443, + "learning_rate": 3.482332483870834e-06, + "loss": 0.516, + "step": 8921 + }, + { + "epoch": 0.6103016622203982, + "grad_norm": 2.4267230639665898, + "learning_rate": 3.48127702933574e-06, + "loss": 0.3409, + "step": 8922 + }, + { + "epoch": 0.6103700663520076, + "grad_norm": 2.5247945620508685, + "learning_rate": 3.4802216493469043e-06, + "loss": 0.4136, + "step": 8923 + }, + { + "epoch": 0.6104384704836172, + "grad_norm": 2.171172577318171, + "learning_rate": 3.4791663439561275e-06, + "loss": 0.45, + "step": 8924 + }, + { + "epoch": 0.6105068746152268, + "grad_norm": 1.6924127751079214, + "learning_rate": 3.4781111132152113e-06, + "loss": 0.3504, + "step": 8925 + }, + { + "epoch": 0.6105752787468363, + "grad_norm": 1.6956011926278185, + "learning_rate": 3.477055957175951e-06, + "loss": 0.3179, + "step": 8926 + }, + { + "epoch": 0.6106436828784458, + "grad_norm": 1.7742245309954736, + "learning_rate": 3.4760008758901378e-06, + "loss": 0.2722, + "step": 8927 + }, + { + "epoch": 0.6107120870100554, + "grad_norm": 2.4105773879342176, + "learning_rate": 3.474945869409561e-06, + "loss": 0.7503, + "step": 8928 + }, + { + "epoch": 0.610780491141665, + "grad_norm": 1.4712182063748702, + "learning_rate": 3.4738909377860063e-06, + "loss": 0.1495, + "step": 8929 + }, + { + "epoch": 0.6108488952732745, + "grad_norm": 1.946947682097042, + "learning_rate": 3.4728360810712523e-06, + "loss": 0.2989, + "step": 8930 + }, + { + "epoch": 0.610917299404884, + "grad_norm": 2.1367324187069943, + "learning_rate": 3.4717812993170774e-06, + "loss": 0.2901, + "step": 8931 + }, + { + "epoch": 0.6109857035364936, + "grad_norm": 2.2598219314721475, + "learning_rate": 3.4707265925752577e-06, + "loss": 0.381, + "step": 8932 + }, + { + "epoch": 0.6110541076681032, + "grad_norm": 2.1535545239048286, + "learning_rate": 3.4696719608975614e-06, + "loss": 0.2682, + "step": 8933 + }, + { + "epoch": 0.6111225117997127, + "grad_norm": 2.0513581109484007, + "learning_rate": 3.4686174043357547e-06, + "loss": 0.1889, + "step": 8934 + }, + { + "epoch": 0.6111909159313222, + "grad_norm": 2.0302747219094477, + "learning_rate": 3.467562922941602e-06, + "loss": 0.3228, + "step": 8935 + }, + { + "epoch": 0.6112593200629318, + "grad_norm": 2.3296917437508995, + "learning_rate": 3.4665085167668606e-06, + "loss": 0.3196, + "step": 8936 + }, + { + "epoch": 0.6113277241945414, + "grad_norm": 11.577468801994687, + "learning_rate": 3.4654541858632852e-06, + "loss": 0.2795, + "step": 8937 + }, + { + "epoch": 0.6113961283261509, + "grad_norm": 2.184149778813357, + "learning_rate": 3.4643999302826305e-06, + "loss": 0.3479, + "step": 8938 + }, + { + "epoch": 0.6114645324577604, + "grad_norm": 1.9868943374211159, + "learning_rate": 3.4633457500766413e-06, + "loss": 0.4306, + "step": 8939 + }, + { + "epoch": 0.61153293658937, + "grad_norm": 4.631205101413309, + "learning_rate": 3.462291645297063e-06, + "loss": 0.3366, + "step": 8940 + }, + { + "epoch": 0.6116013407209796, + "grad_norm": 3.366011438525411, + "learning_rate": 3.4612376159956353e-06, + "loss": 0.1975, + "step": 8941 + }, + { + "epoch": 0.6116697448525891, + "grad_norm": 2.0070104889707268, + "learning_rate": 3.460183662224097e-06, + "loss": 0.1433, + "step": 8942 + }, + { + "epoch": 0.6117381489841986, + "grad_norm": 2.7124478917456476, + "learning_rate": 3.459129784034181e-06, + "loss": 0.5604, + "step": 8943 + }, + { + "epoch": 0.6118065531158082, + "grad_norm": 2.186386799259286, + "learning_rate": 3.4580759814776156e-06, + "loss": 0.2956, + "step": 8944 + }, + { + "epoch": 0.6118749572474177, + "grad_norm": 2.086780863372682, + "learning_rate": 3.457022254606127e-06, + "loss": 0.5752, + "step": 8945 + }, + { + "epoch": 0.6119433613790273, + "grad_norm": 2.5402944422504854, + "learning_rate": 3.4559686034714367e-06, + "loss": 0.1357, + "step": 8946 + }, + { + "epoch": 0.6120117655106369, + "grad_norm": 1.5667764459313367, + "learning_rate": 3.4549150281252635e-06, + "loss": 0.3697, + "step": 8947 + }, + { + "epoch": 0.6120801696422464, + "grad_norm": 2.0073329430763804, + "learning_rate": 3.4538615286193224e-06, + "loss": 0.3419, + "step": 8948 + }, + { + "epoch": 0.6121485737738559, + "grad_norm": 1.8156470363087251, + "learning_rate": 3.4528081050053243e-06, + "loss": 0.2733, + "step": 8949 + }, + { + "epoch": 0.6122169779054655, + "grad_norm": 2.2699143682200478, + "learning_rate": 3.4517547573349725e-06, + "loss": 0.1812, + "step": 8950 + }, + { + "epoch": 0.612285382037075, + "grad_norm": 1.9392959823025449, + "learning_rate": 3.450701485659975e-06, + "loss": 0.3625, + "step": 8951 + }, + { + "epoch": 0.6123537861686846, + "grad_norm": 2.1170185767770073, + "learning_rate": 3.4496482900320314e-06, + "loss": 0.6091, + "step": 8952 + }, + { + "epoch": 0.6124221903002941, + "grad_norm": 1.8550056240929642, + "learning_rate": 3.4485951705028364e-06, + "loss": 0.3637, + "step": 8953 + }, + { + "epoch": 0.6124905944319037, + "grad_norm": 2.3999851154322798, + "learning_rate": 3.447542127124081e-06, + "loss": 0.597, + "step": 8954 + }, + { + "epoch": 0.6125589985635133, + "grad_norm": 2.4534957386918164, + "learning_rate": 3.446489159947456e-06, + "loss": 0.599, + "step": 8955 + }, + { + "epoch": 0.6126274026951228, + "grad_norm": 1.8795495864997398, + "learning_rate": 3.4454362690246447e-06, + "loss": 0.3672, + "step": 8956 + }, + { + "epoch": 0.6126958068267323, + "grad_norm": 2.4074433637257586, + "learning_rate": 3.444383454407326e-06, + "loss": 0.4789, + "step": 8957 + }, + { + "epoch": 0.6127642109583419, + "grad_norm": 3.237778959443895, + "learning_rate": 3.443330716147182e-06, + "loss": 0.1932, + "step": 8958 + }, + { + "epoch": 0.6128326150899515, + "grad_norm": 2.6091014924301446, + "learning_rate": 3.442278054295883e-06, + "loss": 0.559, + "step": 8959 + }, + { + "epoch": 0.6129010192215609, + "grad_norm": 1.4039904057069308, + "learning_rate": 3.4412254689050973e-06, + "loss": 0.1342, + "step": 8960 + }, + { + "epoch": 0.6129694233531705, + "grad_norm": 2.3885089260363355, + "learning_rate": 3.4401729600264934e-06, + "loss": 0.4247, + "step": 8961 + }, + { + "epoch": 0.6130378274847801, + "grad_norm": 2.6361246301496175, + "learning_rate": 3.4391205277117345e-06, + "loss": 0.9354, + "step": 8962 + }, + { + "epoch": 0.6131062316163897, + "grad_norm": 2.2007391909696024, + "learning_rate": 3.438068172012477e-06, + "loss": 0.3753, + "step": 8963 + }, + { + "epoch": 0.6131746357479991, + "grad_norm": 2.163627961321411, + "learning_rate": 3.4370158929803753e-06, + "loss": 0.2704, + "step": 8964 + }, + { + "epoch": 0.6132430398796087, + "grad_norm": 1.0980000330875257, + "learning_rate": 3.4359636906670825e-06, + "loss": 0.0975, + "step": 8965 + }, + { + "epoch": 0.6133114440112183, + "grad_norm": 2.954553591610301, + "learning_rate": 3.434911565124244e-06, + "loss": 0.2253, + "step": 8966 + }, + { + "epoch": 0.6133798481428279, + "grad_norm": 2.2662736161564303, + "learning_rate": 3.4338595164035027e-06, + "loss": 0.331, + "step": 8967 + }, + { + "epoch": 0.6134482522744373, + "grad_norm": 1.9369921287813403, + "learning_rate": 3.4328075445565e-06, + "loss": 0.3649, + "step": 8968 + }, + { + "epoch": 0.6135166564060469, + "grad_norm": 3.060532677574311, + "learning_rate": 3.4317556496348714e-06, + "loss": 0.2956, + "step": 8969 + }, + { + "epoch": 0.6135850605376565, + "grad_norm": 2.2749424898298325, + "learning_rate": 3.4307038316902464e-06, + "loss": 0.6221, + "step": 8970 + }, + { + "epoch": 0.6136534646692661, + "grad_norm": 2.5193086865914456, + "learning_rate": 3.429652090774256e-06, + "loss": 0.4408, + "step": 8971 + }, + { + "epoch": 0.6137218688008755, + "grad_norm": 2.0132602970334137, + "learning_rate": 3.4286004269385254e-06, + "loss": 0.4517, + "step": 8972 + }, + { + "epoch": 0.6137902729324851, + "grad_norm": 1.6694258505325623, + "learning_rate": 3.4275488402346734e-06, + "loss": 0.295, + "step": 8973 + }, + { + "epoch": 0.6138586770640947, + "grad_norm": 1.6062021227948453, + "learning_rate": 3.4264973307143177e-06, + "loss": 0.2922, + "step": 8974 + }, + { + "epoch": 0.6139270811957043, + "grad_norm": 1.9567696554508545, + "learning_rate": 3.425445898429072e-06, + "loss": 0.1901, + "step": 8975 + }, + { + "epoch": 0.6139954853273137, + "grad_norm": 2.159494281350522, + "learning_rate": 3.4243945434305447e-06, + "loss": 0.3149, + "step": 8976 + }, + { + "epoch": 0.6140638894589233, + "grad_norm": 1.9601708733570788, + "learning_rate": 3.4233432657703414e-06, + "loss": 0.3121, + "step": 8977 + }, + { + "epoch": 0.6141322935905329, + "grad_norm": 2.496587784997587, + "learning_rate": 3.4222920655000647e-06, + "loss": 0.4518, + "step": 8978 + }, + { + "epoch": 0.6142006977221424, + "grad_norm": 2.3067301862805656, + "learning_rate": 3.421240942671312e-06, + "loss": 0.456, + "step": 8979 + }, + { + "epoch": 0.6142691018537519, + "grad_norm": 2.4852740832170075, + "learning_rate": 3.4201898973356757e-06, + "loss": 0.266, + "step": 8980 + }, + { + "epoch": 0.6143375059853615, + "grad_norm": 2.0976726311127507, + "learning_rate": 3.4191389295447486e-06, + "loss": 0.44, + "step": 8981 + }, + { + "epoch": 0.6144059101169711, + "grad_norm": 2.1228795621722187, + "learning_rate": 3.418088039350119e-06, + "loss": 0.2852, + "step": 8982 + }, + { + "epoch": 0.6144743142485806, + "grad_norm": 2.148715039446497, + "learning_rate": 3.4170372268033665e-06, + "loss": 0.4689, + "step": 8983 + }, + { + "epoch": 0.6145427183801901, + "grad_norm": 2.084133382402072, + "learning_rate": 3.4159864919560696e-06, + "loss": 0.1716, + "step": 8984 + }, + { + "epoch": 0.6146111225117997, + "grad_norm": 1.3912508284968181, + "learning_rate": 3.4149358348598065e-06, + "loss": 0.191, + "step": 8985 + }, + { + "epoch": 0.6146795266434093, + "grad_norm": 2.5910758630452277, + "learning_rate": 3.4138852555661467e-06, + "loss": 0.3107, + "step": 8986 + }, + { + "epoch": 0.6147479307750188, + "grad_norm": 2.0777675995355014, + "learning_rate": 3.412834754126657e-06, + "loss": 0.463, + "step": 8987 + }, + { + "epoch": 0.6148163349066283, + "grad_norm": 2.63247687580498, + "learning_rate": 3.411784330592902e-06, + "loss": 0.4647, + "step": 8988 + }, + { + "epoch": 0.6148847390382379, + "grad_norm": 5.1434579713041915, + "learning_rate": 3.410733985016441e-06, + "loss": 0.5517, + "step": 8989 + }, + { + "epoch": 0.6149531431698475, + "grad_norm": 2.3225123833407273, + "learning_rate": 3.409683717448831e-06, + "loss": 0.1599, + "step": 8990 + }, + { + "epoch": 0.615021547301457, + "grad_norm": 2.682250073798693, + "learning_rate": 3.4086335279416237e-06, + "loss": 0.6886, + "step": 8991 + }, + { + "epoch": 0.6150899514330666, + "grad_norm": 2.3144009869976223, + "learning_rate": 3.4075834165463678e-06, + "loss": 0.3657, + "step": 8992 + }, + { + "epoch": 0.6151583555646761, + "grad_norm": 1.9432111252255913, + "learning_rate": 3.406533383314608e-06, + "loss": 0.2084, + "step": 8993 + }, + { + "epoch": 0.6152267596962857, + "grad_norm": 2.0455319970933195, + "learning_rate": 3.405483428297883e-06, + "loss": 0.3638, + "step": 8994 + }, + { + "epoch": 0.6152951638278952, + "grad_norm": 2.1310706658857317, + "learning_rate": 3.404433551547733e-06, + "loss": 0.5254, + "step": 8995 + }, + { + "epoch": 0.6153635679595048, + "grad_norm": 2.439793907289147, + "learning_rate": 3.403383753115688e-06, + "loss": 0.2134, + "step": 8996 + }, + { + "epoch": 0.6154319720911143, + "grad_norm": 1.7429544146472762, + "learning_rate": 3.4023340330532785e-06, + "loss": 0.3303, + "step": 8997 + }, + { + "epoch": 0.6155003762227238, + "grad_norm": 2.164195714217521, + "learning_rate": 3.401284391412031e-06, + "loss": 0.3076, + "step": 8998 + }, + { + "epoch": 0.6155687803543334, + "grad_norm": 2.112957497721717, + "learning_rate": 3.4002348282434637e-06, + "loss": 0.5269, + "step": 8999 + }, + { + "epoch": 0.615637184485943, + "grad_norm": 2.859006138757958, + "learning_rate": 3.399185343599096e-06, + "loss": 0.1916, + "step": 9000 + }, + { + "epoch": 0.6157055886175525, + "grad_norm": 1.55945474223893, + "learning_rate": 3.3981359375304425e-06, + "loss": 0.2275, + "step": 9001 + }, + { + "epoch": 0.615773992749162, + "grad_norm": 1.772457318262866, + "learning_rate": 3.397086610089014e-06, + "loss": 0.301, + "step": 9002 + }, + { + "epoch": 0.6158423968807716, + "grad_norm": 1.9219037379465522, + "learning_rate": 3.396037361326313e-06, + "loss": 0.2375, + "step": 9003 + }, + { + "epoch": 0.6159108010123812, + "grad_norm": 1.5856926028959777, + "learning_rate": 3.394988191293845e-06, + "loss": 0.2232, + "step": 9004 + }, + { + "epoch": 0.6159792051439907, + "grad_norm": 2.008896657448029, + "learning_rate": 3.393939100043108e-06, + "loss": 0.464, + "step": 9005 + }, + { + "epoch": 0.6160476092756002, + "grad_norm": 2.169409305935391, + "learning_rate": 3.392890087625594e-06, + "loss": 0.1495, + "step": 9006 + }, + { + "epoch": 0.6161160134072098, + "grad_norm": 1.7727838689696673, + "learning_rate": 3.3918411540927957e-06, + "loss": 0.3005, + "step": 9007 + }, + { + "epoch": 0.6161844175388194, + "grad_norm": 2.767729793422793, + "learning_rate": 3.3907922994961994e-06, + "loss": 0.435, + "step": 9008 + }, + { + "epoch": 0.6162528216704289, + "grad_norm": 2.022519779868946, + "learning_rate": 3.3897435238872877e-06, + "loss": 0.4978, + "step": 9009 + }, + { + "epoch": 0.6163212258020384, + "grad_norm": 1.947013176688046, + "learning_rate": 3.388694827317539e-06, + "loss": 0.5205, + "step": 9010 + }, + { + "epoch": 0.616389629933648, + "grad_norm": 2.0674635880826076, + "learning_rate": 3.387646209838429e-06, + "loss": 0.5486, + "step": 9011 + }, + { + "epoch": 0.6164580340652576, + "grad_norm": 2.104775992461023, + "learning_rate": 3.3865976715014304e-06, + "loss": 0.2417, + "step": 9012 + }, + { + "epoch": 0.6165264381968671, + "grad_norm": 2.545947268668505, + "learning_rate": 3.3855492123580076e-06, + "loss": 0.4488, + "step": 9013 + }, + { + "epoch": 0.6165948423284766, + "grad_norm": 2.539035315734886, + "learning_rate": 3.3845008324596268e-06, + "loss": 0.3184, + "step": 9014 + }, + { + "epoch": 0.6166632464600862, + "grad_norm": 2.382847832589765, + "learning_rate": 3.383452531857746e-06, + "loss": 0.2769, + "step": 9015 + }, + { + "epoch": 0.6167316505916958, + "grad_norm": 1.823954810217725, + "learning_rate": 3.3824043106038207e-06, + "loss": 0.3594, + "step": 9016 + }, + { + "epoch": 0.6168000547233052, + "grad_norm": 2.1291443087964788, + "learning_rate": 3.381356168749304e-06, + "loss": 0.3225, + "step": 9017 + }, + { + "epoch": 0.6168684588549148, + "grad_norm": 2.4210512913947353, + "learning_rate": 3.3803081063456433e-06, + "loss": 0.3987, + "step": 9018 + }, + { + "epoch": 0.6169368629865244, + "grad_norm": 1.5564946162384672, + "learning_rate": 3.37926012344428e-06, + "loss": 0.2789, + "step": 9019 + }, + { + "epoch": 0.617005267118134, + "grad_norm": 2.1648255528126077, + "learning_rate": 3.378212220096656e-06, + "loss": 0.4666, + "step": 9020 + }, + { + "epoch": 0.6170736712497434, + "grad_norm": 1.9566999828910232, + "learning_rate": 3.37716439635421e-06, + "loss": 0.5511, + "step": 9021 + }, + { + "epoch": 0.617142075381353, + "grad_norm": 1.608719599702602, + "learning_rate": 3.376116652268372e-06, + "loss": 0.244, + "step": 9022 + }, + { + "epoch": 0.6172104795129626, + "grad_norm": 2.1878486048990324, + "learning_rate": 3.375068987890569e-06, + "loss": 0.5717, + "step": 9023 + }, + { + "epoch": 0.6172788836445722, + "grad_norm": 2.053879375869606, + "learning_rate": 3.374021403272228e-06, + "loss": 0.3885, + "step": 9024 + }, + { + "epoch": 0.6173472877761816, + "grad_norm": 1.7236597550747599, + "learning_rate": 3.372973898464768e-06, + "loss": 0.4301, + "step": 9025 + }, + { + "epoch": 0.6174156919077912, + "grad_norm": 2.0706363784518858, + "learning_rate": 3.3719264735196056e-06, + "loss": 0.3378, + "step": 9026 + }, + { + "epoch": 0.6174840960394008, + "grad_norm": 1.8952296041551806, + "learning_rate": 3.370879128488154e-06, + "loss": 0.4225, + "step": 9027 + }, + { + "epoch": 0.6175525001710104, + "grad_norm": 2.2059294659594855, + "learning_rate": 3.369831863421822e-06, + "loss": 0.2959, + "step": 9028 + }, + { + "epoch": 0.6176209043026198, + "grad_norm": 1.8428627886030882, + "learning_rate": 3.3687846783720134e-06, + "loss": 0.4772, + "step": 9029 + }, + { + "epoch": 0.6176893084342294, + "grad_norm": 2.3253825183216383, + "learning_rate": 3.3677375733901283e-06, + "loss": 0.2025, + "step": 9030 + }, + { + "epoch": 0.617757712565839, + "grad_norm": 2.086280808789854, + "learning_rate": 3.3666905485275676e-06, + "loss": 0.5743, + "step": 9031 + }, + { + "epoch": 0.6178261166974486, + "grad_norm": 1.7883034890968112, + "learning_rate": 3.365643603835722e-06, + "loss": 0.4156, + "step": 9032 + }, + { + "epoch": 0.617894520829058, + "grad_norm": 2.6601500826188573, + "learning_rate": 3.3645967393659797e-06, + "loss": 0.5802, + "step": 9033 + }, + { + "epoch": 0.6179629249606676, + "grad_norm": 1.7738877619715112, + "learning_rate": 3.363549955169727e-06, + "loss": 0.1824, + "step": 9034 + }, + { + "epoch": 0.6180313290922772, + "grad_norm": 2.10197432641476, + "learning_rate": 3.3625032512983454e-06, + "loss": 0.2669, + "step": 9035 + }, + { + "epoch": 0.6180997332238867, + "grad_norm": 2.5472226929435955, + "learning_rate": 3.3614566278032102e-06, + "loss": 0.6452, + "step": 9036 + }, + { + "epoch": 0.6181681373554962, + "grad_norm": 1.6522558112224814, + "learning_rate": 3.3604100847356974e-06, + "loss": 0.2438, + "step": 9037 + }, + { + "epoch": 0.6182365414871058, + "grad_norm": 2.0218660445803573, + "learning_rate": 3.3593636221471753e-06, + "loss": 0.4817, + "step": 9038 + }, + { + "epoch": 0.6183049456187154, + "grad_norm": 2.692850387092448, + "learning_rate": 3.358317240089008e-06, + "loss": 0.3846, + "step": 9039 + }, + { + "epoch": 0.6183733497503249, + "grad_norm": 1.9155594596727892, + "learning_rate": 3.3572709386125567e-06, + "loss": 0.3217, + "step": 9040 + }, + { + "epoch": 0.6184417538819345, + "grad_norm": 2.090988394739701, + "learning_rate": 3.356224717769183e-06, + "loss": 0.3569, + "step": 9041 + }, + { + "epoch": 0.618510158013544, + "grad_norm": 2.0476142989616575, + "learning_rate": 3.355178577610238e-06, + "loss": 0.4468, + "step": 9042 + }, + { + "epoch": 0.6185785621451536, + "grad_norm": 1.6669290549642994, + "learning_rate": 3.35413251818707e-06, + "loss": 0.3403, + "step": 9043 + }, + { + "epoch": 0.6186469662767631, + "grad_norm": 1.683367764912404, + "learning_rate": 3.3530865395510272e-06, + "loss": 0.1743, + "step": 9044 + }, + { + "epoch": 0.6187153704083727, + "grad_norm": 2.1695217103511437, + "learning_rate": 3.352040641753449e-06, + "loss": 0.4066, + "step": 9045 + }, + { + "epoch": 0.6187837745399822, + "grad_norm": 1.8399846940207345, + "learning_rate": 3.3509948248456735e-06, + "loss": 0.253, + "step": 9046 + }, + { + "epoch": 0.6188521786715918, + "grad_norm": 2.509526258401554, + "learning_rate": 3.349949088879036e-06, + "loss": 0.3943, + "step": 9047 + }, + { + "epoch": 0.6189205828032013, + "grad_norm": 2.300208455984548, + "learning_rate": 3.3489034339048653e-06, + "loss": 0.6735, + "step": 9048 + }, + { + "epoch": 0.6189889869348109, + "grad_norm": 2.5055320448652276, + "learning_rate": 3.347857859974485e-06, + "loss": 0.4018, + "step": 9049 + }, + { + "epoch": 0.6190573910664204, + "grad_norm": 1.489323816556526, + "learning_rate": 3.3468123671392196e-06, + "loss": 0.2551, + "step": 9050 + }, + { + "epoch": 0.61912579519803, + "grad_norm": 2.6026063319918578, + "learning_rate": 3.345766955450388e-06, + "loss": 0.4808, + "step": 9051 + }, + { + "epoch": 0.6191941993296395, + "grad_norm": 2.2524825281953773, + "learning_rate": 3.344721624959302e-06, + "loss": 0.4994, + "step": 9052 + }, + { + "epoch": 0.6192626034612491, + "grad_norm": 2.106353616288213, + "learning_rate": 3.343676375717271e-06, + "loss": 0.1863, + "step": 9053 + }, + { + "epoch": 0.6193310075928586, + "grad_norm": 3.6370673744940287, + "learning_rate": 3.3426312077756033e-06, + "loss": 0.4912, + "step": 9054 + }, + { + "epoch": 0.6193994117244681, + "grad_norm": 1.7019701419493543, + "learning_rate": 3.341586121185599e-06, + "loss": 0.3165, + "step": 9055 + }, + { + "epoch": 0.6194678158560777, + "grad_norm": 2.1262485994214395, + "learning_rate": 3.340541115998555e-06, + "loss": 0.4188, + "step": 9056 + }, + { + "epoch": 0.6195362199876873, + "grad_norm": 2.5933936172985033, + "learning_rate": 3.3394961922657677e-06, + "loss": 0.7608, + "step": 9057 + }, + { + "epoch": 0.6196046241192968, + "grad_norm": 1.9320699293782102, + "learning_rate": 3.3384513500385262e-06, + "loss": 0.4125, + "step": 9058 + }, + { + "epoch": 0.6196730282509063, + "grad_norm": 2.1649815780173043, + "learning_rate": 3.337406589368113e-06, + "loss": 0.6342, + "step": 9059 + }, + { + "epoch": 0.6197414323825159, + "grad_norm": 1.9290680837125564, + "learning_rate": 3.336361910305815e-06, + "loss": 0.3407, + "step": 9060 + }, + { + "epoch": 0.6198098365141255, + "grad_norm": 1.794162286502661, + "learning_rate": 3.335317312902908e-06, + "loss": 0.3095, + "step": 9061 + }, + { + "epoch": 0.619878240645735, + "grad_norm": 2.207667281065008, + "learning_rate": 3.334272797210667e-06, + "loss": 0.3455, + "step": 9062 + }, + { + "epoch": 0.6199466447773445, + "grad_norm": 2.8556504302667745, + "learning_rate": 3.333228363280359e-06, + "loss": 0.1735, + "step": 9063 + }, + { + "epoch": 0.6200150489089541, + "grad_norm": 1.4160468025721449, + "learning_rate": 3.332184011163253e-06, + "loss": 0.16, + "step": 9064 + }, + { + "epoch": 0.6200834530405637, + "grad_norm": 2.0366273787078426, + "learning_rate": 3.3311397409106095e-06, + "loss": 0.4782, + "step": 9065 + }, + { + "epoch": 0.6201518571721732, + "grad_norm": 1.46587838573151, + "learning_rate": 3.3300955525736857e-06, + "loss": 0.2105, + "step": 9066 + }, + { + "epoch": 0.6202202613037827, + "grad_norm": 1.93839913701248, + "learning_rate": 3.329051446203737e-06, + "loss": 0.4265, + "step": 9067 + }, + { + "epoch": 0.6202886654353923, + "grad_norm": 2.288530763181382, + "learning_rate": 3.328007421852012e-06, + "loss": 0.5936, + "step": 9068 + }, + { + "epoch": 0.6203570695670019, + "grad_norm": 2.225892966811849, + "learning_rate": 3.326963479569755e-06, + "loss": 0.5781, + "step": 9069 + }, + { + "epoch": 0.6204254736986115, + "grad_norm": 2.4079676912907035, + "learning_rate": 3.32591961940821e-06, + "loss": 0.2099, + "step": 9070 + }, + { + "epoch": 0.6204938778302209, + "grad_norm": 2.217675069544862, + "learning_rate": 3.324875841418615e-06, + "loss": 0.4209, + "step": 9071 + }, + { + "epoch": 0.6205622819618305, + "grad_norm": 2.3909711774120277, + "learning_rate": 3.3238321456522034e-06, + "loss": 0.4432, + "step": 9072 + }, + { + "epoch": 0.6206306860934401, + "grad_norm": 2.509009221307466, + "learning_rate": 3.322788532160203e-06, + "loss": 0.2822, + "step": 9073 + }, + { + "epoch": 0.6206990902250495, + "grad_norm": 2.092868313260866, + "learning_rate": 3.321745000993841e-06, + "loss": 0.4713, + "step": 9074 + }, + { + "epoch": 0.6207674943566591, + "grad_norm": 2.062651585297027, + "learning_rate": 3.3207015522043394e-06, + "loss": 0.2754, + "step": 9075 + }, + { + "epoch": 0.6208358984882687, + "grad_norm": 3.327863817309158, + "learning_rate": 3.3196581858429134e-06, + "loss": 0.4889, + "step": 9076 + }, + { + "epoch": 0.6209043026198783, + "grad_norm": 1.8974281893273384, + "learning_rate": 3.3186149019607793e-06, + "loss": 0.2758, + "step": 9077 + }, + { + "epoch": 0.6209727067514877, + "grad_norm": 1.812240183229416, + "learning_rate": 3.3175717006091446e-06, + "loss": 0.3622, + "step": 9078 + }, + { + "epoch": 0.6210411108830973, + "grad_norm": 2.932697875557577, + "learning_rate": 3.3165285818392133e-06, + "loss": 0.5073, + "step": 9079 + }, + { + "epoch": 0.6211095150147069, + "grad_norm": 1.998560704213128, + "learning_rate": 3.3154855457021906e-06, + "loss": 0.3545, + "step": 9080 + }, + { + "epoch": 0.6211779191463165, + "grad_norm": 1.7445548357054659, + "learning_rate": 3.314442592249272e-06, + "loss": 0.2699, + "step": 9081 + }, + { + "epoch": 0.621246323277926, + "grad_norm": 2.072009783599743, + "learning_rate": 3.3133997215316505e-06, + "loss": 0.4741, + "step": 9082 + }, + { + "epoch": 0.6213147274095355, + "grad_norm": 2.3249438140602097, + "learning_rate": 3.3123569336005135e-06, + "loss": 0.4131, + "step": 9083 + }, + { + "epoch": 0.6213831315411451, + "grad_norm": 2.09933408130041, + "learning_rate": 3.31131422850705e-06, + "loss": 0.1516, + "step": 9084 + }, + { + "epoch": 0.6214515356727547, + "grad_norm": 1.5710709600061863, + "learning_rate": 3.3102716063024378e-06, + "loss": 0.224, + "step": 9085 + }, + { + "epoch": 0.6215199398043642, + "grad_norm": 1.7374404931341272, + "learning_rate": 3.3092290670378536e-06, + "loss": 0.3537, + "step": 9086 + }, + { + "epoch": 0.6215883439359737, + "grad_norm": 1.8229759769781986, + "learning_rate": 3.3081866107644734e-06, + "loss": 0.3981, + "step": 9087 + }, + { + "epoch": 0.6216567480675833, + "grad_norm": 2.300341118778473, + "learning_rate": 3.307144237533462e-06, + "loss": 0.4741, + "step": 9088 + }, + { + "epoch": 0.6217251521991928, + "grad_norm": 3.5476688975344146, + "learning_rate": 3.306101947395987e-06, + "loss": 0.3286, + "step": 9089 + }, + { + "epoch": 0.6217935563308024, + "grad_norm": 1.4955819975293707, + "learning_rate": 3.305059740403208e-06, + "loss": 0.2354, + "step": 9090 + }, + { + "epoch": 0.6218619604624119, + "grad_norm": 1.8781601977104763, + "learning_rate": 3.3040176166062825e-06, + "loss": 0.3561, + "step": 9091 + }, + { + "epoch": 0.6219303645940215, + "grad_norm": 1.8463425916157377, + "learning_rate": 3.302975576056362e-06, + "loss": 0.438, + "step": 9092 + }, + { + "epoch": 0.621998768725631, + "grad_norm": 1.8509776563923948, + "learning_rate": 3.3019336188045948e-06, + "loss": 0.347, + "step": 9093 + }, + { + "epoch": 0.6220671728572406, + "grad_norm": 1.975359214938829, + "learning_rate": 3.3008917449021262e-06, + "loss": 0.4587, + "step": 9094 + }, + { + "epoch": 0.6221355769888501, + "grad_norm": 1.9175336192676495, + "learning_rate": 3.2998499544000938e-06, + "loss": 0.2041, + "step": 9095 + }, + { + "epoch": 0.6222039811204597, + "grad_norm": 2.4467073490299374, + "learning_rate": 3.2988082473496374e-06, + "loss": 0.4447, + "step": 9096 + }, + { + "epoch": 0.6222723852520692, + "grad_norm": 1.588637290223591, + "learning_rate": 3.2977666238018875e-06, + "loss": 0.3098, + "step": 9097 + }, + { + "epoch": 0.6223407893836788, + "grad_norm": 2.5479907827184407, + "learning_rate": 3.29672508380797e-06, + "loss": 0.3402, + "step": 9098 + }, + { + "epoch": 0.6224091935152883, + "grad_norm": 1.9168474295281122, + "learning_rate": 3.2956836274190107e-06, + "loss": 0.1331, + "step": 9099 + }, + { + "epoch": 0.6224775976468979, + "grad_norm": 2.6647580041569223, + "learning_rate": 3.294642254686129e-06, + "loss": 0.4816, + "step": 9100 + }, + { + "epoch": 0.6225460017785074, + "grad_norm": 1.8528146051210568, + "learning_rate": 3.293600965660442e-06, + "loss": 0.3611, + "step": 9101 + }, + { + "epoch": 0.622614405910117, + "grad_norm": 2.213060646899036, + "learning_rate": 3.2925597603930586e-06, + "loss": 0.2422, + "step": 9102 + }, + { + "epoch": 0.6226828100417265, + "grad_norm": 3.715596448693974, + "learning_rate": 3.2915186389350873e-06, + "loss": 0.3235, + "step": 9103 + }, + { + "epoch": 0.6227512141733361, + "grad_norm": 1.8693637176640001, + "learning_rate": 3.2904776013376316e-06, + "loss": 0.3577, + "step": 9104 + }, + { + "epoch": 0.6228196183049456, + "grad_norm": 2.643388047477857, + "learning_rate": 3.2894366476517904e-06, + "loss": 0.2987, + "step": 9105 + }, + { + "epoch": 0.6228880224365552, + "grad_norm": 2.3451729460386312, + "learning_rate": 3.288395777928659e-06, + "loss": 0.7262, + "step": 9106 + }, + { + "epoch": 0.6229564265681647, + "grad_norm": 2.1327222518551188, + "learning_rate": 3.287354992219328e-06, + "loss": 0.6271, + "step": 9107 + }, + { + "epoch": 0.6230248306997742, + "grad_norm": 2.1360167129383796, + "learning_rate": 3.2863142905748835e-06, + "loss": 0.2974, + "step": 9108 + }, + { + "epoch": 0.6230932348313838, + "grad_norm": 1.8197641815689476, + "learning_rate": 3.2852736730464096e-06, + "loss": 0.338, + "step": 9109 + }, + { + "epoch": 0.6231616389629934, + "grad_norm": 2.0321389820560913, + "learning_rate": 3.2842331396849836e-06, + "loss": 0.6238, + "step": 9110 + }, + { + "epoch": 0.623230043094603, + "grad_norm": 1.8586004270711856, + "learning_rate": 3.283192690541681e-06, + "loss": 0.265, + "step": 9111 + }, + { + "epoch": 0.6232984472262124, + "grad_norm": 2.0557807110017907, + "learning_rate": 3.2821523256675707e-06, + "loss": 0.4701, + "step": 9112 + }, + { + "epoch": 0.623366851357822, + "grad_norm": 1.5433093769391177, + "learning_rate": 3.281112045113721e-06, + "loss": 0.175, + "step": 9113 + }, + { + "epoch": 0.6234352554894316, + "grad_norm": 1.882937537983885, + "learning_rate": 3.280071848931192e-06, + "loss": 0.4246, + "step": 9114 + }, + { + "epoch": 0.6235036596210412, + "grad_norm": 1.6152977475485077, + "learning_rate": 3.2790317371710423e-06, + "loss": 0.2486, + "step": 9115 + }, + { + "epoch": 0.6235720637526506, + "grad_norm": 5.129267009473564, + "learning_rate": 3.277991709884325e-06, + "loss": 0.322, + "step": 9116 + }, + { + "epoch": 0.6236404678842602, + "grad_norm": 2.5378588537805378, + "learning_rate": 3.2769517671220907e-06, + "loss": 0.5897, + "step": 9117 + }, + { + "epoch": 0.6237088720158698, + "grad_norm": 1.4597348143337625, + "learning_rate": 3.2759119089353834e-06, + "loss": 0.1663, + "step": 9118 + }, + { + "epoch": 0.6237772761474794, + "grad_norm": 1.8333167464596891, + "learning_rate": 3.2748721353752445e-06, + "loss": 0.3042, + "step": 9119 + }, + { + "epoch": 0.6238456802790888, + "grad_norm": 2.40039026494718, + "learning_rate": 3.2738324464927128e-06, + "loss": 0.5415, + "step": 9120 + }, + { + "epoch": 0.6239140844106984, + "grad_norm": 2.072314349005271, + "learning_rate": 3.2727928423388207e-06, + "loss": 0.4124, + "step": 9121 + }, + { + "epoch": 0.623982488542308, + "grad_norm": 2.392228106372721, + "learning_rate": 3.2717533229645952e-06, + "loss": 0.3562, + "step": 9122 + }, + { + "epoch": 0.6240508926739176, + "grad_norm": 2.7134540760940267, + "learning_rate": 3.2707138884210633e-06, + "loss": 0.6511, + "step": 9123 + }, + { + "epoch": 0.624119296805527, + "grad_norm": 1.714981855596756, + "learning_rate": 3.2696745387592443e-06, + "loss": 0.208, + "step": 9124 + }, + { + "epoch": 0.6241877009371366, + "grad_norm": 2.828020462717487, + "learning_rate": 3.268635274030154e-06, + "loss": 0.4772, + "step": 9125 + }, + { + "epoch": 0.6242561050687462, + "grad_norm": 2.159982626009535, + "learning_rate": 3.2675960942848062e-06, + "loss": 0.5934, + "step": 9126 + }, + { + "epoch": 0.6243245092003556, + "grad_norm": 2.7796811061397477, + "learning_rate": 3.2665569995742076e-06, + "loss": 0.3838, + "step": 9127 + }, + { + "epoch": 0.6243929133319652, + "grad_norm": 2.9653284332590246, + "learning_rate": 3.265517989949361e-06, + "loss": 0.6543, + "step": 9128 + }, + { + "epoch": 0.6244613174635748, + "grad_norm": 1.460974157352864, + "learning_rate": 3.264479065461266e-06, + "loss": 0.1875, + "step": 9129 + }, + { + "epoch": 0.6245297215951844, + "grad_norm": 2.2965482816776404, + "learning_rate": 3.2634402261609214e-06, + "loss": 0.4304, + "step": 9130 + }, + { + "epoch": 0.6245981257267939, + "grad_norm": 1.5159442530961094, + "learning_rate": 3.262401472099317e-06, + "loss": 0.1701, + "step": 9131 + }, + { + "epoch": 0.6246665298584034, + "grad_norm": 2.1382790400668212, + "learning_rate": 3.2613628033274368e-06, + "loss": 0.3318, + "step": 9132 + }, + { + "epoch": 0.624734933990013, + "grad_norm": 2.7634619104001126, + "learning_rate": 3.2603242198962682e-06, + "loss": 0.5535, + "step": 9133 + }, + { + "epoch": 0.6248033381216226, + "grad_norm": 1.9341837897878844, + "learning_rate": 3.259285721856787e-06, + "loss": 0.4405, + "step": 9134 + }, + { + "epoch": 0.624871742253232, + "grad_norm": 1.705148452933917, + "learning_rate": 3.2582473092599675e-06, + "loss": 0.408, + "step": 9135 + }, + { + "epoch": 0.6249401463848416, + "grad_norm": 1.7185018490336583, + "learning_rate": 3.257208982156782e-06, + "loss": 0.397, + "step": 9136 + }, + { + "epoch": 0.6250085505164512, + "grad_norm": 1.757301052711383, + "learning_rate": 3.256170740598196e-06, + "loss": 0.3754, + "step": 9137 + }, + { + "epoch": 0.6250769546480608, + "grad_norm": 1.3479101743448414, + "learning_rate": 3.255132584635169e-06, + "loss": 0.2597, + "step": 9138 + }, + { + "epoch": 0.6251453587796703, + "grad_norm": 2.9705906432235905, + "learning_rate": 3.25409451431866e-06, + "loss": 0.3729, + "step": 9139 + }, + { + "epoch": 0.6252137629112798, + "grad_norm": 1.5736610333366032, + "learning_rate": 3.253056529699625e-06, + "loss": 0.2021, + "step": 9140 + }, + { + "epoch": 0.6252821670428894, + "grad_norm": 2.2441568598147565, + "learning_rate": 3.2520186308290115e-06, + "loss": 0.6944, + "step": 9141 + }, + { + "epoch": 0.625350571174499, + "grad_norm": 1.6093646576928713, + "learning_rate": 3.2509808177577638e-06, + "loss": 0.1507, + "step": 9142 + }, + { + "epoch": 0.6254189753061085, + "grad_norm": 1.2839505241348854, + "learning_rate": 3.2499430905368247e-06, + "loss": 0.2368, + "step": 9143 + }, + { + "epoch": 0.625487379437718, + "grad_norm": 1.752718950530626, + "learning_rate": 3.248905449217129e-06, + "loss": 0.1313, + "step": 9144 + }, + { + "epoch": 0.6255557835693276, + "grad_norm": 3.3844292017294006, + "learning_rate": 3.247867893849609e-06, + "loss": 0.5871, + "step": 9145 + }, + { + "epoch": 0.6256241877009371, + "grad_norm": 1.7129461699193398, + "learning_rate": 3.2468304244851947e-06, + "loss": 0.262, + "step": 9146 + }, + { + "epoch": 0.6256925918325467, + "grad_norm": 2.9267654087908697, + "learning_rate": 3.245793041174809e-06, + "loss": 0.2497, + "step": 9147 + }, + { + "epoch": 0.6257609959641562, + "grad_norm": 1.9753169179709162, + "learning_rate": 3.244755743969371e-06, + "loss": 0.2179, + "step": 9148 + }, + { + "epoch": 0.6258294000957658, + "grad_norm": 1.5549802312866943, + "learning_rate": 3.2437185329197963e-06, + "loss": 0.2938, + "step": 9149 + }, + { + "epoch": 0.6258978042273753, + "grad_norm": 1.8240064052199534, + "learning_rate": 3.2426814080769985e-06, + "loss": 0.1816, + "step": 9150 + }, + { + "epoch": 0.6259662083589849, + "grad_norm": 1.8150016481152977, + "learning_rate": 3.241644369491884e-06, + "loss": 0.3302, + "step": 9151 + }, + { + "epoch": 0.6260346124905944, + "grad_norm": 2.0936823841108647, + "learning_rate": 3.2406074172153527e-06, + "loss": 0.3085, + "step": 9152 + }, + { + "epoch": 0.626103016622204, + "grad_norm": 3.7711081848121153, + "learning_rate": 3.239570551298308e-06, + "loss": 0.2674, + "step": 9153 + }, + { + "epoch": 0.6261714207538135, + "grad_norm": 1.9650300546221071, + "learning_rate": 3.2385337717916397e-06, + "loss": 0.393, + "step": 9154 + }, + { + "epoch": 0.6262398248854231, + "grad_norm": 2.512483068190881, + "learning_rate": 3.23749707874624e-06, + "loss": 0.4267, + "step": 9155 + }, + { + "epoch": 0.6263082290170326, + "grad_norm": 2.154699037284475, + "learning_rate": 3.236460472212996e-06, + "loss": 0.4438, + "step": 9156 + }, + { + "epoch": 0.6263766331486422, + "grad_norm": 2.3308743813455024, + "learning_rate": 3.2354239522427876e-06, + "loss": 0.4502, + "step": 9157 + }, + { + "epoch": 0.6264450372802517, + "grad_norm": 2.0527636510501717, + "learning_rate": 3.2343875188864904e-06, + "loss": 0.4319, + "step": 9158 + }, + { + "epoch": 0.6265134414118613, + "grad_norm": 2.702064603849944, + "learning_rate": 3.2333511721949816e-06, + "loss": 0.4571, + "step": 9159 + }, + { + "epoch": 0.6265818455434709, + "grad_norm": 2.548506676292244, + "learning_rate": 3.232314912219129e-06, + "loss": 0.4331, + "step": 9160 + }, + { + "epoch": 0.6266502496750804, + "grad_norm": 2.317104137976413, + "learning_rate": 3.2312787390097967e-06, + "loss": 0.4293, + "step": 9161 + }, + { + "epoch": 0.6267186538066899, + "grad_norm": 3.016096093139782, + "learning_rate": 3.2302426526178433e-06, + "loss": 0.4196, + "step": 9162 + }, + { + "epoch": 0.6267870579382995, + "grad_norm": 1.88991048410397, + "learning_rate": 3.229206653094128e-06, + "loss": 0.2524, + "step": 9163 + }, + { + "epoch": 0.626855462069909, + "grad_norm": 3.043698670152239, + "learning_rate": 3.228170740489501e-06, + "loss": 0.2387, + "step": 9164 + }, + { + "epoch": 0.6269238662015185, + "grad_norm": 2.044183934950434, + "learning_rate": 3.22713491485481e-06, + "loss": 0.2669, + "step": 9165 + }, + { + "epoch": 0.6269922703331281, + "grad_norm": 2.5488590983997184, + "learning_rate": 3.226099176240899e-06, + "loss": 0.6264, + "step": 9166 + }, + { + "epoch": 0.6270606744647377, + "grad_norm": 2.1584656982877477, + "learning_rate": 3.225063524698606e-06, + "loss": 0.2706, + "step": 9167 + }, + { + "epoch": 0.6271290785963473, + "grad_norm": 1.80101941699671, + "learning_rate": 3.224027960278765e-06, + "loss": 0.3033, + "step": 9168 + }, + { + "epoch": 0.6271974827279567, + "grad_norm": 2.8070662885135853, + "learning_rate": 3.2229924830322083e-06, + "loss": 0.413, + "step": 9169 + }, + { + "epoch": 0.6272658868595663, + "grad_norm": 1.5636349180277378, + "learning_rate": 3.2219570930097632e-06, + "loss": 0.2091, + "step": 9170 + }, + { + "epoch": 0.6273342909911759, + "grad_norm": 2.249350765391322, + "learning_rate": 3.220921790262249e-06, + "loss": 0.5277, + "step": 9171 + }, + { + "epoch": 0.6274026951227855, + "grad_norm": 1.671313320713634, + "learning_rate": 3.2198865748404838e-06, + "loss": 0.3004, + "step": 9172 + }, + { + "epoch": 0.6274710992543949, + "grad_norm": 1.9783998928132127, + "learning_rate": 3.2188514467952835e-06, + "loss": 0.3642, + "step": 9173 + }, + { + "epoch": 0.6275395033860045, + "grad_norm": 2.7762290015837072, + "learning_rate": 3.217816406177454e-06, + "loss": 0.294, + "step": 9174 + }, + { + "epoch": 0.6276079075176141, + "grad_norm": 2.6679394205137275, + "learning_rate": 3.2167814530378016e-06, + "loss": 0.2062, + "step": 9175 + }, + { + "epoch": 0.6276763116492237, + "grad_norm": 2.0415739765016916, + "learning_rate": 3.2157465874271266e-06, + "loss": 0.3136, + "step": 9176 + }, + { + "epoch": 0.6277447157808331, + "grad_norm": 2.2327998951795722, + "learning_rate": 3.2147118093962258e-06, + "loss": 0.4485, + "step": 9177 + }, + { + "epoch": 0.6278131199124427, + "grad_norm": 1.768728739506222, + "learning_rate": 3.213677118995888e-06, + "loss": 0.494, + "step": 9178 + }, + { + "epoch": 0.6278815240440523, + "grad_norm": 2.078706824805684, + "learning_rate": 3.212642516276905e-06, + "loss": 0.4114, + "step": 9179 + }, + { + "epoch": 0.6279499281756619, + "grad_norm": 2.2400143955185374, + "learning_rate": 3.2116080012900584e-06, + "loss": 0.229, + "step": 9180 + }, + { + "epoch": 0.6280183323072713, + "grad_norm": 2.896992901786454, + "learning_rate": 3.210573574086128e-06, + "loss": 0.4717, + "step": 9181 + }, + { + "epoch": 0.6280867364388809, + "grad_norm": 2.681806805067442, + "learning_rate": 3.2095392347158853e-06, + "loss": 0.515, + "step": 9182 + }, + { + "epoch": 0.6281551405704905, + "grad_norm": 2.346604687734041, + "learning_rate": 3.2085049832301052e-06, + "loss": 0.4936, + "step": 9183 + }, + { + "epoch": 0.6282235447021, + "grad_norm": 3.016506130577432, + "learning_rate": 3.207470819679551e-06, + "loss": 0.593, + "step": 9184 + }, + { + "epoch": 0.6282919488337095, + "grad_norm": 2.15953504104563, + "learning_rate": 3.2064367441149845e-06, + "loss": 0.4754, + "step": 9185 + }, + { + "epoch": 0.6283603529653191, + "grad_norm": 1.7171352430589202, + "learning_rate": 3.205402756587164e-06, + "loss": 0.2102, + "step": 9186 + }, + { + "epoch": 0.6284287570969287, + "grad_norm": 2.1913855444598043, + "learning_rate": 3.2043688571468427e-06, + "loss": 0.38, + "step": 9187 + }, + { + "epoch": 0.6284971612285382, + "grad_norm": 2.413821393537076, + "learning_rate": 3.203335045844767e-06, + "loss": 0.4334, + "step": 9188 + }, + { + "epoch": 0.6285655653601477, + "grad_norm": 2.6445979697349, + "learning_rate": 3.202301322731684e-06, + "loss": 0.34, + "step": 9189 + }, + { + "epoch": 0.6286339694917573, + "grad_norm": 2.62467463213935, + "learning_rate": 3.2012676878583344e-06, + "loss": 0.3145, + "step": 9190 + }, + { + "epoch": 0.6287023736233669, + "grad_norm": 2.5859544067752114, + "learning_rate": 3.200234141275452e-06, + "loss": 0.5411, + "step": 9191 + }, + { + "epoch": 0.6287707777549764, + "grad_norm": 2.485013651781557, + "learning_rate": 3.1992006830337685e-06, + "loss": 0.7675, + "step": 9192 + }, + { + "epoch": 0.6288391818865859, + "grad_norm": 1.5712969621515462, + "learning_rate": 3.1981673131840125e-06, + "loss": 0.2933, + "step": 9193 + }, + { + "epoch": 0.6289075860181955, + "grad_norm": 1.5577240216679127, + "learning_rate": 3.197134031776905e-06, + "loss": 0.1994, + "step": 9194 + }, + { + "epoch": 0.6289759901498051, + "grad_norm": 2.2552755549913064, + "learning_rate": 3.196100838863165e-06, + "loss": 0.3597, + "step": 9195 + }, + { + "epoch": 0.6290443942814146, + "grad_norm": 2.0307935385241622, + "learning_rate": 3.195067734493508e-06, + "loss": 0.3849, + "step": 9196 + }, + { + "epoch": 0.6291127984130241, + "grad_norm": 1.6921877168001862, + "learning_rate": 3.1940347187186403e-06, + "loss": 0.19, + "step": 9197 + }, + { + "epoch": 0.6291812025446337, + "grad_norm": 1.6112459658860045, + "learning_rate": 3.1930017915892707e-06, + "loss": 0.2956, + "step": 9198 + }, + { + "epoch": 0.6292496066762433, + "grad_norm": 1.9461119973679806, + "learning_rate": 3.191968953156098e-06, + "loss": 0.4128, + "step": 9199 + }, + { + "epoch": 0.6293180108078528, + "grad_norm": 1.9940288710407588, + "learning_rate": 3.190936203469822e-06, + "loss": 0.5411, + "step": 9200 + }, + { + "epoch": 0.6293864149394623, + "grad_norm": 2.119231426912961, + "learning_rate": 3.189903542581132e-06, + "loss": 0.5712, + "step": 9201 + }, + { + "epoch": 0.6294548190710719, + "grad_norm": 3.4703819456885254, + "learning_rate": 3.188870970540717e-06, + "loss": 0.4188, + "step": 9202 + }, + { + "epoch": 0.6295232232026814, + "grad_norm": 1.4036236078865392, + "learning_rate": 3.1878384873992614e-06, + "loss": 0.122, + "step": 9203 + }, + { + "epoch": 0.629591627334291, + "grad_norm": 2.3919160074646584, + "learning_rate": 3.186806093207442e-06, + "loss": 0.4215, + "step": 9204 + }, + { + "epoch": 0.6296600314659005, + "grad_norm": 2.1483857165945093, + "learning_rate": 3.1857737880159368e-06, + "loss": 0.1314, + "step": 9205 + }, + { + "epoch": 0.6297284355975101, + "grad_norm": 2.366657776939012, + "learning_rate": 3.184741571875415e-06, + "loss": 0.3861, + "step": 9206 + }, + { + "epoch": 0.6297968397291196, + "grad_norm": 1.63737967445824, + "learning_rate": 3.1837094448365412e-06, + "loss": 0.1705, + "step": 9207 + }, + { + "epoch": 0.6298652438607292, + "grad_norm": 1.6227043194324673, + "learning_rate": 3.1826774069499792e-06, + "loss": 0.3292, + "step": 9208 + }, + { + "epoch": 0.6299336479923388, + "grad_norm": 1.8566578959555213, + "learning_rate": 3.1816454582663856e-06, + "loss": 0.4701, + "step": 9209 + }, + { + "epoch": 0.6300020521239483, + "grad_norm": 1.6499817200572886, + "learning_rate": 3.180613598836415e-06, + "loss": 0.3058, + "step": 9210 + }, + { + "epoch": 0.6300704562555578, + "grad_norm": 1.8887044081307116, + "learning_rate": 3.1795818287107133e-06, + "loss": 0.3651, + "step": 9211 + }, + { + "epoch": 0.6301388603871674, + "grad_norm": 2.015605660279221, + "learning_rate": 3.178550147939927e-06, + "loss": 0.3224, + "step": 9212 + }, + { + "epoch": 0.630207264518777, + "grad_norm": 1.6965948410134661, + "learning_rate": 3.177518556574696e-06, + "loss": 0.287, + "step": 9213 + }, + { + "epoch": 0.6302756686503865, + "grad_norm": 2.1357667256128705, + "learning_rate": 3.176487054665653e-06, + "loss": 0.4454, + "step": 9214 + }, + { + "epoch": 0.630344072781996, + "grad_norm": 2.2700932081134795, + "learning_rate": 3.175455642263432e-06, + "loss": 0.4781, + "step": 9215 + }, + { + "epoch": 0.6304124769136056, + "grad_norm": 1.7464091380957991, + "learning_rate": 3.174424319418659e-06, + "loss": 0.402, + "step": 9216 + }, + { + "epoch": 0.6304808810452152, + "grad_norm": 1.768534739743602, + "learning_rate": 3.173393086181955e-06, + "loss": 0.3093, + "step": 9217 + }, + { + "epoch": 0.6305492851768246, + "grad_norm": 1.887795474754416, + "learning_rate": 3.172361942603938e-06, + "loss": 0.4072, + "step": 9218 + }, + { + "epoch": 0.6306176893084342, + "grad_norm": 1.8795902490231402, + "learning_rate": 3.1713308887352244e-06, + "loss": 0.3997, + "step": 9219 + }, + { + "epoch": 0.6306860934400438, + "grad_norm": 2.7246285331900224, + "learning_rate": 3.170299924626421e-06, + "loss": 0.3462, + "step": 9220 + }, + { + "epoch": 0.6307544975716534, + "grad_norm": 2.143110113595617, + "learning_rate": 3.1692690503281326e-06, + "loss": 0.4126, + "step": 9221 + }, + { + "epoch": 0.6308229017032628, + "grad_norm": 2.0077612310016337, + "learning_rate": 3.1682382658909607e-06, + "loss": 0.4949, + "step": 9222 + }, + { + "epoch": 0.6308913058348724, + "grad_norm": 2.109668720070235, + "learning_rate": 3.1672075713654993e-06, + "loss": 0.2815, + "step": 9223 + }, + { + "epoch": 0.630959709966482, + "grad_norm": 2.032292930916126, + "learning_rate": 3.16617696680234e-06, + "loss": 0.3433, + "step": 9224 + }, + { + "epoch": 0.6310281140980916, + "grad_norm": 2.862324509280344, + "learning_rate": 3.165146452252072e-06, + "loss": 0.4372, + "step": 9225 + }, + { + "epoch": 0.631096518229701, + "grad_norm": 1.567644215114331, + "learning_rate": 3.164116027765276e-06, + "loss": 0.1379, + "step": 9226 + }, + { + "epoch": 0.6311649223613106, + "grad_norm": 2.6366590535505687, + "learning_rate": 3.1630856933925302e-06, + "loss": 0.4277, + "step": 9227 + }, + { + "epoch": 0.6312333264929202, + "grad_norm": 2.2295739137605852, + "learning_rate": 3.162055449184408e-06, + "loss": 0.6017, + "step": 9228 + }, + { + "epoch": 0.6313017306245298, + "grad_norm": 1.8627244769262454, + "learning_rate": 3.161025295191482e-06, + "loss": 0.3774, + "step": 9229 + }, + { + "epoch": 0.6313701347561392, + "grad_norm": 3.8330546965434804, + "learning_rate": 3.159995231464314e-06, + "loss": 0.4847, + "step": 9230 + }, + { + "epoch": 0.6314385388877488, + "grad_norm": 1.6594632935247238, + "learning_rate": 3.158965258053465e-06, + "loss": 0.2992, + "step": 9231 + }, + { + "epoch": 0.6315069430193584, + "grad_norm": 2.650770987961049, + "learning_rate": 3.1579353750094922e-06, + "loss": 0.2939, + "step": 9232 + }, + { + "epoch": 0.631575347150968, + "grad_norm": 3.0770237690699016, + "learning_rate": 3.156905582382946e-06, + "loss": 0.3268, + "step": 9233 + }, + { + "epoch": 0.6316437512825774, + "grad_norm": 2.994657114848658, + "learning_rate": 3.1558758802243734e-06, + "loss": 0.3651, + "step": 9234 + }, + { + "epoch": 0.631712155414187, + "grad_norm": 1.6953642085755074, + "learning_rate": 3.1548462685843185e-06, + "loss": 0.3163, + "step": 9235 + }, + { + "epoch": 0.6317805595457966, + "grad_norm": 2.1941437443973695, + "learning_rate": 3.15381674751332e-06, + "loss": 0.4576, + "step": 9236 + }, + { + "epoch": 0.6318489636774061, + "grad_norm": 2.017401980219374, + "learning_rate": 3.152787317061908e-06, + "loss": 0.4927, + "step": 9237 + }, + { + "epoch": 0.6319173678090156, + "grad_norm": 2.401378505908158, + "learning_rate": 3.1517579772806144e-06, + "loss": 0.5733, + "step": 9238 + }, + { + "epoch": 0.6319857719406252, + "grad_norm": 2.0507576350714394, + "learning_rate": 3.150728728219966e-06, + "loss": 0.3923, + "step": 9239 + }, + { + "epoch": 0.6320541760722348, + "grad_norm": 1.8954025231450962, + "learning_rate": 3.1496995699304823e-06, + "loss": 0.2018, + "step": 9240 + }, + { + "epoch": 0.6321225802038443, + "grad_norm": 4.3006682281472575, + "learning_rate": 3.1486705024626773e-06, + "loss": 0.2836, + "step": 9241 + }, + { + "epoch": 0.6321909843354538, + "grad_norm": 1.9416931984143357, + "learning_rate": 3.1476415258670656e-06, + "loss": 0.2033, + "step": 9242 + }, + { + "epoch": 0.6322593884670634, + "grad_norm": 2.283850011238195, + "learning_rate": 3.146612640194152e-06, + "loss": 0.4555, + "step": 9243 + }, + { + "epoch": 0.632327792598673, + "grad_norm": 2.047858235025882, + "learning_rate": 3.145583845494439e-06, + "loss": 0.4918, + "step": 9244 + }, + { + "epoch": 0.6323961967302825, + "grad_norm": 1.7665941481806349, + "learning_rate": 3.144555141818427e-06, + "loss": 0.2239, + "step": 9245 + }, + { + "epoch": 0.632464600861892, + "grad_norm": 2.494972878778857, + "learning_rate": 3.1435265292166084e-06, + "loss": 0.4629, + "step": 9246 + }, + { + "epoch": 0.6325330049935016, + "grad_norm": 3.1810951210197773, + "learning_rate": 3.1424980077394717e-06, + "loss": 0.3542, + "step": 9247 + }, + { + "epoch": 0.6326014091251112, + "grad_norm": 2.3569656600651694, + "learning_rate": 3.1414695774375014e-06, + "loss": 0.4581, + "step": 9248 + }, + { + "epoch": 0.6326698132567207, + "grad_norm": 2.3827093258392846, + "learning_rate": 3.1404412383611815e-06, + "loss": 0.4393, + "step": 9249 + }, + { + "epoch": 0.6327382173883302, + "grad_norm": 2.469017611859492, + "learning_rate": 3.139412990560985e-06, + "loss": 0.553, + "step": 9250 + }, + { + "epoch": 0.6328066215199398, + "grad_norm": 1.488182620287202, + "learning_rate": 3.138384834087382e-06, + "loss": 0.1562, + "step": 9251 + }, + { + "epoch": 0.6328750256515494, + "grad_norm": 2.262202377528627, + "learning_rate": 3.137356768990843e-06, + "loss": 0.5168, + "step": 9252 + }, + { + "epoch": 0.6329434297831589, + "grad_norm": 2.4554050109581675, + "learning_rate": 3.136328795321828e-06, + "loss": 0.3503, + "step": 9253 + }, + { + "epoch": 0.6330118339147685, + "grad_norm": 2.058469004895958, + "learning_rate": 3.1353009131307943e-06, + "loss": 0.4253, + "step": 9254 + }, + { + "epoch": 0.633080238046378, + "grad_norm": 2.416084399589775, + "learning_rate": 3.1342731224681965e-06, + "loss": 0.4126, + "step": 9255 + }, + { + "epoch": 0.6331486421779875, + "grad_norm": 2.0579004513679107, + "learning_rate": 3.1332454233844844e-06, + "loss": 0.3951, + "step": 9256 + }, + { + "epoch": 0.6332170463095971, + "grad_norm": 2.161977241841189, + "learning_rate": 3.132217815930099e-06, + "loss": 0.202, + "step": 9257 + }, + { + "epoch": 0.6332854504412067, + "grad_norm": 1.6510673715493784, + "learning_rate": 3.131190300155484e-06, + "loss": 0.308, + "step": 9258 + }, + { + "epoch": 0.6333538545728162, + "grad_norm": 2.448145757601291, + "learning_rate": 3.130162876111074e-06, + "loss": 0.4388, + "step": 9259 + }, + { + "epoch": 0.6334222587044257, + "grad_norm": 2.2285071012583013, + "learning_rate": 3.1291355438473e-06, + "loss": 0.2898, + "step": 9260 + }, + { + "epoch": 0.6334906628360353, + "grad_norm": 1.4597048940012176, + "learning_rate": 3.128108303414586e-06, + "loss": 0.1896, + "step": 9261 + }, + { + "epoch": 0.6335590669676449, + "grad_norm": 2.861801165644468, + "learning_rate": 3.127081154863357e-06, + "loss": 0.3035, + "step": 9262 + }, + { + "epoch": 0.6336274710992544, + "grad_norm": 2.1374145992686033, + "learning_rate": 3.126054098244029e-06, + "loss": 0.3687, + "step": 9263 + }, + { + "epoch": 0.6336958752308639, + "grad_norm": 2.34183522527884, + "learning_rate": 3.125027133607015e-06, + "loss": 0.1974, + "step": 9264 + }, + { + "epoch": 0.6337642793624735, + "grad_norm": 2.120669960132241, + "learning_rate": 3.1240002610027233e-06, + "loss": 0.2416, + "step": 9265 + }, + { + "epoch": 0.6338326834940831, + "grad_norm": 2.3504405117335647, + "learning_rate": 3.122973480481558e-06, + "loss": 0.3397, + "step": 9266 + }, + { + "epoch": 0.6339010876256926, + "grad_norm": 2.013088052884907, + "learning_rate": 3.1219467920939174e-06, + "loss": 0.1625, + "step": 9267 + }, + { + "epoch": 0.6339694917573021, + "grad_norm": 2.9135266749724047, + "learning_rate": 3.1209201958901975e-06, + "loss": 0.5977, + "step": 9268 + }, + { + "epoch": 0.6340378958889117, + "grad_norm": 2.1914596891092266, + "learning_rate": 3.119893691920789e-06, + "loss": 0.3113, + "step": 9269 + }, + { + "epoch": 0.6341063000205213, + "grad_norm": 1.864271584525377, + "learning_rate": 3.1188672802360776e-06, + "loss": 0.1739, + "step": 9270 + }, + { + "epoch": 0.6341747041521308, + "grad_norm": 1.9777863863533376, + "learning_rate": 3.117840960886443e-06, + "loss": 0.4994, + "step": 9271 + }, + { + "epoch": 0.6342431082837403, + "grad_norm": 1.7215840469427317, + "learning_rate": 3.116814733922264e-06, + "loss": 0.1988, + "step": 9272 + }, + { + "epoch": 0.6343115124153499, + "grad_norm": 1.8514452144641207, + "learning_rate": 3.115788599393911e-06, + "loss": 0.2489, + "step": 9273 + }, + { + "epoch": 0.6343799165469595, + "grad_norm": 1.5446636500167916, + "learning_rate": 3.114762557351752e-06, + "loss": 0.147, + "step": 9274 + }, + { + "epoch": 0.6344483206785689, + "grad_norm": 2.0650549005225605, + "learning_rate": 3.113736607846151e-06, + "loss": 0.3405, + "step": 9275 + }, + { + "epoch": 0.6345167248101785, + "grad_norm": 2.282618928413971, + "learning_rate": 3.112710750927466e-06, + "loss": 0.6007, + "step": 9276 + }, + { + "epoch": 0.6345851289417881, + "grad_norm": 2.1375566440182063, + "learning_rate": 3.111684986646049e-06, + "loss": 0.5487, + "step": 9277 + }, + { + "epoch": 0.6346535330733977, + "grad_norm": 2.0506314649831703, + "learning_rate": 3.1106593150522522e-06, + "loss": 0.4099, + "step": 9278 + }, + { + "epoch": 0.6347219372050071, + "grad_norm": 1.987543167689109, + "learning_rate": 3.1096337361964213e-06, + "loss": 0.4243, + "step": 9279 + }, + { + "epoch": 0.6347903413366167, + "grad_norm": 2.0894972540702037, + "learning_rate": 3.108608250128895e-06, + "loss": 0.1181, + "step": 9280 + }, + { + "epoch": 0.6348587454682263, + "grad_norm": 1.7517519806692392, + "learning_rate": 3.1075828569000083e-06, + "loss": 0.2971, + "step": 9281 + }, + { + "epoch": 0.6349271495998359, + "grad_norm": 2.1755168178820536, + "learning_rate": 3.1065575565600936e-06, + "loss": 0.5572, + "step": 9282 + }, + { + "epoch": 0.6349955537314453, + "grad_norm": 2.2755793742860453, + "learning_rate": 3.105532349159478e-06, + "loss": 0.5352, + "step": 9283 + }, + { + "epoch": 0.6350639578630549, + "grad_norm": 2.9296159252324543, + "learning_rate": 3.1045072347484822e-06, + "loss": 0.3037, + "step": 9284 + }, + { + "epoch": 0.6351323619946645, + "grad_norm": 2.0416728298587286, + "learning_rate": 3.1034822133774258e-06, + "loss": 0.2651, + "step": 9285 + }, + { + "epoch": 0.6352007661262741, + "grad_norm": 1.7666908271130155, + "learning_rate": 3.1024572850966197e-06, + "loss": 0.1754, + "step": 9286 + }, + { + "epoch": 0.6352691702578835, + "grad_norm": 2.165909230514616, + "learning_rate": 3.1014324499563713e-06, + "loss": 0.6041, + "step": 9287 + }, + { + "epoch": 0.6353375743894931, + "grad_norm": 1.6387802379184069, + "learning_rate": 3.100407708006987e-06, + "loss": 0.3051, + "step": 9288 + }, + { + "epoch": 0.6354059785211027, + "grad_norm": 2.133234352619504, + "learning_rate": 3.099383059298767e-06, + "loss": 0.3646, + "step": 9289 + }, + { + "epoch": 0.6354743826527123, + "grad_norm": 1.3050160965027267, + "learning_rate": 3.098358503882005e-06, + "loss": 0.189, + "step": 9290 + }, + { + "epoch": 0.6355427867843217, + "grad_norm": 1.9101920095267617, + "learning_rate": 3.0973340418069887e-06, + "loss": 0.3259, + "step": 9291 + }, + { + "epoch": 0.6356111909159313, + "grad_norm": 1.6767915556794444, + "learning_rate": 3.096309673124007e-06, + "loss": 0.2891, + "step": 9292 + }, + { + "epoch": 0.6356795950475409, + "grad_norm": 2.620842695649067, + "learning_rate": 3.09528539788334e-06, + "loss": 0.6935, + "step": 9293 + }, + { + "epoch": 0.6357479991791504, + "grad_norm": 1.5891571530609268, + "learning_rate": 3.0942612161352607e-06, + "loss": 0.1391, + "step": 9294 + }, + { + "epoch": 0.63581640331076, + "grad_norm": 2.709059193739269, + "learning_rate": 3.0932371279300455e-06, + "loss": 0.2382, + "step": 9295 + }, + { + "epoch": 0.6358848074423695, + "grad_norm": 1.7942758337725393, + "learning_rate": 3.0922131333179587e-06, + "loss": 0.3049, + "step": 9296 + }, + { + "epoch": 0.6359532115739791, + "grad_norm": 2.1349543342352115, + "learning_rate": 3.0911892323492643e-06, + "loss": 0.582, + "step": 9297 + }, + { + "epoch": 0.6360216157055886, + "grad_norm": 2.3341585804557354, + "learning_rate": 3.0901654250742195e-06, + "loss": 0.3693, + "step": 9298 + }, + { + "epoch": 0.6360900198371982, + "grad_norm": 2.1373635178767394, + "learning_rate": 3.0891417115430794e-06, + "loss": 0.4396, + "step": 9299 + }, + { + "epoch": 0.6361584239688077, + "grad_norm": 8.828473838733256, + "learning_rate": 3.0881180918060914e-06, + "loss": 0.6467, + "step": 9300 + }, + { + "epoch": 0.6362268281004173, + "grad_norm": 2.8386428476668133, + "learning_rate": 3.087094565913499e-06, + "loss": 0.7122, + "step": 9301 + }, + { + "epoch": 0.6362952322320268, + "grad_norm": 1.8035722399363117, + "learning_rate": 3.0860711339155437e-06, + "loss": 0.3702, + "step": 9302 + }, + { + "epoch": 0.6363636363636364, + "grad_norm": 2.656090658681895, + "learning_rate": 3.0850477958624582e-06, + "loss": 0.7354, + "step": 9303 + }, + { + "epoch": 0.6364320404952459, + "grad_norm": 2.239852353875487, + "learning_rate": 3.084024551804474e-06, + "loss": 0.1776, + "step": 9304 + }, + { + "epoch": 0.6365004446268555, + "grad_norm": 1.9315408154381153, + "learning_rate": 3.0830014017918185e-06, + "loss": 0.3417, + "step": 9305 + }, + { + "epoch": 0.636568848758465, + "grad_norm": 1.8577396349658195, + "learning_rate": 3.0819783458747097e-06, + "loss": 0.4111, + "step": 9306 + }, + { + "epoch": 0.6366372528900746, + "grad_norm": 1.7790457614276751, + "learning_rate": 3.0809553841033657e-06, + "loss": 0.3464, + "step": 9307 + }, + { + "epoch": 0.6367056570216841, + "grad_norm": 1.866199842706475, + "learning_rate": 3.079932516527999e-06, + "loss": 0.3116, + "step": 9308 + }, + { + "epoch": 0.6367740611532937, + "grad_norm": 2.529729568998723, + "learning_rate": 3.078909743198817e-06, + "loss": 0.3562, + "step": 9309 + }, + { + "epoch": 0.6368424652849032, + "grad_norm": 1.945034872872617, + "learning_rate": 3.07788706416602e-06, + "loss": 0.1337, + "step": 9310 + }, + { + "epoch": 0.6369108694165128, + "grad_norm": 2.0877762880055517, + "learning_rate": 3.0768644794798096e-06, + "loss": 0.4796, + "step": 9311 + }, + { + "epoch": 0.6369792735481223, + "grad_norm": 2.446829693838856, + "learning_rate": 3.0758419891903766e-06, + "loss": 0.7514, + "step": 9312 + }, + { + "epoch": 0.6370476776797318, + "grad_norm": 1.602847077258339, + "learning_rate": 3.07481959334791e-06, + "loss": 0.2986, + "step": 9313 + }, + { + "epoch": 0.6371160818113414, + "grad_norm": 2.128046397579499, + "learning_rate": 3.073797292002595e-06, + "loss": 0.3769, + "step": 9314 + }, + { + "epoch": 0.637184485942951, + "grad_norm": 2.1690852340366416, + "learning_rate": 3.07277508520461e-06, + "loss": 0.5366, + "step": 9315 + }, + { + "epoch": 0.6372528900745605, + "grad_norm": 1.817756708661867, + "learning_rate": 3.07175297300413e-06, + "loss": 0.3756, + "step": 9316 + }, + { + "epoch": 0.63732129420617, + "grad_norm": 6.161895525768796, + "learning_rate": 3.0707309554513244e-06, + "loss": 0.6085, + "step": 9317 + }, + { + "epoch": 0.6373896983377796, + "grad_norm": 1.684071838047218, + "learning_rate": 3.0697090325963618e-06, + "loss": 0.3709, + "step": 9318 + }, + { + "epoch": 0.6374581024693892, + "grad_norm": 1.971388683411652, + "learning_rate": 3.0686872044894014e-06, + "loss": 0.3394, + "step": 9319 + }, + { + "epoch": 0.6375265066009987, + "grad_norm": 1.989134584785305, + "learning_rate": 3.0676654711805977e-06, + "loss": 0.2274, + "step": 9320 + }, + { + "epoch": 0.6375949107326082, + "grad_norm": 2.1121497165728056, + "learning_rate": 3.066643832720105e-06, + "loss": 0.4984, + "step": 9321 + }, + { + "epoch": 0.6376633148642178, + "grad_norm": 1.9562972818699624, + "learning_rate": 3.065622289158069e-06, + "loss": 0.3302, + "step": 9322 + }, + { + "epoch": 0.6377317189958274, + "grad_norm": 2.6657058256357034, + "learning_rate": 3.0646008405446315e-06, + "loss": 0.7464, + "step": 9323 + }, + { + "epoch": 0.637800123127437, + "grad_norm": 2.2918193282228234, + "learning_rate": 3.0635794869299305e-06, + "loss": 0.3042, + "step": 9324 + }, + { + "epoch": 0.6378685272590464, + "grad_norm": 2.122632934728683, + "learning_rate": 3.0625582283641007e-06, + "loss": 0.5722, + "step": 9325 + }, + { + "epoch": 0.637936931390656, + "grad_norm": 2.1162669232069793, + "learning_rate": 3.061537064897267e-06, + "loss": 0.3564, + "step": 9326 + }, + { + "epoch": 0.6380053355222656, + "grad_norm": 1.3412033223049433, + "learning_rate": 3.060515996579554e-06, + "loss": 0.1693, + "step": 9327 + }, + { + "epoch": 0.6380737396538751, + "grad_norm": 2.5307863010903198, + "learning_rate": 3.0594950234610836e-06, + "loss": 0.4166, + "step": 9328 + }, + { + "epoch": 0.6381421437854846, + "grad_norm": 2.3392001558109716, + "learning_rate": 3.0584741455919686e-06, + "loss": 0.1763, + "step": 9329 + }, + { + "epoch": 0.6382105479170942, + "grad_norm": 2.6279946003427295, + "learning_rate": 3.0574533630223162e-06, + "loss": 0.5619, + "step": 9330 + }, + { + "epoch": 0.6382789520487038, + "grad_norm": 2.384836194293161, + "learning_rate": 3.0564326758022343e-06, + "loss": 0.4896, + "step": 9331 + }, + { + "epoch": 0.6383473561803132, + "grad_norm": 2.493115076106038, + "learning_rate": 3.0554120839818225e-06, + "loss": 0.3429, + "step": 9332 + }, + { + "epoch": 0.6384157603119228, + "grad_norm": 2.791470212334371, + "learning_rate": 3.0543915876111754e-06, + "loss": 0.3344, + "step": 9333 + }, + { + "epoch": 0.6384841644435324, + "grad_norm": 4.02068009191572, + "learning_rate": 3.0533711867403844e-06, + "loss": 0.4306, + "step": 9334 + }, + { + "epoch": 0.638552568575142, + "grad_norm": 1.9410129739793534, + "learning_rate": 3.0523508814195356e-06, + "loss": 0.3065, + "step": 9335 + }, + { + "epoch": 0.6386209727067514, + "grad_norm": 2.5441033458863074, + "learning_rate": 3.0513306716987106e-06, + "loss": 0.226, + "step": 9336 + }, + { + "epoch": 0.638689376838361, + "grad_norm": 2.493772976991737, + "learning_rate": 3.0503105576279846e-06, + "loss": 0.2066, + "step": 9337 + }, + { + "epoch": 0.6387577809699706, + "grad_norm": 2.0297225711250495, + "learning_rate": 3.0492905392574335e-06, + "loss": 0.4175, + "step": 9338 + }, + { + "epoch": 0.6388261851015802, + "grad_norm": 2.0201865632707334, + "learning_rate": 3.0482706166371236e-06, + "loss": 0.5297, + "step": 9339 + }, + { + "epoch": 0.6388945892331896, + "grad_norm": 2.390428622289421, + "learning_rate": 3.0472507898171146e-06, + "loss": 0.2508, + "step": 9340 + }, + { + "epoch": 0.6389629933647992, + "grad_norm": 2.6785152362424323, + "learning_rate": 3.0462310588474674e-06, + "loss": 0.783, + "step": 9341 + }, + { + "epoch": 0.6390313974964088, + "grad_norm": 1.9532209754251149, + "learning_rate": 3.045211423778235e-06, + "loss": 0.4548, + "step": 9342 + }, + { + "epoch": 0.6390998016280184, + "grad_norm": 1.5246306613117373, + "learning_rate": 3.044191884659464e-06, + "loss": 0.2653, + "step": 9343 + }, + { + "epoch": 0.6391682057596278, + "grad_norm": 1.8687644827163497, + "learning_rate": 3.0431724415412013e-06, + "loss": 0.4105, + "step": 9344 + }, + { + "epoch": 0.6392366098912374, + "grad_norm": 2.2964801319159034, + "learning_rate": 3.0421530944734844e-06, + "loss": 0.4225, + "step": 9345 + }, + { + "epoch": 0.639305014022847, + "grad_norm": 2.3142067307285092, + "learning_rate": 3.041133843506347e-06, + "loss": 0.3383, + "step": 9346 + }, + { + "epoch": 0.6393734181544565, + "grad_norm": 2.2059763792745564, + "learning_rate": 3.040114688689819e-06, + "loss": 0.4861, + "step": 9347 + }, + { + "epoch": 0.639441822286066, + "grad_norm": 2.1416262770034273, + "learning_rate": 3.0390956300739276e-06, + "loss": 0.5096, + "step": 9348 + }, + { + "epoch": 0.6395102264176756, + "grad_norm": 2.0970583228690223, + "learning_rate": 3.0380766677086925e-06, + "loss": 0.4056, + "step": 9349 + }, + { + "epoch": 0.6395786305492852, + "grad_norm": 1.9189644111284698, + "learning_rate": 3.0370578016441277e-06, + "loss": 0.3217, + "step": 9350 + }, + { + "epoch": 0.6396470346808947, + "grad_norm": 2.2296367569985707, + "learning_rate": 3.036039031930245e-06, + "loss": 0.3208, + "step": 9351 + }, + { + "epoch": 0.6397154388125043, + "grad_norm": 3.126417639568294, + "learning_rate": 3.0350203586170513e-06, + "loss": 0.5184, + "step": 9352 + }, + { + "epoch": 0.6397838429441138, + "grad_norm": 1.7613343963545238, + "learning_rate": 3.034001781754546e-06, + "loss": 0.333, + "step": 9353 + }, + { + "epoch": 0.6398522470757234, + "grad_norm": 3.19898379514584, + "learning_rate": 3.032983301392728e-06, + "loss": 0.5033, + "step": 9354 + }, + { + "epoch": 0.6399206512073329, + "grad_norm": 2.012722757403156, + "learning_rate": 3.031964917581588e-06, + "loss": 0.2992, + "step": 9355 + }, + { + "epoch": 0.6399890553389425, + "grad_norm": 1.7059833837062748, + "learning_rate": 3.0309466303711114e-06, + "loss": 0.2749, + "step": 9356 + }, + { + "epoch": 0.640057459470552, + "grad_norm": 2.1278893273050565, + "learning_rate": 3.029928439811284e-06, + "loss": 0.3385, + "step": 9357 + }, + { + "epoch": 0.6401258636021616, + "grad_norm": 2.888958992143149, + "learning_rate": 3.028910345952083e-06, + "loss": 0.2459, + "step": 9358 + }, + { + "epoch": 0.6401942677337711, + "grad_norm": 3.281641618882646, + "learning_rate": 3.02789234884348e-06, + "loss": 0.244, + "step": 9359 + }, + { + "epoch": 0.6402626718653807, + "grad_norm": 2.0605434110350216, + "learning_rate": 3.0268744485354432e-06, + "loss": 0.3121, + "step": 9360 + }, + { + "epoch": 0.6403310759969902, + "grad_norm": 1.7907701113954988, + "learning_rate": 3.0258566450779375e-06, + "loss": 0.3293, + "step": 9361 + }, + { + "epoch": 0.6403994801285998, + "grad_norm": 2.0359126678673864, + "learning_rate": 3.024838938520921e-06, + "loss": 0.3792, + "step": 9362 + }, + { + "epoch": 0.6404678842602093, + "grad_norm": 2.56735395471164, + "learning_rate": 3.023821328914346e-06, + "loss": 0.8239, + "step": 9363 + }, + { + "epoch": 0.6405362883918189, + "grad_norm": 3.1587402226757537, + "learning_rate": 3.022803816308164e-06, + "loss": 0.471, + "step": 9364 + }, + { + "epoch": 0.6406046925234284, + "grad_norm": 1.803058136439092, + "learning_rate": 3.021786400752319e-06, + "loss": 0.215, + "step": 9365 + }, + { + "epoch": 0.6406730966550379, + "grad_norm": 2.9281093670276066, + "learning_rate": 3.0207690822967472e-06, + "loss": 0.192, + "step": 9366 + }, + { + "epoch": 0.6407415007866475, + "grad_norm": 2.0524836858084763, + "learning_rate": 3.0197518609913883e-06, + "loss": 0.2256, + "step": 9367 + }, + { + "epoch": 0.6408099049182571, + "grad_norm": 2.2864688787600933, + "learning_rate": 3.018734736886171e-06, + "loss": 0.2699, + "step": 9368 + }, + { + "epoch": 0.6408783090498666, + "grad_norm": 2.0088019029858772, + "learning_rate": 3.0177177100310205e-06, + "loss": 0.253, + "step": 9369 + }, + { + "epoch": 0.6409467131814761, + "grad_norm": 1.6762219451985925, + "learning_rate": 3.0167007804758565e-06, + "loss": 0.3049, + "step": 9370 + }, + { + "epoch": 0.6410151173130857, + "grad_norm": 1.874745218530187, + "learning_rate": 3.015683948270596e-06, + "loss": 0.3071, + "step": 9371 + }, + { + "epoch": 0.6410835214446953, + "grad_norm": 2.0088584249980626, + "learning_rate": 3.01466721346515e-06, + "loss": 0.3737, + "step": 9372 + }, + { + "epoch": 0.6411519255763048, + "grad_norm": 2.432713173251673, + "learning_rate": 3.0136505761094226e-06, + "loss": 0.4353, + "step": 9373 + }, + { + "epoch": 0.6412203297079143, + "grad_norm": 2.5202718666429, + "learning_rate": 3.012634036253318e-06, + "loss": 0.5763, + "step": 9374 + }, + { + "epoch": 0.6412887338395239, + "grad_norm": 1.9399116104798912, + "learning_rate": 3.011617593946732e-06, + "loss": 0.4272, + "step": 9375 + }, + { + "epoch": 0.6413571379711335, + "grad_norm": 1.5329856426741837, + "learning_rate": 3.0106012492395543e-06, + "loss": 0.2834, + "step": 9376 + }, + { + "epoch": 0.641425542102743, + "grad_norm": 2.044609548769795, + "learning_rate": 3.0095850021816754e-06, + "loss": 0.1721, + "step": 9377 + }, + { + "epoch": 0.6414939462343525, + "grad_norm": 2.9605665150640355, + "learning_rate": 3.0085688528229763e-06, + "loss": 0.2569, + "step": 9378 + }, + { + "epoch": 0.6415623503659621, + "grad_norm": 1.9961104569603771, + "learning_rate": 3.007552801213335e-06, + "loss": 0.4836, + "step": 9379 + }, + { + "epoch": 0.6416307544975717, + "grad_norm": 2.8326828253092433, + "learning_rate": 3.0065368474026226e-06, + "loss": 0.2724, + "step": 9380 + }, + { + "epoch": 0.6416991586291813, + "grad_norm": 2.2139125797445893, + "learning_rate": 3.005520991440709e-06, + "loss": 0.3541, + "step": 9381 + }, + { + "epoch": 0.6417675627607907, + "grad_norm": 1.921476163805877, + "learning_rate": 3.004505233377456e-06, + "loss": 0.3547, + "step": 9382 + }, + { + "epoch": 0.6418359668924003, + "grad_norm": 1.417482379382479, + "learning_rate": 3.0034895732627224e-06, + "loss": 0.2624, + "step": 9383 + }, + { + "epoch": 0.6419043710240099, + "grad_norm": 2.2288278332596785, + "learning_rate": 3.0024740111463613e-06, + "loss": 0.4726, + "step": 9384 + }, + { + "epoch": 0.6419727751556193, + "grad_norm": 1.9606686973991998, + "learning_rate": 3.001458547078222e-06, + "loss": 0.4903, + "step": 9385 + }, + { + "epoch": 0.6420411792872289, + "grad_norm": 2.2622756865252267, + "learning_rate": 3.000443181108147e-06, + "loss": 0.6089, + "step": 9386 + }, + { + "epoch": 0.6421095834188385, + "grad_norm": 2.2324746326743936, + "learning_rate": 2.999427913285976e-06, + "loss": 0.4431, + "step": 9387 + }, + { + "epoch": 0.6421779875504481, + "grad_norm": 2.040665219540048, + "learning_rate": 2.9984127436615456e-06, + "loss": 0.3128, + "step": 9388 + }, + { + "epoch": 0.6422463916820575, + "grad_norm": 2.0257065029160994, + "learning_rate": 2.997397672284683e-06, + "loss": 0.271, + "step": 9389 + }, + { + "epoch": 0.6423147958136671, + "grad_norm": 2.1082725116252075, + "learning_rate": 2.9963826992052126e-06, + "loss": 0.5618, + "step": 9390 + }, + { + "epoch": 0.6423831999452767, + "grad_norm": 1.8621602675035407, + "learning_rate": 2.9953678244729556e-06, + "loss": 0.4706, + "step": 9391 + }, + { + "epoch": 0.6424516040768863, + "grad_norm": 1.2787511782813215, + "learning_rate": 2.9943530481377255e-06, + "loss": 0.1058, + "step": 9392 + }, + { + "epoch": 0.6425200082084958, + "grad_norm": 1.6634067846383107, + "learning_rate": 2.9933383702493325e-06, + "loss": 0.2236, + "step": 9393 + }, + { + "epoch": 0.6425884123401053, + "grad_norm": 2.141589260805269, + "learning_rate": 2.992323790857583e-06, + "loss": 0.1429, + "step": 9394 + }, + { + "epoch": 0.6426568164717149, + "grad_norm": 1.3955802898659475, + "learning_rate": 2.9913093100122758e-06, + "loss": 0.1606, + "step": 9395 + }, + { + "epoch": 0.6427252206033245, + "grad_norm": 2.1010981192039484, + "learning_rate": 2.9902949277632075e-06, + "loss": 0.4625, + "step": 9396 + }, + { + "epoch": 0.642793624734934, + "grad_norm": 1.626313643698204, + "learning_rate": 2.989280644160169e-06, + "loss": 0.3233, + "step": 9397 + }, + { + "epoch": 0.6428620288665435, + "grad_norm": 2.0482296355906784, + "learning_rate": 2.9882664592529476e-06, + "loss": 0.387, + "step": 9398 + }, + { + "epoch": 0.6429304329981531, + "grad_norm": 2.348457161725263, + "learning_rate": 2.987252373091322e-06, + "loss": 0.7056, + "step": 9399 + }, + { + "epoch": 0.6429988371297627, + "grad_norm": 2.660958044630249, + "learning_rate": 2.986238385725069e-06, + "loss": 0.2978, + "step": 9400 + }, + { + "epoch": 0.6430672412613722, + "grad_norm": 2.5423798003614424, + "learning_rate": 2.9852244972039614e-06, + "loss": 0.4246, + "step": 9401 + }, + { + "epoch": 0.6431356453929817, + "grad_norm": 1.9272159296675655, + "learning_rate": 2.984210707577764e-06, + "loss": 0.3779, + "step": 9402 + }, + { + "epoch": 0.6432040495245913, + "grad_norm": 1.452523453087317, + "learning_rate": 2.98319701689624e-06, + "loss": 0.2824, + "step": 9403 + }, + { + "epoch": 0.6432724536562008, + "grad_norm": 2.021450661640719, + "learning_rate": 2.9821834252091455e-06, + "loss": 0.5057, + "step": 9404 + }, + { + "epoch": 0.6433408577878104, + "grad_norm": 1.817944832821948, + "learning_rate": 2.9811699325662306e-06, + "loss": 0.1605, + "step": 9405 + }, + { + "epoch": 0.6434092619194199, + "grad_norm": 1.9846852165200624, + "learning_rate": 2.980156539017245e-06, + "loss": 0.3305, + "step": 9406 + }, + { + "epoch": 0.6434776660510295, + "grad_norm": 1.7859521756774854, + "learning_rate": 2.9791432446119302e-06, + "loss": 0.2899, + "step": 9407 + }, + { + "epoch": 0.643546070182639, + "grad_norm": 2.4979258874103127, + "learning_rate": 2.9781300494000244e-06, + "loss": 0.4633, + "step": 9408 + }, + { + "epoch": 0.6436144743142486, + "grad_norm": 2.8644182307897066, + "learning_rate": 2.977116953431258e-06, + "loss": 0.173, + "step": 9409 + }, + { + "epoch": 0.6436828784458581, + "grad_norm": 2.322529552403098, + "learning_rate": 2.9761039567553617e-06, + "loss": 0.3015, + "step": 9410 + }, + { + "epoch": 0.6437512825774677, + "grad_norm": 2.6103245612109807, + "learning_rate": 2.975091059422056e-06, + "loss": 0.6027, + "step": 9411 + }, + { + "epoch": 0.6438196867090772, + "grad_norm": 2.3843569268657827, + "learning_rate": 2.9740782614810582e-06, + "loss": 0.3979, + "step": 9412 + }, + { + "epoch": 0.6438880908406868, + "grad_norm": 1.8417538974638392, + "learning_rate": 2.9730655629820836e-06, + "loss": 0.3451, + "step": 9413 + }, + { + "epoch": 0.6439564949722963, + "grad_norm": 2.518061630661693, + "learning_rate": 2.9720529639748392e-06, + "loss": 0.3845, + "step": 9414 + }, + { + "epoch": 0.6440248991039059, + "grad_norm": 2.1946312434446664, + "learning_rate": 2.9710404645090273e-06, + "loss": 0.4594, + "step": 9415 + }, + { + "epoch": 0.6440933032355154, + "grad_norm": 1.801506574942189, + "learning_rate": 2.970028064634346e-06, + "loss": 0.332, + "step": 9416 + }, + { + "epoch": 0.644161707367125, + "grad_norm": 2.329183747383047, + "learning_rate": 2.9690157644004915e-06, + "loss": 0.3482, + "step": 9417 + }, + { + "epoch": 0.6442301114987345, + "grad_norm": 2.1157795070649184, + "learning_rate": 2.9680035638571513e-06, + "loss": 0.3139, + "step": 9418 + }, + { + "epoch": 0.6442985156303441, + "grad_norm": 2.7323949620282812, + "learning_rate": 2.9669914630540074e-06, + "loss": 0.3959, + "step": 9419 + }, + { + "epoch": 0.6443669197619536, + "grad_norm": 1.803018301677857, + "learning_rate": 2.9659794620407413e-06, + "loss": 0.3339, + "step": 9420 + }, + { + "epoch": 0.6444353238935632, + "grad_norm": 2.4732104549678313, + "learning_rate": 2.9649675608670247e-06, + "loss": 0.5686, + "step": 9421 + }, + { + "epoch": 0.6445037280251728, + "grad_norm": 2.036280649064582, + "learning_rate": 2.963955759582526e-06, + "loss": 0.5049, + "step": 9422 + }, + { + "epoch": 0.6445721321567822, + "grad_norm": 1.7082008134995075, + "learning_rate": 2.962944058236912e-06, + "loss": 0.2149, + "step": 9423 + }, + { + "epoch": 0.6446405362883918, + "grad_norm": 2.1429500135380763, + "learning_rate": 2.9619324568798415e-06, + "loss": 0.49, + "step": 9424 + }, + { + "epoch": 0.6447089404200014, + "grad_norm": 1.9549935214918264, + "learning_rate": 2.960920955560965e-06, + "loss": 0.2065, + "step": 9425 + }, + { + "epoch": 0.644777344551611, + "grad_norm": 2.476551940365972, + "learning_rate": 2.959909554329935e-06, + "loss": 0.4589, + "step": 9426 + }, + { + "epoch": 0.6448457486832204, + "grad_norm": 2.3034088855250077, + "learning_rate": 2.9588982532363964e-06, + "loss": 0.4233, + "step": 9427 + }, + { + "epoch": 0.64491415281483, + "grad_norm": 1.3729533873159052, + "learning_rate": 2.9578870523299885e-06, + "loss": 0.1961, + "step": 9428 + }, + { + "epoch": 0.6449825569464396, + "grad_norm": 2.1514627415531327, + "learning_rate": 2.9568759516603445e-06, + "loss": 0.425, + "step": 9429 + }, + { + "epoch": 0.6450509610780492, + "grad_norm": 1.6145875274439654, + "learning_rate": 2.9558649512770954e-06, + "loss": 0.2737, + "step": 9430 + }, + { + "epoch": 0.6451193652096586, + "grad_norm": 2.352626642138534, + "learning_rate": 2.9548540512298663e-06, + "loss": 0.3529, + "step": 9431 + }, + { + "epoch": 0.6451877693412682, + "grad_norm": 2.08993367866188, + "learning_rate": 2.9538432515682747e-06, + "loss": 0.419, + "step": 9432 + }, + { + "epoch": 0.6452561734728778, + "grad_norm": 2.118932877302887, + "learning_rate": 2.9528325523419386e-06, + "loss": 0.5932, + "step": 9433 + }, + { + "epoch": 0.6453245776044874, + "grad_norm": 2.7170229282303264, + "learning_rate": 2.951821953600466e-06, + "loss": 0.7862, + "step": 9434 + }, + { + "epoch": 0.6453929817360968, + "grad_norm": 2.2271486012881003, + "learning_rate": 2.950811455393462e-06, + "loss": 0.2923, + "step": 9435 + }, + { + "epoch": 0.6454613858677064, + "grad_norm": 2.052545261110527, + "learning_rate": 2.9498010577705256e-06, + "loss": 0.392, + "step": 9436 + }, + { + "epoch": 0.645529789999316, + "grad_norm": 2.200933252244106, + "learning_rate": 2.948790760781256e-06, + "loss": 0.5309, + "step": 9437 + }, + { + "epoch": 0.6455981941309256, + "grad_norm": 2.453141461456854, + "learning_rate": 2.9477805644752407e-06, + "loss": 0.4174, + "step": 9438 + }, + { + "epoch": 0.645666598262535, + "grad_norm": 1.543561758695882, + "learning_rate": 2.946770468902064e-06, + "loss": 0.2629, + "step": 9439 + }, + { + "epoch": 0.6457350023941446, + "grad_norm": 1.7546009310719555, + "learning_rate": 2.94576047411131e-06, + "loss": 0.3057, + "step": 9440 + }, + { + "epoch": 0.6458034065257542, + "grad_norm": 1.5879060362083242, + "learning_rate": 2.9447505801525507e-06, + "loss": 0.3799, + "step": 9441 + }, + { + "epoch": 0.6458718106573637, + "grad_norm": 1.3412141714616759, + "learning_rate": 2.9437407870753564e-06, + "loss": 0.1286, + "step": 9442 + }, + { + "epoch": 0.6459402147889732, + "grad_norm": 1.5451802641678303, + "learning_rate": 2.9427310949292952e-06, + "loss": 0.2214, + "step": 9443 + }, + { + "epoch": 0.6460086189205828, + "grad_norm": 2.2573798137917436, + "learning_rate": 2.9417215037639266e-06, + "loss": 0.448, + "step": 9444 + }, + { + "epoch": 0.6460770230521924, + "grad_norm": 2.9170986037355706, + "learning_rate": 2.9407120136288046e-06, + "loss": 0.454, + "step": 9445 + }, + { + "epoch": 0.6461454271838019, + "grad_norm": 2.262798980326909, + "learning_rate": 2.9397026245734805e-06, + "loss": 0.2437, + "step": 9446 + }, + { + "epoch": 0.6462138313154114, + "grad_norm": 2.099566699636707, + "learning_rate": 2.9386933366475024e-06, + "loss": 0.2944, + "step": 9447 + }, + { + "epoch": 0.646282235447021, + "grad_norm": 1.8716498327078517, + "learning_rate": 2.9376841499004094e-06, + "loss": 0.3363, + "step": 9448 + }, + { + "epoch": 0.6463506395786306, + "grad_norm": 1.9166004326682053, + "learning_rate": 2.9366750643817365e-06, + "loss": 0.3131, + "step": 9449 + }, + { + "epoch": 0.6464190437102401, + "grad_norm": 1.6682183795817995, + "learning_rate": 2.935666080141015e-06, + "loss": 0.3203, + "step": 9450 + }, + { + "epoch": 0.6464874478418496, + "grad_norm": 2.155735396746155, + "learning_rate": 2.9346571972277715e-06, + "loss": 0.2079, + "step": 9451 + }, + { + "epoch": 0.6465558519734592, + "grad_norm": 1.968167748712296, + "learning_rate": 2.933648415691526e-06, + "loss": 0.3108, + "step": 9452 + }, + { + "epoch": 0.6466242561050688, + "grad_norm": 1.4417780316572804, + "learning_rate": 2.9326397355817947e-06, + "loss": 0.2337, + "step": 9453 + }, + { + "epoch": 0.6466926602366783, + "grad_norm": 2.0022517730508804, + "learning_rate": 2.931631156948088e-06, + "loss": 0.3963, + "step": 9454 + }, + { + "epoch": 0.6467610643682878, + "grad_norm": 1.8143095877449678, + "learning_rate": 2.930622679839911e-06, + "loss": 0.1954, + "step": 9455 + }, + { + "epoch": 0.6468294684998974, + "grad_norm": 2.6058797211925637, + "learning_rate": 2.9296143043067666e-06, + "loss": 0.5281, + "step": 9456 + }, + { + "epoch": 0.646897872631507, + "grad_norm": 2.069044062871216, + "learning_rate": 2.9286060303981507e-06, + "loss": 0.1682, + "step": 9457 + }, + { + "epoch": 0.6469662767631165, + "grad_norm": 1.7035972067461906, + "learning_rate": 2.9275978581635534e-06, + "loss": 0.3314, + "step": 9458 + }, + { + "epoch": 0.647034680894726, + "grad_norm": 2.399324482979456, + "learning_rate": 2.92658978765246e-06, + "loss": 0.2764, + "step": 9459 + }, + { + "epoch": 0.6471030850263356, + "grad_norm": 1.8753904452698267, + "learning_rate": 2.9255818189143537e-06, + "loss": 0.3894, + "step": 9460 + }, + { + "epoch": 0.6471714891579451, + "grad_norm": 1.958013474606059, + "learning_rate": 2.924573951998708e-06, + "loss": 0.442, + "step": 9461 + }, + { + "epoch": 0.6472398932895547, + "grad_norm": 2.5050201021461476, + "learning_rate": 2.9235661869549947e-06, + "loss": 0.3103, + "step": 9462 + }, + { + "epoch": 0.6473082974211642, + "grad_norm": 2.1645771612740847, + "learning_rate": 2.92255852383268e-06, + "loss": 0.4262, + "step": 9463 + }, + { + "epoch": 0.6473767015527738, + "grad_norm": 2.404211183782621, + "learning_rate": 2.9215509626812257e-06, + "loss": 0.3824, + "step": 9464 + }, + { + "epoch": 0.6474451056843833, + "grad_norm": 1.88021399020738, + "learning_rate": 2.920543503550084e-06, + "loss": 0.3125, + "step": 9465 + }, + { + "epoch": 0.6475135098159929, + "grad_norm": 2.0126482214455477, + "learning_rate": 2.9195361464887107e-06, + "loss": 0.3921, + "step": 9466 + }, + { + "epoch": 0.6475819139476024, + "grad_norm": 1.6078199493932581, + "learning_rate": 2.9185288915465492e-06, + "loss": 0.2319, + "step": 9467 + }, + { + "epoch": 0.647650318079212, + "grad_norm": 1.8466413343653902, + "learning_rate": 2.9175217387730424e-06, + "loss": 0.3812, + "step": 9468 + }, + { + "epoch": 0.6477187222108215, + "grad_norm": 1.823108425045419, + "learning_rate": 2.9165146882176242e-06, + "loss": 0.1552, + "step": 9469 + }, + { + "epoch": 0.6477871263424311, + "grad_norm": 2.0317617138623154, + "learning_rate": 2.915507739929727e-06, + "loss": 0.5088, + "step": 9470 + }, + { + "epoch": 0.6478555304740407, + "grad_norm": 2.1702762867135683, + "learning_rate": 2.9145008939587756e-06, + "loss": 0.3651, + "step": 9471 + }, + { + "epoch": 0.6479239346056502, + "grad_norm": 2.276137829341511, + "learning_rate": 2.913494150354191e-06, + "loss": 0.5172, + "step": 9472 + }, + { + "epoch": 0.6479923387372597, + "grad_norm": 1.9594897134714808, + "learning_rate": 2.912487509165387e-06, + "loss": 0.4207, + "step": 9473 + }, + { + "epoch": 0.6480607428688693, + "grad_norm": 1.7179963583677262, + "learning_rate": 2.9114809704417783e-06, + "loss": 0.2335, + "step": 9474 + }, + { + "epoch": 0.6481291470004789, + "grad_norm": 2.0537485435768574, + "learning_rate": 2.9104745342327673e-06, + "loss": 0.5261, + "step": 9475 + }, + { + "epoch": 0.6481975511320883, + "grad_norm": 2.1126291675692417, + "learning_rate": 2.9094682005877584e-06, + "loss": 0.5207, + "step": 9476 + }, + { + "epoch": 0.6482659552636979, + "grad_norm": 2.137153532436254, + "learning_rate": 2.908461969556145e-06, + "loss": 0.5691, + "step": 9477 + }, + { + "epoch": 0.6483343593953075, + "grad_norm": 2.7017407704316847, + "learning_rate": 2.9074558411873183e-06, + "loss": 0.247, + "step": 9478 + }, + { + "epoch": 0.6484027635269171, + "grad_norm": 2.388084120970859, + "learning_rate": 2.906449815530664e-06, + "loss": 0.508, + "step": 9479 + }, + { + "epoch": 0.6484711676585265, + "grad_norm": 3.2978521860453633, + "learning_rate": 2.90544389263556e-06, + "loss": 0.6964, + "step": 9480 + }, + { + "epoch": 0.6485395717901361, + "grad_norm": 2.8464629992136175, + "learning_rate": 2.904438072551386e-06, + "loss": 0.32, + "step": 9481 + }, + { + "epoch": 0.6486079759217457, + "grad_norm": 2.2790973186984638, + "learning_rate": 2.9034323553275114e-06, + "loss": 0.7282, + "step": 9482 + }, + { + "epoch": 0.6486763800533553, + "grad_norm": 2.406175637025022, + "learning_rate": 2.9024267410133004e-06, + "loss": 0.3934, + "step": 9483 + }, + { + "epoch": 0.6487447841849647, + "grad_norm": 1.779934194261238, + "learning_rate": 2.9014212296581146e-06, + "loss": 0.3011, + "step": 9484 + }, + { + "epoch": 0.6488131883165743, + "grad_norm": 1.929280234228593, + "learning_rate": 2.9004158213113065e-06, + "loss": 0.3253, + "step": 9485 + }, + { + "epoch": 0.6488815924481839, + "grad_norm": 2.27833001191535, + "learning_rate": 2.899410516022231e-06, + "loss": 0.6066, + "step": 9486 + }, + { + "epoch": 0.6489499965797935, + "grad_norm": 2.035409475962634, + "learning_rate": 2.898405313840228e-06, + "loss": 0.5305, + "step": 9487 + }, + { + "epoch": 0.6490184007114029, + "grad_norm": 2.320932051920734, + "learning_rate": 2.8974002148146442e-06, + "loss": 0.6441, + "step": 9488 + }, + { + "epoch": 0.6490868048430125, + "grad_norm": 2.0496574883314693, + "learning_rate": 2.89639521899481e-06, + "loss": 0.2269, + "step": 9489 + }, + { + "epoch": 0.6491552089746221, + "grad_norm": 2.3336264746880597, + "learning_rate": 2.895390326430057e-06, + "loss": 0.6555, + "step": 9490 + }, + { + "epoch": 0.6492236131062317, + "grad_norm": 1.596239095671532, + "learning_rate": 2.89438553716971e-06, + "loss": 0.2905, + "step": 9491 + }, + { + "epoch": 0.6492920172378411, + "grad_norm": 1.9272998602253149, + "learning_rate": 2.893380851263089e-06, + "loss": 0.4096, + "step": 9492 + }, + { + "epoch": 0.6493604213694507, + "grad_norm": 2.5260671607759626, + "learning_rate": 2.892376268759506e-06, + "loss": 0.3689, + "step": 9493 + }, + { + "epoch": 0.6494288255010603, + "grad_norm": 2.1338619717290803, + "learning_rate": 2.8913717897082754e-06, + "loss": 0.4658, + "step": 9494 + }, + { + "epoch": 0.6494972296326698, + "grad_norm": 3.751445523447744, + "learning_rate": 2.890367414158698e-06, + "loss": 0.3885, + "step": 9495 + }, + { + "epoch": 0.6495656337642793, + "grad_norm": 1.8615675469010378, + "learning_rate": 2.8893631421600775e-06, + "loss": 0.285, + "step": 9496 + }, + { + "epoch": 0.6496340378958889, + "grad_norm": 1.6290580349467116, + "learning_rate": 2.8883589737617057e-06, + "loss": 0.1731, + "step": 9497 + }, + { + "epoch": 0.6497024420274985, + "grad_norm": 2.0628743157142067, + "learning_rate": 2.8873549090128727e-06, + "loss": 0.4709, + "step": 9498 + }, + { + "epoch": 0.649770846159108, + "grad_norm": 1.3724477879958454, + "learning_rate": 2.8863509479628626e-06, + "loss": 0.1711, + "step": 9499 + }, + { + "epoch": 0.6498392502907175, + "grad_norm": 1.7605507313356215, + "learning_rate": 2.8853470906609526e-06, + "loss": 0.3796, + "step": 9500 + }, + { + "epoch": 0.6499076544223271, + "grad_norm": 1.5101826080848006, + "learning_rate": 2.884343337156421e-06, + "loss": 0.1884, + "step": 9501 + }, + { + "epoch": 0.6499760585539367, + "grad_norm": 2.340808637619346, + "learning_rate": 2.883339687498535e-06, + "loss": 0.3548, + "step": 9502 + }, + { + "epoch": 0.6500444626855462, + "grad_norm": 1.4801782056206358, + "learning_rate": 2.882336141736558e-06, + "loss": 0.2504, + "step": 9503 + }, + { + "epoch": 0.6501128668171557, + "grad_norm": 2.2817821535202487, + "learning_rate": 2.8813326999197494e-06, + "loss": 0.2705, + "step": 9504 + }, + { + "epoch": 0.6501812709487653, + "grad_norm": 1.8683542193693168, + "learning_rate": 2.880329362097361e-06, + "loss": 0.3631, + "step": 9505 + }, + { + "epoch": 0.6502496750803749, + "grad_norm": 2.1497652528656235, + "learning_rate": 2.8793261283186453e-06, + "loss": 0.3007, + "step": 9506 + }, + { + "epoch": 0.6503180792119844, + "grad_norm": 2.2674853480874018, + "learning_rate": 2.8783229986328414e-06, + "loss": 0.6106, + "step": 9507 + }, + { + "epoch": 0.650386483343594, + "grad_norm": 1.9645872763968009, + "learning_rate": 2.8773199730891925e-06, + "loss": 0.3118, + "step": 9508 + }, + { + "epoch": 0.6504548874752035, + "grad_norm": 2.9229150032986735, + "learning_rate": 2.8763170517369297e-06, + "loss": 0.3159, + "step": 9509 + }, + { + "epoch": 0.6505232916068131, + "grad_norm": 1.728880187560857, + "learning_rate": 2.8753142346252814e-06, + "loss": 0.2617, + "step": 9510 + }, + { + "epoch": 0.6505916957384226, + "grad_norm": 2.21419323176522, + "learning_rate": 2.87431152180347e-06, + "loss": 0.4776, + "step": 9511 + }, + { + "epoch": 0.6506600998700321, + "grad_norm": 2.48240087156617, + "learning_rate": 2.8733089133207143e-06, + "loss": 0.3784, + "step": 9512 + }, + { + "epoch": 0.6507285040016417, + "grad_norm": 1.7799777121702265, + "learning_rate": 2.872306409226225e-06, + "loss": 0.2659, + "step": 9513 + }, + { + "epoch": 0.6507969081332512, + "grad_norm": 2.0254994347533164, + "learning_rate": 2.871304009569213e-06, + "loss": 0.4014, + "step": 9514 + }, + { + "epoch": 0.6508653122648608, + "grad_norm": 3.229846548474432, + "learning_rate": 2.8703017143988775e-06, + "loss": 0.5355, + "step": 9515 + }, + { + "epoch": 0.6509337163964704, + "grad_norm": 2.2311193115364465, + "learning_rate": 2.8692995237644204e-06, + "loss": 0.3056, + "step": 9516 + }, + { + "epoch": 0.6510021205280799, + "grad_norm": 1.838944144317569, + "learning_rate": 2.8682974377150327e-06, + "loss": 0.3772, + "step": 9517 + }, + { + "epoch": 0.6510705246596894, + "grad_norm": 1.8115643314156875, + "learning_rate": 2.8672954562998993e-06, + "loss": 0.3161, + "step": 9518 + }, + { + "epoch": 0.651138928791299, + "grad_norm": 1.8125281701694078, + "learning_rate": 2.8662935795682046e-06, + "loss": 0.3863, + "step": 9519 + }, + { + "epoch": 0.6512073329229086, + "grad_norm": 1.4268458010927114, + "learning_rate": 2.865291807569122e-06, + "loss": 0.1669, + "step": 9520 + }, + { + "epoch": 0.6512757370545181, + "grad_norm": 2.2267133628197042, + "learning_rate": 2.8642901403518275e-06, + "loss": 0.4833, + "step": 9521 + }, + { + "epoch": 0.6513441411861276, + "grad_norm": 1.9210805064858327, + "learning_rate": 2.863288577965487e-06, + "loss": 0.288, + "step": 9522 + }, + { + "epoch": 0.6514125453177372, + "grad_norm": 2.1938403409848233, + "learning_rate": 2.862287120459261e-06, + "loss": 0.3654, + "step": 9523 + }, + { + "epoch": 0.6514809494493468, + "grad_norm": 2.290902286987931, + "learning_rate": 2.8612857678823053e-06, + "loss": 0.6196, + "step": 9524 + }, + { + "epoch": 0.6515493535809563, + "grad_norm": 2.339041016555825, + "learning_rate": 2.86028452028377e-06, + "loss": 0.514, + "step": 9525 + }, + { + "epoch": 0.6516177577125658, + "grad_norm": 2.347587762173227, + "learning_rate": 2.859283377712806e-06, + "loss": 0.6383, + "step": 9526 + }, + { + "epoch": 0.6516861618441754, + "grad_norm": 2.1426349169872907, + "learning_rate": 2.858282340218548e-06, + "loss": 0.4245, + "step": 9527 + }, + { + "epoch": 0.651754565975785, + "grad_norm": 3.030311616195451, + "learning_rate": 2.8572814078501375e-06, + "loss": 0.444, + "step": 9528 + }, + { + "epoch": 0.6518229701073945, + "grad_norm": 2.5218048217113114, + "learning_rate": 2.856280580656703e-06, + "loss": 0.3751, + "step": 9529 + }, + { + "epoch": 0.651891374239004, + "grad_norm": 2.264683353472388, + "learning_rate": 2.8552798586873694e-06, + "loss": 0.3338, + "step": 9530 + }, + { + "epoch": 0.6519597783706136, + "grad_norm": 2.2695385478127257, + "learning_rate": 2.8542792419912567e-06, + "loss": 0.3484, + "step": 9531 + }, + { + "epoch": 0.6520281825022232, + "grad_norm": 2.038241268539302, + "learning_rate": 2.8532787306174807e-06, + "loss": 0.3377, + "step": 9532 + }, + { + "epoch": 0.6520965866338326, + "grad_norm": 2.2034638428347555, + "learning_rate": 2.8522783246151487e-06, + "loss": 0.3664, + "step": 9533 + }, + { + "epoch": 0.6521649907654422, + "grad_norm": 2.0765629253860625, + "learning_rate": 2.8512780240333706e-06, + "loss": 0.336, + "step": 9534 + }, + { + "epoch": 0.6522333948970518, + "grad_norm": 1.4936249385338574, + "learning_rate": 2.8502778289212403e-06, + "loss": 0.2086, + "step": 9535 + }, + { + "epoch": 0.6523017990286614, + "grad_norm": 1.860949270593939, + "learning_rate": 2.849277739327857e-06, + "loss": 0.2335, + "step": 9536 + }, + { + "epoch": 0.6523702031602708, + "grad_norm": 8.066816726844127, + "learning_rate": 2.8482777553023084e-06, + "loss": 0.3524, + "step": 9537 + }, + { + "epoch": 0.6524386072918804, + "grad_norm": 1.9813537990142618, + "learning_rate": 2.8472778768936782e-06, + "loss": 0.3425, + "step": 9538 + }, + { + "epoch": 0.65250701142349, + "grad_norm": 2.476630026441254, + "learning_rate": 2.8462781041510446e-06, + "loss": 0.2993, + "step": 9539 + }, + { + "epoch": 0.6525754155550996, + "grad_norm": 1.963660450756624, + "learning_rate": 2.8452784371234806e-06, + "loss": 0.3634, + "step": 9540 + }, + { + "epoch": 0.652643819686709, + "grad_norm": 1.7853387200987323, + "learning_rate": 2.844278875860057e-06, + "loss": 0.3233, + "step": 9541 + }, + { + "epoch": 0.6527122238183186, + "grad_norm": 2.7138915004238275, + "learning_rate": 2.8432794204098367e-06, + "loss": 0.6353, + "step": 9542 + }, + { + "epoch": 0.6527806279499282, + "grad_norm": 1.8053479905723224, + "learning_rate": 2.842280070821877e-06, + "loss": 0.1718, + "step": 9543 + }, + { + "epoch": 0.6528490320815378, + "grad_norm": 1.7280991823085539, + "learning_rate": 2.8412808271452312e-06, + "loss": 0.3369, + "step": 9544 + }, + { + "epoch": 0.6529174362131472, + "grad_norm": 2.109270668981773, + "learning_rate": 2.840281689428945e-06, + "loss": 0.3607, + "step": 9545 + }, + { + "epoch": 0.6529858403447568, + "grad_norm": 2.0119275882800665, + "learning_rate": 2.839282657722064e-06, + "loss": 0.2776, + "step": 9546 + }, + { + "epoch": 0.6530542444763664, + "grad_norm": 2.0256943139349475, + "learning_rate": 2.8382837320736223e-06, + "loss": 0.3346, + "step": 9547 + }, + { + "epoch": 0.653122648607976, + "grad_norm": 1.9792801038122998, + "learning_rate": 2.837284912532656e-06, + "loss": 0.4343, + "step": 9548 + }, + { + "epoch": 0.6531910527395854, + "grad_norm": 2.2586595866775605, + "learning_rate": 2.83628619914819e-06, + "loss": 0.4457, + "step": 9549 + }, + { + "epoch": 0.653259456871195, + "grad_norm": 1.8792555676624294, + "learning_rate": 2.8352875919692456e-06, + "loss": 0.3466, + "step": 9550 + }, + { + "epoch": 0.6533278610028046, + "grad_norm": 2.2691932006250757, + "learning_rate": 2.8342890910448394e-06, + "loss": 0.4775, + "step": 9551 + }, + { + "epoch": 0.6533962651344141, + "grad_norm": 2.112107555627789, + "learning_rate": 2.8332906964239804e-06, + "loss": 0.2684, + "step": 9552 + }, + { + "epoch": 0.6534646692660236, + "grad_norm": 1.6911748106666624, + "learning_rate": 2.8322924081556795e-06, + "loss": 0.2419, + "step": 9553 + }, + { + "epoch": 0.6535330733976332, + "grad_norm": 1.9247851319967175, + "learning_rate": 2.8312942262889356e-06, + "loss": 0.3613, + "step": 9554 + }, + { + "epoch": 0.6536014775292428, + "grad_norm": 2.411518979116724, + "learning_rate": 2.83029615087274e-06, + "loss": 0.6143, + "step": 9555 + }, + { + "epoch": 0.6536698816608523, + "grad_norm": 2.096072730528789, + "learning_rate": 2.8292981819560895e-06, + "loss": 0.3845, + "step": 9556 + }, + { + "epoch": 0.6537382857924618, + "grad_norm": 2.8223304714686064, + "learning_rate": 2.8283003195879667e-06, + "loss": 0.8432, + "step": 9557 + }, + { + "epoch": 0.6538066899240714, + "grad_norm": 2.2587924137577144, + "learning_rate": 2.8273025638173513e-06, + "loss": 0.5237, + "step": 9558 + }, + { + "epoch": 0.653875094055681, + "grad_norm": 2.2201019855756376, + "learning_rate": 2.8263049146932153e-06, + "loss": 0.5216, + "step": 9559 + }, + { + "epoch": 0.6539434981872905, + "grad_norm": 1.865667656310205, + "learning_rate": 2.825307372264533e-06, + "loss": 0.2909, + "step": 9560 + }, + { + "epoch": 0.6540119023189, + "grad_norm": 1.5103939865864946, + "learning_rate": 2.824309936580267e-06, + "loss": 0.1607, + "step": 9561 + }, + { + "epoch": 0.6540803064505096, + "grad_norm": 2.2225027785589577, + "learning_rate": 2.8233126076893745e-06, + "loss": 0.2905, + "step": 9562 + }, + { + "epoch": 0.6541487105821192, + "grad_norm": 2.6168060878848305, + "learning_rate": 2.8223153856408116e-06, + "loss": 0.5071, + "step": 9563 + }, + { + "epoch": 0.6542171147137287, + "grad_norm": 1.2892882987424084, + "learning_rate": 2.821318270483523e-06, + "loss": 0.1055, + "step": 9564 + }, + { + "epoch": 0.6542855188453383, + "grad_norm": 2.2257521108244323, + "learning_rate": 2.8203212622664556e-06, + "loss": 0.1594, + "step": 9565 + }, + { + "epoch": 0.6543539229769478, + "grad_norm": 1.9052411364579565, + "learning_rate": 2.819324361038545e-06, + "loss": 0.3979, + "step": 9566 + }, + { + "epoch": 0.6544223271085574, + "grad_norm": 1.938255372387843, + "learning_rate": 2.818327566848727e-06, + "loss": 0.3383, + "step": 9567 + }, + { + "epoch": 0.6544907312401669, + "grad_norm": 1.8948428615436812, + "learning_rate": 2.8173308797459277e-06, + "loss": 0.4089, + "step": 9568 + }, + { + "epoch": 0.6545591353717765, + "grad_norm": 1.8673322433967496, + "learning_rate": 2.8163342997790675e-06, + "loss": 0.2976, + "step": 9569 + }, + { + "epoch": 0.654627539503386, + "grad_norm": 1.6112818208456956, + "learning_rate": 2.815337826997066e-06, + "loss": 0.2562, + "step": 9570 + }, + { + "epoch": 0.6546959436349955, + "grad_norm": 1.823164898910169, + "learning_rate": 2.814341461448833e-06, + "loss": 0.4204, + "step": 9571 + }, + { + "epoch": 0.6547643477666051, + "grad_norm": 1.7206663892030891, + "learning_rate": 2.8133452031832738e-06, + "loss": 0.2997, + "step": 9572 + }, + { + "epoch": 0.6548327518982147, + "grad_norm": 1.9401211620499548, + "learning_rate": 2.812349052249293e-06, + "loss": 0.5109, + "step": 9573 + }, + { + "epoch": 0.6549011560298242, + "grad_norm": 2.06012552663022, + "learning_rate": 2.811353008695783e-06, + "loss": 0.4029, + "step": 9574 + }, + { + "epoch": 0.6549695601614337, + "grad_norm": 2.3722074052088353, + "learning_rate": 2.810357072571638e-06, + "loss": 0.4676, + "step": 9575 + }, + { + "epoch": 0.6550379642930433, + "grad_norm": 1.708986110141418, + "learning_rate": 2.8093612439257412e-06, + "loss": 0.2505, + "step": 9576 + }, + { + "epoch": 0.6551063684246529, + "grad_norm": 1.8521625858173993, + "learning_rate": 2.808365522806974e-06, + "loss": 0.4918, + "step": 9577 + }, + { + "epoch": 0.6551747725562624, + "grad_norm": 1.8820630081556617, + "learning_rate": 2.807369909264209e-06, + "loss": 0.3006, + "step": 9578 + }, + { + "epoch": 0.6552431766878719, + "grad_norm": 1.6298527151977904, + "learning_rate": 2.8063744033463157e-06, + "loss": 0.1291, + "step": 9579 + }, + { + "epoch": 0.6553115808194815, + "grad_norm": 1.8093515625195018, + "learning_rate": 2.8053790051021617e-06, + "loss": 0.2591, + "step": 9580 + }, + { + "epoch": 0.6553799849510911, + "grad_norm": 2.074931006261213, + "learning_rate": 2.804383714580604e-06, + "loss": 0.5536, + "step": 9581 + }, + { + "epoch": 0.6554483890827006, + "grad_norm": 2.0401375191223026, + "learning_rate": 2.8033885318304956e-06, + "loss": 0.3228, + "step": 9582 + }, + { + "epoch": 0.6555167932143101, + "grad_norm": 1.7400001299723762, + "learning_rate": 2.8023934569006857e-06, + "loss": 0.298, + "step": 9583 + }, + { + "epoch": 0.6555851973459197, + "grad_norm": 1.9269209733218227, + "learning_rate": 2.8013984898400148e-06, + "loss": 0.3188, + "step": 9584 + }, + { + "epoch": 0.6556536014775293, + "grad_norm": 1.9983938965503214, + "learning_rate": 2.8004036306973247e-06, + "loss": 0.562, + "step": 9585 + }, + { + "epoch": 0.6557220056091388, + "grad_norm": 2.4721110316845047, + "learning_rate": 2.799408879521445e-06, + "loss": 0.4177, + "step": 9586 + }, + { + "epoch": 0.6557904097407483, + "grad_norm": 2.2745386124769715, + "learning_rate": 2.7984142363612055e-06, + "loss": 0.5854, + "step": 9587 + }, + { + "epoch": 0.6558588138723579, + "grad_norm": 2.340171268830362, + "learning_rate": 2.797419701265427e-06, + "loss": 0.3711, + "step": 9588 + }, + { + "epoch": 0.6559272180039675, + "grad_norm": 1.865907805140562, + "learning_rate": 2.796425274282925e-06, + "loss": 0.3846, + "step": 9589 + }, + { + "epoch": 0.6559956221355769, + "grad_norm": 2.0904182535703693, + "learning_rate": 2.795430955462512e-06, + "loss": 0.4274, + "step": 9590 + }, + { + "epoch": 0.6560640262671865, + "grad_norm": 2.2241380307738754, + "learning_rate": 2.7944367448529937e-06, + "loss": 0.4559, + "step": 9591 + }, + { + "epoch": 0.6561324303987961, + "grad_norm": 2.163119375689452, + "learning_rate": 2.793442642503168e-06, + "loss": 0.3906, + "step": 9592 + }, + { + "epoch": 0.6562008345304057, + "grad_norm": 2.288368409549106, + "learning_rate": 2.792448648461835e-06, + "loss": 0.3773, + "step": 9593 + }, + { + "epoch": 0.6562692386620151, + "grad_norm": 2.2587332778421447, + "learning_rate": 2.7914547627777807e-06, + "loss": 0.2095, + "step": 9594 + }, + { + "epoch": 0.6563376427936247, + "grad_norm": 1.619661769585286, + "learning_rate": 2.790460985499793e-06, + "loss": 0.2791, + "step": 9595 + }, + { + "epoch": 0.6564060469252343, + "grad_norm": 2.158840206240683, + "learning_rate": 2.7894673166766507e-06, + "loss": 0.2091, + "step": 9596 + }, + { + "epoch": 0.6564744510568439, + "grad_norm": 1.7498072449931255, + "learning_rate": 2.788473756357126e-06, + "loss": 0.2097, + "step": 9597 + }, + { + "epoch": 0.6565428551884533, + "grad_norm": 2.3130421776156447, + "learning_rate": 2.7874803045899895e-06, + "loss": 0.2751, + "step": 9598 + }, + { + "epoch": 0.6566112593200629, + "grad_norm": 2.037415453116199, + "learning_rate": 2.7864869614240013e-06, + "loss": 0.3364, + "step": 9599 + }, + { + "epoch": 0.6566796634516725, + "grad_norm": 1.7034209108078486, + "learning_rate": 2.7854937269079246e-06, + "loss": 0.294, + "step": 9600 + }, + { + "epoch": 0.6567480675832821, + "grad_norm": 2.437600152085587, + "learning_rate": 2.784500601090509e-06, + "loss": 0.6099, + "step": 9601 + }, + { + "epoch": 0.6568164717148915, + "grad_norm": 1.89810267976984, + "learning_rate": 2.783507584020503e-06, + "loss": 0.4596, + "step": 9602 + }, + { + "epoch": 0.6568848758465011, + "grad_norm": 1.9583614610282034, + "learning_rate": 2.782514675746647e-06, + "loss": 0.5105, + "step": 9603 + }, + { + "epoch": 0.6569532799781107, + "grad_norm": 1.960707005028306, + "learning_rate": 2.7815218763176777e-06, + "loss": 0.4499, + "step": 9604 + }, + { + "epoch": 0.6570216841097202, + "grad_norm": 1.8563693701769357, + "learning_rate": 2.7805291857823304e-06, + "loss": 0.4646, + "step": 9605 + }, + { + "epoch": 0.6570900882413298, + "grad_norm": 1.7067200924879335, + "learning_rate": 2.7795366041893246e-06, + "loss": 0.1339, + "step": 9606 + }, + { + "epoch": 0.6571584923729393, + "grad_norm": 2.6315287876132807, + "learning_rate": 2.778544131587388e-06, + "loss": 0.4068, + "step": 9607 + }, + { + "epoch": 0.6572268965045489, + "grad_norm": 1.5923009366888576, + "learning_rate": 2.777551768025232e-06, + "loss": 0.2651, + "step": 9608 + }, + { + "epoch": 0.6572953006361584, + "grad_norm": 2.2757698889048386, + "learning_rate": 2.7765595135515677e-06, + "loss": 0.4869, + "step": 9609 + }, + { + "epoch": 0.657363704767768, + "grad_norm": 2.354651111506402, + "learning_rate": 2.7755673682150997e-06, + "loss": 0.3559, + "step": 9610 + }, + { + "epoch": 0.6574321088993775, + "grad_norm": 2.032105991759176, + "learning_rate": 2.7745753320645265e-06, + "loss": 0.5325, + "step": 9611 + }, + { + "epoch": 0.6575005130309871, + "grad_norm": 3.0562006527089864, + "learning_rate": 2.7735834051485406e-06, + "loss": 0.3041, + "step": 9612 + }, + { + "epoch": 0.6575689171625966, + "grad_norm": 1.8087787224618523, + "learning_rate": 2.772591587515834e-06, + "loss": 0.2431, + "step": 9613 + }, + { + "epoch": 0.6576373212942062, + "grad_norm": 1.8005784224603603, + "learning_rate": 2.771599879215087e-06, + "loss": 0.3763, + "step": 9614 + }, + { + "epoch": 0.6577057254258157, + "grad_norm": 2.1193725634802467, + "learning_rate": 2.7706082802949807e-06, + "loss": 0.4204, + "step": 9615 + }, + { + "epoch": 0.6577741295574253, + "grad_norm": 1.2005706957421605, + "learning_rate": 2.7696167908041856e-06, + "loss": 0.1691, + "step": 9616 + }, + { + "epoch": 0.6578425336890348, + "grad_norm": 2.18985985441407, + "learning_rate": 2.7686254107913686e-06, + "loss": 0.4427, + "step": 9617 + }, + { + "epoch": 0.6579109378206444, + "grad_norm": 2.156937524126404, + "learning_rate": 2.7676341403051916e-06, + "loss": 0.5241, + "step": 9618 + }, + { + "epoch": 0.6579793419522539, + "grad_norm": 2.1052943849286776, + "learning_rate": 2.7666429793943087e-06, + "loss": 0.4571, + "step": 9619 + }, + { + "epoch": 0.6580477460838635, + "grad_norm": 1.8711741973689253, + "learning_rate": 2.7656519281073756e-06, + "loss": 0.2551, + "step": 9620 + }, + { + "epoch": 0.658116150215473, + "grad_norm": 2.2291073995507267, + "learning_rate": 2.764660986493035e-06, + "loss": 0.473, + "step": 9621 + }, + { + "epoch": 0.6581845543470826, + "grad_norm": 2.011499681859391, + "learning_rate": 2.7636701545999283e-06, + "loss": 0.2769, + "step": 9622 + }, + { + "epoch": 0.6582529584786921, + "grad_norm": 2.0661684018354443, + "learning_rate": 2.762679432476689e-06, + "loss": 0.2779, + "step": 9623 + }, + { + "epoch": 0.6583213626103016, + "grad_norm": 1.7635841812519055, + "learning_rate": 2.761688820171945e-06, + "loss": 0.4472, + "step": 9624 + }, + { + "epoch": 0.6583897667419112, + "grad_norm": 2.4614199089853517, + "learning_rate": 2.760698317734325e-06, + "loss": 0.3733, + "step": 9625 + }, + { + "epoch": 0.6584581708735208, + "grad_norm": 2.19984276085005, + "learning_rate": 2.7597079252124425e-06, + "loss": 0.4579, + "step": 9626 + }, + { + "epoch": 0.6585265750051303, + "grad_norm": 1.8731331909784665, + "learning_rate": 2.7587176426549155e-06, + "loss": 0.3164, + "step": 9627 + }, + { + "epoch": 0.6585949791367398, + "grad_norm": 1.5713869281304238, + "learning_rate": 2.7577274701103495e-06, + "loss": 0.1903, + "step": 9628 + }, + { + "epoch": 0.6586633832683494, + "grad_norm": 2.0424094884304447, + "learning_rate": 2.7567374076273466e-06, + "loss": 0.4897, + "step": 9629 + }, + { + "epoch": 0.658731787399959, + "grad_norm": 1.8447672454241408, + "learning_rate": 2.755747455254505e-06, + "loss": 0.3486, + "step": 9630 + }, + { + "epoch": 0.6588001915315685, + "grad_norm": 1.9562005470959873, + "learning_rate": 2.754757613040415e-06, + "loss": 0.4583, + "step": 9631 + }, + { + "epoch": 0.658868595663178, + "grad_norm": 2.043059994353967, + "learning_rate": 2.753767881033663e-06, + "loss": 0.3432, + "step": 9632 + }, + { + "epoch": 0.6589369997947876, + "grad_norm": 2.6072103493146583, + "learning_rate": 2.7527782592828312e-06, + "loss": 0.3685, + "step": 9633 + }, + { + "epoch": 0.6590054039263972, + "grad_norm": 1.6127934307175107, + "learning_rate": 2.7517887478364925e-06, + "loss": 0.1694, + "step": 9634 + }, + { + "epoch": 0.6590738080580067, + "grad_norm": 1.8484984367690054, + "learning_rate": 2.7507993467432205e-06, + "loss": 0.4557, + "step": 9635 + }, + { + "epoch": 0.6591422121896162, + "grad_norm": 1.8635317316176059, + "learning_rate": 2.749810056051577e-06, + "loss": 0.3898, + "step": 9636 + }, + { + "epoch": 0.6592106163212258, + "grad_norm": 2.1725769829581596, + "learning_rate": 2.7488208758101233e-06, + "loss": 0.5068, + "step": 9637 + }, + { + "epoch": 0.6592790204528354, + "grad_norm": 2.1148704786411847, + "learning_rate": 2.7478318060674113e-06, + "loss": 0.5795, + "step": 9638 + }, + { + "epoch": 0.659347424584445, + "grad_norm": 2.084176080640021, + "learning_rate": 2.7468428468719877e-06, + "loss": 0.5621, + "step": 9639 + }, + { + "epoch": 0.6594158287160544, + "grad_norm": 1.6667857832860302, + "learning_rate": 2.7458539982724e-06, + "loss": 0.2449, + "step": 9640 + }, + { + "epoch": 0.659484232847664, + "grad_norm": 1.912342824897508, + "learning_rate": 2.744865260317183e-06, + "loss": 0.3587, + "step": 9641 + }, + { + "epoch": 0.6595526369792736, + "grad_norm": 2.6454951728527303, + "learning_rate": 2.743876633054868e-06, + "loss": 0.3977, + "step": 9642 + }, + { + "epoch": 0.659621041110883, + "grad_norm": 2.331164002302471, + "learning_rate": 2.7428881165339837e-06, + "loss": 0.4733, + "step": 9643 + }, + { + "epoch": 0.6596894452424926, + "grad_norm": 1.715853908429198, + "learning_rate": 2.741899710803048e-06, + "loss": 0.2587, + "step": 9644 + }, + { + "epoch": 0.6597578493741022, + "grad_norm": 2.355555986652045, + "learning_rate": 2.740911415910581e-06, + "loss": 0.4951, + "step": 9645 + }, + { + "epoch": 0.6598262535057118, + "grad_norm": 3.019318305399061, + "learning_rate": 2.7399232319050885e-06, + "loss": 0.2577, + "step": 9646 + }, + { + "epoch": 0.6598946576373212, + "grad_norm": 1.3487389458529462, + "learning_rate": 2.738935158835079e-06, + "loss": 0.1688, + "step": 9647 + }, + { + "epoch": 0.6599630617689308, + "grad_norm": 4.867174690437911, + "learning_rate": 2.7379471967490513e-06, + "loss": 0.5737, + "step": 9648 + }, + { + "epoch": 0.6600314659005404, + "grad_norm": 2.7269443583405604, + "learning_rate": 2.736959345695498e-06, + "loss": 0.3134, + "step": 9649 + }, + { + "epoch": 0.66009987003215, + "grad_norm": 1.6371982573090091, + "learning_rate": 2.735971605722908e-06, + "loss": 0.25, + "step": 9650 + }, + { + "epoch": 0.6601682741637594, + "grad_norm": 2.551660062112573, + "learning_rate": 2.734983976879763e-06, + "loss": 0.3153, + "step": 9651 + }, + { + "epoch": 0.660236678295369, + "grad_norm": 2.0198018613980677, + "learning_rate": 2.733996459214544e-06, + "loss": 0.2754, + "step": 9652 + }, + { + "epoch": 0.6603050824269786, + "grad_norm": 2.2252441887052687, + "learning_rate": 2.733009052775721e-06, + "loss": 0.4186, + "step": 9653 + }, + { + "epoch": 0.6603734865585882, + "grad_norm": 2.2194686853975907, + "learning_rate": 2.7320217576117593e-06, + "loss": 0.3485, + "step": 9654 + }, + { + "epoch": 0.6604418906901977, + "grad_norm": 1.9533678437896036, + "learning_rate": 2.731034573771123e-06, + "loss": 0.4727, + "step": 9655 + }, + { + "epoch": 0.6605102948218072, + "grad_norm": 2.236658903088103, + "learning_rate": 2.7300475013022666e-06, + "loss": 0.4655, + "step": 9656 + }, + { + "epoch": 0.6605786989534168, + "grad_norm": 1.9275944242529268, + "learning_rate": 2.729060540253641e-06, + "loss": 0.3138, + "step": 9657 + }, + { + "epoch": 0.6606471030850264, + "grad_norm": 2.058561082323798, + "learning_rate": 2.728073690673687e-06, + "loss": 0.3559, + "step": 9658 + }, + { + "epoch": 0.6607155072166359, + "grad_norm": 1.8939177473449555, + "learning_rate": 2.7270869526108507e-06, + "loss": 0.2616, + "step": 9659 + }, + { + "epoch": 0.6607839113482454, + "grad_norm": 1.4927751771965905, + "learning_rate": 2.7261003261135623e-06, + "loss": 0.1661, + "step": 9660 + }, + { + "epoch": 0.660852315479855, + "grad_norm": 1.9009657613850526, + "learning_rate": 2.72511381123025e-06, + "loss": 0.5225, + "step": 9661 + }, + { + "epoch": 0.6609207196114645, + "grad_norm": 2.014995014448364, + "learning_rate": 2.724127408009337e-06, + "loss": 0.5405, + "step": 9662 + }, + { + "epoch": 0.6609891237430741, + "grad_norm": 1.6969549555408825, + "learning_rate": 2.7231411164992394e-06, + "loss": 0.1851, + "step": 9663 + }, + { + "epoch": 0.6610575278746836, + "grad_norm": 1.6502183903315177, + "learning_rate": 2.722154936748372e-06, + "loss": 0.2258, + "step": 9664 + }, + { + "epoch": 0.6611259320062932, + "grad_norm": 5.290544119237508, + "learning_rate": 2.7211688688051387e-06, + "loss": 0.6823, + "step": 9665 + }, + { + "epoch": 0.6611943361379027, + "grad_norm": 2.080614212514655, + "learning_rate": 2.7201829127179435e-06, + "loss": 0.2442, + "step": 9666 + }, + { + "epoch": 0.6612627402695123, + "grad_norm": 1.6311683818889908, + "learning_rate": 2.7191970685351797e-06, + "loss": 0.2535, + "step": 9667 + }, + { + "epoch": 0.6613311444011218, + "grad_norm": 1.8677624651171625, + "learning_rate": 2.718211336305238e-06, + "loss": 0.4204, + "step": 9668 + }, + { + "epoch": 0.6613995485327314, + "grad_norm": 2.210927269039139, + "learning_rate": 2.7172257160765024e-06, + "loss": 0.4095, + "step": 9669 + }, + { + "epoch": 0.6614679526643409, + "grad_norm": 1.7961864064240927, + "learning_rate": 2.716240207897352e-06, + "loss": 0.2999, + "step": 9670 + }, + { + "epoch": 0.6615363567959505, + "grad_norm": 2.168095559200428, + "learning_rate": 2.7152548118161588e-06, + "loss": 0.3287, + "step": 9671 + }, + { + "epoch": 0.66160476092756, + "grad_norm": 1.830069373973984, + "learning_rate": 2.7142695278812934e-06, + "loss": 0.2052, + "step": 9672 + }, + { + "epoch": 0.6616731650591696, + "grad_norm": 1.5931400646606397, + "learning_rate": 2.7132843561411165e-06, + "loss": 0.1939, + "step": 9673 + }, + { + "epoch": 0.6617415691907791, + "grad_norm": 2.2283034762772034, + "learning_rate": 2.712299296643987e-06, + "loss": 0.3888, + "step": 9674 + }, + { + "epoch": 0.6618099733223887, + "grad_norm": 1.8662363526726162, + "learning_rate": 2.7113143494382553e-06, + "loss": 0.2008, + "step": 9675 + }, + { + "epoch": 0.6618783774539982, + "grad_norm": 1.7267761443508767, + "learning_rate": 2.7103295145722676e-06, + "loss": 0.3363, + "step": 9676 + }, + { + "epoch": 0.6619467815856078, + "grad_norm": 2.118194840147372, + "learning_rate": 2.709344792094364e-06, + "loss": 0.3731, + "step": 9677 + }, + { + "epoch": 0.6620151857172173, + "grad_norm": 2.4043577234546376, + "learning_rate": 2.7083601820528773e-06, + "loss": 0.4728, + "step": 9678 + }, + { + "epoch": 0.6620835898488269, + "grad_norm": 1.4418141271261289, + "learning_rate": 2.7073756844961407e-06, + "loss": 0.1873, + "step": 9679 + }, + { + "epoch": 0.6621519939804364, + "grad_norm": 2.319195494565719, + "learning_rate": 2.7063912994724774e-06, + "loss": 0.3641, + "step": 9680 + }, + { + "epoch": 0.6622203981120459, + "grad_norm": 2.1474262458528353, + "learning_rate": 2.705407027030204e-06, + "loss": 0.4417, + "step": 9681 + }, + { + "epoch": 0.6622888022436555, + "grad_norm": 2.376060449865191, + "learning_rate": 2.704422867217634e-06, + "loss": 0.6064, + "step": 9682 + }, + { + "epoch": 0.6623572063752651, + "grad_norm": 1.9621856030635594, + "learning_rate": 2.703438820083074e-06, + "loss": 0.3145, + "step": 9683 + }, + { + "epoch": 0.6624256105068747, + "grad_norm": 2.0987300539630915, + "learning_rate": 2.702454885674828e-06, + "loss": 0.4587, + "step": 9684 + }, + { + "epoch": 0.6624940146384841, + "grad_norm": 2.044213209534697, + "learning_rate": 2.701471064041189e-06, + "loss": 0.1472, + "step": 9685 + }, + { + "epoch": 0.6625624187700937, + "grad_norm": 1.742428929659667, + "learning_rate": 2.7004873552304516e-06, + "loss": 0.3598, + "step": 9686 + }, + { + "epoch": 0.6626308229017033, + "grad_norm": 2.7463909004677944, + "learning_rate": 2.699503759290899e-06, + "loss": 0.3724, + "step": 9687 + }, + { + "epoch": 0.6626992270333129, + "grad_norm": 2.080982645578809, + "learning_rate": 2.698520276270811e-06, + "loss": 0.5297, + "step": 9688 + }, + { + "epoch": 0.6627676311649223, + "grad_norm": 2.0854045315623475, + "learning_rate": 2.697536906218462e-06, + "loss": 0.4676, + "step": 9689 + }, + { + "epoch": 0.6628360352965319, + "grad_norm": 2.1295922250349957, + "learning_rate": 2.6965536491821198e-06, + "loss": 0.2108, + "step": 9690 + }, + { + "epoch": 0.6629044394281415, + "grad_norm": 2.8227841966709915, + "learning_rate": 2.6955705052100457e-06, + "loss": 0.4051, + "step": 9691 + }, + { + "epoch": 0.6629728435597511, + "grad_norm": 1.8636898496453023, + "learning_rate": 2.6945874743505006e-06, + "loss": 0.3871, + "step": 9692 + }, + { + "epoch": 0.6630412476913605, + "grad_norm": 2.8407947149661035, + "learning_rate": 2.6936045566517333e-06, + "loss": 0.1499, + "step": 9693 + }, + { + "epoch": 0.6631096518229701, + "grad_norm": 2.4396029411839684, + "learning_rate": 2.6926217521619946e-06, + "loss": 0.424, + "step": 9694 + }, + { + "epoch": 0.6631780559545797, + "grad_norm": 2.2877821221983776, + "learning_rate": 2.6916390609295217e-06, + "loss": 0.4026, + "step": 9695 + }, + { + "epoch": 0.6632464600861893, + "grad_norm": 2.8103553595382382, + "learning_rate": 2.6906564830025506e-06, + "loss": 0.5062, + "step": 9696 + }, + { + "epoch": 0.6633148642177987, + "grad_norm": 1.7604265866385713, + "learning_rate": 2.6896740184293113e-06, + "loss": 0.3853, + "step": 9697 + }, + { + "epoch": 0.6633832683494083, + "grad_norm": 2.6452101564297164, + "learning_rate": 2.688691667258026e-06, + "loss": 0.9587, + "step": 9698 + }, + { + "epoch": 0.6634516724810179, + "grad_norm": 3.074462338095147, + "learning_rate": 2.6877094295369167e-06, + "loss": 0.4081, + "step": 9699 + }, + { + "epoch": 0.6635200766126274, + "grad_norm": 2.4823823470983664, + "learning_rate": 2.686727305314194e-06, + "loss": 0.6307, + "step": 9700 + }, + { + "epoch": 0.6635884807442369, + "grad_norm": 2.284194623320155, + "learning_rate": 2.6857452946380663e-06, + "loss": 0.3624, + "step": 9701 + }, + { + "epoch": 0.6636568848758465, + "grad_norm": 1.9281403675425246, + "learning_rate": 2.6847633975567346e-06, + "loss": 0.3553, + "step": 9702 + }, + { + "epoch": 0.6637252890074561, + "grad_norm": 1.7339924141185532, + "learning_rate": 2.6837816141183937e-06, + "loss": 0.2971, + "step": 9703 + }, + { + "epoch": 0.6637936931390656, + "grad_norm": 2.575956889473726, + "learning_rate": 2.682799944371238e-06, + "loss": 0.1633, + "step": 9704 + }, + { + "epoch": 0.6638620972706751, + "grad_norm": 2.274677790366757, + "learning_rate": 2.6818183883634492e-06, + "loss": 0.4262, + "step": 9705 + }, + { + "epoch": 0.6639305014022847, + "grad_norm": 2.4596868455450025, + "learning_rate": 2.6808369461432098e-06, + "loss": 0.3357, + "step": 9706 + }, + { + "epoch": 0.6639989055338943, + "grad_norm": 2.7489688712035285, + "learning_rate": 2.6798556177586923e-06, + "loss": 0.2698, + "step": 9707 + }, + { + "epoch": 0.6640673096655038, + "grad_norm": 1.701947667093655, + "learning_rate": 2.678874403258065e-06, + "loss": 0.2795, + "step": 9708 + }, + { + "epoch": 0.6641357137971133, + "grad_norm": 2.1337985669236446, + "learning_rate": 2.6778933026894906e-06, + "loss": 0.2615, + "step": 9709 + }, + { + "epoch": 0.6642041179287229, + "grad_norm": 1.9154966470056243, + "learning_rate": 2.6769123161011255e-06, + "loss": 0.4182, + "step": 9710 + }, + { + "epoch": 0.6642725220603325, + "grad_norm": 1.7551546623917618, + "learning_rate": 2.675931443541121e-06, + "loss": 0.3003, + "step": 9711 + }, + { + "epoch": 0.664340926191942, + "grad_norm": 1.643709815734395, + "learning_rate": 2.6749506850576257e-06, + "loss": 0.3449, + "step": 9712 + }, + { + "epoch": 0.6644093303235515, + "grad_norm": 1.8147339126251176, + "learning_rate": 2.6739700406987767e-06, + "loss": 0.2864, + "step": 9713 + }, + { + "epoch": 0.6644777344551611, + "grad_norm": 2.639538761616666, + "learning_rate": 2.6729895105127124e-06, + "loss": 0.3896, + "step": 9714 + }, + { + "epoch": 0.6645461385867707, + "grad_norm": 1.9258013989243699, + "learning_rate": 2.672009094547559e-06, + "loss": 0.3874, + "step": 9715 + }, + { + "epoch": 0.6646145427183802, + "grad_norm": 2.586836880331379, + "learning_rate": 2.6710287928514415e-06, + "loss": 0.4255, + "step": 9716 + }, + { + "epoch": 0.6646829468499897, + "grad_norm": 2.433344355214209, + "learning_rate": 2.670048605472478e-06, + "loss": 0.3574, + "step": 9717 + }, + { + "epoch": 0.6647513509815993, + "grad_norm": 1.8407572606466938, + "learning_rate": 2.6690685324587774e-06, + "loss": 0.2089, + "step": 9718 + }, + { + "epoch": 0.6648197551132088, + "grad_norm": 2.1547673421538733, + "learning_rate": 2.6680885738584512e-06, + "loss": 0.4968, + "step": 9719 + }, + { + "epoch": 0.6648881592448184, + "grad_norm": 2.1370745542210434, + "learning_rate": 2.6671087297195986e-06, + "loss": 0.3655, + "step": 9720 + }, + { + "epoch": 0.664956563376428, + "grad_norm": 2.098420321862188, + "learning_rate": 2.6661290000903152e-06, + "loss": 0.481, + "step": 9721 + }, + { + "epoch": 0.6650249675080375, + "grad_norm": 2.3183943079796774, + "learning_rate": 2.6651493850186906e-06, + "loss": 0.2106, + "step": 9722 + }, + { + "epoch": 0.665093371639647, + "grad_norm": 1.8838623056366879, + "learning_rate": 2.664169884552806e-06, + "loss": 0.1837, + "step": 9723 + }, + { + "epoch": 0.6651617757712566, + "grad_norm": 2.18221592980354, + "learning_rate": 2.6631904987407462e-06, + "loss": 0.4582, + "step": 9724 + }, + { + "epoch": 0.6652301799028661, + "grad_norm": 2.0399296988266755, + "learning_rate": 2.662211227630579e-06, + "loss": 0.4733, + "step": 9725 + }, + { + "epoch": 0.6652985840344757, + "grad_norm": 2.3435668492281803, + "learning_rate": 2.661232071270376e-06, + "loss": 0.4226, + "step": 9726 + }, + { + "epoch": 0.6653669881660852, + "grad_norm": 2.2432881289006854, + "learning_rate": 2.660253029708196e-06, + "loss": 0.488, + "step": 9727 + }, + { + "epoch": 0.6654353922976948, + "grad_norm": 1.6812167813164003, + "learning_rate": 2.6592741029920966e-06, + "loss": 0.3, + "step": 9728 + }, + { + "epoch": 0.6655037964293044, + "grad_norm": 2.358637031190914, + "learning_rate": 2.6582952911701266e-06, + "loss": 0.4127, + "step": 9729 + }, + { + "epoch": 0.6655722005609139, + "grad_norm": 2.1499772293803443, + "learning_rate": 2.657316594290332e-06, + "loss": 0.5003, + "step": 9730 + }, + { + "epoch": 0.6656406046925234, + "grad_norm": 1.758813149362995, + "learning_rate": 2.6563380124007505e-06, + "loss": 0.2497, + "step": 9731 + }, + { + "epoch": 0.665709008824133, + "grad_norm": 1.508991714250342, + "learning_rate": 2.655359545549418e-06, + "loss": 0.2992, + "step": 9732 + }, + { + "epoch": 0.6657774129557426, + "grad_norm": 1.7949772972541398, + "learning_rate": 2.654381193784359e-06, + "loss": 0.3265, + "step": 9733 + }, + { + "epoch": 0.665845817087352, + "grad_norm": 2.377421618943556, + "learning_rate": 2.6534029571536e-06, + "loss": 0.3323, + "step": 9734 + }, + { + "epoch": 0.6659142212189616, + "grad_norm": 2.238810970875507, + "learning_rate": 2.6524248357051564e-06, + "loss": 0.392, + "step": 9735 + }, + { + "epoch": 0.6659826253505712, + "grad_norm": 1.6737472621476541, + "learning_rate": 2.6514468294870364e-06, + "loss": 0.2579, + "step": 9736 + }, + { + "epoch": 0.6660510294821808, + "grad_norm": 2.222047598421699, + "learning_rate": 2.650468938547248e-06, + "loss": 0.3236, + "step": 9737 + }, + { + "epoch": 0.6661194336137902, + "grad_norm": 1.6881897585119592, + "learning_rate": 2.6494911629337883e-06, + "loss": 0.2382, + "step": 9738 + }, + { + "epoch": 0.6661878377453998, + "grad_norm": 1.99536262753938, + "learning_rate": 2.6485135026946545e-06, + "loss": 0.478, + "step": 9739 + }, + { + "epoch": 0.6662562418770094, + "grad_norm": 2.2129876232520793, + "learning_rate": 2.647535957877833e-06, + "loss": 0.3451, + "step": 9740 + }, + { + "epoch": 0.666324646008619, + "grad_norm": 1.5930392887199933, + "learning_rate": 2.6465585285313066e-06, + "loss": 0.259, + "step": 9741 + }, + { + "epoch": 0.6663930501402284, + "grad_norm": 1.9823738656079546, + "learning_rate": 2.645581214703053e-06, + "loss": 0.4065, + "step": 9742 + }, + { + "epoch": 0.666461454271838, + "grad_norm": 2.0839346582118594, + "learning_rate": 2.64460401644104e-06, + "loss": 0.5279, + "step": 9743 + }, + { + "epoch": 0.6665298584034476, + "grad_norm": 2.544981714981135, + "learning_rate": 2.6436269337932385e-06, + "loss": 0.7113, + "step": 9744 + }, + { + "epoch": 0.6665982625350572, + "grad_norm": 1.995713610200634, + "learning_rate": 2.6426499668076032e-06, + "loss": 0.3736, + "step": 9745 + }, + { + "epoch": 0.6666666666666666, + "grad_norm": 2.384763269321737, + "learning_rate": 2.6416731155320934e-06, + "loss": 0.4223, + "step": 9746 + }, + { + "epoch": 0.6667350707982762, + "grad_norm": 2.1885937136368683, + "learning_rate": 2.6406963800146558e-06, + "loss": 0.5521, + "step": 9747 + }, + { + "epoch": 0.6668034749298858, + "grad_norm": 1.5347187738993568, + "learning_rate": 2.639719760303232e-06, + "loss": 0.1778, + "step": 9748 + }, + { + "epoch": 0.6668718790614954, + "grad_norm": 2.1697087977299607, + "learning_rate": 2.6387432564457604e-06, + "loss": 0.4236, + "step": 9749 + }, + { + "epoch": 0.6669402831931048, + "grad_norm": 1.6176784065149283, + "learning_rate": 2.6377668684901704e-06, + "loss": 0.1743, + "step": 9750 + }, + { + "epoch": 0.6670086873247144, + "grad_norm": 2.279565568137829, + "learning_rate": 2.636790596484391e-06, + "loss": 0.5265, + "step": 9751 + }, + { + "epoch": 0.667077091456324, + "grad_norm": 1.6342486449389642, + "learning_rate": 2.6358144404763414e-06, + "loss": 0.3289, + "step": 9752 + }, + { + "epoch": 0.6671454955879335, + "grad_norm": 2.0989915302450126, + "learning_rate": 2.634838400513934e-06, + "loss": 0.4628, + "step": 9753 + }, + { + "epoch": 0.667213899719543, + "grad_norm": 2.2566610984144253, + "learning_rate": 2.6338624766450805e-06, + "loss": 0.3287, + "step": 9754 + }, + { + "epoch": 0.6672823038511526, + "grad_norm": 3.3582238553773895, + "learning_rate": 2.6328866689176824e-06, + "loss": 0.3942, + "step": 9755 + }, + { + "epoch": 0.6673507079827622, + "grad_norm": 2.846453052744704, + "learning_rate": 2.631910977379638e-06, + "loss": 0.4956, + "step": 9756 + }, + { + "epoch": 0.6674191121143717, + "grad_norm": 2.37476146905581, + "learning_rate": 2.6309354020788376e-06, + "loss": 0.5974, + "step": 9757 + }, + { + "epoch": 0.6674875162459812, + "grad_norm": 1.9829469708836835, + "learning_rate": 2.629959943063166e-06, + "loss": 0.2484, + "step": 9758 + }, + { + "epoch": 0.6675559203775908, + "grad_norm": 2.205939341818725, + "learning_rate": 2.6289846003805073e-06, + "loss": 0.2006, + "step": 9759 + }, + { + "epoch": 0.6676243245092004, + "grad_norm": 2.0995859189164894, + "learning_rate": 2.628009374078734e-06, + "loss": 0.2942, + "step": 9760 + }, + { + "epoch": 0.6676927286408099, + "grad_norm": 2.388779538738999, + "learning_rate": 2.6270342642057144e-06, + "loss": 0.4796, + "step": 9761 + }, + { + "epoch": 0.6677611327724194, + "grad_norm": 2.5395108927429106, + "learning_rate": 2.6260592708093103e-06, + "loss": 0.6436, + "step": 9762 + }, + { + "epoch": 0.667829536904029, + "grad_norm": 2.47453821647479, + "learning_rate": 2.6250843939373827e-06, + "loss": 0.6351, + "step": 9763 + }, + { + "epoch": 0.6678979410356386, + "grad_norm": 2.092114785022622, + "learning_rate": 2.624109633637782e-06, + "loss": 0.4481, + "step": 9764 + }, + { + "epoch": 0.6679663451672481, + "grad_norm": 2.141501289104098, + "learning_rate": 2.6231349899583514e-06, + "loss": 0.4215, + "step": 9765 + }, + { + "epoch": 0.6680347492988576, + "grad_norm": 1.2501403729775458, + "learning_rate": 2.6221604629469346e-06, + "loss": 0.1539, + "step": 9766 + }, + { + "epoch": 0.6681031534304672, + "grad_norm": 1.8092924955726752, + "learning_rate": 2.6211860526513647e-06, + "loss": 0.4125, + "step": 9767 + }, + { + "epoch": 0.6681715575620768, + "grad_norm": 1.7582910310220128, + "learning_rate": 2.620211759119471e-06, + "loss": 0.2249, + "step": 9768 + }, + { + "epoch": 0.6682399616936863, + "grad_norm": 2.3149101803968173, + "learning_rate": 2.619237582399076e-06, + "loss": 0.3173, + "step": 9769 + }, + { + "epoch": 0.6683083658252958, + "grad_norm": 2.185605540097584, + "learning_rate": 2.6182635225379955e-06, + "loss": 0.4739, + "step": 9770 + }, + { + "epoch": 0.6683767699569054, + "grad_norm": 1.9013283543865402, + "learning_rate": 2.6172895795840435e-06, + "loss": 0.3052, + "step": 9771 + }, + { + "epoch": 0.6684451740885149, + "grad_norm": 2.284529145098077, + "learning_rate": 2.6163157535850238e-06, + "loss": 0.2449, + "step": 9772 + }, + { + "epoch": 0.6685135782201245, + "grad_norm": 1.8661438247693987, + "learning_rate": 2.6153420445887394e-06, + "loss": 0.1496, + "step": 9773 + }, + { + "epoch": 0.668581982351734, + "grad_norm": 1.7472422021071519, + "learning_rate": 2.614368452642983e-06, + "loss": 0.4106, + "step": 9774 + }, + { + "epoch": 0.6686503864833436, + "grad_norm": 1.6444027685142795, + "learning_rate": 2.6133949777955424e-06, + "loss": 0.2025, + "step": 9775 + }, + { + "epoch": 0.6687187906149531, + "grad_norm": 1.9319724018838702, + "learning_rate": 2.612421620094202e-06, + "loss": 0.3173, + "step": 9776 + }, + { + "epoch": 0.6687871947465627, + "grad_norm": 2.578207615771315, + "learning_rate": 2.611448379586735e-06, + "loss": 0.5514, + "step": 9777 + }, + { + "epoch": 0.6688555988781723, + "grad_norm": 2.2960040815768616, + "learning_rate": 2.6104752563209186e-06, + "loss": 0.3647, + "step": 9778 + }, + { + "epoch": 0.6689240030097818, + "grad_norm": 1.9687684792453657, + "learning_rate": 2.6095022503445155e-06, + "loss": 0.2851, + "step": 9779 + }, + { + "epoch": 0.6689924071413913, + "grad_norm": 2.335610893151787, + "learning_rate": 2.6085293617052847e-06, + "loss": 0.3385, + "step": 9780 + }, + { + "epoch": 0.6690608112730009, + "grad_norm": 1.2555935579847266, + "learning_rate": 2.6075565904509816e-06, + "loss": 0.1442, + "step": 9781 + }, + { + "epoch": 0.6691292154046105, + "grad_norm": 1.6656253218874164, + "learning_rate": 2.606583936629352e-06, + "loss": 0.1695, + "step": 9782 + }, + { + "epoch": 0.66919761953622, + "grad_norm": 1.3338527251442722, + "learning_rate": 2.6056114002881428e-06, + "loss": 0.1447, + "step": 9783 + }, + { + "epoch": 0.6692660236678295, + "grad_norm": 2.402379424907996, + "learning_rate": 2.6046389814750862e-06, + "loss": 0.4998, + "step": 9784 + }, + { + "epoch": 0.6693344277994391, + "grad_norm": 1.446473990085542, + "learning_rate": 2.603666680237917e-06, + "loss": 0.1978, + "step": 9785 + }, + { + "epoch": 0.6694028319310487, + "grad_norm": 2.25144065987896, + "learning_rate": 2.6026944966243594e-06, + "loss": 0.4829, + "step": 9786 + }, + { + "epoch": 0.6694712360626582, + "grad_norm": 1.5780209858769396, + "learning_rate": 2.6017224306821324e-06, + "loss": 0.1657, + "step": 9787 + }, + { + "epoch": 0.6695396401942677, + "grad_norm": 3.0327853905686153, + "learning_rate": 2.6007504824589503e-06, + "loss": 0.4117, + "step": 9788 + }, + { + "epoch": 0.6696080443258773, + "grad_norm": 2.371574309165467, + "learning_rate": 2.5997786520025192e-06, + "loss": 0.4651, + "step": 9789 + }, + { + "epoch": 0.6696764484574869, + "grad_norm": 1.5711804216664667, + "learning_rate": 2.5988069393605407e-06, + "loss": 0.2129, + "step": 9790 + }, + { + "epoch": 0.6697448525890963, + "grad_norm": 2.182440403685018, + "learning_rate": 2.5978353445807154e-06, + "loss": 0.6679, + "step": 9791 + }, + { + "epoch": 0.6698132567207059, + "grad_norm": 1.4193907220466073, + "learning_rate": 2.5968638677107292e-06, + "loss": 0.1072, + "step": 9792 + }, + { + "epoch": 0.6698816608523155, + "grad_norm": 1.9532760828711349, + "learning_rate": 2.5958925087982705e-06, + "loss": 0.4093, + "step": 9793 + }, + { + "epoch": 0.6699500649839251, + "grad_norm": 1.568610649834822, + "learning_rate": 2.5949212678910173e-06, + "loss": 0.1566, + "step": 9794 + }, + { + "epoch": 0.6700184691155345, + "grad_norm": 1.7788797930264952, + "learning_rate": 2.593950145036642e-06, + "loss": 0.3456, + "step": 9795 + }, + { + "epoch": 0.6700868732471441, + "grad_norm": 2.8344601022210685, + "learning_rate": 2.592979140282812e-06, + "loss": 0.7261, + "step": 9796 + }, + { + "epoch": 0.6701552773787537, + "grad_norm": 1.9914659839110307, + "learning_rate": 2.5920082536771873e-06, + "loss": 0.3127, + "step": 9797 + }, + { + "epoch": 0.6702236815103633, + "grad_norm": 2.092357873710142, + "learning_rate": 2.591037485267428e-06, + "loss": 0.5027, + "step": 9798 + }, + { + "epoch": 0.6702920856419727, + "grad_norm": 3.7698219305831495, + "learning_rate": 2.5900668351011815e-06, + "loss": 0.2337, + "step": 9799 + }, + { + "epoch": 0.6703604897735823, + "grad_norm": 2.3113097401104223, + "learning_rate": 2.5890963032260923e-06, + "loss": 0.2213, + "step": 9800 + }, + { + "epoch": 0.6704288939051919, + "grad_norm": 2.226586950736731, + "learning_rate": 2.5881258896897987e-06, + "loss": 0.405, + "step": 9801 + }, + { + "epoch": 0.6704972980368015, + "grad_norm": 2.0191256265199438, + "learning_rate": 2.5871555945399306e-06, + "loss": 0.4473, + "step": 9802 + }, + { + "epoch": 0.6705657021684109, + "grad_norm": 1.9769563018767273, + "learning_rate": 2.58618541782412e-06, + "loss": 0.4175, + "step": 9803 + }, + { + "epoch": 0.6706341063000205, + "grad_norm": 2.252419660799423, + "learning_rate": 2.585215359589983e-06, + "loss": 0.3927, + "step": 9804 + }, + { + "epoch": 0.6707025104316301, + "grad_norm": 2.01502175877847, + "learning_rate": 2.5842454198851396e-06, + "loss": 0.3743, + "step": 9805 + }, + { + "epoch": 0.6707709145632397, + "grad_norm": 3.493394220761272, + "learning_rate": 2.5832755987571963e-06, + "loss": 0.3307, + "step": 9806 + }, + { + "epoch": 0.6708393186948491, + "grad_norm": 2.248066349425755, + "learning_rate": 2.5823058962537567e-06, + "loss": 0.5151, + "step": 9807 + }, + { + "epoch": 0.6709077228264587, + "grad_norm": 1.8247121760311384, + "learning_rate": 2.581336312422419e-06, + "loss": 0.2526, + "step": 9808 + }, + { + "epoch": 0.6709761269580683, + "grad_norm": 2.3622282977012103, + "learning_rate": 2.580366847310774e-06, + "loss": 0.4804, + "step": 9809 + }, + { + "epoch": 0.6710445310896778, + "grad_norm": 2.1281996998161237, + "learning_rate": 2.5793975009664066e-06, + "loss": 0.5416, + "step": 9810 + }, + { + "epoch": 0.6711129352212873, + "grad_norm": 2.137497071492827, + "learning_rate": 2.5784282734369014e-06, + "loss": 0.5961, + "step": 9811 + }, + { + "epoch": 0.6711813393528969, + "grad_norm": 2.427150824976854, + "learning_rate": 2.5774591647698276e-06, + "loss": 0.397, + "step": 9812 + }, + { + "epoch": 0.6712497434845065, + "grad_norm": 2.0944594633100495, + "learning_rate": 2.5764901750127587e-06, + "loss": 0.327, + "step": 9813 + }, + { + "epoch": 0.671318147616116, + "grad_norm": 2.237775276075852, + "learning_rate": 2.5755213042132545e-06, + "loss": 0.4699, + "step": 9814 + }, + { + "epoch": 0.6713865517477255, + "grad_norm": 2.111878287230115, + "learning_rate": 2.574552552418873e-06, + "loss": 0.1666, + "step": 9815 + }, + { + "epoch": 0.6714549558793351, + "grad_norm": 2.1217114021707233, + "learning_rate": 2.573583919677164e-06, + "loss": 0.1983, + "step": 9816 + }, + { + "epoch": 0.6715233600109447, + "grad_norm": 1.7109042560709515, + "learning_rate": 2.572615406035672e-06, + "loss": 0.2209, + "step": 9817 + }, + { + "epoch": 0.6715917641425542, + "grad_norm": 1.7217547031965363, + "learning_rate": 2.571647011541939e-06, + "loss": 0.2856, + "step": 9818 + }, + { + "epoch": 0.6716601682741637, + "grad_norm": 2.1510917082008154, + "learning_rate": 2.570678736243497e-06, + "loss": 0.4911, + "step": 9819 + }, + { + "epoch": 0.6717285724057733, + "grad_norm": 2.2863139944782067, + "learning_rate": 2.569710580187874e-06, + "loss": 0.1797, + "step": 9820 + }, + { + "epoch": 0.6717969765373829, + "grad_norm": 2.0026238225130175, + "learning_rate": 2.5687425434225914e-06, + "loss": 0.4696, + "step": 9821 + }, + { + "epoch": 0.6718653806689924, + "grad_norm": 2.41078037298786, + "learning_rate": 2.567774625995163e-06, + "loss": 0.5309, + "step": 9822 + }, + { + "epoch": 0.671933784800602, + "grad_norm": 2.0497429826162463, + "learning_rate": 2.5668068279531035e-06, + "loss": 0.421, + "step": 9823 + }, + { + "epoch": 0.6720021889322115, + "grad_norm": 1.8848451449365269, + "learning_rate": 2.565839149343913e-06, + "loss": 0.4183, + "step": 9824 + }, + { + "epoch": 0.6720705930638211, + "grad_norm": 2.568851403511184, + "learning_rate": 2.5648715902150924e-06, + "loss": 0.5721, + "step": 9825 + }, + { + "epoch": 0.6721389971954306, + "grad_norm": 1.7510568629083314, + "learning_rate": 2.5639041506141338e-06, + "loss": 0.2593, + "step": 9826 + }, + { + "epoch": 0.6722074013270402, + "grad_norm": 3.424459352423876, + "learning_rate": 2.562936830588523e-06, + "loss": 0.3702, + "step": 9827 + }, + { + "epoch": 0.6722758054586497, + "grad_norm": 2.1865058274643965, + "learning_rate": 2.561969630185741e-06, + "loss": 0.5301, + "step": 9828 + }, + { + "epoch": 0.6723442095902592, + "grad_norm": 2.0324030831117263, + "learning_rate": 2.561002549453263e-06, + "loss": 0.4482, + "step": 9829 + }, + { + "epoch": 0.6724126137218688, + "grad_norm": 1.8600922010755065, + "learning_rate": 2.5600355884385564e-06, + "loss": 0.2805, + "step": 9830 + }, + { + "epoch": 0.6724810178534784, + "grad_norm": 2.3557279200652155, + "learning_rate": 2.559068747189086e-06, + "loss": 0.4358, + "step": 9831 + }, + { + "epoch": 0.6725494219850879, + "grad_norm": 2.406608307158155, + "learning_rate": 2.558102025752308e-06, + "loss": 0.2985, + "step": 9832 + }, + { + "epoch": 0.6726178261166974, + "grad_norm": 1.7751373555202217, + "learning_rate": 2.5571354241756757e-06, + "loss": 0.225, + "step": 9833 + }, + { + "epoch": 0.672686230248307, + "grad_norm": 2.1415186221043188, + "learning_rate": 2.5561689425066332e-06, + "loss": 0.242, + "step": 9834 + }, + { + "epoch": 0.6727546343799166, + "grad_norm": 1.9972038280372924, + "learning_rate": 2.5552025807926206e-06, + "loss": 0.2938, + "step": 9835 + }, + { + "epoch": 0.6728230385115261, + "grad_norm": 2.3711581025039883, + "learning_rate": 2.554236339081071e-06, + "loss": 0.7642, + "step": 9836 + }, + { + "epoch": 0.6728914426431356, + "grad_norm": 1.669153364594004, + "learning_rate": 2.55327021741941e-06, + "loss": 0.3027, + "step": 9837 + }, + { + "epoch": 0.6729598467747452, + "grad_norm": 2.3877010165967465, + "learning_rate": 2.552304215855064e-06, + "loss": 0.4793, + "step": 9838 + }, + { + "epoch": 0.6730282509063548, + "grad_norm": 1.907986303562212, + "learning_rate": 2.551338334435447e-06, + "loss": 0.3991, + "step": 9839 + }, + { + "epoch": 0.6730966550379643, + "grad_norm": 1.9188621444733727, + "learning_rate": 2.5503725732079687e-06, + "loss": 0.5082, + "step": 9840 + }, + { + "epoch": 0.6731650591695738, + "grad_norm": 2.888111944914128, + "learning_rate": 2.5494069322200344e-06, + "loss": 0.2959, + "step": 9841 + }, + { + "epoch": 0.6732334633011834, + "grad_norm": 2.2741824368323456, + "learning_rate": 2.5484414115190385e-06, + "loss": 0.6113, + "step": 9842 + }, + { + "epoch": 0.673301867432793, + "grad_norm": 1.613867378442599, + "learning_rate": 2.5474760111523796e-06, + "loss": 0.1498, + "step": 9843 + }, + { + "epoch": 0.6733702715644025, + "grad_norm": 2.0390032215198195, + "learning_rate": 2.5465107311674387e-06, + "loss": 0.5147, + "step": 9844 + }, + { + "epoch": 0.673438675696012, + "grad_norm": 1.9733422701126897, + "learning_rate": 2.545545571611601e-06, + "loss": 0.1743, + "step": 9845 + }, + { + "epoch": 0.6735070798276216, + "grad_norm": 1.7009014304459567, + "learning_rate": 2.5445805325322394e-06, + "loss": 0.3154, + "step": 9846 + }, + { + "epoch": 0.6735754839592312, + "grad_norm": 2.125583586077663, + "learning_rate": 2.543615613976722e-06, + "loss": 0.3507, + "step": 9847 + }, + { + "epoch": 0.6736438880908406, + "grad_norm": 2.1565591479945505, + "learning_rate": 2.542650815992412e-06, + "loss": 0.3122, + "step": 9848 + }, + { + "epoch": 0.6737122922224502, + "grad_norm": 2.2703322016575913, + "learning_rate": 2.541686138626667e-06, + "loss": 0.7102, + "step": 9849 + }, + { + "epoch": 0.6737806963540598, + "grad_norm": 5.0270894185596005, + "learning_rate": 2.540721581926834e-06, + "loss": 0.1856, + "step": 9850 + }, + { + "epoch": 0.6738491004856694, + "grad_norm": 1.678585143199943, + "learning_rate": 2.5397571459402647e-06, + "loss": 0.3247, + "step": 9851 + }, + { + "epoch": 0.6739175046172788, + "grad_norm": 2.0618828370314515, + "learning_rate": 2.538792830714293e-06, + "loss": 0.2921, + "step": 9852 + }, + { + "epoch": 0.6739859087488884, + "grad_norm": 2.642378129178849, + "learning_rate": 2.5378286362962555e-06, + "loss": 0.5174, + "step": 9853 + }, + { + "epoch": 0.674054312880498, + "grad_norm": 2.3974806091371703, + "learning_rate": 2.5368645627334786e-06, + "loss": 0.3392, + "step": 9854 + }, + { + "epoch": 0.6741227170121076, + "grad_norm": 2.2164025178594837, + "learning_rate": 2.535900610073283e-06, + "loss": 0.2932, + "step": 9855 + }, + { + "epoch": 0.674191121143717, + "grad_norm": 2.466075396043793, + "learning_rate": 2.5349367783629852e-06, + "loss": 0.7692, + "step": 9856 + }, + { + "epoch": 0.6742595252753266, + "grad_norm": 2.7677396566972825, + "learning_rate": 2.533973067649892e-06, + "loss": 0.5641, + "step": 9857 + }, + { + "epoch": 0.6743279294069362, + "grad_norm": 1.7003025347595189, + "learning_rate": 2.5330094779813102e-06, + "loss": 0.1699, + "step": 9858 + }, + { + "epoch": 0.6743963335385458, + "grad_norm": 2.0586273364343874, + "learning_rate": 2.532046009404537e-06, + "loss": 0.319, + "step": 9859 + }, + { + "epoch": 0.6744647376701552, + "grad_norm": 1.866709879516206, + "learning_rate": 2.5310826619668626e-06, + "loss": 0.2506, + "step": 9860 + }, + { + "epoch": 0.6745331418017648, + "grad_norm": 1.5483209878909594, + "learning_rate": 2.5301194357155718e-06, + "loss": 0.2266, + "step": 9861 + }, + { + "epoch": 0.6746015459333744, + "grad_norm": 1.4246943367678766, + "learning_rate": 2.5291563306979484e-06, + "loss": 0.2862, + "step": 9862 + }, + { + "epoch": 0.6746699500649839, + "grad_norm": 1.8392091737983518, + "learning_rate": 2.528193346961263e-06, + "loss": 0.2669, + "step": 9863 + }, + { + "epoch": 0.6747383541965934, + "grad_norm": 3.253502682208862, + "learning_rate": 2.5272304845527828e-06, + "loss": 0.2902, + "step": 9864 + }, + { + "epoch": 0.674806758328203, + "grad_norm": 2.320593511859398, + "learning_rate": 2.5262677435197736e-06, + "loss": 0.3889, + "step": 9865 + }, + { + "epoch": 0.6748751624598126, + "grad_norm": 2.000279556048423, + "learning_rate": 2.5253051239094883e-06, + "loss": 0.4098, + "step": 9866 + }, + { + "epoch": 0.6749435665914221, + "grad_norm": 2.238397841554244, + "learning_rate": 2.5243426257691784e-06, + "loss": 0.5197, + "step": 9867 + }, + { + "epoch": 0.6750119707230317, + "grad_norm": 1.6578402082768495, + "learning_rate": 2.523380249146087e-06, + "loss": 0.2199, + "step": 9868 + }, + { + "epoch": 0.6750803748546412, + "grad_norm": 2.0137738739435984, + "learning_rate": 2.5224179940874507e-06, + "loss": 0.2963, + "step": 9869 + }, + { + "epoch": 0.6751487789862508, + "grad_norm": 1.5265622834954162, + "learning_rate": 2.521455860640505e-06, + "loss": 0.158, + "step": 9870 + }, + { + "epoch": 0.6752171831178603, + "grad_norm": 1.973419449724419, + "learning_rate": 2.520493848852472e-06, + "loss": 0.3975, + "step": 9871 + }, + { + "epoch": 0.6752855872494699, + "grad_norm": 2.4167648670918735, + "learning_rate": 2.519531958770577e-06, + "loss": 0.6503, + "step": 9872 + }, + { + "epoch": 0.6753539913810794, + "grad_norm": 1.4167337004378735, + "learning_rate": 2.5185701904420313e-06, + "loss": 0.25, + "step": 9873 + }, + { + "epoch": 0.675422395512689, + "grad_norm": 1.8678787549994016, + "learning_rate": 2.5176085439140443e-06, + "loss": 0.2119, + "step": 9874 + }, + { + "epoch": 0.6754907996442985, + "grad_norm": 2.3701805354620746, + "learning_rate": 2.5166470192338156e-06, + "loss": 0.603, + "step": 9875 + }, + { + "epoch": 0.6755592037759081, + "grad_norm": 2.0302695769272834, + "learning_rate": 2.5156856164485423e-06, + "loss": 0.4096, + "step": 9876 + }, + { + "epoch": 0.6756276079075176, + "grad_norm": 2.042420848073841, + "learning_rate": 2.5147243356054173e-06, + "loss": 0.3005, + "step": 9877 + }, + { + "epoch": 0.6756960120391272, + "grad_norm": 2.4263497575595827, + "learning_rate": 2.5137631767516234e-06, + "loss": 0.468, + "step": 9878 + }, + { + "epoch": 0.6757644161707367, + "grad_norm": 2.039975688967112, + "learning_rate": 2.5128021399343385e-06, + "loss": 0.5428, + "step": 9879 + }, + { + "epoch": 0.6758328203023463, + "grad_norm": 1.564671109522993, + "learning_rate": 2.511841225200735e-06, + "loss": 0.1709, + "step": 9880 + }, + { + "epoch": 0.6759012244339558, + "grad_norm": 2.4519714267273143, + "learning_rate": 2.5108804325979774e-06, + "loss": 0.4284, + "step": 9881 + }, + { + "epoch": 0.6759696285655653, + "grad_norm": 2.164978329327601, + "learning_rate": 2.5099197621732303e-06, + "loss": 0.5824, + "step": 9882 + }, + { + "epoch": 0.6760380326971749, + "grad_norm": 2.490421494578247, + "learning_rate": 2.5089592139736434e-06, + "loss": 0.44, + "step": 9883 + }, + { + "epoch": 0.6761064368287845, + "grad_norm": 2.6399203325211613, + "learning_rate": 2.5079987880463698e-06, + "loss": 0.2282, + "step": 9884 + }, + { + "epoch": 0.676174840960394, + "grad_norm": 1.8762257185641222, + "learning_rate": 2.5070384844385485e-06, + "loss": 0.2989, + "step": 9885 + }, + { + "epoch": 0.6762432450920035, + "grad_norm": 2.2950561743865405, + "learning_rate": 2.506078303197318e-06, + "loss": 0.3647, + "step": 9886 + }, + { + "epoch": 0.6763116492236131, + "grad_norm": 2.2165987169874524, + "learning_rate": 2.505118244369806e-06, + "loss": 0.3386, + "step": 9887 + }, + { + "epoch": 0.6763800533552227, + "grad_norm": 2.0001239514683995, + "learning_rate": 2.5041583080031396e-06, + "loss": 0.3195, + "step": 9888 + }, + { + "epoch": 0.6764484574868322, + "grad_norm": 1.8988582278315695, + "learning_rate": 2.5031984941444334e-06, + "loss": 0.4136, + "step": 9889 + }, + { + "epoch": 0.6765168616184417, + "grad_norm": 5.484068294111031, + "learning_rate": 2.502238802840804e-06, + "loss": 0.2731, + "step": 9890 + }, + { + "epoch": 0.6765852657500513, + "grad_norm": 2.594173959759996, + "learning_rate": 2.5012792341393534e-06, + "loss": 0.2699, + "step": 9891 + }, + { + "epoch": 0.6766536698816609, + "grad_norm": 2.1036235919754778, + "learning_rate": 2.5003197880871866e-06, + "loss": 0.4293, + "step": 9892 + }, + { + "epoch": 0.6767220740132704, + "grad_norm": 2.2459246264930433, + "learning_rate": 2.499360464731395e-06, + "loss": 0.5359, + "step": 9893 + }, + { + "epoch": 0.6767904781448799, + "grad_norm": 2.2137702479801584, + "learning_rate": 2.4984012641190674e-06, + "loss": 0.5296, + "step": 9894 + }, + { + "epoch": 0.6768588822764895, + "grad_norm": 2.266991821246965, + "learning_rate": 2.497442186297286e-06, + "loss": 0.512, + "step": 9895 + }, + { + "epoch": 0.6769272864080991, + "grad_norm": 1.8166220547035306, + "learning_rate": 2.4964832313131248e-06, + "loss": 0.3396, + "step": 9896 + }, + { + "epoch": 0.6769956905397086, + "grad_norm": 2.0391807741085866, + "learning_rate": 2.495524399213658e-06, + "loss": 0.3174, + "step": 9897 + }, + { + "epoch": 0.6770640946713181, + "grad_norm": 2.468287497064169, + "learning_rate": 2.4945656900459474e-06, + "loss": 0.5821, + "step": 9898 + }, + { + "epoch": 0.6771324988029277, + "grad_norm": 2.3488623239275555, + "learning_rate": 2.4936071038570514e-06, + "loss": 0.464, + "step": 9899 + }, + { + "epoch": 0.6772009029345373, + "grad_norm": 2.06816407777464, + "learning_rate": 2.492648640694022e-06, + "loss": 0.4246, + "step": 9900 + }, + { + "epoch": 0.6772693070661467, + "grad_norm": 1.372577701029847, + "learning_rate": 2.4916903006039033e-06, + "loss": 0.1874, + "step": 9901 + }, + { + "epoch": 0.6773377111977563, + "grad_norm": 2.419738220684342, + "learning_rate": 2.4907320836337396e-06, + "loss": 0.3646, + "step": 9902 + }, + { + "epoch": 0.6774061153293659, + "grad_norm": 1.9906380031646067, + "learning_rate": 2.48977398983056e-06, + "loss": 0.4358, + "step": 9903 + }, + { + "epoch": 0.6774745194609755, + "grad_norm": 1.5989223397477164, + "learning_rate": 2.4888160192413963e-06, + "loss": 0.2659, + "step": 9904 + }, + { + "epoch": 0.677542923592585, + "grad_norm": 2.1784335669018375, + "learning_rate": 2.4878581719132696e-06, + "loss": 0.3533, + "step": 9905 + }, + { + "epoch": 0.6776113277241945, + "grad_norm": 2.4548856618820527, + "learning_rate": 2.486900447893194e-06, + "loss": 0.6535, + "step": 9906 + }, + { + "epoch": 0.6776797318558041, + "grad_norm": 1.9275490034927214, + "learning_rate": 2.4859428472281804e-06, + "loss": 0.1347, + "step": 9907 + }, + { + "epoch": 0.6777481359874137, + "grad_norm": 2.0295474546659142, + "learning_rate": 2.4849853699652323e-06, + "loss": 0.5214, + "step": 9908 + }, + { + "epoch": 0.6778165401190231, + "grad_norm": 1.9099057741290495, + "learning_rate": 2.4840280161513447e-06, + "loss": 0.2785, + "step": 9909 + }, + { + "epoch": 0.6778849442506327, + "grad_norm": 1.9192130525674147, + "learning_rate": 2.483070785833514e-06, + "loss": 0.276, + "step": 9910 + }, + { + "epoch": 0.6779533483822423, + "grad_norm": 2.101739553346169, + "learning_rate": 2.4821136790587207e-06, + "loss": 0.4134, + "step": 9911 + }, + { + "epoch": 0.6780217525138519, + "grad_norm": 1.8521760093135586, + "learning_rate": 2.4811566958739497e-06, + "loss": 0.2343, + "step": 9912 + }, + { + "epoch": 0.6780901566454614, + "grad_norm": 2.391385711611635, + "learning_rate": 2.480199836326171e-06, + "loss": 0.6788, + "step": 9913 + }, + { + "epoch": 0.6781585607770709, + "grad_norm": 1.9853157194085351, + "learning_rate": 2.4792431004623516e-06, + "loss": 0.2865, + "step": 9914 + }, + { + "epoch": 0.6782269649086805, + "grad_norm": 2.668409057074258, + "learning_rate": 2.4782864883294545e-06, + "loss": 0.4641, + "step": 9915 + }, + { + "epoch": 0.6782953690402901, + "grad_norm": 2.9299605077060984, + "learning_rate": 2.477329999974431e-06, + "loss": 0.2596, + "step": 9916 + }, + { + "epoch": 0.6783637731718996, + "grad_norm": 1.705998651799422, + "learning_rate": 2.476373635444235e-06, + "loss": 0.3028, + "step": 9917 + }, + { + "epoch": 0.6784321773035091, + "grad_norm": 2.536161825772778, + "learning_rate": 2.4754173947858084e-06, + "loss": 0.6484, + "step": 9918 + }, + { + "epoch": 0.6785005814351187, + "grad_norm": 2.336612466750053, + "learning_rate": 2.4744612780460863e-06, + "loss": 0.3293, + "step": 9919 + }, + { + "epoch": 0.6785689855667282, + "grad_norm": 1.9593613281868143, + "learning_rate": 2.473505285272e-06, + "loss": 0.3687, + "step": 9920 + }, + { + "epoch": 0.6786373896983378, + "grad_norm": 2.2560405740997416, + "learning_rate": 2.4725494165104735e-06, + "loss": 0.4578, + "step": 9921 + }, + { + "epoch": 0.6787057938299473, + "grad_norm": 1.4481318323091918, + "learning_rate": 2.4715936718084284e-06, + "loss": 0.1609, + "step": 9922 + }, + { + "epoch": 0.6787741979615569, + "grad_norm": 2.514161946171405, + "learning_rate": 2.4706380512127736e-06, + "loss": 0.4929, + "step": 9923 + }, + { + "epoch": 0.6788426020931664, + "grad_norm": 3.258003814714937, + "learning_rate": 2.469682554770419e-06, + "loss": 0.3953, + "step": 9924 + }, + { + "epoch": 0.678911006224776, + "grad_norm": 2.01318270923596, + "learning_rate": 2.4687271825282634e-06, + "loss": 0.269, + "step": 9925 + }, + { + "epoch": 0.6789794103563855, + "grad_norm": 1.8724529370835026, + "learning_rate": 2.4677719345332013e-06, + "loss": 0.3109, + "step": 9926 + }, + { + "epoch": 0.6790478144879951, + "grad_norm": 1.7166920696118149, + "learning_rate": 2.4668168108321205e-06, + "loss": 0.0924, + "step": 9927 + }, + { + "epoch": 0.6791162186196046, + "grad_norm": 2.2893453494877005, + "learning_rate": 2.465861811471903e-06, + "loss": 0.4857, + "step": 9928 + }, + { + "epoch": 0.6791846227512142, + "grad_norm": 2.496382471260129, + "learning_rate": 2.4649069364994234e-06, + "loss": 0.6462, + "step": 9929 + }, + { + "epoch": 0.6792530268828237, + "grad_norm": 2.0863168850045835, + "learning_rate": 2.463952185961555e-06, + "loss": 0.4444, + "step": 9930 + }, + { + "epoch": 0.6793214310144333, + "grad_norm": 2.549751225659431, + "learning_rate": 2.462997559905158e-06, + "loss": 0.6906, + "step": 9931 + }, + { + "epoch": 0.6793898351460428, + "grad_norm": 2.1618782202461477, + "learning_rate": 2.4620430583770936e-06, + "loss": 0.4076, + "step": 9932 + }, + { + "epoch": 0.6794582392776524, + "grad_norm": 2.523962137584414, + "learning_rate": 2.4610886814242114e-06, + "loss": 0.3846, + "step": 9933 + }, + { + "epoch": 0.6795266434092619, + "grad_norm": 1.8274357914831996, + "learning_rate": 2.4601344290933572e-06, + "loss": 0.153, + "step": 9934 + }, + { + "epoch": 0.6795950475408715, + "grad_norm": 2.4275142112575243, + "learning_rate": 2.4591803014313703e-06, + "loss": 0.7149, + "step": 9935 + }, + { + "epoch": 0.679663451672481, + "grad_norm": 2.420079802623896, + "learning_rate": 2.4582262984850813e-06, + "loss": 0.5824, + "step": 9936 + }, + { + "epoch": 0.6797318558040906, + "grad_norm": 2.4826188868418617, + "learning_rate": 2.457272420301322e-06, + "loss": 0.6677, + "step": 9937 + }, + { + "epoch": 0.6798002599357001, + "grad_norm": 2.6851041005730685, + "learning_rate": 2.4563186669269114e-06, + "loss": 0.3216, + "step": 9938 + }, + { + "epoch": 0.6798686640673096, + "grad_norm": 3.392348009913037, + "learning_rate": 2.455365038408663e-06, + "loss": 0.2553, + "step": 9939 + }, + { + "epoch": 0.6799370681989192, + "grad_norm": 1.0264869895870985, + "learning_rate": 2.454411534793387e-06, + "loss": 0.1199, + "step": 9940 + }, + { + "epoch": 0.6800054723305288, + "grad_norm": 1.8976913210194615, + "learning_rate": 2.453458156127883e-06, + "loss": 0.1939, + "step": 9941 + }, + { + "epoch": 0.6800738764621383, + "grad_norm": 1.8652125971456475, + "learning_rate": 2.452504902458952e-06, + "loss": 0.3969, + "step": 9942 + }, + { + "epoch": 0.6801422805937478, + "grad_norm": 2.1639434974533756, + "learning_rate": 2.4515517738333803e-06, + "loss": 0.3721, + "step": 9943 + }, + { + "epoch": 0.6802106847253574, + "grad_norm": 2.106828713928497, + "learning_rate": 2.450598770297956e-06, + "loss": 0.4376, + "step": 9944 + }, + { + "epoch": 0.680279088856967, + "grad_norm": 1.723268002382971, + "learning_rate": 2.449645891899455e-06, + "loss": 0.3125, + "step": 9945 + }, + { + "epoch": 0.6803474929885766, + "grad_norm": 1.5904558863795606, + "learning_rate": 2.4486931386846485e-06, + "loss": 0.2786, + "step": 9946 + }, + { + "epoch": 0.680415897120186, + "grad_norm": 1.889345822019079, + "learning_rate": 2.447740510700303e-06, + "loss": 0.3119, + "step": 9947 + }, + { + "epoch": 0.6804843012517956, + "grad_norm": 2.6588715183141467, + "learning_rate": 2.446788007993179e-06, + "loss": 0.2737, + "step": 9948 + }, + { + "epoch": 0.6805527053834052, + "grad_norm": 2.6117356866075716, + "learning_rate": 2.445835630610026e-06, + "loss": 0.3255, + "step": 9949 + }, + { + "epoch": 0.6806211095150148, + "grad_norm": 1.9582998538368193, + "learning_rate": 2.4448833785975963e-06, + "loss": 0.3383, + "step": 9950 + }, + { + "epoch": 0.6806895136466242, + "grad_norm": 3.8401394145214867, + "learning_rate": 2.443931252002627e-06, + "loss": 0.4947, + "step": 9951 + }, + { + "epoch": 0.6807579177782338, + "grad_norm": 2.0077802046913353, + "learning_rate": 2.442979250871857e-06, + "loss": 0.4145, + "step": 9952 + }, + { + "epoch": 0.6808263219098434, + "grad_norm": 2.1041940287582985, + "learning_rate": 2.442027375252013e-06, + "loss": 0.4111, + "step": 9953 + }, + { + "epoch": 0.680894726041453, + "grad_norm": 1.9973681799024168, + "learning_rate": 2.4410756251898178e-06, + "loss": 0.4219, + "step": 9954 + }, + { + "epoch": 0.6809631301730624, + "grad_norm": 2.4686593268438424, + "learning_rate": 2.4401240007319876e-06, + "loss": 0.326, + "step": 9955 + }, + { + "epoch": 0.681031534304672, + "grad_norm": 2.3030513580846974, + "learning_rate": 2.4391725019252314e-06, + "loss": 0.2608, + "step": 9956 + }, + { + "epoch": 0.6810999384362816, + "grad_norm": 1.6939721764117073, + "learning_rate": 2.4382211288162566e-06, + "loss": 0.2711, + "step": 9957 + }, + { + "epoch": 0.681168342567891, + "grad_norm": 2.64369375471161, + "learning_rate": 2.4372698814517598e-06, + "loss": 0.3913, + "step": 9958 + }, + { + "epoch": 0.6812367466995006, + "grad_norm": 1.9312585234839728, + "learning_rate": 2.4363187598784323e-06, + "loss": 0.3487, + "step": 9959 + }, + { + "epoch": 0.6813051508311102, + "grad_norm": 1.8217722233986857, + "learning_rate": 2.435367764142958e-06, + "loss": 0.2353, + "step": 9960 + }, + { + "epoch": 0.6813735549627198, + "grad_norm": 1.9142540777438741, + "learning_rate": 2.4344168942920206e-06, + "loss": 0.4268, + "step": 9961 + }, + { + "epoch": 0.6814419590943293, + "grad_norm": 1.9859224503469892, + "learning_rate": 2.4334661503722908e-06, + "loss": 0.4709, + "step": 9962 + }, + { + "epoch": 0.6815103632259388, + "grad_norm": 4.980024753302217, + "learning_rate": 2.4325155324304338e-06, + "loss": 0.3977, + "step": 9963 + }, + { + "epoch": 0.6815787673575484, + "grad_norm": 2.250813330567687, + "learning_rate": 2.431565040513115e-06, + "loss": 0.5165, + "step": 9964 + }, + { + "epoch": 0.681647171489158, + "grad_norm": 2.0809366075445377, + "learning_rate": 2.430614674666987e-06, + "loss": 0.2098, + "step": 9965 + }, + { + "epoch": 0.6817155756207675, + "grad_norm": 1.9943096376955807, + "learning_rate": 2.4296644349386977e-06, + "loss": 0.465, + "step": 9966 + }, + { + "epoch": 0.681783979752377, + "grad_norm": 1.8490035606727027, + "learning_rate": 2.4287143213748903e-06, + "loss": 0.2632, + "step": 9967 + }, + { + "epoch": 0.6818523838839866, + "grad_norm": 3.095944743549856, + "learning_rate": 2.4277643340221985e-06, + "loss": 0.1861, + "step": 9968 + }, + { + "epoch": 0.6819207880155962, + "grad_norm": 1.8583998406986746, + "learning_rate": 2.4268144729272565e-06, + "loss": 0.237, + "step": 9969 + }, + { + "epoch": 0.6819891921472057, + "grad_norm": 2.8622085466640863, + "learning_rate": 2.425864738136684e-06, + "loss": 0.3182, + "step": 9970 + }, + { + "epoch": 0.6820575962788152, + "grad_norm": 2.852347299933543, + "learning_rate": 2.4249151296971026e-06, + "loss": 0.5067, + "step": 9971 + }, + { + "epoch": 0.6821260004104248, + "grad_norm": 1.738837755252592, + "learning_rate": 2.4239656476551215e-06, + "loss": 0.2431, + "step": 9972 + }, + { + "epoch": 0.6821944045420344, + "grad_norm": 2.6990859886455603, + "learning_rate": 2.423016292057346e-06, + "loss": 0.4296, + "step": 9973 + }, + { + "epoch": 0.6822628086736439, + "grad_norm": 2.2801522461667743, + "learning_rate": 2.4220670629503746e-06, + "loss": 0.458, + "step": 9974 + }, + { + "epoch": 0.6823312128052534, + "grad_norm": 2.4686246671128322, + "learning_rate": 2.421117960380799e-06, + "loss": 0.2823, + "step": 9975 + }, + { + "epoch": 0.682399616936863, + "grad_norm": 2.0439131616213526, + "learning_rate": 2.4201689843952085e-06, + "loss": 0.3438, + "step": 9976 + }, + { + "epoch": 0.6824680210684725, + "grad_norm": 2.636456415760233, + "learning_rate": 2.4192201350401828e-06, + "loss": 0.2447, + "step": 9977 + }, + { + "epoch": 0.6825364252000821, + "grad_norm": 2.068857619275324, + "learning_rate": 2.4182714123622953e-06, + "loss": 0.2504, + "step": 9978 + }, + { + "epoch": 0.6826048293316916, + "grad_norm": 1.7331370153880727, + "learning_rate": 2.4173228164081135e-06, + "loss": 0.1691, + "step": 9979 + }, + { + "epoch": 0.6826732334633012, + "grad_norm": 1.8252115182580655, + "learning_rate": 2.416374347224198e-06, + "loss": 0.3299, + "step": 9980 + }, + { + "epoch": 0.6827416375949107, + "grad_norm": 1.8347941551727276, + "learning_rate": 2.4154260048571072e-06, + "loss": 0.2144, + "step": 9981 + }, + { + "epoch": 0.6828100417265203, + "grad_norm": 1.8253832782031802, + "learning_rate": 2.4144777893533874e-06, + "loss": 0.2774, + "step": 9982 + }, + { + "epoch": 0.6828784458581298, + "grad_norm": 2.183227487872824, + "learning_rate": 2.413529700759585e-06, + "loss": 0.5787, + "step": 9983 + }, + { + "epoch": 0.6829468499897394, + "grad_norm": 2.195544926295997, + "learning_rate": 2.4125817391222344e-06, + "loss": 0.2505, + "step": 9984 + }, + { + "epoch": 0.6830152541213489, + "grad_norm": 1.8598548045707575, + "learning_rate": 2.411633904487867e-06, + "loss": 0.3751, + "step": 9985 + }, + { + "epoch": 0.6830836582529585, + "grad_norm": 1.994280960570038, + "learning_rate": 2.410686196903006e-06, + "loss": 0.3012, + "step": 9986 + }, + { + "epoch": 0.683152062384568, + "grad_norm": 1.794603983775628, + "learning_rate": 2.4097386164141706e-06, + "loss": 0.2935, + "step": 9987 + }, + { + "epoch": 0.6832204665161776, + "grad_norm": 2.507026989042537, + "learning_rate": 2.40879116306787e-06, + "loss": 0.4727, + "step": 9988 + }, + { + "epoch": 0.6832888706477871, + "grad_norm": 1.8258080471946783, + "learning_rate": 2.4078438369106145e-06, + "loss": 0.2168, + "step": 9989 + }, + { + "epoch": 0.6833572747793967, + "grad_norm": 1.9882725863464703, + "learning_rate": 2.406896637988898e-06, + "loss": 0.2874, + "step": 9990 + }, + { + "epoch": 0.6834256789110063, + "grad_norm": 2.587309839107176, + "learning_rate": 2.4059495663492187e-06, + "loss": 0.448, + "step": 9991 + }, + { + "epoch": 0.6834940830426157, + "grad_norm": 1.9826559144725602, + "learning_rate": 2.4050026220380617e-06, + "loss": 0.3238, + "step": 9992 + }, + { + "epoch": 0.6835624871742253, + "grad_norm": 2.1573875957737756, + "learning_rate": 2.404055805101907e-06, + "loss": 0.4971, + "step": 9993 + }, + { + "epoch": 0.6836308913058349, + "grad_norm": 2.5616129409795843, + "learning_rate": 2.403109115587229e-06, + "loss": 0.6181, + "step": 9994 + }, + { + "epoch": 0.6836992954374445, + "grad_norm": 1.9975513503811209, + "learning_rate": 2.4021625535404935e-06, + "loss": 0.2575, + "step": 9995 + }, + { + "epoch": 0.6837676995690539, + "grad_norm": 2.1790493540310556, + "learning_rate": 2.401216119008167e-06, + "loss": 0.3094, + "step": 9996 + }, + { + "epoch": 0.6838361037006635, + "grad_norm": 1.8238530885054611, + "learning_rate": 2.400269812036703e-06, + "loss": 0.2532, + "step": 9997 + }, + { + "epoch": 0.6839045078322731, + "grad_norm": 1.9195989218781955, + "learning_rate": 2.39932363267255e-06, + "loss": 0.3005, + "step": 9998 + }, + { + "epoch": 0.6839729119638827, + "grad_norm": 2.3140415002667227, + "learning_rate": 2.3983775809621525e-06, + "loss": 0.5639, + "step": 9999 + }, + { + "epoch": 0.6840413160954921, + "grad_norm": 1.9473342079933378, + "learning_rate": 2.3974316569519437e-06, + "loss": 0.3449, + "step": 10000 + }, + { + "epoch": 0.6841097202271017, + "grad_norm": 2.6304781551889715, + "learning_rate": 2.396485860688359e-06, + "loss": 0.4648, + "step": 10001 + }, + { + "epoch": 0.6841781243587113, + "grad_norm": 2.1052377676726777, + "learning_rate": 2.395540192217819e-06, + "loss": 0.4229, + "step": 10002 + }, + { + "epoch": 0.6842465284903209, + "grad_norm": 2.1041347120395497, + "learning_rate": 2.3945946515867445e-06, + "loss": 0.3883, + "step": 10003 + }, + { + "epoch": 0.6843149326219303, + "grad_norm": 1.7943342157988644, + "learning_rate": 2.3936492388415462e-06, + "loss": 0.2607, + "step": 10004 + }, + { + "epoch": 0.6843833367535399, + "grad_norm": 1.7411367728997678, + "learning_rate": 2.392703954028629e-06, + "loss": 0.2003, + "step": 10005 + }, + { + "epoch": 0.6844517408851495, + "grad_norm": 2.335055985504528, + "learning_rate": 2.3917587971943924e-06, + "loss": 0.4492, + "step": 10006 + }, + { + "epoch": 0.6845201450167591, + "grad_norm": 2.77651982486626, + "learning_rate": 2.390813768385229e-06, + "loss": 0.2982, + "step": 10007 + }, + { + "epoch": 0.6845885491483685, + "grad_norm": 2.311728052221768, + "learning_rate": 2.3898688676475237e-06, + "loss": 0.3896, + "step": 10008 + }, + { + "epoch": 0.6846569532799781, + "grad_norm": 1.6400767409364643, + "learning_rate": 2.3889240950276602e-06, + "loss": 0.2437, + "step": 10009 + }, + { + "epoch": 0.6847253574115877, + "grad_norm": 2.243602213335841, + "learning_rate": 2.3879794505720094e-06, + "loss": 0.4706, + "step": 10010 + }, + { + "epoch": 0.6847937615431972, + "grad_norm": 2.0421342476495705, + "learning_rate": 2.387034934326942e-06, + "loss": 0.4538, + "step": 10011 + }, + { + "epoch": 0.6848621656748067, + "grad_norm": 1.9564169187904572, + "learning_rate": 2.3860905463388184e-06, + "loss": 0.3749, + "step": 10012 + }, + { + "epoch": 0.6849305698064163, + "grad_norm": 2.047972301294681, + "learning_rate": 2.3851462866539927e-06, + "loss": 0.3426, + "step": 10013 + }, + { + "epoch": 0.6849989739380259, + "grad_norm": 2.3879281459002817, + "learning_rate": 2.3842021553188146e-06, + "loss": 0.5632, + "step": 10014 + }, + { + "epoch": 0.6850673780696354, + "grad_norm": 1.5423606301922912, + "learning_rate": 2.383258152379624e-06, + "loss": 0.1836, + "step": 10015 + }, + { + "epoch": 0.6851357822012449, + "grad_norm": 1.6613974609903097, + "learning_rate": 2.3823142778827617e-06, + "loss": 0.3068, + "step": 10016 + }, + { + "epoch": 0.6852041863328545, + "grad_norm": 2.1631931252541015, + "learning_rate": 2.381370531874555e-06, + "loss": 0.5253, + "step": 10017 + }, + { + "epoch": 0.6852725904644641, + "grad_norm": 1.7131390886169313, + "learning_rate": 2.380426914401328e-06, + "loss": 0.3268, + "step": 10018 + }, + { + "epoch": 0.6853409945960736, + "grad_norm": 2.0819635147613624, + "learning_rate": 2.3794834255093977e-06, + "loss": 0.2215, + "step": 10019 + }, + { + "epoch": 0.6854093987276831, + "grad_norm": 1.9504959149842267, + "learning_rate": 2.3785400652450735e-06, + "loss": 0.3883, + "step": 10020 + }, + { + "epoch": 0.6854778028592927, + "grad_norm": 2.2178681058176135, + "learning_rate": 2.3775968336546635e-06, + "loss": 0.2317, + "step": 10021 + }, + { + "epoch": 0.6855462069909023, + "grad_norm": 1.9458287446466889, + "learning_rate": 2.3766537307844628e-06, + "loss": 0.3694, + "step": 10022 + }, + { + "epoch": 0.6856146111225118, + "grad_norm": 2.2972424660838424, + "learning_rate": 2.3757107566807663e-06, + "loss": 0.5311, + "step": 10023 + }, + { + "epoch": 0.6856830152541213, + "grad_norm": 1.6028213205911082, + "learning_rate": 2.374767911389859e-06, + "loss": 0.3211, + "step": 10024 + }, + { + "epoch": 0.6857514193857309, + "grad_norm": 2.148672999566301, + "learning_rate": 2.3738251949580194e-06, + "loss": 0.5296, + "step": 10025 + }, + { + "epoch": 0.6858198235173405, + "grad_norm": 2.177433420777322, + "learning_rate": 2.3728826074315212e-06, + "loss": 0.4602, + "step": 10026 + }, + { + "epoch": 0.68588822764895, + "grad_norm": 1.2683199517761412, + "learning_rate": 2.3719401488566307e-06, + "loss": 0.1839, + "step": 10027 + }, + { + "epoch": 0.6859566317805595, + "grad_norm": 2.3014055969042593, + "learning_rate": 2.370997819279607e-06, + "loss": 0.6669, + "step": 10028 + }, + { + "epoch": 0.6860250359121691, + "grad_norm": 2.039531069297961, + "learning_rate": 2.3700556187467067e-06, + "loss": 0.2257, + "step": 10029 + }, + { + "epoch": 0.6860934400437786, + "grad_norm": 2.059590983027069, + "learning_rate": 2.369113547304175e-06, + "loss": 0.524, + "step": 10030 + }, + { + "epoch": 0.6861618441753882, + "grad_norm": 2.2677183871790345, + "learning_rate": 2.3681716049982567e-06, + "loss": 0.4665, + "step": 10031 + }, + { + "epoch": 0.6862302483069977, + "grad_norm": 1.6512270190612834, + "learning_rate": 2.367229791875185e-06, + "loss": 0.2701, + "step": 10032 + }, + { + "epoch": 0.6862986524386073, + "grad_norm": 2.466320955422612, + "learning_rate": 2.366288107981189e-06, + "loss": 0.4147, + "step": 10033 + }, + { + "epoch": 0.6863670565702168, + "grad_norm": 2.0522572101634116, + "learning_rate": 2.3653465533624897e-06, + "loss": 0.4938, + "step": 10034 + }, + { + "epoch": 0.6864354607018264, + "grad_norm": 1.7636591761946148, + "learning_rate": 2.3644051280653034e-06, + "loss": 0.2494, + "step": 10035 + }, + { + "epoch": 0.686503864833436, + "grad_norm": 1.7677395193698051, + "learning_rate": 2.3634638321358416e-06, + "loss": 0.3407, + "step": 10036 + }, + { + "epoch": 0.6865722689650455, + "grad_norm": 2.1681333152928244, + "learning_rate": 2.362522665620307e-06, + "loss": 0.4547, + "step": 10037 + }, + { + "epoch": 0.686640673096655, + "grad_norm": 2.1760674430509885, + "learning_rate": 2.3615816285648967e-06, + "loss": 0.3278, + "step": 10038 + }, + { + "epoch": 0.6867090772282646, + "grad_norm": 2.6602656584211544, + "learning_rate": 2.3606407210158007e-06, + "loss": 0.472, + "step": 10039 + }, + { + "epoch": 0.6867774813598742, + "grad_norm": 1.963517289216946, + "learning_rate": 2.359699943019201e-06, + "loss": 0.3448, + "step": 10040 + }, + { + "epoch": 0.6868458854914837, + "grad_norm": 2.0703158468543297, + "learning_rate": 2.35875929462128e-06, + "loss": 0.4255, + "step": 10041 + }, + { + "epoch": 0.6869142896230932, + "grad_norm": 1.7338467558174047, + "learning_rate": 2.357818775868205e-06, + "loss": 0.2361, + "step": 10042 + }, + { + "epoch": 0.6869826937547028, + "grad_norm": 1.8193321494177195, + "learning_rate": 2.3568783868061462e-06, + "loss": 0.3976, + "step": 10043 + }, + { + "epoch": 0.6870510978863124, + "grad_norm": 2.199809196716147, + "learning_rate": 2.3559381274812592e-06, + "loss": 0.5055, + "step": 10044 + }, + { + "epoch": 0.6871195020179219, + "grad_norm": 2.64970402549236, + "learning_rate": 2.3549979979396977e-06, + "loss": 0.2789, + "step": 10045 + }, + { + "epoch": 0.6871879061495314, + "grad_norm": 2.553622645060754, + "learning_rate": 2.354057998227606e-06, + "loss": 0.6744, + "step": 10046 + }, + { + "epoch": 0.687256310281141, + "grad_norm": 2.81417832631309, + "learning_rate": 2.353118128391125e-06, + "loss": 0.727, + "step": 10047 + }, + { + "epoch": 0.6873247144127506, + "grad_norm": 2.8639951807279904, + "learning_rate": 2.3521783884763866e-06, + "loss": 0.4456, + "step": 10048 + }, + { + "epoch": 0.68739311854436, + "grad_norm": 2.1918922130277583, + "learning_rate": 2.3512387785295204e-06, + "loss": 0.5176, + "step": 10049 + }, + { + "epoch": 0.6874615226759696, + "grad_norm": 2.4805651938845177, + "learning_rate": 2.3502992985966445e-06, + "loss": 0.3065, + "step": 10050 + }, + { + "epoch": 0.6875299268075792, + "grad_norm": 1.8225869879743586, + "learning_rate": 2.3493599487238755e-06, + "loss": 0.2908, + "step": 10051 + }, + { + "epoch": 0.6875983309391888, + "grad_norm": 4.3325729512997295, + "learning_rate": 2.34842072895732e-06, + "loss": 0.3768, + "step": 10052 + }, + { + "epoch": 0.6876667350707982, + "grad_norm": 1.8978621745286415, + "learning_rate": 2.34748163934308e-06, + "loss": 0.3318, + "step": 10053 + }, + { + "epoch": 0.6877351392024078, + "grad_norm": 1.9547626076896807, + "learning_rate": 2.34654267992725e-06, + "loss": 0.4274, + "step": 10054 + }, + { + "epoch": 0.6878035433340174, + "grad_norm": 2.4394567503340987, + "learning_rate": 2.3456038507559166e-06, + "loss": 0.2091, + "step": 10055 + }, + { + "epoch": 0.687871947465627, + "grad_norm": 2.0556800388612397, + "learning_rate": 2.3446651518751664e-06, + "loss": 0.2603, + "step": 10056 + }, + { + "epoch": 0.6879403515972364, + "grad_norm": 2.718494216900867, + "learning_rate": 2.3437265833310722e-06, + "loss": 0.9099, + "step": 10057 + }, + { + "epoch": 0.688008755728846, + "grad_norm": 2.9693494342160234, + "learning_rate": 2.342788145169705e-06, + "loss": 0.3453, + "step": 10058 + }, + { + "epoch": 0.6880771598604556, + "grad_norm": 1.9203640728350155, + "learning_rate": 2.3418498374371266e-06, + "loss": 0.3085, + "step": 10059 + }, + { + "epoch": 0.6881455639920652, + "grad_norm": 2.2101594748933633, + "learning_rate": 2.340911660179393e-06, + "loss": 0.3487, + "step": 10060 + }, + { + "epoch": 0.6882139681236746, + "grad_norm": 2.691314036833995, + "learning_rate": 2.3399736134425573e-06, + "loss": 0.373, + "step": 10061 + }, + { + "epoch": 0.6882823722552842, + "grad_norm": 2.406958952415074, + "learning_rate": 2.3390356972726597e-06, + "loss": 0.2743, + "step": 10062 + }, + { + "epoch": 0.6883507763868938, + "grad_norm": 1.829741146692195, + "learning_rate": 2.3380979117157423e-06, + "loss": 0.2892, + "step": 10063 + }, + { + "epoch": 0.6884191805185034, + "grad_norm": 2.23681087031051, + "learning_rate": 2.3371602568178324e-06, + "loss": 0.4706, + "step": 10064 + }, + { + "epoch": 0.6884875846501128, + "grad_norm": 2.0430937522734496, + "learning_rate": 2.336222732624957e-06, + "loss": 0.4065, + "step": 10065 + }, + { + "epoch": 0.6885559887817224, + "grad_norm": 1.9822138033508958, + "learning_rate": 2.3352853391831325e-06, + "loss": 0.3494, + "step": 10066 + }, + { + "epoch": 0.688624392913332, + "grad_norm": 2.5717599961257225, + "learning_rate": 2.3343480765383696e-06, + "loss": 0.7348, + "step": 10067 + }, + { + "epoch": 0.6886927970449415, + "grad_norm": 2.3784682523658707, + "learning_rate": 2.3334109447366775e-06, + "loss": 0.4596, + "step": 10068 + }, + { + "epoch": 0.688761201176551, + "grad_norm": 2.2015690618651416, + "learning_rate": 2.3324739438240514e-06, + "loss": 0.3285, + "step": 10069 + }, + { + "epoch": 0.6888296053081606, + "grad_norm": 2.1812863450725013, + "learning_rate": 2.3315370738464863e-06, + "loss": 0.2442, + "step": 10070 + }, + { + "epoch": 0.6888980094397702, + "grad_norm": 2.5794683944472494, + "learning_rate": 2.3306003348499683e-06, + "loss": 0.2568, + "step": 10071 + }, + { + "epoch": 0.6889664135713797, + "grad_norm": 1.689812651730328, + "learning_rate": 2.329663726880476e-06, + "loss": 0.1605, + "step": 10072 + }, + { + "epoch": 0.6890348177029892, + "grad_norm": 2.624449587064341, + "learning_rate": 2.3287272499839832e-06, + "loss": 0.3103, + "step": 10073 + }, + { + "epoch": 0.6891032218345988, + "grad_norm": 2.243177567517319, + "learning_rate": 2.3277909042064546e-06, + "loss": 0.415, + "step": 10074 + }, + { + "epoch": 0.6891716259662084, + "grad_norm": 2.7594468620421244, + "learning_rate": 2.326854689593854e-06, + "loss": 0.4611, + "step": 10075 + }, + { + "epoch": 0.6892400300978179, + "grad_norm": 1.9633324462736692, + "learning_rate": 2.3259186061921333e-06, + "loss": 0.4386, + "step": 10076 + }, + { + "epoch": 0.6893084342294274, + "grad_norm": 1.5174208649170067, + "learning_rate": 2.3249826540472413e-06, + "loss": 0.198, + "step": 10077 + }, + { + "epoch": 0.689376838361037, + "grad_norm": 2.410980800751576, + "learning_rate": 2.3240468332051176e-06, + "loss": 0.6766, + "step": 10078 + }, + { + "epoch": 0.6894452424926466, + "grad_norm": 2.254588693210442, + "learning_rate": 2.3231111437116954e-06, + "loss": 0.4671, + "step": 10079 + }, + { + "epoch": 0.6895136466242561, + "grad_norm": 1.7923496149735512, + "learning_rate": 2.3221755856129074e-06, + "loss": 0.2022, + "step": 10080 + }, + { + "epoch": 0.6895820507558656, + "grad_norm": 1.655442918411209, + "learning_rate": 2.32124015895467e-06, + "loss": 0.2777, + "step": 10081 + }, + { + "epoch": 0.6896504548874752, + "grad_norm": 2.4884147608224794, + "learning_rate": 2.3203048637829032e-06, + "loss": 0.3278, + "step": 10082 + }, + { + "epoch": 0.6897188590190848, + "grad_norm": 1.9820465296305576, + "learning_rate": 2.3193697001435145e-06, + "loss": 0.4591, + "step": 10083 + }, + { + "epoch": 0.6897872631506943, + "grad_norm": 2.136153277065179, + "learning_rate": 2.318434668082405e-06, + "loss": 0.3659, + "step": 10084 + }, + { + "epoch": 0.6898556672823039, + "grad_norm": 2.249589207024445, + "learning_rate": 2.3174997676454712e-06, + "loss": 0.2186, + "step": 10085 + }, + { + "epoch": 0.6899240714139134, + "grad_norm": 2.0443813334605707, + "learning_rate": 2.3165649988786026e-06, + "loss": 0.4524, + "step": 10086 + }, + { + "epoch": 0.6899924755455229, + "grad_norm": 1.669460277450116, + "learning_rate": 2.3156303618276806e-06, + "loss": 0.3023, + "step": 10087 + }, + { + "epoch": 0.6900608796771325, + "grad_norm": 2.4866746899161734, + "learning_rate": 2.314695856538584e-06, + "loss": 0.4295, + "step": 10088 + }, + { + "epoch": 0.6901292838087421, + "grad_norm": 2.408226737024983, + "learning_rate": 2.313761483057181e-06, + "loss": 0.4762, + "step": 10089 + }, + { + "epoch": 0.6901976879403516, + "grad_norm": 2.811216994210986, + "learning_rate": 2.3128272414293374e-06, + "loss": 0.3339, + "step": 10090 + }, + { + "epoch": 0.6902660920719611, + "grad_norm": 2.8029109885896055, + "learning_rate": 2.311893131700909e-06, + "loss": 0.5583, + "step": 10091 + }, + { + "epoch": 0.6903344962035707, + "grad_norm": 2.9898262467979677, + "learning_rate": 2.3109591539177466e-06, + "loss": 0.4624, + "step": 10092 + }, + { + "epoch": 0.6904029003351803, + "grad_norm": 1.7303454180785194, + "learning_rate": 2.3100253081256944e-06, + "loss": 0.3072, + "step": 10093 + }, + { + "epoch": 0.6904713044667898, + "grad_norm": 1.9352142902798246, + "learning_rate": 2.3090915943705878e-06, + "loss": 0.3849, + "step": 10094 + }, + { + "epoch": 0.6905397085983993, + "grad_norm": 2.1934229908631915, + "learning_rate": 2.3081580126982617e-06, + "loss": 0.4845, + "step": 10095 + }, + { + "epoch": 0.6906081127300089, + "grad_norm": 2.2826769559439306, + "learning_rate": 2.3072245631545394e-06, + "loss": 0.2581, + "step": 10096 + }, + { + "epoch": 0.6906765168616185, + "grad_norm": 1.7735668126825552, + "learning_rate": 2.3062912457852387e-06, + "loss": 0.2375, + "step": 10097 + }, + { + "epoch": 0.690744920993228, + "grad_norm": 2.3828660552463243, + "learning_rate": 2.3053580606361716e-06, + "loss": 0.5485, + "step": 10098 + }, + { + "epoch": 0.6908133251248375, + "grad_norm": 2.4829873222130354, + "learning_rate": 2.304425007753141e-06, + "loss": 0.6727, + "step": 10099 + }, + { + "epoch": 0.6908817292564471, + "grad_norm": 1.6977215627508344, + "learning_rate": 2.3034920871819493e-06, + "loss": 0.1769, + "step": 10100 + }, + { + "epoch": 0.6909501333880567, + "grad_norm": 2.0951884438516823, + "learning_rate": 2.3025592989683867e-06, + "loss": 0.4234, + "step": 10101 + }, + { + "epoch": 0.6910185375196662, + "grad_norm": 2.5943557077975736, + "learning_rate": 2.30162664315824e-06, + "loss": 0.3134, + "step": 10102 + }, + { + "epoch": 0.6910869416512757, + "grad_norm": 2.234077793975572, + "learning_rate": 2.3006941197972886e-06, + "loss": 0.3058, + "step": 10103 + }, + { + "epoch": 0.6911553457828853, + "grad_norm": 2.263547424897759, + "learning_rate": 2.2997617289313047e-06, + "loss": 0.4973, + "step": 10104 + }, + { + "epoch": 0.6912237499144949, + "grad_norm": 2.0584250096865433, + "learning_rate": 2.2988294706060545e-06, + "loss": 0.268, + "step": 10105 + }, + { + "epoch": 0.6912921540461043, + "grad_norm": 2.2790758929636508, + "learning_rate": 2.2978973448672975e-06, + "loss": 0.3439, + "step": 10106 + }, + { + "epoch": 0.6913605581777139, + "grad_norm": 1.872682705064918, + "learning_rate": 2.2969653517607855e-06, + "loss": 0.4711, + "step": 10107 + }, + { + "epoch": 0.6914289623093235, + "grad_norm": 2.4992159239743166, + "learning_rate": 2.296033491332268e-06, + "loss": 0.2415, + "step": 10108 + }, + { + "epoch": 0.6914973664409331, + "grad_norm": 1.9427046319825723, + "learning_rate": 2.295101763627483e-06, + "loss": 0.4214, + "step": 10109 + }, + { + "epoch": 0.6915657705725425, + "grad_norm": 3.3304981134647282, + "learning_rate": 2.2941701686921667e-06, + "loss": 0.5549, + "step": 10110 + }, + { + "epoch": 0.6916341747041521, + "grad_norm": 2.2390991633861512, + "learning_rate": 2.2932387065720453e-06, + "loss": 0.454, + "step": 10111 + }, + { + "epoch": 0.6917025788357617, + "grad_norm": 1.6216782722526695, + "learning_rate": 2.2923073773128385e-06, + "loss": 0.1855, + "step": 10112 + }, + { + "epoch": 0.6917709829673713, + "grad_norm": 2.161991540029014, + "learning_rate": 2.2913761809602607e-06, + "loss": 0.4647, + "step": 10113 + }, + { + "epoch": 0.6918393870989807, + "grad_norm": 2.2484492414961865, + "learning_rate": 2.2904451175600186e-06, + "loss": 0.5108, + "step": 10114 + }, + { + "epoch": 0.6919077912305903, + "grad_norm": 2.718452788144545, + "learning_rate": 2.2895141871578164e-06, + "loss": 0.5789, + "step": 10115 + }, + { + "epoch": 0.6919761953621999, + "grad_norm": 2.4001635947697757, + "learning_rate": 2.288583389799346e-06, + "loss": 0.4075, + "step": 10116 + }, + { + "epoch": 0.6920445994938095, + "grad_norm": 1.8119141026330656, + "learning_rate": 2.2876527255302965e-06, + "loss": 0.2558, + "step": 10117 + }, + { + "epoch": 0.6921130036254189, + "grad_norm": 2.5028043277229353, + "learning_rate": 2.286722194396349e-06, + "loss": 0.6107, + "step": 10118 + }, + { + "epoch": 0.6921814077570285, + "grad_norm": 2.70113684465335, + "learning_rate": 2.285791796443176e-06, + "loss": 0.5373, + "step": 10119 + }, + { + "epoch": 0.6922498118886381, + "grad_norm": 2.794266731256826, + "learning_rate": 2.284861531716451e-06, + "loss": 0.4888, + "step": 10120 + }, + { + "epoch": 0.6923182160202476, + "grad_norm": 2.409546847715036, + "learning_rate": 2.2839314002618317e-06, + "loss": 0.7129, + "step": 10121 + }, + { + "epoch": 0.6923866201518571, + "grad_norm": 2.5645376686866985, + "learning_rate": 2.2830014021249765e-06, + "loss": 0.828, + "step": 10122 + }, + { + "epoch": 0.6924550242834667, + "grad_norm": 2.1541101231215976, + "learning_rate": 2.2820715373515328e-06, + "loss": 0.6475, + "step": 10123 + }, + { + "epoch": 0.6925234284150763, + "grad_norm": 2.4260867935368866, + "learning_rate": 2.2811418059871428e-06, + "loss": 0.5476, + "step": 10124 + }, + { + "epoch": 0.6925918325466858, + "grad_norm": 1.6239306801314826, + "learning_rate": 2.280212208077442e-06, + "loss": 0.2827, + "step": 10125 + }, + { + "epoch": 0.6926602366782953, + "grad_norm": 2.0370643954555114, + "learning_rate": 2.2792827436680597e-06, + "loss": 0.4962, + "step": 10126 + }, + { + "epoch": 0.6927286408099049, + "grad_norm": 2.3923881330764956, + "learning_rate": 2.278353412804617e-06, + "loss": 0.483, + "step": 10127 + }, + { + "epoch": 0.6927970449415145, + "grad_norm": 1.6575797698731995, + "learning_rate": 2.277424215532733e-06, + "loss": 0.2276, + "step": 10128 + }, + { + "epoch": 0.692865449073124, + "grad_norm": 2.4791023335537923, + "learning_rate": 2.2764951518980138e-06, + "loss": 0.3078, + "step": 10129 + }, + { + "epoch": 0.6929338532047336, + "grad_norm": 1.6686149971176565, + "learning_rate": 2.275566221946067e-06, + "loss": 0.1261, + "step": 10130 + }, + { + "epoch": 0.6930022573363431, + "grad_norm": 2.928812817019145, + "learning_rate": 2.274637425722485e-06, + "loss": 0.8399, + "step": 10131 + }, + { + "epoch": 0.6930706614679527, + "grad_norm": 2.5924307932624284, + "learning_rate": 2.2737087632728594e-06, + "loss": 0.2697, + "step": 10132 + }, + { + "epoch": 0.6931390655995622, + "grad_norm": 2.064176809826035, + "learning_rate": 2.2727802346427724e-06, + "loss": 0.503, + "step": 10133 + }, + { + "epoch": 0.6932074697311718, + "grad_norm": 2.2262723231709116, + "learning_rate": 2.2718518398778e-06, + "loss": 0.6169, + "step": 10134 + }, + { + "epoch": 0.6932758738627813, + "grad_norm": 2.0679870176643353, + "learning_rate": 2.2709235790235147e-06, + "loss": 0.445, + "step": 10135 + }, + { + "epoch": 0.6933442779943909, + "grad_norm": 2.3656471830912795, + "learning_rate": 2.2699954521254785e-06, + "loss": 0.5164, + "step": 10136 + }, + { + "epoch": 0.6934126821260004, + "grad_norm": 2.091356674592853, + "learning_rate": 2.269067459229249e-06, + "loss": 0.3343, + "step": 10137 + }, + { + "epoch": 0.69348108625761, + "grad_norm": 1.96966935444916, + "learning_rate": 2.268139600380376e-06, + "loss": 0.2878, + "step": 10138 + }, + { + "epoch": 0.6935494903892195, + "grad_norm": 2.6619566036057547, + "learning_rate": 2.2672118756244014e-06, + "loss": 0.6058, + "step": 10139 + }, + { + "epoch": 0.693617894520829, + "grad_norm": 2.34321254289921, + "learning_rate": 2.2662842850068668e-06, + "loss": 0.3, + "step": 10140 + }, + { + "epoch": 0.6936862986524386, + "grad_norm": 2.2162725626791544, + "learning_rate": 2.2653568285732978e-06, + "loss": 0.274, + "step": 10141 + }, + { + "epoch": 0.6937547027840482, + "grad_norm": 2.1561579891973897, + "learning_rate": 2.264429506369224e-06, + "loss": 0.5099, + "step": 10142 + }, + { + "epoch": 0.6938231069156577, + "grad_norm": 1.6818390661601919, + "learning_rate": 2.2635023184401596e-06, + "loss": 0.1475, + "step": 10143 + }, + { + "epoch": 0.6938915110472672, + "grad_norm": 1.9678176382699548, + "learning_rate": 2.262575264831615e-06, + "loss": 0.2772, + "step": 10144 + }, + { + "epoch": 0.6939599151788768, + "grad_norm": 2.0231040215840848, + "learning_rate": 2.2616483455890966e-06, + "loss": 0.3925, + "step": 10145 + }, + { + "epoch": 0.6940283193104864, + "grad_norm": 2.6711297696929797, + "learning_rate": 2.2607215607580997e-06, + "loss": 0.4247, + "step": 10146 + }, + { + "epoch": 0.6940967234420959, + "grad_norm": 2.1816907345883347, + "learning_rate": 2.259794910384115e-06, + "loss": 0.3368, + "step": 10147 + }, + { + "epoch": 0.6941651275737054, + "grad_norm": 1.5783314994695954, + "learning_rate": 2.2588683945126307e-06, + "loss": 0.2445, + "step": 10148 + }, + { + "epoch": 0.694233531705315, + "grad_norm": 2.1557437986706964, + "learning_rate": 2.2579420131891194e-06, + "loss": 0.3906, + "step": 10149 + }, + { + "epoch": 0.6943019358369246, + "grad_norm": 1.6732766167589632, + "learning_rate": 2.257015766459058e-06, + "loss": 0.2963, + "step": 10150 + }, + { + "epoch": 0.6943703399685341, + "grad_norm": 1.7536462535882198, + "learning_rate": 2.2560896543679086e-06, + "loss": 0.3619, + "step": 10151 + }, + { + "epoch": 0.6944387441001436, + "grad_norm": 1.8785618567533693, + "learning_rate": 2.255163676961129e-06, + "loss": 0.4033, + "step": 10152 + }, + { + "epoch": 0.6945071482317532, + "grad_norm": 1.9986958759475546, + "learning_rate": 2.254237834284171e-06, + "loss": 0.2231, + "step": 10153 + }, + { + "epoch": 0.6945755523633628, + "grad_norm": 2.391129920345383, + "learning_rate": 2.253312126382477e-06, + "loss": 0.5939, + "step": 10154 + }, + { + "epoch": 0.6946439564949723, + "grad_norm": 2.2478557475594068, + "learning_rate": 2.2523865533014896e-06, + "loss": 0.5709, + "step": 10155 + }, + { + "epoch": 0.6947123606265818, + "grad_norm": 2.2952630945943, + "learning_rate": 2.251461115086638e-06, + "loss": 0.4002, + "step": 10156 + }, + { + "epoch": 0.6947807647581914, + "grad_norm": 2.5584259444092625, + "learning_rate": 2.2505358117833474e-06, + "loss": 0.2951, + "step": 10157 + }, + { + "epoch": 0.694849168889801, + "grad_norm": 2.4464168187814783, + "learning_rate": 2.249610643437037e-06, + "loss": 0.7146, + "step": 10158 + }, + { + "epoch": 0.6949175730214104, + "grad_norm": 2.138598774396611, + "learning_rate": 2.2486856100931146e-06, + "loss": 0.3765, + "step": 10159 + }, + { + "epoch": 0.69498597715302, + "grad_norm": 1.9593828581389288, + "learning_rate": 2.2477607117969913e-06, + "loss": 0.3694, + "step": 10160 + }, + { + "epoch": 0.6950543812846296, + "grad_norm": 2.179624877905182, + "learning_rate": 2.2468359485940604e-06, + "loss": 0.4249, + "step": 10161 + }, + { + "epoch": 0.6951227854162392, + "grad_norm": 2.0208464864584625, + "learning_rate": 2.245911320529718e-06, + "loss": 0.345, + "step": 10162 + }, + { + "epoch": 0.6951911895478486, + "grad_norm": 1.3914253203570017, + "learning_rate": 2.2449868276493468e-06, + "loss": 0.1186, + "step": 10163 + }, + { + "epoch": 0.6952595936794582, + "grad_norm": 2.585109854841268, + "learning_rate": 2.2440624699983265e-06, + "loss": 0.4937, + "step": 10164 + }, + { + "epoch": 0.6953279978110678, + "grad_norm": 2.5194609639617966, + "learning_rate": 2.2431382476220285e-06, + "loss": 0.301, + "step": 10165 + }, + { + "epoch": 0.6953964019426774, + "grad_norm": 1.85911721758681, + "learning_rate": 2.242214160565816e-06, + "loss": 0.1732, + "step": 10166 + }, + { + "epoch": 0.6954648060742868, + "grad_norm": 3.017681566952997, + "learning_rate": 2.2412902088750516e-06, + "loss": 0.6469, + "step": 10167 + }, + { + "epoch": 0.6955332102058964, + "grad_norm": 1.9998975258145206, + "learning_rate": 2.2403663925950837e-06, + "loss": 0.1598, + "step": 10168 + }, + { + "epoch": 0.695601614337506, + "grad_norm": 1.818469649596265, + "learning_rate": 2.2394427117712613e-06, + "loss": 0.2732, + "step": 10169 + }, + { + "epoch": 0.6956700184691156, + "grad_norm": 2.771330449438288, + "learning_rate": 2.238519166448921e-06, + "loss": 0.3409, + "step": 10170 + }, + { + "epoch": 0.695738422600725, + "grad_norm": 2.877457169910844, + "learning_rate": 2.2375957566733953e-06, + "loss": 0.4766, + "step": 10171 + }, + { + "epoch": 0.6958068267323346, + "grad_norm": 2.8624565737566146, + "learning_rate": 2.2366724824900086e-06, + "loss": 0.4735, + "step": 10172 + }, + { + "epoch": 0.6958752308639442, + "grad_norm": 2.472578183067191, + "learning_rate": 2.2357493439440796e-06, + "loss": 0.3354, + "step": 10173 + }, + { + "epoch": 0.6959436349955538, + "grad_norm": 2.2810419719376585, + "learning_rate": 2.2348263410809225e-06, + "loss": 0.2242, + "step": 10174 + }, + { + "epoch": 0.6960120391271633, + "grad_norm": 2.4550609845979574, + "learning_rate": 2.2339034739458416e-06, + "loss": 0.4862, + "step": 10175 + }, + { + "epoch": 0.6960804432587728, + "grad_norm": 2.0291961850546105, + "learning_rate": 2.2329807425841354e-06, + "loss": 0.2944, + "step": 10176 + }, + { + "epoch": 0.6961488473903824, + "grad_norm": 2.057972374500816, + "learning_rate": 2.2320581470410964e-06, + "loss": 0.2488, + "step": 10177 + }, + { + "epoch": 0.6962172515219919, + "grad_norm": 2.6546507901891765, + "learning_rate": 2.2311356873620076e-06, + "loss": 0.472, + "step": 10178 + }, + { + "epoch": 0.6962856556536015, + "grad_norm": 1.512159546946331, + "learning_rate": 2.2302133635921524e-06, + "loss": 0.1129, + "step": 10179 + }, + { + "epoch": 0.696354059785211, + "grad_norm": 2.0918225341134957, + "learning_rate": 2.229291175776798e-06, + "loss": 0.4637, + "step": 10180 + }, + { + "epoch": 0.6964224639168206, + "grad_norm": 1.8426315341929613, + "learning_rate": 2.2283691239612147e-06, + "loss": 0.3343, + "step": 10181 + }, + { + "epoch": 0.6964908680484301, + "grad_norm": 2.014472698584814, + "learning_rate": 2.227447208190659e-06, + "loss": 0.4186, + "step": 10182 + }, + { + "epoch": 0.6965592721800397, + "grad_norm": 2.608956651787349, + "learning_rate": 2.226525428510382e-06, + "loss": 0.4764, + "step": 10183 + }, + { + "epoch": 0.6966276763116492, + "grad_norm": 2.639862977629873, + "learning_rate": 2.2256037849656304e-06, + "loss": 0.4652, + "step": 10184 + }, + { + "epoch": 0.6966960804432588, + "grad_norm": 2.229396849777311, + "learning_rate": 2.2246822776016425e-06, + "loss": 0.2585, + "step": 10185 + }, + { + "epoch": 0.6967644845748683, + "grad_norm": 2.2699787598195975, + "learning_rate": 2.2237609064636485e-06, + "loss": 0.4851, + "step": 10186 + }, + { + "epoch": 0.6968328887064779, + "grad_norm": 2.1866616822564877, + "learning_rate": 2.2228396715968776e-06, + "loss": 0.4031, + "step": 10187 + }, + { + "epoch": 0.6969012928380874, + "grad_norm": 2.303653303964087, + "learning_rate": 2.2219185730465442e-06, + "loss": 0.3657, + "step": 10188 + }, + { + "epoch": 0.696969696969697, + "grad_norm": 2.168718864646164, + "learning_rate": 2.220997610857864e-06, + "loss": 0.3733, + "step": 10189 + }, + { + "epoch": 0.6970381011013065, + "grad_norm": 1.972584464400552, + "learning_rate": 2.2200767850760412e-06, + "loss": 0.4344, + "step": 10190 + }, + { + "epoch": 0.6971065052329161, + "grad_norm": 2.6693719326815337, + "learning_rate": 2.2191560957462737e-06, + "loss": 0.2313, + "step": 10191 + }, + { + "epoch": 0.6971749093645256, + "grad_norm": 4.547067815094141, + "learning_rate": 2.2182355429137535e-06, + "loss": 0.2118, + "step": 10192 + }, + { + "epoch": 0.6972433134961352, + "grad_norm": 2.817261956067247, + "learning_rate": 2.2173151266236644e-06, + "loss": 0.5535, + "step": 10193 + }, + { + "epoch": 0.6973117176277447, + "grad_norm": 2.130696284331717, + "learning_rate": 2.2163948469211875e-06, + "loss": 0.3979, + "step": 10194 + }, + { + "epoch": 0.6973801217593543, + "grad_norm": 2.141308069883634, + "learning_rate": 2.215474703851494e-06, + "loss": 0.3347, + "step": 10195 + }, + { + "epoch": 0.6974485258909638, + "grad_norm": 2.552129491401588, + "learning_rate": 2.214554697459748e-06, + "loss": 0.8838, + "step": 10196 + }, + { + "epoch": 0.6975169300225733, + "grad_norm": 2.7450786792791373, + "learning_rate": 2.213634827791109e-06, + "loss": 0.4493, + "step": 10197 + }, + { + "epoch": 0.6975853341541829, + "grad_norm": 1.5392243450288865, + "learning_rate": 2.2127150948907254e-06, + "loss": 0.1359, + "step": 10198 + }, + { + "epoch": 0.6976537382857925, + "grad_norm": 1.56451454946873, + "learning_rate": 2.2117954988037467e-06, + "loss": 0.2356, + "step": 10199 + }, + { + "epoch": 0.697722142417402, + "grad_norm": 2.171908014718931, + "learning_rate": 2.2108760395753064e-06, + "loss": 0.3962, + "step": 10200 + }, + { + "epoch": 0.6977905465490115, + "grad_norm": 2.608606762209528, + "learning_rate": 2.2099567172505414e-06, + "loss": 0.4354, + "step": 10201 + }, + { + "epoch": 0.6978589506806211, + "grad_norm": 2.3542829283697895, + "learning_rate": 2.209037531874573e-06, + "loss": 0.4966, + "step": 10202 + }, + { + "epoch": 0.6979273548122307, + "grad_norm": 2.075880997414412, + "learning_rate": 2.2081184834925208e-06, + "loss": 0.3296, + "step": 10203 + }, + { + "epoch": 0.6979957589438402, + "grad_norm": 2.3186825990520457, + "learning_rate": 2.2071995721494945e-06, + "loss": 0.4356, + "step": 10204 + }, + { + "epoch": 0.6980641630754497, + "grad_norm": 2.571567970851373, + "learning_rate": 2.2062807978905997e-06, + "loss": 0.4185, + "step": 10205 + }, + { + "epoch": 0.6981325672070593, + "grad_norm": 2.4824088131540702, + "learning_rate": 2.2053621607609325e-06, + "loss": 0.3655, + "step": 10206 + }, + { + "epoch": 0.6982009713386689, + "grad_norm": 3.5818127947204808, + "learning_rate": 2.2044436608055876e-06, + "loss": 0.6645, + "step": 10207 + }, + { + "epoch": 0.6982693754702785, + "grad_norm": 2.0233725239104694, + "learning_rate": 2.2035252980696446e-06, + "loss": 0.4968, + "step": 10208 + }, + { + "epoch": 0.6983377796018879, + "grad_norm": 2.6765752686275595, + "learning_rate": 2.2026070725981867e-06, + "loss": 0.5745, + "step": 10209 + }, + { + "epoch": 0.6984061837334975, + "grad_norm": 2.130275628877546, + "learning_rate": 2.201688984436282e-06, + "loss": 0.3532, + "step": 10210 + }, + { + "epoch": 0.6984745878651071, + "grad_norm": 2.0171935609568363, + "learning_rate": 2.200771033628995e-06, + "loss": 0.3607, + "step": 10211 + }, + { + "epoch": 0.6985429919967167, + "grad_norm": 2.7580076586262217, + "learning_rate": 2.199853220221383e-06, + "loss": 0.6253, + "step": 10212 + }, + { + "epoch": 0.6986113961283261, + "grad_norm": 2.403237897505347, + "learning_rate": 2.198935544258494e-06, + "loss": 0.5061, + "step": 10213 + }, + { + "epoch": 0.6986798002599357, + "grad_norm": 1.8030548265510662, + "learning_rate": 2.198018005785378e-06, + "loss": 0.4372, + "step": 10214 + }, + { + "epoch": 0.6987482043915453, + "grad_norm": 2.5853404770587756, + "learning_rate": 2.197100604847068e-06, + "loss": 0.2245, + "step": 10215 + }, + { + "epoch": 0.6988166085231547, + "grad_norm": 2.5304080363448422, + "learning_rate": 2.196183341488595e-06, + "loss": 0.4868, + "step": 10216 + }, + { + "epoch": 0.6988850126547643, + "grad_norm": 2.4790961124628037, + "learning_rate": 2.195266215754984e-06, + "loss": 0.6972, + "step": 10217 + }, + { + "epoch": 0.6989534167863739, + "grad_norm": 2.0153930294009066, + "learning_rate": 2.1943492276912486e-06, + "loss": 0.4821, + "step": 10218 + }, + { + "epoch": 0.6990218209179835, + "grad_norm": 2.450641662883464, + "learning_rate": 2.193432377342404e-06, + "loss": 0.4183, + "step": 10219 + }, + { + "epoch": 0.699090225049593, + "grad_norm": 2.235006063216261, + "learning_rate": 2.1925156647534483e-06, + "loss": 0.4841, + "step": 10220 + }, + { + "epoch": 0.6991586291812025, + "grad_norm": 1.9687366863587725, + "learning_rate": 2.1915990899693834e-06, + "loss": 0.3185, + "step": 10221 + }, + { + "epoch": 0.6992270333128121, + "grad_norm": 2.1464394173326573, + "learning_rate": 2.1906826530351965e-06, + "loss": 0.1144, + "step": 10222 + }, + { + "epoch": 0.6992954374444217, + "grad_norm": 1.805212827905459, + "learning_rate": 2.1897663539958715e-06, + "loss": 0.1361, + "step": 10223 + }, + { + "epoch": 0.6993638415760312, + "grad_norm": 1.8296038848769824, + "learning_rate": 2.188850192896384e-06, + "loss": 0.3152, + "step": 10224 + }, + { + "epoch": 0.6994322457076407, + "grad_norm": 2.9623035913056546, + "learning_rate": 2.1879341697817043e-06, + "loss": 0.4179, + "step": 10225 + }, + { + "epoch": 0.6995006498392503, + "grad_norm": 2.9436897542988554, + "learning_rate": 2.1870182846967924e-06, + "loss": 0.1501, + "step": 10226 + }, + { + "epoch": 0.6995690539708599, + "grad_norm": 2.3681414038763786, + "learning_rate": 2.186102537686609e-06, + "loss": 0.5479, + "step": 10227 + }, + { + "epoch": 0.6996374581024694, + "grad_norm": 1.9675771042633892, + "learning_rate": 2.1851869287961e-06, + "loss": 0.3333, + "step": 10228 + }, + { + "epoch": 0.6997058622340789, + "grad_norm": 2.2740473175171068, + "learning_rate": 2.18427145807021e-06, + "loss": 0.3445, + "step": 10229 + }, + { + "epoch": 0.6997742663656885, + "grad_norm": 2.0914715069524608, + "learning_rate": 2.1833561255538745e-06, + "loss": 0.2649, + "step": 10230 + }, + { + "epoch": 0.6998426704972981, + "grad_norm": 2.3333366644500613, + "learning_rate": 2.182440931292021e-06, + "loss": 0.3102, + "step": 10231 + }, + { + "epoch": 0.6999110746289076, + "grad_norm": 1.779456503079791, + "learning_rate": 2.181525875329573e-06, + "loss": 0.4268, + "step": 10232 + }, + { + "epoch": 0.6999794787605171, + "grad_norm": 2.328142140472603, + "learning_rate": 2.180610957711444e-06, + "loss": 0.5518, + "step": 10233 + }, + { + "epoch": 0.7000478828921267, + "grad_norm": 2.7660310145481843, + "learning_rate": 2.179696178482545e-06, + "loss": 0.6228, + "step": 10234 + }, + { + "epoch": 0.7001162870237362, + "grad_norm": 2.2888058788722234, + "learning_rate": 2.1787815376877764e-06, + "loss": 0.4807, + "step": 10235 + }, + { + "epoch": 0.7001846911553458, + "grad_norm": 2.048056494446318, + "learning_rate": 2.1778670353720337e-06, + "loss": 0.4253, + "step": 10236 + }, + { + "epoch": 0.7002530952869553, + "grad_norm": 2.3326839511250417, + "learning_rate": 2.1769526715802053e-06, + "loss": 0.5797, + "step": 10237 + }, + { + "epoch": 0.7003214994185649, + "grad_norm": 2.625702741198909, + "learning_rate": 2.17603844635717e-06, + "loss": 0.3162, + "step": 10238 + }, + { + "epoch": 0.7003899035501744, + "grad_norm": 1.440859199755159, + "learning_rate": 2.175124359747806e-06, + "loss": 0.1758, + "step": 10239 + }, + { + "epoch": 0.700458307681784, + "grad_norm": 2.8948037949717866, + "learning_rate": 2.174210411796979e-06, + "loss": 0.5596, + "step": 10240 + }, + { + "epoch": 0.7005267118133935, + "grad_norm": 2.205003566511994, + "learning_rate": 2.173296602549551e-06, + "loss": 0.4518, + "step": 10241 + }, + { + "epoch": 0.7005951159450031, + "grad_norm": 2.7466299257976883, + "learning_rate": 2.1723829320503765e-06, + "loss": 0.2243, + "step": 10242 + }, + { + "epoch": 0.7006635200766126, + "grad_norm": 1.8022313351209027, + "learning_rate": 2.1714694003443022e-06, + "loss": 0.1541, + "step": 10243 + }, + { + "epoch": 0.7007319242082222, + "grad_norm": 2.3175735780924342, + "learning_rate": 2.170556007476168e-06, + "loss": 0.5857, + "step": 10244 + }, + { + "epoch": 0.7008003283398317, + "grad_norm": 1.9928280642031624, + "learning_rate": 2.169642753490808e-06, + "loss": 0.2407, + "step": 10245 + }, + { + "epoch": 0.7008687324714413, + "grad_norm": 2.764497882889813, + "learning_rate": 2.168729638433049e-06, + "loss": 0.6322, + "step": 10246 + }, + { + "epoch": 0.7009371366030508, + "grad_norm": 3.2042474780332086, + "learning_rate": 2.1678166623477116e-06, + "loss": 0.3402, + "step": 10247 + }, + { + "epoch": 0.7010055407346604, + "grad_norm": 3.0400545369216085, + "learning_rate": 2.1669038252796074e-06, + "loss": 0.3459, + "step": 10248 + }, + { + "epoch": 0.70107394486627, + "grad_norm": 2.5386768458996114, + "learning_rate": 2.165991127273547e-06, + "loss": 0.359, + "step": 10249 + }, + { + "epoch": 0.7011423489978794, + "grad_norm": 2.3077838335471528, + "learning_rate": 2.165078568374327e-06, + "loss": 0.4747, + "step": 10250 + }, + { + "epoch": 0.701210753129489, + "grad_norm": 2.185415009994049, + "learning_rate": 2.16416614862674e-06, + "loss": 0.2265, + "step": 10251 + }, + { + "epoch": 0.7012791572610986, + "grad_norm": 14.25945751158789, + "learning_rate": 2.1632538680755728e-06, + "loss": 0.2708, + "step": 10252 + }, + { + "epoch": 0.7013475613927082, + "grad_norm": 2.1001919953816626, + "learning_rate": 2.1623417267656026e-06, + "loss": 0.3105, + "step": 10253 + }, + { + "epoch": 0.7014159655243176, + "grad_norm": 1.51160558380484, + "learning_rate": 2.1614297247416053e-06, + "loss": 0.2789, + "step": 10254 + }, + { + "epoch": 0.7014843696559272, + "grad_norm": 2.334432021112333, + "learning_rate": 2.160517862048344e-06, + "loss": 0.4596, + "step": 10255 + }, + { + "epoch": 0.7015527737875368, + "grad_norm": 2.0528980193527757, + "learning_rate": 2.1596061387305783e-06, + "loss": 0.3153, + "step": 10256 + }, + { + "epoch": 0.7016211779191464, + "grad_norm": 2.3091916828413437, + "learning_rate": 2.158694554833059e-06, + "loss": 0.58, + "step": 10257 + }, + { + "epoch": 0.7016895820507558, + "grad_norm": 2.4017240639957187, + "learning_rate": 2.15778311040053e-06, + "loss": 0.5955, + "step": 10258 + }, + { + "epoch": 0.7017579861823654, + "grad_norm": 1.7130903608208279, + "learning_rate": 2.1568718054777322e-06, + "loss": 0.1586, + "step": 10259 + }, + { + "epoch": 0.701826390313975, + "grad_norm": 1.9936820613562438, + "learning_rate": 2.155960640109394e-06, + "loss": 0.247, + "step": 10260 + }, + { + "epoch": 0.7018947944455846, + "grad_norm": 2.7165280228957576, + "learning_rate": 2.1550496143402433e-06, + "loss": 0.394, + "step": 10261 + }, + { + "epoch": 0.701963198577194, + "grad_norm": 2.1086926769556094, + "learning_rate": 2.1541387282149957e-06, + "loss": 0.2991, + "step": 10262 + }, + { + "epoch": 0.7020316027088036, + "grad_norm": 2.5369225683590244, + "learning_rate": 2.153227981778361e-06, + "loss": 0.4583, + "step": 10263 + }, + { + "epoch": 0.7021000068404132, + "grad_norm": 1.5789497175089402, + "learning_rate": 2.1523173750750447e-06, + "loss": 0.2813, + "step": 10264 + }, + { + "epoch": 0.7021684109720228, + "grad_norm": 2.1398401910056593, + "learning_rate": 2.151406908149741e-06, + "loss": 0.2523, + "step": 10265 + }, + { + "epoch": 0.7022368151036322, + "grad_norm": 2.2168119186361457, + "learning_rate": 2.1504965810471435e-06, + "loss": 0.521, + "step": 10266 + }, + { + "epoch": 0.7023052192352418, + "grad_norm": 2.803193855503149, + "learning_rate": 2.1495863938119325e-06, + "loss": 0.2991, + "step": 10267 + }, + { + "epoch": 0.7023736233668514, + "grad_norm": 2.289723505919144, + "learning_rate": 2.1486763464887874e-06, + "loss": 0.346, + "step": 10268 + }, + { + "epoch": 0.7024420274984609, + "grad_norm": 2.189094500474564, + "learning_rate": 2.1477664391223763e-06, + "loss": 0.497, + "step": 10269 + }, + { + "epoch": 0.7025104316300704, + "grad_norm": 1.958234953607371, + "learning_rate": 2.146856671757361e-06, + "loss": 0.4191, + "step": 10270 + }, + { + "epoch": 0.70257883576168, + "grad_norm": 3.173962197626406, + "learning_rate": 2.145947044438399e-06, + "loss": 0.311, + "step": 10271 + }, + { + "epoch": 0.7026472398932896, + "grad_norm": 2.111125798725043, + "learning_rate": 2.1450375572101357e-06, + "loss": 0.2484, + "step": 10272 + }, + { + "epoch": 0.702715644024899, + "grad_norm": 2.002852889903799, + "learning_rate": 2.1441282101172175e-06, + "loss": 0.5506, + "step": 10273 + }, + { + "epoch": 0.7027840481565086, + "grad_norm": 2.5956081203643917, + "learning_rate": 2.143219003204277e-06, + "loss": 0.5095, + "step": 10274 + }, + { + "epoch": 0.7028524522881182, + "grad_norm": 2.05831245657595, + "learning_rate": 2.1423099365159437e-06, + "loss": 0.2993, + "step": 10275 + }, + { + "epoch": 0.7029208564197278, + "grad_norm": 2.0228672402299126, + "learning_rate": 2.1414010100968387e-06, + "loss": 0.3263, + "step": 10276 + }, + { + "epoch": 0.7029892605513373, + "grad_norm": 1.9668198918470343, + "learning_rate": 2.1404922239915735e-06, + "loss": 0.2643, + "step": 10277 + }, + { + "epoch": 0.7030576646829468, + "grad_norm": 2.860231370311917, + "learning_rate": 2.1395835782447605e-06, + "loss": 0.7292, + "step": 10278 + }, + { + "epoch": 0.7031260688145564, + "grad_norm": 1.5756282365199028, + "learning_rate": 2.138675072900997e-06, + "loss": 0.1514, + "step": 10279 + }, + { + "epoch": 0.703194472946166, + "grad_norm": 2.0951641772724274, + "learning_rate": 2.1377667080048794e-06, + "loss": 0.2492, + "step": 10280 + }, + { + "epoch": 0.7032628770777755, + "grad_norm": 2.7981807069087834, + "learning_rate": 2.1368584836009933e-06, + "loss": 0.5405, + "step": 10281 + }, + { + "epoch": 0.703331281209385, + "grad_norm": 1.2292121900918276, + "learning_rate": 2.1359503997339195e-06, + "loss": 0.1266, + "step": 10282 + }, + { + "epoch": 0.7033996853409946, + "grad_norm": 1.9212103485242407, + "learning_rate": 2.1350424564482297e-06, + "loss": 0.3957, + "step": 10283 + }, + { + "epoch": 0.7034680894726042, + "grad_norm": 2.037262600835314, + "learning_rate": 2.134134653788491e-06, + "loss": 0.5639, + "step": 10284 + }, + { + "epoch": 0.7035364936042137, + "grad_norm": 3.29225351840444, + "learning_rate": 2.133226991799261e-06, + "loss": 0.5979, + "step": 10285 + }, + { + "epoch": 0.7036048977358232, + "grad_norm": 2.408485327998464, + "learning_rate": 2.132319470525095e-06, + "loss": 0.2888, + "step": 10286 + }, + { + "epoch": 0.7036733018674328, + "grad_norm": 2.3285574671366875, + "learning_rate": 2.1314120900105363e-06, + "loss": 0.422, + "step": 10287 + }, + { + "epoch": 0.7037417059990423, + "grad_norm": 2.3898201503879126, + "learning_rate": 2.1305048503001256e-06, + "loss": 0.5739, + "step": 10288 + }, + { + "epoch": 0.7038101101306519, + "grad_norm": 2.166016374868224, + "learning_rate": 2.1295977514383937e-06, + "loss": 0.3464, + "step": 10289 + }, + { + "epoch": 0.7038785142622614, + "grad_norm": 3.218368155783106, + "learning_rate": 2.1286907934698653e-06, + "loss": 0.2523, + "step": 10290 + }, + { + "epoch": 0.703946918393871, + "grad_norm": 2.0460406779732345, + "learning_rate": 2.1277839764390583e-06, + "loss": 0.2871, + "step": 10291 + }, + { + "epoch": 0.7040153225254805, + "grad_norm": 2.256825782558253, + "learning_rate": 2.1268773003904815e-06, + "loss": 0.5441, + "step": 10292 + }, + { + "epoch": 0.7040837266570901, + "grad_norm": 2.572639820453023, + "learning_rate": 2.125970765368643e-06, + "loss": 0.5641, + "step": 10293 + }, + { + "epoch": 0.7041521307886996, + "grad_norm": 1.998758623690249, + "learning_rate": 2.1250643714180373e-06, + "loss": 0.387, + "step": 10294 + }, + { + "epoch": 0.7042205349203092, + "grad_norm": 2.7356784220781036, + "learning_rate": 2.124158118583156e-06, + "loss": 0.6762, + "step": 10295 + }, + { + "epoch": 0.7042889390519187, + "grad_norm": 2.109141909118526, + "learning_rate": 2.123252006908481e-06, + "loss": 0.4672, + "step": 10296 + }, + { + "epoch": 0.7043573431835283, + "grad_norm": 2.733628438122786, + "learning_rate": 2.1223460364384875e-06, + "loss": 0.3361, + "step": 10297 + }, + { + "epoch": 0.7044257473151379, + "grad_norm": 1.8279206985039034, + "learning_rate": 2.121440207217649e-06, + "loss": 0.348, + "step": 10298 + }, + { + "epoch": 0.7044941514467474, + "grad_norm": 2.9057787851593004, + "learning_rate": 2.1205345192904224e-06, + "loss": 0.2215, + "step": 10299 + }, + { + "epoch": 0.7045625555783569, + "grad_norm": 1.8502426191346903, + "learning_rate": 2.1196289727012692e-06, + "loss": 0.1469, + "step": 10300 + }, + { + "epoch": 0.7046309597099665, + "grad_norm": 2.6374329849967837, + "learning_rate": 2.118723567494635e-06, + "loss": 0.592, + "step": 10301 + }, + { + "epoch": 0.704699363841576, + "grad_norm": 2.9271399853383824, + "learning_rate": 2.1178183037149613e-06, + "loss": 0.5566, + "step": 10302 + }, + { + "epoch": 0.7047677679731856, + "grad_norm": 2.0239227589399458, + "learning_rate": 2.1169131814066832e-06, + "loss": 0.4402, + "step": 10303 + }, + { + "epoch": 0.7048361721047951, + "grad_norm": 2.6308703404617897, + "learning_rate": 2.116008200614229e-06, + "loss": 0.4194, + "step": 10304 + }, + { + "epoch": 0.7049045762364047, + "grad_norm": 2.2965947324032805, + "learning_rate": 2.115103361382017e-06, + "loss": 0.5044, + "step": 10305 + }, + { + "epoch": 0.7049729803680143, + "grad_norm": 2.5591639225592187, + "learning_rate": 2.114198663754465e-06, + "loss": 0.2553, + "step": 10306 + }, + { + "epoch": 0.7050413844996237, + "grad_norm": 1.584593480562749, + "learning_rate": 2.1132941077759754e-06, + "loss": 0.2232, + "step": 10307 + }, + { + "epoch": 0.7051097886312333, + "grad_norm": 2.1341526306621246, + "learning_rate": 2.112389693490953e-06, + "loss": 0.3705, + "step": 10308 + }, + { + "epoch": 0.7051781927628429, + "grad_norm": 2.4275163904043335, + "learning_rate": 2.111485420943789e-06, + "loss": 0.5986, + "step": 10309 + }, + { + "epoch": 0.7052465968944525, + "grad_norm": 2.0277901894771624, + "learning_rate": 2.1105812901788686e-06, + "loss": 0.1595, + "step": 10310 + }, + { + "epoch": 0.7053150010260619, + "grad_norm": 2.784165333175987, + "learning_rate": 2.1096773012405714e-06, + "loss": 0.3638, + "step": 10311 + }, + { + "epoch": 0.7053834051576715, + "grad_norm": 3.044119434926635, + "learning_rate": 2.1087734541732675e-06, + "loss": 0.2425, + "step": 10312 + }, + { + "epoch": 0.7054518092892811, + "grad_norm": 2.589394543385489, + "learning_rate": 2.1078697490213258e-06, + "loss": 0.2696, + "step": 10313 + }, + { + "epoch": 0.7055202134208907, + "grad_norm": 2.3172493312249087, + "learning_rate": 2.106966185829103e-06, + "loss": 0.5349, + "step": 10314 + }, + { + "epoch": 0.7055886175525001, + "grad_norm": 2.7308452930504794, + "learning_rate": 2.1060627646409503e-06, + "loss": 0.4455, + "step": 10315 + }, + { + "epoch": 0.7056570216841097, + "grad_norm": 2.2339458586988923, + "learning_rate": 2.1051594855012115e-06, + "loss": 0.4979, + "step": 10316 + }, + { + "epoch": 0.7057254258157193, + "grad_norm": 1.7815547042532922, + "learning_rate": 2.1042563484542226e-06, + "loss": 0.2064, + "step": 10317 + }, + { + "epoch": 0.7057938299473289, + "grad_norm": 1.7556223725285327, + "learning_rate": 2.1033533535443175e-06, + "loss": 0.263, + "step": 10318 + }, + { + "epoch": 0.7058622340789383, + "grad_norm": 2.07693105115168, + "learning_rate": 2.1024505008158153e-06, + "loss": 0.3463, + "step": 10319 + }, + { + "epoch": 0.7059306382105479, + "grad_norm": 2.149060741538665, + "learning_rate": 2.1015477903130367e-06, + "loss": 0.4407, + "step": 10320 + }, + { + "epoch": 0.7059990423421575, + "grad_norm": 2.158646302441213, + "learning_rate": 2.1006452220802888e-06, + "loss": 0.2906, + "step": 10321 + }, + { + "epoch": 0.7060674464737671, + "grad_norm": 2.284812718840501, + "learning_rate": 2.0997427961618746e-06, + "loss": 0.4674, + "step": 10322 + }, + { + "epoch": 0.7061358506053765, + "grad_norm": 2.075485101549404, + "learning_rate": 2.098840512602089e-06, + "loss": 0.3632, + "step": 10323 + }, + { + "epoch": 0.7062042547369861, + "grad_norm": 1.7925375875019596, + "learning_rate": 2.0979383714452202e-06, + "loss": 0.1427, + "step": 10324 + }, + { + "epoch": 0.7062726588685957, + "grad_norm": 2.0577704649088218, + "learning_rate": 2.097036372735549e-06, + "loss": 0.3866, + "step": 10325 + }, + { + "epoch": 0.7063410630002052, + "grad_norm": 2.2224532639746224, + "learning_rate": 2.096134516517352e-06, + "loss": 0.3762, + "step": 10326 + }, + { + "epoch": 0.7064094671318147, + "grad_norm": 2.037879118138315, + "learning_rate": 2.0952328028348937e-06, + "loss": 0.4207, + "step": 10327 + }, + { + "epoch": 0.7064778712634243, + "grad_norm": 2.007011198015672, + "learning_rate": 2.0943312317324386e-06, + "loss": 0.5928, + "step": 10328 + }, + { + "epoch": 0.7065462753950339, + "grad_norm": 2.1354771341896273, + "learning_rate": 2.093429803254238e-06, + "loss": 0.41, + "step": 10329 + }, + { + "epoch": 0.7066146795266434, + "grad_norm": 2.6899498984216135, + "learning_rate": 2.092528517444539e-06, + "loss": 0.6418, + "step": 10330 + }, + { + "epoch": 0.7066830836582529, + "grad_norm": 1.8238193454559712, + "learning_rate": 2.091627374347579e-06, + "loss": 0.2444, + "step": 10331 + }, + { + "epoch": 0.7067514877898625, + "grad_norm": 2.2085062617545024, + "learning_rate": 2.090726374007591e-06, + "loss": 0.5825, + "step": 10332 + }, + { + "epoch": 0.7068198919214721, + "grad_norm": 1.1362574057342945, + "learning_rate": 2.089825516468803e-06, + "loss": 0.1528, + "step": 10333 + }, + { + "epoch": 0.7068882960530816, + "grad_norm": 2.370098370138253, + "learning_rate": 2.0889248017754317e-06, + "loss": 0.4863, + "step": 10334 + }, + { + "epoch": 0.7069567001846911, + "grad_norm": 1.726163824580166, + "learning_rate": 2.088024229971688e-06, + "loss": 0.2743, + "step": 10335 + }, + { + "epoch": 0.7070251043163007, + "grad_norm": 2.4185771238414917, + "learning_rate": 2.0871238011017775e-06, + "loss": 0.4451, + "step": 10336 + }, + { + "epoch": 0.7070935084479103, + "grad_norm": 1.9550350119670312, + "learning_rate": 2.0862235152098947e-06, + "loss": 0.4194, + "step": 10337 + }, + { + "epoch": 0.7071619125795198, + "grad_norm": 1.9606902433951785, + "learning_rate": 2.0853233723402343e-06, + "loss": 0.3694, + "step": 10338 + }, + { + "epoch": 0.7072303167111293, + "grad_norm": 2.3674877026751133, + "learning_rate": 2.084423372536976e-06, + "loss": 0.2957, + "step": 10339 + }, + { + "epoch": 0.7072987208427389, + "grad_norm": 2.049610468589879, + "learning_rate": 2.0835235158442995e-06, + "loss": 0.5394, + "step": 10340 + }, + { + "epoch": 0.7073671249743485, + "grad_norm": 1.9514771032776306, + "learning_rate": 2.0826238023063726e-06, + "loss": 0.4474, + "step": 10341 + }, + { + "epoch": 0.707435529105958, + "grad_norm": 2.878692851963194, + "learning_rate": 2.0817242319673568e-06, + "loss": 0.2781, + "step": 10342 + }, + { + "epoch": 0.7075039332375676, + "grad_norm": 1.939728593423265, + "learning_rate": 2.080824804871408e-06, + "loss": 0.1913, + "step": 10343 + }, + { + "epoch": 0.7075723373691771, + "grad_norm": 2.17892251212096, + "learning_rate": 2.0799255210626744e-06, + "loss": 0.4228, + "step": 10344 + }, + { + "epoch": 0.7076407415007866, + "grad_norm": 1.9405035456984028, + "learning_rate": 2.0790263805852954e-06, + "loss": 0.3225, + "step": 10345 + }, + { + "epoch": 0.7077091456323962, + "grad_norm": 1.9626690943573797, + "learning_rate": 2.0781273834834093e-06, + "loss": 0.2524, + "step": 10346 + }, + { + "epoch": 0.7077775497640058, + "grad_norm": 2.752384008567395, + "learning_rate": 2.077228529801138e-06, + "loss": 0.3685, + "step": 10347 + }, + { + "epoch": 0.7078459538956153, + "grad_norm": 1.8814138821099449, + "learning_rate": 2.0763298195826064e-06, + "loss": 0.4034, + "step": 10348 + }, + { + "epoch": 0.7079143580272248, + "grad_norm": 2.1569188573308575, + "learning_rate": 2.0754312528719263e-06, + "loss": 0.4627, + "step": 10349 + }, + { + "epoch": 0.7079827621588344, + "grad_norm": 2.4187912720601625, + "learning_rate": 2.0745328297132024e-06, + "loss": 0.426, + "step": 10350 + }, + { + "epoch": 0.708051166290444, + "grad_norm": 1.8752375611173795, + "learning_rate": 2.0736345501505335e-06, + "loss": 0.2416, + "step": 10351 + }, + { + "epoch": 0.7081195704220535, + "grad_norm": 1.5979867280929947, + "learning_rate": 2.072736414228011e-06, + "loss": 0.2829, + "step": 10352 + }, + { + "epoch": 0.708187974553663, + "grad_norm": 1.993711083429972, + "learning_rate": 2.0718384219897226e-06, + "loss": 0.4754, + "step": 10353 + }, + { + "epoch": 0.7082563786852726, + "grad_norm": 2.138213839320293, + "learning_rate": 2.0709405734797443e-06, + "loss": 0.4876, + "step": 10354 + }, + { + "epoch": 0.7083247828168822, + "grad_norm": 2.408760829182598, + "learning_rate": 2.070042868742147e-06, + "loss": 0.67, + "step": 10355 + }, + { + "epoch": 0.7083931869484917, + "grad_norm": 2.0831614999637864, + "learning_rate": 2.0691453078209942e-06, + "loss": 0.4137, + "step": 10356 + }, + { + "epoch": 0.7084615910801012, + "grad_norm": 1.707087365282173, + "learning_rate": 2.0682478907603415e-06, + "loss": 0.2538, + "step": 10357 + }, + { + "epoch": 0.7085299952117108, + "grad_norm": 1.9460446309032133, + "learning_rate": 2.067350617604241e-06, + "loss": 0.2035, + "step": 10358 + }, + { + "epoch": 0.7085983993433204, + "grad_norm": 2.117248045136561, + "learning_rate": 2.0664534883967315e-06, + "loss": 0.5212, + "step": 10359 + }, + { + "epoch": 0.7086668034749299, + "grad_norm": 1.6030482564742, + "learning_rate": 2.065556503181853e-06, + "loss": 0.157, + "step": 10360 + }, + { + "epoch": 0.7087352076065394, + "grad_norm": 2.5416906341350307, + "learning_rate": 2.064659662003631e-06, + "loss": 0.8873, + "step": 10361 + }, + { + "epoch": 0.708803611738149, + "grad_norm": 2.231365177376749, + "learning_rate": 2.0637629649060885e-06, + "loss": 0.5152, + "step": 10362 + }, + { + "epoch": 0.7088720158697586, + "grad_norm": 2.0422836964022464, + "learning_rate": 2.062866411933238e-06, + "loss": 0.3151, + "step": 10363 + }, + { + "epoch": 0.708940420001368, + "grad_norm": 2.1629639326352086, + "learning_rate": 2.0619700031290857e-06, + "loss": 0.1635, + "step": 10364 + }, + { + "epoch": 0.7090088241329776, + "grad_norm": 2.5569438244553377, + "learning_rate": 2.061073738537635e-06, + "loss": 0.6221, + "step": 10365 + }, + { + "epoch": 0.7090772282645872, + "grad_norm": 1.7457135317772428, + "learning_rate": 2.0601776182028747e-06, + "loss": 0.2951, + "step": 10366 + }, + { + "epoch": 0.7091456323961968, + "grad_norm": 1.9218191317059563, + "learning_rate": 2.059281642168796e-06, + "loss": 0.2132, + "step": 10367 + }, + { + "epoch": 0.7092140365278062, + "grad_norm": 2.1006523203413883, + "learning_rate": 2.058385810479374e-06, + "loss": 0.4185, + "step": 10368 + }, + { + "epoch": 0.7092824406594158, + "grad_norm": 3.094324752131513, + "learning_rate": 2.057490123178582e-06, + "loss": 0.635, + "step": 10369 + }, + { + "epoch": 0.7093508447910254, + "grad_norm": 1.9554371232673045, + "learning_rate": 2.0565945803103835e-06, + "loss": 0.2269, + "step": 10370 + }, + { + "epoch": 0.709419248922635, + "grad_norm": 2.3101251625237254, + "learning_rate": 2.055699181918735e-06, + "loss": 0.2769, + "step": 10371 + }, + { + "epoch": 0.7094876530542444, + "grad_norm": 1.726591515620912, + "learning_rate": 2.0548039280475906e-06, + "loss": 0.1621, + "step": 10372 + }, + { + "epoch": 0.709556057185854, + "grad_norm": 1.8332183790110783, + "learning_rate": 2.053908818740891e-06, + "loss": 0.325, + "step": 10373 + }, + { + "epoch": 0.7096244613174636, + "grad_norm": 1.468198997545751, + "learning_rate": 2.0530138540425737e-06, + "loss": 0.1169, + "step": 10374 + }, + { + "epoch": 0.7096928654490732, + "grad_norm": 2.0951869005283696, + "learning_rate": 2.052119033996567e-06, + "loss": 0.2545, + "step": 10375 + }, + { + "epoch": 0.7097612695806826, + "grad_norm": 2.1553671591344603, + "learning_rate": 2.051224358646792e-06, + "loss": 0.2923, + "step": 10376 + }, + { + "epoch": 0.7098296737122922, + "grad_norm": 2.8973499015821385, + "learning_rate": 2.050329828037166e-06, + "loss": 0.948, + "step": 10377 + }, + { + "epoch": 0.7098980778439018, + "grad_norm": 3.164694955982199, + "learning_rate": 2.049435442211594e-06, + "loss": 0.5789, + "step": 10378 + }, + { + "epoch": 0.7099664819755113, + "grad_norm": 2.5039680044857757, + "learning_rate": 2.048541201213981e-06, + "loss": 0.6551, + "step": 10379 + }, + { + "epoch": 0.7100348861071208, + "grad_norm": 1.576594413078946, + "learning_rate": 2.047647105088218e-06, + "loss": 0.1947, + "step": 10380 + }, + { + "epoch": 0.7101032902387304, + "grad_norm": 2.1198113114073918, + "learning_rate": 2.046753153878192e-06, + "loss": 0.3923, + "step": 10381 + }, + { + "epoch": 0.71017169437034, + "grad_norm": 1.6912070811225828, + "learning_rate": 2.045859347627782e-06, + "loss": 0.2126, + "step": 10382 + }, + { + "epoch": 0.7102400985019495, + "grad_norm": 1.6367187044785523, + "learning_rate": 2.044965686380861e-06, + "loss": 0.2693, + "step": 10383 + }, + { + "epoch": 0.710308502633559, + "grad_norm": 1.7620370074130143, + "learning_rate": 2.0440721701812914e-06, + "loss": 0.2751, + "step": 10384 + }, + { + "epoch": 0.7103769067651686, + "grad_norm": 2.331531097222108, + "learning_rate": 2.0431787990729364e-06, + "loss": 0.2136, + "step": 10385 + }, + { + "epoch": 0.7104453108967782, + "grad_norm": 1.818822152387782, + "learning_rate": 2.042285573099642e-06, + "loss": 0.3857, + "step": 10386 + }, + { + "epoch": 0.7105137150283877, + "grad_norm": 1.4280673628931335, + "learning_rate": 2.0413924923052558e-06, + "loss": 0.128, + "step": 10387 + }, + { + "epoch": 0.7105821191599972, + "grad_norm": 2.9356184320617973, + "learning_rate": 2.040499556733614e-06, + "loss": 0.4029, + "step": 10388 + }, + { + "epoch": 0.7106505232916068, + "grad_norm": 1.5450730135325295, + "learning_rate": 2.0396067664285446e-06, + "loss": 0.1702, + "step": 10389 + }, + { + "epoch": 0.7107189274232164, + "grad_norm": 1.7417657067880274, + "learning_rate": 2.0387141214338705e-06, + "loss": 0.2974, + "step": 10390 + }, + { + "epoch": 0.7107873315548259, + "grad_norm": 2.386096205114016, + "learning_rate": 2.0378216217934054e-06, + "loss": 0.5983, + "step": 10391 + }, + { + "epoch": 0.7108557356864355, + "grad_norm": 1.7522775866279516, + "learning_rate": 2.036929267550962e-06, + "loss": 0.3237, + "step": 10392 + }, + { + "epoch": 0.710924139818045, + "grad_norm": 1.860639662974752, + "learning_rate": 2.0360370587503374e-06, + "loss": 0.2694, + "step": 10393 + }, + { + "epoch": 0.7109925439496546, + "grad_norm": 2.9997672547019714, + "learning_rate": 2.035144995435328e-06, + "loss": 0.2961, + "step": 10394 + }, + { + "epoch": 0.7110609480812641, + "grad_norm": 2.572540453897665, + "learning_rate": 2.034253077649719e-06, + "loss": 0.2267, + "step": 10395 + }, + { + "epoch": 0.7111293522128737, + "grad_norm": 2.0356738773974787, + "learning_rate": 2.033361305437288e-06, + "loss": 0.2701, + "step": 10396 + }, + { + "epoch": 0.7111977563444832, + "grad_norm": 1.902142649210023, + "learning_rate": 2.0324696788418118e-06, + "loss": 0.3462, + "step": 10397 + }, + { + "epoch": 0.7112661604760927, + "grad_norm": 1.801554135437974, + "learning_rate": 2.0315781979070526e-06, + "loss": 0.268, + "step": 10398 + }, + { + "epoch": 0.7113345646077023, + "grad_norm": 2.722768970055293, + "learning_rate": 2.030686862676771e-06, + "loss": 0.5181, + "step": 10399 + }, + { + "epoch": 0.7114029687393119, + "grad_norm": 1.8159319350630492, + "learning_rate": 2.0297956731947175e-06, + "loss": 0.2489, + "step": 10400 + }, + { + "epoch": 0.7114713728709214, + "grad_norm": 2.7894408346070185, + "learning_rate": 2.028904629504635e-06, + "loss": 0.3723, + "step": 10401 + }, + { + "epoch": 0.7115397770025309, + "grad_norm": 2.120100452668058, + "learning_rate": 2.0280137316502607e-06, + "loss": 0.5037, + "step": 10402 + }, + { + "epoch": 0.7116081811341405, + "grad_norm": 2.5959165555697545, + "learning_rate": 2.0271229796753238e-06, + "loss": 0.844, + "step": 10403 + }, + { + "epoch": 0.7116765852657501, + "grad_norm": 3.3117420660963304, + "learning_rate": 2.026232373623545e-06, + "loss": 0.4421, + "step": 10404 + }, + { + "epoch": 0.7117449893973596, + "grad_norm": 2.5307093855377927, + "learning_rate": 2.0253419135386437e-06, + "loss": 0.3876, + "step": 10405 + }, + { + "epoch": 0.7118133935289691, + "grad_norm": 2.1642020418093497, + "learning_rate": 2.0244515994643237e-06, + "loss": 0.3885, + "step": 10406 + }, + { + "epoch": 0.7118817976605787, + "grad_norm": 2.27616624178434, + "learning_rate": 2.02356143144429e-06, + "loss": 0.4852, + "step": 10407 + }, + { + "epoch": 0.7119502017921883, + "grad_norm": 2.470006634996746, + "learning_rate": 2.022671409522234e-06, + "loss": 0.4437, + "step": 10408 + }, + { + "epoch": 0.7120186059237978, + "grad_norm": 1.2982585260319524, + "learning_rate": 2.021781533741843e-06, + "loss": 0.1849, + "step": 10409 + }, + { + "epoch": 0.7120870100554073, + "grad_norm": 2.176066210456023, + "learning_rate": 2.020891804146796e-06, + "loss": 0.4232, + "step": 10410 + }, + { + "epoch": 0.7121554141870169, + "grad_norm": 2.453999685350412, + "learning_rate": 2.020002220780763e-06, + "loss": 0.5185, + "step": 10411 + }, + { + "epoch": 0.7122238183186265, + "grad_norm": 1.788345368095366, + "learning_rate": 2.019112783687413e-06, + "loss": 0.2869, + "step": 10412 + }, + { + "epoch": 0.712292222450236, + "grad_norm": 2.1346984461817975, + "learning_rate": 2.018223492910402e-06, + "loss": 0.4697, + "step": 10413 + }, + { + "epoch": 0.7123606265818455, + "grad_norm": 1.9922979548632052, + "learning_rate": 2.0173343484933817e-06, + "loss": 0.2847, + "step": 10414 + }, + { + "epoch": 0.7124290307134551, + "grad_norm": 2.0451658553321157, + "learning_rate": 2.016445350479994e-06, + "loss": 0.4854, + "step": 10415 + }, + { + "epoch": 0.7124974348450647, + "grad_norm": 2.212290942692521, + "learning_rate": 2.015556498913874e-06, + "loss": 0.658, + "step": 10416 + }, + { + "epoch": 0.7125658389766741, + "grad_norm": 2.1407450203766936, + "learning_rate": 2.014667793838655e-06, + "loss": 0.4278, + "step": 10417 + }, + { + "epoch": 0.7126342431082837, + "grad_norm": 3.6464499914819304, + "learning_rate": 2.013779235297955e-06, + "loss": 0.4103, + "step": 10418 + }, + { + "epoch": 0.7127026472398933, + "grad_norm": 2.3457316095939778, + "learning_rate": 2.012890823335392e-06, + "loss": 0.5758, + "step": 10419 + }, + { + "epoch": 0.7127710513715029, + "grad_norm": 1.7912433026116035, + "learning_rate": 2.012002557994572e-06, + "loss": 0.335, + "step": 10420 + }, + { + "epoch": 0.7128394555031123, + "grad_norm": 1.4444606848358026, + "learning_rate": 2.0111144393190956e-06, + "loss": 0.1391, + "step": 10421 + }, + { + "epoch": 0.7129078596347219, + "grad_norm": 2.6289155911737367, + "learning_rate": 2.0102264673525552e-06, + "loss": 0.4149, + "step": 10422 + }, + { + "epoch": 0.7129762637663315, + "grad_norm": 3.156769955174502, + "learning_rate": 2.0093386421385375e-06, + "loss": 0.2257, + "step": 10423 + }, + { + "epoch": 0.7130446678979411, + "grad_norm": 2.0623153401564016, + "learning_rate": 2.0084509637206196e-06, + "loss": 0.3782, + "step": 10424 + }, + { + "epoch": 0.7131130720295505, + "grad_norm": 1.9958737470467964, + "learning_rate": 2.007563432142376e-06, + "loss": 0.124, + "step": 10425 + }, + { + "epoch": 0.7131814761611601, + "grad_norm": 2.2986921660764534, + "learning_rate": 2.0066760474473675e-06, + "loss": 0.5965, + "step": 10426 + }, + { + "epoch": 0.7132498802927697, + "grad_norm": 2.0607506601526238, + "learning_rate": 2.005788809679155e-06, + "loss": 0.2355, + "step": 10427 + }, + { + "epoch": 0.7133182844243793, + "grad_norm": 1.8265923408700875, + "learning_rate": 2.0049017188812863e-06, + "loss": 0.2323, + "step": 10428 + }, + { + "epoch": 0.7133866885559887, + "grad_norm": 1.8139789976078655, + "learning_rate": 2.0040147750973047e-06, + "loss": 0.2838, + "step": 10429 + }, + { + "epoch": 0.7134550926875983, + "grad_norm": 1.9994024412222289, + "learning_rate": 2.003127978370746e-06, + "loss": 0.4854, + "step": 10430 + }, + { + "epoch": 0.7135234968192079, + "grad_norm": 2.1927284562976337, + "learning_rate": 2.0022413287451354e-06, + "loss": 0.2924, + "step": 10431 + }, + { + "epoch": 0.7135919009508175, + "grad_norm": 2.1052269699470973, + "learning_rate": 2.001354826263998e-06, + "loss": 0.2132, + "step": 10432 + }, + { + "epoch": 0.713660305082427, + "grad_norm": 2.1402803408077715, + "learning_rate": 2.0004684709708466e-06, + "loss": 0.2715, + "step": 10433 + }, + { + "epoch": 0.7137287092140365, + "grad_norm": 3.123019678165614, + "learning_rate": 1.999582262909187e-06, + "loss": 0.5839, + "step": 10434 + }, + { + "epoch": 0.7137971133456461, + "grad_norm": 2.9674255026550673, + "learning_rate": 1.998696202122519e-06, + "loss": 0.3656, + "step": 10435 + }, + { + "epoch": 0.7138655174772556, + "grad_norm": 1.8999754715892219, + "learning_rate": 1.9978102886543327e-06, + "loss": 0.4812, + "step": 10436 + }, + { + "epoch": 0.7139339216088652, + "grad_norm": 2.335576823786308, + "learning_rate": 1.9969245225481164e-06, + "loss": 0.1724, + "step": 10437 + }, + { + "epoch": 0.7140023257404747, + "grad_norm": 2.360497567613314, + "learning_rate": 1.996038903847345e-06, + "loss": 0.5679, + "step": 10438 + }, + { + "epoch": 0.7140707298720843, + "grad_norm": 2.6768569065081924, + "learning_rate": 1.9951534325954913e-06, + "loss": 0.4558, + "step": 10439 + }, + { + "epoch": 0.7141391340036938, + "grad_norm": 2.3765198775864063, + "learning_rate": 1.9942681088360178e-06, + "loss": 0.5337, + "step": 10440 + }, + { + "epoch": 0.7142075381353034, + "grad_norm": 1.876940684520447, + "learning_rate": 1.99338293261238e-06, + "loss": 0.4591, + "step": 10441 + }, + { + "epoch": 0.7142759422669129, + "grad_norm": 1.968040747270111, + "learning_rate": 1.9924979039680266e-06, + "loss": 0.2029, + "step": 10442 + }, + { + "epoch": 0.7143443463985225, + "grad_norm": 1.9101348204863613, + "learning_rate": 1.9916130229463994e-06, + "loss": 0.1926, + "step": 10443 + }, + { + "epoch": 0.714412750530132, + "grad_norm": 2.056012295836481, + "learning_rate": 1.9907282895909303e-06, + "loss": 0.4859, + "step": 10444 + }, + { + "epoch": 0.7144811546617416, + "grad_norm": 2.2259632843550707, + "learning_rate": 1.9898437039450497e-06, + "loss": 0.2214, + "step": 10445 + }, + { + "epoch": 0.7145495587933511, + "grad_norm": 1.402522614954617, + "learning_rate": 1.988959266052175e-06, + "loss": 0.1771, + "step": 10446 + }, + { + "epoch": 0.7146179629249607, + "grad_norm": 2.3516193459062182, + "learning_rate": 1.9880749759557213e-06, + "loss": 0.2718, + "step": 10447 + }, + { + "epoch": 0.7146863670565702, + "grad_norm": 2.7758943448270657, + "learning_rate": 1.9871908336990918e-06, + "loss": 0.2307, + "step": 10448 + }, + { + "epoch": 0.7147547711881798, + "grad_norm": 2.312304018067093, + "learning_rate": 1.9863068393256853e-06, + "loss": 0.4744, + "step": 10449 + }, + { + "epoch": 0.7148231753197893, + "grad_norm": 2.4195128149772036, + "learning_rate": 1.985422992878892e-06, + "loss": 0.5333, + "step": 10450 + }, + { + "epoch": 0.7148915794513989, + "grad_norm": 1.7071643009316815, + "learning_rate": 1.9845392944020936e-06, + "loss": 0.1691, + "step": 10451 + }, + { + "epoch": 0.7149599835830084, + "grad_norm": 1.4265207854874884, + "learning_rate": 1.98365574393867e-06, + "loss": 0.1679, + "step": 10452 + }, + { + "epoch": 0.715028387714618, + "grad_norm": 1.6274273422828405, + "learning_rate": 1.982772341531988e-06, + "loss": 0.1663, + "step": 10453 + }, + { + "epoch": 0.7150967918462275, + "grad_norm": 1.331906108627156, + "learning_rate": 1.9818890872254102e-06, + "loss": 0.1291, + "step": 10454 + }, + { + "epoch": 0.715165195977837, + "grad_norm": 2.4298116339468994, + "learning_rate": 1.98100598106229e-06, + "loss": 0.6551, + "step": 10455 + }, + { + "epoch": 0.7152336001094466, + "grad_norm": 1.938895776422024, + "learning_rate": 1.980123023085973e-06, + "loss": 0.3554, + "step": 10456 + }, + { + "epoch": 0.7153020042410562, + "grad_norm": 1.5032228725475076, + "learning_rate": 1.9792402133398026e-06, + "loss": 0.1957, + "step": 10457 + }, + { + "epoch": 0.7153704083726657, + "grad_norm": 2.3423717451760524, + "learning_rate": 1.978357551867108e-06, + "loss": 0.4557, + "step": 10458 + }, + { + "epoch": 0.7154388125042752, + "grad_norm": 2.1741480733258696, + "learning_rate": 1.9774750387112176e-06, + "loss": 0.54, + "step": 10459 + }, + { + "epoch": 0.7155072166358848, + "grad_norm": 2.456247338082422, + "learning_rate": 1.9765926739154476e-06, + "loss": 0.4228, + "step": 10460 + }, + { + "epoch": 0.7155756207674944, + "grad_norm": 2.8749175889186867, + "learning_rate": 1.97571045752311e-06, + "loss": 0.384, + "step": 10461 + }, + { + "epoch": 0.715644024899104, + "grad_norm": 1.7249273955227744, + "learning_rate": 1.9748283895775067e-06, + "loss": 0.1773, + "step": 10462 + }, + { + "epoch": 0.7157124290307134, + "grad_norm": 1.963138075077269, + "learning_rate": 1.973946470121933e-06, + "loss": 0.3148, + "step": 10463 + }, + { + "epoch": 0.715780833162323, + "grad_norm": 2.011822840395763, + "learning_rate": 1.9730646991996806e-06, + "loss": 0.2847, + "step": 10464 + }, + { + "epoch": 0.7158492372939326, + "grad_norm": 1.827845859816658, + "learning_rate": 1.9721830768540294e-06, + "loss": 0.4292, + "step": 10465 + }, + { + "epoch": 0.7159176414255422, + "grad_norm": 2.156846889864262, + "learning_rate": 1.9713016031282527e-06, + "loss": 0.253, + "step": 10466 + }, + { + "epoch": 0.7159860455571516, + "grad_norm": 1.8260204354390408, + "learning_rate": 1.970420278065621e-06, + "loss": 0.3209, + "step": 10467 + }, + { + "epoch": 0.7160544496887612, + "grad_norm": 2.080845525604902, + "learning_rate": 1.9695391017093907e-06, + "loss": 0.4118, + "step": 10468 + }, + { + "epoch": 0.7161228538203708, + "grad_norm": 2.051296830534803, + "learning_rate": 1.968658074102816e-06, + "loss": 0.3142, + "step": 10469 + }, + { + "epoch": 0.7161912579519804, + "grad_norm": 2.057029332363975, + "learning_rate": 1.967777195289139e-06, + "loss": 0.3466, + "step": 10470 + }, + { + "epoch": 0.7162596620835898, + "grad_norm": 2.1232621856693714, + "learning_rate": 1.9668964653116018e-06, + "loss": 0.4007, + "step": 10471 + }, + { + "epoch": 0.7163280662151994, + "grad_norm": 1.6732520911675162, + "learning_rate": 1.9660158842134323e-06, + "loss": 0.1975, + "step": 10472 + }, + { + "epoch": 0.716396470346809, + "grad_norm": 2.047023360139122, + "learning_rate": 1.9651354520378546e-06, + "loss": 0.3288, + "step": 10473 + }, + { + "epoch": 0.7164648744784184, + "grad_norm": 1.7674357404318013, + "learning_rate": 1.9642551688280844e-06, + "loss": 0.279, + "step": 10474 + }, + { + "epoch": 0.716533278610028, + "grad_norm": 2.9115039341903435, + "learning_rate": 1.963375034627328e-06, + "loss": 0.4517, + "step": 10475 + }, + { + "epoch": 0.7166016827416376, + "grad_norm": 2.1845213637753447, + "learning_rate": 1.9624950494787896e-06, + "loss": 0.5136, + "step": 10476 + }, + { + "epoch": 0.7166700868732472, + "grad_norm": 2.321517618536001, + "learning_rate": 1.9616152134256616e-06, + "loss": 0.6365, + "step": 10477 + }, + { + "epoch": 0.7167384910048566, + "grad_norm": 1.5768538681213984, + "learning_rate": 1.9607355265111323e-06, + "loss": 0.1952, + "step": 10478 + }, + { + "epoch": 0.7168068951364662, + "grad_norm": 1.8473648099514361, + "learning_rate": 1.9598559887783797e-06, + "loss": 0.3528, + "step": 10479 + }, + { + "epoch": 0.7168752992680758, + "grad_norm": 2.4654386742491607, + "learning_rate": 1.958976600270576e-06, + "loss": 0.6426, + "step": 10480 + }, + { + "epoch": 0.7169437033996854, + "grad_norm": 1.6821622670800618, + "learning_rate": 1.9580973610308858e-06, + "loss": 0.2251, + "step": 10481 + }, + { + "epoch": 0.7170121075312949, + "grad_norm": 2.187050848217138, + "learning_rate": 1.9572182711024664e-06, + "loss": 0.3469, + "step": 10482 + }, + { + "epoch": 0.7170805116629044, + "grad_norm": 2.298505301807771, + "learning_rate": 1.9563393305284655e-06, + "loss": 0.416, + "step": 10483 + }, + { + "epoch": 0.717148915794514, + "grad_norm": 2.141203593281771, + "learning_rate": 1.9554605393520297e-06, + "loss": 0.3806, + "step": 10484 + }, + { + "epoch": 0.7172173199261236, + "grad_norm": 1.981794455325235, + "learning_rate": 1.954581897616291e-06, + "loss": 0.4086, + "step": 10485 + }, + { + "epoch": 0.717285724057733, + "grad_norm": 2.1699611066264106, + "learning_rate": 1.9537034053643813e-06, + "loss": 0.3906, + "step": 10486 + }, + { + "epoch": 0.7173541281893426, + "grad_norm": 2.098807574806445, + "learning_rate": 1.952825062639418e-06, + "loss": 0.2935, + "step": 10487 + }, + { + "epoch": 0.7174225323209522, + "grad_norm": 2.423984550205337, + "learning_rate": 1.9519468694845155e-06, + "loss": 0.3698, + "step": 10488 + }, + { + "epoch": 0.7174909364525618, + "grad_norm": 1.9288590965826713, + "learning_rate": 1.95106882594278e-06, + "loss": 0.2375, + "step": 10489 + }, + { + "epoch": 0.7175593405841713, + "grad_norm": 1.990110241853143, + "learning_rate": 1.9501909320573074e-06, + "loss": 0.302, + "step": 10490 + }, + { + "epoch": 0.7176277447157808, + "grad_norm": 2.2517048681544765, + "learning_rate": 1.949313187871193e-06, + "loss": 0.3293, + "step": 10491 + }, + { + "epoch": 0.7176961488473904, + "grad_norm": 1.775337314959398, + "learning_rate": 1.9484355934275197e-06, + "loss": 0.3089, + "step": 10492 + }, + { + "epoch": 0.7177645529789999, + "grad_norm": 2.1358828853665446, + "learning_rate": 1.947558148769362e-06, + "loss": 0.3853, + "step": 10493 + }, + { + "epoch": 0.7178329571106095, + "grad_norm": 1.8675563799021613, + "learning_rate": 1.9466808539397913e-06, + "loss": 0.2856, + "step": 10494 + }, + { + "epoch": 0.717901361242219, + "grad_norm": 2.591256505752839, + "learning_rate": 1.945803708981867e-06, + "loss": 0.3927, + "step": 10495 + }, + { + "epoch": 0.7179697653738286, + "grad_norm": 1.9636785529474428, + "learning_rate": 1.9449267139386468e-06, + "loss": 0.3456, + "step": 10496 + }, + { + "epoch": 0.7180381695054381, + "grad_norm": 2.6319763372619964, + "learning_rate": 1.9440498688531744e-06, + "loss": 0.5994, + "step": 10497 + }, + { + "epoch": 0.7181065736370477, + "grad_norm": 3.795782607116821, + "learning_rate": 1.9431731737684934e-06, + "loss": 0.2321, + "step": 10498 + }, + { + "epoch": 0.7181749777686572, + "grad_norm": 2.266904870628692, + "learning_rate": 1.942296628727634e-06, + "loss": 0.2807, + "step": 10499 + }, + { + "epoch": 0.7182433819002668, + "grad_norm": 1.9555642955449555, + "learning_rate": 1.941420233773622e-06, + "loss": 0.3566, + "step": 10500 + }, + { + "epoch": 0.7183117860318763, + "grad_norm": 2.3878086180219182, + "learning_rate": 1.940543988949474e-06, + "loss": 0.4259, + "step": 10501 + }, + { + "epoch": 0.7183801901634859, + "grad_norm": 2.3893376886117315, + "learning_rate": 1.939667894298202e-06, + "loss": 0.5388, + "step": 10502 + }, + { + "epoch": 0.7184485942950954, + "grad_norm": 2.849479389951838, + "learning_rate": 1.9387919498628054e-06, + "loss": 0.292, + "step": 10503 + }, + { + "epoch": 0.718516998426705, + "grad_norm": 2.0802918122901657, + "learning_rate": 1.9379161556862843e-06, + "loss": 0.4472, + "step": 10504 + }, + { + "epoch": 0.7185854025583145, + "grad_norm": 1.5170058419624954, + "learning_rate": 1.937040511811623e-06, + "loss": 0.1369, + "step": 10505 + }, + { + "epoch": 0.7186538066899241, + "grad_norm": 2.439896420347905, + "learning_rate": 1.936165018281806e-06, + "loss": 0.4439, + "step": 10506 + }, + { + "epoch": 0.7187222108215336, + "grad_norm": 1.4060295620970606, + "learning_rate": 1.935289675139805e-06, + "loss": 0.154, + "step": 10507 + }, + { + "epoch": 0.7187906149531431, + "grad_norm": 1.6940554823709844, + "learning_rate": 1.9344144824285867e-06, + "loss": 0.3317, + "step": 10508 + }, + { + "epoch": 0.7188590190847527, + "grad_norm": 2.073816296315046, + "learning_rate": 1.9335394401911085e-06, + "loss": 0.5242, + "step": 10509 + }, + { + "epoch": 0.7189274232163623, + "grad_norm": 1.5273410280013158, + "learning_rate": 1.932664548470321e-06, + "loss": 0.1177, + "step": 10510 + }, + { + "epoch": 0.7189958273479718, + "grad_norm": 1.5770952812189054, + "learning_rate": 1.9317898073091707e-06, + "loss": 0.1338, + "step": 10511 + }, + { + "epoch": 0.7190642314795813, + "grad_norm": 2.2384164113616127, + "learning_rate": 1.9309152167505933e-06, + "loss": 0.3758, + "step": 10512 + }, + { + "epoch": 0.7191326356111909, + "grad_norm": 2.5924195588708545, + "learning_rate": 1.9300407768375172e-06, + "loss": 0.5782, + "step": 10513 + }, + { + "epoch": 0.7192010397428005, + "grad_norm": 2.0159539686153294, + "learning_rate": 1.9291664876128654e-06, + "loss": 0.2983, + "step": 10514 + }, + { + "epoch": 0.71926944387441, + "grad_norm": 2.0845832286038912, + "learning_rate": 1.928292349119549e-06, + "loss": 0.3528, + "step": 10515 + }, + { + "epoch": 0.7193378480060195, + "grad_norm": 2.6288063404432545, + "learning_rate": 1.9274183614004784e-06, + "loss": 0.5426, + "step": 10516 + }, + { + "epoch": 0.7194062521376291, + "grad_norm": 2.4159542806819982, + "learning_rate": 1.9265445244985503e-06, + "loss": 0.1922, + "step": 10517 + }, + { + "epoch": 0.7194746562692387, + "grad_norm": 1.9733957811779939, + "learning_rate": 1.925670838456661e-06, + "loss": 0.2128, + "step": 10518 + }, + { + "epoch": 0.7195430604008483, + "grad_norm": 4.064724569866511, + "learning_rate": 1.924797303317692e-06, + "loss": 0.6442, + "step": 10519 + }, + { + "epoch": 0.7196114645324577, + "grad_norm": 2.5316314527583526, + "learning_rate": 1.923923919124522e-06, + "loss": 0.7231, + "step": 10520 + }, + { + "epoch": 0.7196798686640673, + "grad_norm": 2.012719588709411, + "learning_rate": 1.9230506859200197e-06, + "loss": 0.2759, + "step": 10521 + }, + { + "epoch": 0.7197482727956769, + "grad_norm": 1.9299837340647856, + "learning_rate": 1.9221776037470485e-06, + "loss": 0.291, + "step": 10522 + }, + { + "epoch": 0.7198166769272865, + "grad_norm": 2.179507969398219, + "learning_rate": 1.9213046726484615e-06, + "loss": 0.2963, + "step": 10523 + }, + { + "epoch": 0.7198850810588959, + "grad_norm": 1.6365070706869467, + "learning_rate": 1.9204318926671095e-06, + "loss": 0.3334, + "step": 10524 + }, + { + "epoch": 0.7199534851905055, + "grad_norm": 2.3390519389727626, + "learning_rate": 1.91955926384583e-06, + "loss": 0.4557, + "step": 10525 + }, + { + "epoch": 0.7200218893221151, + "grad_norm": 4.614353245954246, + "learning_rate": 1.9186867862274585e-06, + "loss": 0.2252, + "step": 10526 + }, + { + "epoch": 0.7200902934537246, + "grad_norm": 2.714562754188304, + "learning_rate": 1.9178144598548186e-06, + "loss": 0.5019, + "step": 10527 + }, + { + "epoch": 0.7201586975853341, + "grad_norm": 1.985284041397853, + "learning_rate": 1.9169422847707296e-06, + "loss": 0.186, + "step": 10528 + }, + { + "epoch": 0.7202271017169437, + "grad_norm": 2.505324350447747, + "learning_rate": 1.9160702610180005e-06, + "loss": 0.4155, + "step": 10529 + }, + { + "epoch": 0.7202955058485533, + "grad_norm": 1.6040224048122624, + "learning_rate": 1.9151983886394336e-06, + "loss": 0.1489, + "step": 10530 + }, + { + "epoch": 0.7203639099801628, + "grad_norm": 2.6981630593293677, + "learning_rate": 1.9143266676778283e-06, + "loss": 0.3338, + "step": 10531 + }, + { + "epoch": 0.7204323141117723, + "grad_norm": 2.2671391590414722, + "learning_rate": 1.91345509817597e-06, + "loss": 0.494, + "step": 10532 + }, + { + "epoch": 0.7205007182433819, + "grad_norm": 1.8730063886513133, + "learning_rate": 1.9125836801766413e-06, + "loss": 0.3611, + "step": 10533 + }, + { + "epoch": 0.7205691223749915, + "grad_norm": 2.167778721634621, + "learning_rate": 1.9117124137226134e-06, + "loss": 0.2805, + "step": 10534 + }, + { + "epoch": 0.720637526506601, + "grad_norm": 2.3744558225403085, + "learning_rate": 1.9108412988566528e-06, + "loss": 0.3584, + "step": 10535 + }, + { + "epoch": 0.7207059306382105, + "grad_norm": 2.886239049386009, + "learning_rate": 1.9099703356215193e-06, + "loss": 0.3663, + "step": 10536 + }, + { + "epoch": 0.7207743347698201, + "grad_norm": 2.381387315630673, + "learning_rate": 1.9090995240599623e-06, + "loss": 0.3269, + "step": 10537 + }, + { + "epoch": 0.7208427389014297, + "grad_norm": 2.1071713981935356, + "learning_rate": 1.908228864214728e-06, + "loss": 0.4005, + "step": 10538 + }, + { + "epoch": 0.7209111430330392, + "grad_norm": 2.483127124650086, + "learning_rate": 1.9073583561285507e-06, + "loss": 0.6159, + "step": 10539 + }, + { + "epoch": 0.7209795471646487, + "grad_norm": 1.760300418487919, + "learning_rate": 1.9064879998441599e-06, + "loss": 0.2661, + "step": 10540 + }, + { + "epoch": 0.7210479512962583, + "grad_norm": 1.9252378691437353, + "learning_rate": 1.9056177954042764e-06, + "loss": 0.1877, + "step": 10541 + }, + { + "epoch": 0.7211163554278679, + "grad_norm": 2.3572454062516313, + "learning_rate": 1.9047477428516143e-06, + "loss": 0.348, + "step": 10542 + }, + { + "epoch": 0.7211847595594774, + "grad_norm": 2.3845680523634436, + "learning_rate": 1.9038778422288778e-06, + "loss": 0.6273, + "step": 10543 + }, + { + "epoch": 0.7212531636910869, + "grad_norm": 2.3963320807307196, + "learning_rate": 1.9030080935787698e-06, + "loss": 0.5697, + "step": 10544 + }, + { + "epoch": 0.7213215678226965, + "grad_norm": 3.217825108577806, + "learning_rate": 1.9021384969439776e-06, + "loss": 0.5595, + "step": 10545 + }, + { + "epoch": 0.721389971954306, + "grad_norm": 2.2311443768673485, + "learning_rate": 1.9012690523671895e-06, + "loss": 0.4566, + "step": 10546 + }, + { + "epoch": 0.7214583760859156, + "grad_norm": 2.924388722181482, + "learning_rate": 1.90039975989108e-06, + "loss": 0.4152, + "step": 10547 + }, + { + "epoch": 0.7215267802175251, + "grad_norm": 2.259882118631563, + "learning_rate": 1.899530619558318e-06, + "loss": 0.2337, + "step": 10548 + }, + { + "epoch": 0.7215951843491347, + "grad_norm": 2.779848216976676, + "learning_rate": 1.8986616314115652e-06, + "loss": 0.3979, + "step": 10549 + }, + { + "epoch": 0.7216635884807442, + "grad_norm": 1.8554578720577, + "learning_rate": 1.8977927954934739e-06, + "loss": 0.3312, + "step": 10550 + }, + { + "epoch": 0.7217319926123538, + "grad_norm": 2.228092437602507, + "learning_rate": 1.8969241118466947e-06, + "loss": 0.513, + "step": 10551 + }, + { + "epoch": 0.7218003967439633, + "grad_norm": 2.2405912818393383, + "learning_rate": 1.8960555805138641e-06, + "loss": 0.538, + "step": 10552 + }, + { + "epoch": 0.7218688008755729, + "grad_norm": 2.320343004533509, + "learning_rate": 1.895187201537615e-06, + "loss": 0.483, + "step": 10553 + }, + { + "epoch": 0.7219372050071824, + "grad_norm": 5.112998459167218, + "learning_rate": 1.8943189749605705e-06, + "loss": 0.4654, + "step": 10554 + }, + { + "epoch": 0.722005609138792, + "grad_norm": 2.2991618948830834, + "learning_rate": 1.8934509008253465e-06, + "loss": 0.6336, + "step": 10555 + }, + { + "epoch": 0.7220740132704015, + "grad_norm": 1.7895086889794125, + "learning_rate": 1.8925829791745554e-06, + "loss": 0.2527, + "step": 10556 + }, + { + "epoch": 0.7221424174020111, + "grad_norm": 1.6112722814882274, + "learning_rate": 1.8917152100507952e-06, + "loss": 0.3446, + "step": 10557 + }, + { + "epoch": 0.7222108215336206, + "grad_norm": 2.2639620564506986, + "learning_rate": 1.890847593496664e-06, + "loss": 0.2939, + "step": 10558 + }, + { + "epoch": 0.7222792256652302, + "grad_norm": 1.9985946657597555, + "learning_rate": 1.8899801295547476e-06, + "loss": 0.1716, + "step": 10559 + }, + { + "epoch": 0.7223476297968398, + "grad_norm": 2.259076345525041, + "learning_rate": 1.8891128182676238e-06, + "loss": 0.4457, + "step": 10560 + }, + { + "epoch": 0.7224160339284493, + "grad_norm": 2.416735252826667, + "learning_rate": 1.8882456596778653e-06, + "loss": 0.4174, + "step": 10561 + }, + { + "epoch": 0.7224844380600588, + "grad_norm": 2.1543672468877584, + "learning_rate": 1.8873786538280347e-06, + "loss": 0.6449, + "step": 10562 + }, + { + "epoch": 0.7225528421916684, + "grad_norm": 2.286605544916138, + "learning_rate": 1.8865118007606924e-06, + "loss": 0.4671, + "step": 10563 + }, + { + "epoch": 0.722621246323278, + "grad_norm": 2.4619269001863233, + "learning_rate": 1.8856451005183858e-06, + "loss": 0.3185, + "step": 10564 + }, + { + "epoch": 0.7226896504548874, + "grad_norm": 2.5035313818107316, + "learning_rate": 1.8847785531436546e-06, + "loss": 0.4521, + "step": 10565 + }, + { + "epoch": 0.722758054586497, + "grad_norm": 2.34618745261662, + "learning_rate": 1.883912158679037e-06, + "loss": 0.5275, + "step": 10566 + }, + { + "epoch": 0.7228264587181066, + "grad_norm": 1.847644995399054, + "learning_rate": 1.8830459171670584e-06, + "loss": 0.27, + "step": 10567 + }, + { + "epoch": 0.7228948628497162, + "grad_norm": 2.276580407799407, + "learning_rate": 1.882179828650238e-06, + "loss": 0.4491, + "step": 10568 + }, + { + "epoch": 0.7229632669813256, + "grad_norm": 2.094772950127855, + "learning_rate": 1.881313893171085e-06, + "loss": 0.2932, + "step": 10569 + }, + { + "epoch": 0.7230316711129352, + "grad_norm": 2.279744252262594, + "learning_rate": 1.8804481107721085e-06, + "loss": 0.2113, + "step": 10570 + }, + { + "epoch": 0.7231000752445448, + "grad_norm": 2.4166967592887003, + "learning_rate": 1.8795824814958025e-06, + "loss": 0.3961, + "step": 10571 + }, + { + "epoch": 0.7231684793761544, + "grad_norm": 2.3556035890463947, + "learning_rate": 1.8787170053846566e-06, + "loss": 0.2786, + "step": 10572 + }, + { + "epoch": 0.7232368835077638, + "grad_norm": 2.420114738900894, + "learning_rate": 1.8778516824811527e-06, + "loss": 0.5231, + "step": 10573 + }, + { + "epoch": 0.7233052876393734, + "grad_norm": 1.5844469441589146, + "learning_rate": 1.8769865128277638e-06, + "loss": 0.2666, + "step": 10574 + }, + { + "epoch": 0.723373691770983, + "grad_norm": 2.578193524923843, + "learning_rate": 1.8761214964669582e-06, + "loss": 0.2981, + "step": 10575 + }, + { + "epoch": 0.7234420959025926, + "grad_norm": 2.61949231590692, + "learning_rate": 1.875256633441196e-06, + "loss": 0.376, + "step": 10576 + }, + { + "epoch": 0.723510500034202, + "grad_norm": 1.5517523821285002, + "learning_rate": 1.8743919237929247e-06, + "loss": 0.1675, + "step": 10577 + }, + { + "epoch": 0.7235789041658116, + "grad_norm": 1.903570887755453, + "learning_rate": 1.8735273675645932e-06, + "loss": 0.2736, + "step": 10578 + }, + { + "epoch": 0.7236473082974212, + "grad_norm": 2.3863223257039627, + "learning_rate": 1.872662964798636e-06, + "loss": 0.5814, + "step": 10579 + }, + { + "epoch": 0.7237157124290308, + "grad_norm": 2.181871611986723, + "learning_rate": 1.8717987155374828e-06, + "loss": 0.4323, + "step": 10580 + }, + { + "epoch": 0.7237841165606402, + "grad_norm": 1.835492412914379, + "learning_rate": 1.8709346198235539e-06, + "loss": 0.3822, + "step": 10581 + }, + { + "epoch": 0.7238525206922498, + "grad_norm": 2.095048466203759, + "learning_rate": 1.8700706776992627e-06, + "loss": 0.3113, + "step": 10582 + }, + { + "epoch": 0.7239209248238594, + "grad_norm": 2.3518912253528486, + "learning_rate": 1.8692068892070182e-06, + "loss": 0.1987, + "step": 10583 + }, + { + "epoch": 0.7239893289554689, + "grad_norm": 2.3020095978805246, + "learning_rate": 1.8683432543892167e-06, + "loss": 0.5843, + "step": 10584 + }, + { + "epoch": 0.7240577330870784, + "grad_norm": 2.0743542248006515, + "learning_rate": 1.8674797732882522e-06, + "loss": 0.3018, + "step": 10585 + }, + { + "epoch": 0.724126137218688, + "grad_norm": 2.264325670383586, + "learning_rate": 1.8666164459465075e-06, + "loss": 0.3837, + "step": 10586 + }, + { + "epoch": 0.7241945413502976, + "grad_norm": 1.70999504939835, + "learning_rate": 1.8657532724063587e-06, + "loss": 0.3335, + "step": 10587 + }, + { + "epoch": 0.7242629454819071, + "grad_norm": 2.1050374822609403, + "learning_rate": 1.8648902527101742e-06, + "loss": 0.3913, + "step": 10588 + }, + { + "epoch": 0.7243313496135166, + "grad_norm": 2.448125056602043, + "learning_rate": 1.864027386900314e-06, + "loss": 0.4106, + "step": 10589 + }, + { + "epoch": 0.7243997537451262, + "grad_norm": 1.784849015222875, + "learning_rate": 1.8631646750191346e-06, + "loss": 0.3401, + "step": 10590 + }, + { + "epoch": 0.7244681578767358, + "grad_norm": 2.3664627224621513, + "learning_rate": 1.86230211710898e-06, + "loss": 0.2879, + "step": 10591 + }, + { + "epoch": 0.7245365620083453, + "grad_norm": 2.2748183942977787, + "learning_rate": 1.8614397132121903e-06, + "loss": 0.3175, + "step": 10592 + }, + { + "epoch": 0.7246049661399548, + "grad_norm": 2.781798222977221, + "learning_rate": 1.860577463371095e-06, + "loss": 0.3256, + "step": 10593 + }, + { + "epoch": 0.7246733702715644, + "grad_norm": 2.2044749689430976, + "learning_rate": 1.8597153676280162e-06, + "loss": 0.3485, + "step": 10594 + }, + { + "epoch": 0.724741774403174, + "grad_norm": 1.8039663987934498, + "learning_rate": 1.8588534260252726e-06, + "loss": 0.2142, + "step": 10595 + }, + { + "epoch": 0.7248101785347835, + "grad_norm": 1.937072583707064, + "learning_rate": 1.8579916386051699e-06, + "loss": 0.2972, + "step": 10596 + }, + { + "epoch": 0.724878582666393, + "grad_norm": 2.0109846571970587, + "learning_rate": 1.8571300054100123e-06, + "loss": 0.327, + "step": 10597 + }, + { + "epoch": 0.7249469867980026, + "grad_norm": 1.5541904987326678, + "learning_rate": 1.8562685264820906e-06, + "loss": 0.1531, + "step": 10598 + }, + { + "epoch": 0.7250153909296122, + "grad_norm": 1.776408995586726, + "learning_rate": 1.8554072018636903e-06, + "loss": 0.1517, + "step": 10599 + }, + { + "epoch": 0.7250837950612217, + "grad_norm": 2.258468136746543, + "learning_rate": 1.8545460315970902e-06, + "loss": 0.1734, + "step": 10600 + }, + { + "epoch": 0.7251521991928312, + "grad_norm": 1.4479671533505194, + "learning_rate": 1.8536850157245595e-06, + "loss": 0.1539, + "step": 10601 + }, + { + "epoch": 0.7252206033244408, + "grad_norm": 1.9588993848960448, + "learning_rate": 1.8528241542883601e-06, + "loss": 0.582, + "step": 10602 + }, + { + "epoch": 0.7252890074560503, + "grad_norm": 2.411813027405441, + "learning_rate": 1.8519634473307508e-06, + "loss": 0.679, + "step": 10603 + }, + { + "epoch": 0.7253574115876599, + "grad_norm": 2.4175656896343174, + "learning_rate": 1.8511028948939752e-06, + "loss": 0.4246, + "step": 10604 + }, + { + "epoch": 0.7254258157192695, + "grad_norm": 1.9000328409343223, + "learning_rate": 1.850242497020277e-06, + "loss": 0.2213, + "step": 10605 + }, + { + "epoch": 0.725494219850879, + "grad_norm": 2.4788304963846395, + "learning_rate": 1.8493822537518874e-06, + "loss": 0.3419, + "step": 10606 + }, + { + "epoch": 0.7255626239824885, + "grad_norm": 1.7384996512722337, + "learning_rate": 1.848522165131031e-06, + "loss": 0.3923, + "step": 10607 + }, + { + "epoch": 0.7256310281140981, + "grad_norm": 1.9441118786637055, + "learning_rate": 1.8476622311999243e-06, + "loss": 0.4001, + "step": 10608 + }, + { + "epoch": 0.7256994322457077, + "grad_norm": 2.192464814213678, + "learning_rate": 1.8468024520007766e-06, + "loss": 0.3181, + "step": 10609 + }, + { + "epoch": 0.7257678363773172, + "grad_norm": 2.5720892507554143, + "learning_rate": 1.8459428275757929e-06, + "loss": 0.6991, + "step": 10610 + }, + { + "epoch": 0.7258362405089267, + "grad_norm": 2.53038258466209, + "learning_rate": 1.8450833579671662e-06, + "loss": 0.4746, + "step": 10611 + }, + { + "epoch": 0.7259046446405363, + "grad_norm": 1.9186989895504474, + "learning_rate": 1.8442240432170832e-06, + "loss": 0.3938, + "step": 10612 + }, + { + "epoch": 0.7259730487721459, + "grad_norm": 2.2674845594870896, + "learning_rate": 1.843364883367723e-06, + "loss": 0.4534, + "step": 10613 + }, + { + "epoch": 0.7260414529037554, + "grad_norm": 2.0467920213725734, + "learning_rate": 1.842505878461256e-06, + "loss": 0.2281, + "step": 10614 + }, + { + "epoch": 0.7261098570353649, + "grad_norm": 1.6017677826856598, + "learning_rate": 1.841647028539849e-06, + "loss": 0.2804, + "step": 10615 + }, + { + "epoch": 0.7261782611669745, + "grad_norm": 1.5363764875778931, + "learning_rate": 1.8407883336456567e-06, + "loss": 0.2036, + "step": 10616 + }, + { + "epoch": 0.7262466652985841, + "grad_norm": 1.9253167526552315, + "learning_rate": 1.83992979382083e-06, + "loss": 0.4061, + "step": 10617 + }, + { + "epoch": 0.7263150694301936, + "grad_norm": 1.8529061741471433, + "learning_rate": 1.8390714091075084e-06, + "loss": 0.4179, + "step": 10618 + }, + { + "epoch": 0.7263834735618031, + "grad_norm": 2.3491151437345077, + "learning_rate": 1.8382131795478265e-06, + "loss": 0.5927, + "step": 10619 + }, + { + "epoch": 0.7264518776934127, + "grad_norm": 2.249735922681095, + "learning_rate": 1.8373551051839096e-06, + "loss": 0.5706, + "step": 10620 + }, + { + "epoch": 0.7265202818250223, + "grad_norm": 1.826456237521412, + "learning_rate": 1.8364971860578761e-06, + "loss": 0.2316, + "step": 10621 + }, + { + "epoch": 0.7265886859566317, + "grad_norm": 2.331228018853923, + "learning_rate": 1.8356394222118363e-06, + "loss": 0.4545, + "step": 10622 + }, + { + "epoch": 0.7266570900882413, + "grad_norm": 2.4534859968733853, + "learning_rate": 1.8347818136878947e-06, + "loss": 0.5371, + "step": 10623 + }, + { + "epoch": 0.7267254942198509, + "grad_norm": 1.78337842324072, + "learning_rate": 1.8339243605281454e-06, + "loss": 0.2552, + "step": 10624 + }, + { + "epoch": 0.7267938983514605, + "grad_norm": 1.5169030557347611, + "learning_rate": 1.8330670627746794e-06, + "loss": 0.2625, + "step": 10625 + }, + { + "epoch": 0.7268623024830699, + "grad_norm": 2.0717190409055304, + "learning_rate": 1.8322099204695743e-06, + "loss": 0.2991, + "step": 10626 + }, + { + "epoch": 0.7269307066146795, + "grad_norm": 2.181598809779821, + "learning_rate": 1.8313529336549034e-06, + "loss": 0.45, + "step": 10627 + }, + { + "epoch": 0.7269991107462891, + "grad_norm": 2.4082731807570834, + "learning_rate": 1.830496102372732e-06, + "loss": 0.1558, + "step": 10628 + }, + { + "epoch": 0.7270675148778987, + "grad_norm": 1.8741012962591261, + "learning_rate": 1.8296394266651158e-06, + "loss": 0.2325, + "step": 10629 + }, + { + "epoch": 0.7271359190095081, + "grad_norm": 1.916513905104291, + "learning_rate": 1.8287829065741069e-06, + "loss": 0.2532, + "step": 10630 + }, + { + "epoch": 0.7272043231411177, + "grad_norm": 2.133294802059457, + "learning_rate": 1.8279265421417474e-06, + "loss": 0.4559, + "step": 10631 + }, + { + "epoch": 0.7272727272727273, + "grad_norm": 2.186342369034553, + "learning_rate": 1.827070333410071e-06, + "loss": 0.2873, + "step": 10632 + }, + { + "epoch": 0.7273411314043369, + "grad_norm": 2.321926775595062, + "learning_rate": 1.8262142804211042e-06, + "loss": 0.271, + "step": 10633 + }, + { + "epoch": 0.7274095355359463, + "grad_norm": 1.8667846987403633, + "learning_rate": 1.8253583832168648e-06, + "loss": 0.374, + "step": 10634 + }, + { + "epoch": 0.7274779396675559, + "grad_norm": 2.704781691631662, + "learning_rate": 1.824502641839368e-06, + "loss": 0.3494, + "step": 10635 + }, + { + "epoch": 0.7275463437991655, + "grad_norm": 2.434651660745989, + "learning_rate": 1.8236470563306141e-06, + "loss": 0.322, + "step": 10636 + }, + { + "epoch": 0.7276147479307751, + "grad_norm": 2.1115308976528993, + "learning_rate": 1.8227916267326029e-06, + "loss": 0.5202, + "step": 10637 + }, + { + "epoch": 0.7276831520623845, + "grad_norm": 2.3705692182415175, + "learning_rate": 1.8219363530873214e-06, + "loss": 0.3273, + "step": 10638 + }, + { + "epoch": 0.7277515561939941, + "grad_norm": 2.0471805679304085, + "learning_rate": 1.8210812354367501e-06, + "loss": 0.2291, + "step": 10639 + }, + { + "epoch": 0.7278199603256037, + "grad_norm": 6.094777631920061, + "learning_rate": 1.8202262738228627e-06, + "loss": 0.2871, + "step": 10640 + }, + { + "epoch": 0.7278883644572132, + "grad_norm": 2.1159612898305444, + "learning_rate": 1.819371468287625e-06, + "loss": 0.4448, + "step": 10641 + }, + { + "epoch": 0.7279567685888227, + "grad_norm": 2.6162213833989445, + "learning_rate": 1.8185168188729929e-06, + "loss": 0.3337, + "step": 10642 + }, + { + "epoch": 0.7280251727204323, + "grad_norm": 1.90681631243632, + "learning_rate": 1.8176623256209203e-06, + "loss": 0.2731, + "step": 10643 + }, + { + "epoch": 0.7280935768520419, + "grad_norm": 2.307076241028365, + "learning_rate": 1.8168079885733459e-06, + "loss": 0.3605, + "step": 10644 + }, + { + "epoch": 0.7281619809836514, + "grad_norm": 2.6075034882420645, + "learning_rate": 1.815953807772209e-06, + "loss": 0.4286, + "step": 10645 + }, + { + "epoch": 0.728230385115261, + "grad_norm": 2.5771057310392864, + "learning_rate": 1.8150997832594347e-06, + "loss": 0.2704, + "step": 10646 + }, + { + "epoch": 0.7282987892468705, + "grad_norm": 2.315334593237173, + "learning_rate": 1.8142459150769431e-06, + "loss": 0.4584, + "step": 10647 + }, + { + "epoch": 0.7283671933784801, + "grad_norm": 2.5038613063926665, + "learning_rate": 1.8133922032666456e-06, + "loss": 0.5459, + "step": 10648 + }, + { + "epoch": 0.7284355975100896, + "grad_norm": 3.2602956168440453, + "learning_rate": 1.8125386478704449e-06, + "loss": 0.1801, + "step": 10649 + }, + { + "epoch": 0.7285040016416992, + "grad_norm": 1.8541763040652894, + "learning_rate": 1.8116852489302417e-06, + "loss": 0.324, + "step": 10650 + }, + { + "epoch": 0.7285724057733087, + "grad_norm": 2.7408539227317066, + "learning_rate": 1.810832006487922e-06, + "loss": 0.3964, + "step": 10651 + }, + { + "epoch": 0.7286408099049183, + "grad_norm": 1.9889764788142108, + "learning_rate": 1.809978920585368e-06, + "loss": 0.3019, + "step": 10652 + }, + { + "epoch": 0.7287092140365278, + "grad_norm": 2.2657264094841407, + "learning_rate": 1.8091259912644531e-06, + "loss": 0.3612, + "step": 10653 + }, + { + "epoch": 0.7287776181681374, + "grad_norm": 1.621859087158521, + "learning_rate": 1.8082732185670415e-06, + "loss": 0.3329, + "step": 10654 + }, + { + "epoch": 0.7288460222997469, + "grad_norm": 2.219380463765583, + "learning_rate": 1.807420602534995e-06, + "loss": 0.5032, + "step": 10655 + }, + { + "epoch": 0.7289144264313564, + "grad_norm": 2.021048464745138, + "learning_rate": 1.80656814321016e-06, + "loss": 0.3843, + "step": 10656 + }, + { + "epoch": 0.728982830562966, + "grad_norm": 2.1354047860226, + "learning_rate": 1.8057158406343834e-06, + "loss": 0.4832, + "step": 10657 + }, + { + "epoch": 0.7290512346945756, + "grad_norm": 1.6286002496045988, + "learning_rate": 1.804863694849498e-06, + "loss": 0.2681, + "step": 10658 + }, + { + "epoch": 0.7291196388261851, + "grad_norm": 2.8978054869052814, + "learning_rate": 1.8040117058973317e-06, + "loss": 0.4417, + "step": 10659 + }, + { + "epoch": 0.7291880429577946, + "grad_norm": 1.510302957551608, + "learning_rate": 1.8031598738197042e-06, + "loss": 0.1353, + "step": 10660 + }, + { + "epoch": 0.7292564470894042, + "grad_norm": 1.9297300635211265, + "learning_rate": 1.8023081986584262e-06, + "loss": 0.3096, + "step": 10661 + }, + { + "epoch": 0.7293248512210138, + "grad_norm": 2.41900966714226, + "learning_rate": 1.8014566804553046e-06, + "loss": 0.5332, + "step": 10662 + }, + { + "epoch": 0.7293932553526233, + "grad_norm": 1.5617034002048749, + "learning_rate": 1.8006053192521345e-06, + "loss": 0.2387, + "step": 10663 + }, + { + "epoch": 0.7294616594842328, + "grad_norm": 5.26285754187523, + "learning_rate": 1.7997541150907039e-06, + "loss": 0.4362, + "step": 10664 + }, + { + "epoch": 0.7295300636158424, + "grad_norm": 2.1559726357019504, + "learning_rate": 1.7989030680127967e-06, + "loss": 0.1803, + "step": 10665 + }, + { + "epoch": 0.729598467747452, + "grad_norm": 2.2265574204236365, + "learning_rate": 1.7980521780601846e-06, + "loss": 0.3348, + "step": 10666 + }, + { + "epoch": 0.7296668718790615, + "grad_norm": 1.8687360837985985, + "learning_rate": 1.7972014452746339e-06, + "loss": 0.3664, + "step": 10667 + }, + { + "epoch": 0.729735276010671, + "grad_norm": 2.010098717967403, + "learning_rate": 1.796350869697902e-06, + "loss": 0.2498, + "step": 10668 + }, + { + "epoch": 0.7298036801422806, + "grad_norm": 1.7266004135318918, + "learning_rate": 1.7955004513717382e-06, + "loss": 0.2237, + "step": 10669 + }, + { + "epoch": 0.7298720842738902, + "grad_norm": 2.043777972942785, + "learning_rate": 1.7946501903378887e-06, + "loss": 0.4956, + "step": 10670 + }, + { + "epoch": 0.7299404884054997, + "grad_norm": 3.49006675484219, + "learning_rate": 1.7938000866380862e-06, + "loss": 0.7829, + "step": 10671 + }, + { + "epoch": 0.7300088925371092, + "grad_norm": 2.2110011896995476, + "learning_rate": 1.7929501403140576e-06, + "loss": 0.4511, + "step": 10672 + }, + { + "epoch": 0.7300772966687188, + "grad_norm": 1.8259396981730915, + "learning_rate": 1.7921003514075214e-06, + "loss": 0.2762, + "step": 10673 + }, + { + "epoch": 0.7301457008003284, + "grad_norm": 2.406769516831774, + "learning_rate": 1.7912507199601925e-06, + "loss": 0.641, + "step": 10674 + }, + { + "epoch": 0.7302141049319378, + "grad_norm": 2.0310921167489906, + "learning_rate": 1.790401246013773e-06, + "loss": 0.4917, + "step": 10675 + }, + { + "epoch": 0.7302825090635474, + "grad_norm": 2.3809666597698844, + "learning_rate": 1.7895519296099577e-06, + "loss": 0.4308, + "step": 10676 + }, + { + "epoch": 0.730350913195157, + "grad_norm": 1.6204020242320887, + "learning_rate": 1.7887027707904387e-06, + "loss": 0.2115, + "step": 10677 + }, + { + "epoch": 0.7304193173267666, + "grad_norm": 1.9593728194764581, + "learning_rate": 1.787853769596895e-06, + "loss": 0.3907, + "step": 10678 + }, + { + "epoch": 0.730487721458376, + "grad_norm": 2.4970820683637838, + "learning_rate": 1.7870049260709992e-06, + "loss": 0.4311, + "step": 10679 + }, + { + "epoch": 0.7305561255899856, + "grad_norm": 1.9613342719636073, + "learning_rate": 1.786156240254417e-06, + "loss": 0.2845, + "step": 10680 + }, + { + "epoch": 0.7306245297215952, + "grad_norm": 2.0198859930678283, + "learning_rate": 1.7853077121888046e-06, + "loss": 0.415, + "step": 10681 + }, + { + "epoch": 0.7306929338532048, + "grad_norm": 3.0197645050402833, + "learning_rate": 1.7844593419158146e-06, + "loss": 0.9122, + "step": 10682 + }, + { + "epoch": 0.7307613379848142, + "grad_norm": 2.5052441431183965, + "learning_rate": 1.7836111294770864e-06, + "loss": 0.5203, + "step": 10683 + }, + { + "epoch": 0.7308297421164238, + "grad_norm": 1.641303371780431, + "learning_rate": 1.782763074914257e-06, + "loss": 0.287, + "step": 10684 + }, + { + "epoch": 0.7308981462480334, + "grad_norm": 2.7534422355024923, + "learning_rate": 1.7819151782689525e-06, + "loss": 0.5972, + "step": 10685 + }, + { + "epoch": 0.730966550379643, + "grad_norm": 2.230711020224331, + "learning_rate": 1.781067439582791e-06, + "loss": 0.4479, + "step": 10686 + }, + { + "epoch": 0.7310349545112524, + "grad_norm": 2.538552241772297, + "learning_rate": 1.7802198588973829e-06, + "loss": 0.6176, + "step": 10687 + }, + { + "epoch": 0.731103358642862, + "grad_norm": 2.2456601064280517, + "learning_rate": 1.7793724362543314e-06, + "loss": 0.3675, + "step": 10688 + }, + { + "epoch": 0.7311717627744716, + "grad_norm": 2.015887978867571, + "learning_rate": 1.778525171695234e-06, + "loss": 0.1417, + "step": 10689 + }, + { + "epoch": 0.7312401669060812, + "grad_norm": 2.207890253931876, + "learning_rate": 1.7776780652616776e-06, + "loss": 0.5429, + "step": 10690 + }, + { + "epoch": 0.7313085710376906, + "grad_norm": 1.7860349333209884, + "learning_rate": 1.7768311169952424e-06, + "loss": 0.158, + "step": 10691 + }, + { + "epoch": 0.7313769751693002, + "grad_norm": 1.4763300450327619, + "learning_rate": 1.7759843269375004e-06, + "loss": 0.2676, + "step": 10692 + }, + { + "epoch": 0.7314453793009098, + "grad_norm": 1.9836429011418006, + "learning_rate": 1.7751376951300147e-06, + "loss": 0.3447, + "step": 10693 + }, + { + "epoch": 0.7315137834325193, + "grad_norm": 1.9143863248349993, + "learning_rate": 1.7742912216143455e-06, + "loss": 0.3082, + "step": 10694 + }, + { + "epoch": 0.7315821875641288, + "grad_norm": 2.3741930293365434, + "learning_rate": 1.7734449064320375e-06, + "loss": 0.6786, + "step": 10695 + }, + { + "epoch": 0.7316505916957384, + "grad_norm": 2.11564801258548, + "learning_rate": 1.7725987496246366e-06, + "loss": 0.4118, + "step": 10696 + }, + { + "epoch": 0.731718995827348, + "grad_norm": 3.655347663845274, + "learning_rate": 1.771752751233674e-06, + "loss": 0.4428, + "step": 10697 + }, + { + "epoch": 0.7317873999589575, + "grad_norm": 2.7037922360934687, + "learning_rate": 1.7709069113006756e-06, + "loss": 0.1652, + "step": 10698 + }, + { + "epoch": 0.731855804090567, + "grad_norm": 2.0694061454611283, + "learning_rate": 1.7700612298671587e-06, + "loss": 0.3903, + "step": 10699 + }, + { + "epoch": 0.7319242082221766, + "grad_norm": 2.239559241468385, + "learning_rate": 1.7692157069746342e-06, + "loss": 0.559, + "step": 10700 + }, + { + "epoch": 0.7319926123537862, + "grad_norm": 1.9168070792685887, + "learning_rate": 1.768370342664602e-06, + "loss": 0.3508, + "step": 10701 + }, + { + "epoch": 0.7320610164853957, + "grad_norm": 2.3950121771556625, + "learning_rate": 1.7675251369785612e-06, + "loss": 0.3476, + "step": 10702 + }, + { + "epoch": 0.7321294206170053, + "grad_norm": 2.63059404900112, + "learning_rate": 1.7666800899579939e-06, + "loss": 0.626, + "step": 10703 + }, + { + "epoch": 0.7321978247486148, + "grad_norm": 2.144304802061636, + "learning_rate": 1.765835201644383e-06, + "loss": 0.2662, + "step": 10704 + }, + { + "epoch": 0.7322662288802244, + "grad_norm": 2.763762277802758, + "learning_rate": 1.7649904720791982e-06, + "loss": 0.3843, + "step": 10705 + }, + { + "epoch": 0.7323346330118339, + "grad_norm": 1.9798683595727224, + "learning_rate": 1.7641459013039024e-06, + "loss": 0.4614, + "step": 10706 + }, + { + "epoch": 0.7324030371434435, + "grad_norm": 2.038581934017726, + "learning_rate": 1.763301489359952e-06, + "loss": 0.242, + "step": 10707 + }, + { + "epoch": 0.732471441275053, + "grad_norm": 1.6995481478745456, + "learning_rate": 1.762457236288792e-06, + "loss": 0.193, + "step": 10708 + }, + { + "epoch": 0.7325398454066626, + "grad_norm": 2.113418891922965, + "learning_rate": 1.7616131421318672e-06, + "loss": 0.4021, + "step": 10709 + }, + { + "epoch": 0.7326082495382721, + "grad_norm": 2.1242262960442333, + "learning_rate": 1.760769206930607e-06, + "loss": 0.5448, + "step": 10710 + }, + { + "epoch": 0.7326766536698817, + "grad_norm": 2.5354547545289563, + "learning_rate": 1.7599254307264362e-06, + "loss": 0.5653, + "step": 10711 + }, + { + "epoch": 0.7327450578014912, + "grad_norm": 2.656052559639189, + "learning_rate": 1.7590818135607718e-06, + "loss": 0.2403, + "step": 10712 + }, + { + "epoch": 0.7328134619331007, + "grad_norm": 2.093526038247776, + "learning_rate": 1.7582383554750204e-06, + "loss": 0.3986, + "step": 10713 + }, + { + "epoch": 0.7328818660647103, + "grad_norm": 2.4632195926865803, + "learning_rate": 1.7573950565105863e-06, + "loss": 0.4071, + "step": 10714 + }, + { + "epoch": 0.7329502701963199, + "grad_norm": 1.7820955578457056, + "learning_rate": 1.7565519167088595e-06, + "loss": 0.2067, + "step": 10715 + }, + { + "epoch": 0.7330186743279294, + "grad_norm": 2.8597833102174035, + "learning_rate": 1.755708936111229e-06, + "loss": 0.5921, + "step": 10716 + }, + { + "epoch": 0.7330870784595389, + "grad_norm": 2.450897827634752, + "learning_rate": 1.75486611475907e-06, + "loss": 0.6457, + "step": 10717 + }, + { + "epoch": 0.7331554825911485, + "grad_norm": 2.5826922333877547, + "learning_rate": 1.7540234526937532e-06, + "loss": 0.5394, + "step": 10718 + }, + { + "epoch": 0.7332238867227581, + "grad_norm": 1.6162596588627975, + "learning_rate": 1.7531809499566399e-06, + "loss": 0.2527, + "step": 10719 + }, + { + "epoch": 0.7332922908543676, + "grad_norm": 2.367774147890106, + "learning_rate": 1.752338606589084e-06, + "loss": 0.4132, + "step": 10720 + }, + { + "epoch": 0.7333606949859771, + "grad_norm": 2.2655971393845302, + "learning_rate": 1.751496422632431e-06, + "loss": 0.2077, + "step": 10721 + }, + { + "epoch": 0.7334290991175867, + "grad_norm": 2.544006487500718, + "learning_rate": 1.7506543981280216e-06, + "loss": 0.4119, + "step": 10722 + }, + { + "epoch": 0.7334975032491963, + "grad_norm": 2.2265216790188336, + "learning_rate": 1.7498125331171838e-06, + "loss": 0.3321, + "step": 10723 + }, + { + "epoch": 0.7335659073808058, + "grad_norm": 2.749297550824957, + "learning_rate": 1.7489708276412432e-06, + "loss": 0.5106, + "step": 10724 + }, + { + "epoch": 0.7336343115124153, + "grad_norm": 2.1516502876351487, + "learning_rate": 1.748129281741514e-06, + "loss": 0.2842, + "step": 10725 + }, + { + "epoch": 0.7337027156440249, + "grad_norm": 3.009955888461378, + "learning_rate": 1.7472878954593026e-06, + "loss": 0.635, + "step": 10726 + }, + { + "epoch": 0.7337711197756345, + "grad_norm": 1.7428225997596591, + "learning_rate": 1.7464466688359082e-06, + "loss": 0.2235, + "step": 10727 + }, + { + "epoch": 0.733839523907244, + "grad_norm": 3.032070019500412, + "learning_rate": 1.745605601912621e-06, + "loss": 0.3815, + "step": 10728 + }, + { + "epoch": 0.7339079280388535, + "grad_norm": 3.4341286859878215, + "learning_rate": 1.7447646947307277e-06, + "loss": 0.4521, + "step": 10729 + }, + { + "epoch": 0.7339763321704631, + "grad_norm": 1.878589500327889, + "learning_rate": 1.7439239473315027e-06, + "loss": 0.4093, + "step": 10730 + }, + { + "epoch": 0.7340447363020727, + "grad_norm": 2.353575165181447, + "learning_rate": 1.743083359756214e-06, + "loss": 0.4, + "step": 10731 + }, + { + "epoch": 0.7341131404336821, + "grad_norm": 2.1467227820748596, + "learning_rate": 1.7422429320461215e-06, + "loss": 0.4755, + "step": 10732 + }, + { + "epoch": 0.7341815445652917, + "grad_norm": 2.1044968536396027, + "learning_rate": 1.7414026642424753e-06, + "loss": 0.2456, + "step": 10733 + }, + { + "epoch": 0.7342499486969013, + "grad_norm": 1.8329518735923442, + "learning_rate": 1.7405625563865242e-06, + "loss": 0.422, + "step": 10734 + }, + { + "epoch": 0.7343183528285109, + "grad_norm": 5.597963823811479, + "learning_rate": 1.7397226085195002e-06, + "loss": 0.3771, + "step": 10735 + }, + { + "epoch": 0.7343867569601203, + "grad_norm": 2.4530659504483507, + "learning_rate": 1.7388828206826363e-06, + "loss": 0.5385, + "step": 10736 + }, + { + "epoch": 0.7344551610917299, + "grad_norm": 2.2327897410370725, + "learning_rate": 1.7380431929171515e-06, + "loss": 0.5024, + "step": 10737 + }, + { + "epoch": 0.7345235652233395, + "grad_norm": 2.7179345190442534, + "learning_rate": 1.7372037252642581e-06, + "loss": 0.4961, + "step": 10738 + }, + { + "epoch": 0.7345919693549491, + "grad_norm": 2.623442181917594, + "learning_rate": 1.7363644177651623e-06, + "loss": 0.3399, + "step": 10739 + }, + { + "epoch": 0.7346603734865585, + "grad_norm": 2.2215049591103546, + "learning_rate": 1.735525270461061e-06, + "loss": 0.2653, + "step": 10740 + }, + { + "epoch": 0.7347287776181681, + "grad_norm": 2.24340865479441, + "learning_rate": 1.734686283393141e-06, + "loss": 0.2306, + "step": 10741 + }, + { + "epoch": 0.7347971817497777, + "grad_norm": 2.383595872526121, + "learning_rate": 1.7338474566025881e-06, + "loss": 0.4252, + "step": 10742 + }, + { + "epoch": 0.7348655858813873, + "grad_norm": 2.2904252302169206, + "learning_rate": 1.7330087901305725e-06, + "loss": 0.4102, + "step": 10743 + }, + { + "epoch": 0.7349339900129968, + "grad_norm": 3.2429691337671605, + "learning_rate": 1.7321702840182636e-06, + "loss": 0.3896, + "step": 10744 + }, + { + "epoch": 0.7350023941446063, + "grad_norm": 2.814001683833507, + "learning_rate": 1.7313319383068172e-06, + "loss": 0.2681, + "step": 10745 + }, + { + "epoch": 0.7350707982762159, + "grad_norm": 2.136211335014758, + "learning_rate": 1.7304937530373833e-06, + "loss": 0.2383, + "step": 10746 + }, + { + "epoch": 0.7351392024078255, + "grad_norm": 3.3980268208330418, + "learning_rate": 1.7296557282511045e-06, + "loss": 0.3389, + "step": 10747 + }, + { + "epoch": 0.735207606539435, + "grad_norm": 2.1561908541044885, + "learning_rate": 1.7288178639891134e-06, + "loss": 0.4916, + "step": 10748 + }, + { + "epoch": 0.7352760106710445, + "grad_norm": 2.228438196333264, + "learning_rate": 1.7279801602925394e-06, + "loss": 0.4528, + "step": 10749 + }, + { + "epoch": 0.7353444148026541, + "grad_norm": 1.4808841892418423, + "learning_rate": 1.7271426172024992e-06, + "loss": 0.2249, + "step": 10750 + }, + { + "epoch": 0.7354128189342636, + "grad_norm": 8.809694100480616, + "learning_rate": 1.7263052347601045e-06, + "loss": 0.3904, + "step": 10751 + }, + { + "epoch": 0.7354812230658732, + "grad_norm": 2.534619184835481, + "learning_rate": 1.7254680130064567e-06, + "loss": 0.347, + "step": 10752 + }, + { + "epoch": 0.7355496271974827, + "grad_norm": 2.2759921216191463, + "learning_rate": 1.7246309519826498e-06, + "loss": 0.4195, + "step": 10753 + }, + { + "epoch": 0.7356180313290923, + "grad_norm": 2.129957922757176, + "learning_rate": 1.7237940517297746e-06, + "loss": 0.1982, + "step": 10754 + }, + { + "epoch": 0.7356864354607018, + "grad_norm": 2.32237749014779, + "learning_rate": 1.722957312288906e-06, + "loss": 0.4882, + "step": 10755 + }, + { + "epoch": 0.7357548395923114, + "grad_norm": 2.329787032305751, + "learning_rate": 1.7221207337011187e-06, + "loss": 0.2015, + "step": 10756 + }, + { + "epoch": 0.7358232437239209, + "grad_norm": 2.3321020796097733, + "learning_rate": 1.7212843160074749e-06, + "loss": 0.2322, + "step": 10757 + }, + { + "epoch": 0.7358916478555305, + "grad_norm": 3.6289402659354595, + "learning_rate": 1.7204480592490292e-06, + "loss": 0.4587, + "step": 10758 + }, + { + "epoch": 0.73596005198714, + "grad_norm": 2.0029196141443957, + "learning_rate": 1.7196119634668296e-06, + "loss": 0.3975, + "step": 10759 + }, + { + "epoch": 0.7360284561187496, + "grad_norm": 1.9381576429482519, + "learning_rate": 1.7187760287019157e-06, + "loss": 0.2836, + "step": 10760 + }, + { + "epoch": 0.7360968602503591, + "grad_norm": 4.5134391245964185, + "learning_rate": 1.7179402549953177e-06, + "loss": 0.1665, + "step": 10761 + }, + { + "epoch": 0.7361652643819687, + "grad_norm": 2.145376021687394, + "learning_rate": 1.7171046423880622e-06, + "loss": 0.4528, + "step": 10762 + }, + { + "epoch": 0.7362336685135782, + "grad_norm": 2.0690288774944334, + "learning_rate": 1.7162691909211625e-06, + "loss": 0.2769, + "step": 10763 + }, + { + "epoch": 0.7363020726451878, + "grad_norm": 1.6149291081008417, + "learning_rate": 1.7154339006356297e-06, + "loss": 0.1589, + "step": 10764 + }, + { + "epoch": 0.7363704767767973, + "grad_norm": 2.667107606483458, + "learning_rate": 1.7145987715724617e-06, + "loss": 0.3961, + "step": 10765 + }, + { + "epoch": 0.7364388809084069, + "grad_norm": 2.273272950814536, + "learning_rate": 1.7137638037726513e-06, + "loss": 0.2803, + "step": 10766 + }, + { + "epoch": 0.7365072850400164, + "grad_norm": 2.2779016743732905, + "learning_rate": 1.7129289972771818e-06, + "loss": 0.3851, + "step": 10767 + }, + { + "epoch": 0.736575689171626, + "grad_norm": 2.5181546139390054, + "learning_rate": 1.7120943521270295e-06, + "loss": 0.5501, + "step": 10768 + }, + { + "epoch": 0.7366440933032355, + "grad_norm": 3.499468763089668, + "learning_rate": 1.7112598683631642e-06, + "loss": 0.2911, + "step": 10769 + }, + { + "epoch": 0.736712497434845, + "grad_norm": 2.082846140321554, + "learning_rate": 1.7104255460265467e-06, + "loss": 0.4804, + "step": 10770 + }, + { + "epoch": 0.7367809015664546, + "grad_norm": 2.6465899907394963, + "learning_rate": 1.709591385158128e-06, + "loss": 0.2247, + "step": 10771 + }, + { + "epoch": 0.7368493056980642, + "grad_norm": 1.7786143068802742, + "learning_rate": 1.7087573857988522e-06, + "loss": 0.2493, + "step": 10772 + }, + { + "epoch": 0.7369177098296738, + "grad_norm": 1.834923013454605, + "learning_rate": 1.7079235479896582e-06, + "loss": 0.2179, + "step": 10773 + }, + { + "epoch": 0.7369861139612832, + "grad_norm": 2.7514343639130874, + "learning_rate": 1.707089871771474e-06, + "loss": 0.1881, + "step": 10774 + }, + { + "epoch": 0.7370545180928928, + "grad_norm": 2.085192979679935, + "learning_rate": 1.7062563571852186e-06, + "loss": 0.3114, + "step": 10775 + }, + { + "epoch": 0.7371229222245024, + "grad_norm": 2.085662327164606, + "learning_rate": 1.7054230042718078e-06, + "loss": 0.3508, + "step": 10776 + }, + { + "epoch": 0.737191326356112, + "grad_norm": 2.2729622911185916, + "learning_rate": 1.7045898130721456e-06, + "loss": 0.4818, + "step": 10777 + }, + { + "epoch": 0.7372597304877214, + "grad_norm": 2.133136473143398, + "learning_rate": 1.703756783627129e-06, + "loss": 0.5405, + "step": 10778 + }, + { + "epoch": 0.737328134619331, + "grad_norm": 2.230002399601522, + "learning_rate": 1.7029239159776468e-06, + "loss": 0.3991, + "step": 10779 + }, + { + "epoch": 0.7373965387509406, + "grad_norm": 2.7393860486047332, + "learning_rate": 1.7020912101645781e-06, + "loss": 0.4542, + "step": 10780 + }, + { + "epoch": 0.7374649428825502, + "grad_norm": 2.230518427360742, + "learning_rate": 1.7012586662288e-06, + "loss": 0.6751, + "step": 10781 + }, + { + "epoch": 0.7375333470141596, + "grad_norm": 2.3022006179023666, + "learning_rate": 1.7004262842111745e-06, + "loss": 0.6587, + "step": 10782 + }, + { + "epoch": 0.7376017511457692, + "grad_norm": 2.081829297432446, + "learning_rate": 1.6995940641525621e-06, + "loss": 0.4583, + "step": 10783 + }, + { + "epoch": 0.7376701552773788, + "grad_norm": 2.073367287108887, + "learning_rate": 1.698762006093811e-06, + "loss": 0.2698, + "step": 10784 + }, + { + "epoch": 0.7377385594089882, + "grad_norm": 1.8910838547635018, + "learning_rate": 1.6979301100757616e-06, + "loss": 0.3374, + "step": 10785 + }, + { + "epoch": 0.7378069635405978, + "grad_norm": 2.312294229167542, + "learning_rate": 1.6970983761392478e-06, + "loss": 0.2999, + "step": 10786 + }, + { + "epoch": 0.7378753676722074, + "grad_norm": 1.4865233986062263, + "learning_rate": 1.696266804325094e-06, + "loss": 0.1784, + "step": 10787 + }, + { + "epoch": 0.737943771803817, + "grad_norm": 1.756029920601701, + "learning_rate": 1.6954353946741202e-06, + "loss": 0.2718, + "step": 10788 + }, + { + "epoch": 0.7380121759354265, + "grad_norm": 2.6160894225444227, + "learning_rate": 1.6946041472271356e-06, + "loss": 0.2701, + "step": 10789 + }, + { + "epoch": 0.738080580067036, + "grad_norm": 2.617821084379921, + "learning_rate": 1.6937730620249404e-06, + "loss": 0.4969, + "step": 10790 + }, + { + "epoch": 0.7381489841986456, + "grad_norm": 2.0434014717306024, + "learning_rate": 1.692942139108329e-06, + "loss": 0.3989, + "step": 10791 + }, + { + "epoch": 0.7382173883302552, + "grad_norm": 2.2422912254665044, + "learning_rate": 1.6921113785180858e-06, + "loss": 0.5929, + "step": 10792 + }, + { + "epoch": 0.7382857924618647, + "grad_norm": 2.553912894029397, + "learning_rate": 1.691280780294991e-06, + "loss": 0.2582, + "step": 10793 + }, + { + "epoch": 0.7383541965934742, + "grad_norm": 2.2336798166649725, + "learning_rate": 1.6904503444798116e-06, + "loss": 0.4362, + "step": 10794 + }, + { + "epoch": 0.7384226007250838, + "grad_norm": 2.078797109807545, + "learning_rate": 1.6896200711133126e-06, + "loss": 0.3827, + "step": 10795 + }, + { + "epoch": 0.7384910048566934, + "grad_norm": 2.4541089216820993, + "learning_rate": 1.6887899602362462e-06, + "loss": 0.5546, + "step": 10796 + }, + { + "epoch": 0.7385594089883029, + "grad_norm": 2.0901085944231403, + "learning_rate": 1.687960011889358e-06, + "loss": 0.5259, + "step": 10797 + }, + { + "epoch": 0.7386278131199124, + "grad_norm": 2.911907959182671, + "learning_rate": 1.6871302261133864e-06, + "loss": 0.3451, + "step": 10798 + }, + { + "epoch": 0.738696217251522, + "grad_norm": 2.2623121244184308, + "learning_rate": 1.686300602949061e-06, + "loss": 0.4634, + "step": 10799 + }, + { + "epoch": 0.7387646213831316, + "grad_norm": 1.986898331020681, + "learning_rate": 1.6854711424371017e-06, + "loss": 0.288, + "step": 10800 + }, + { + "epoch": 0.7388330255147411, + "grad_norm": 2.203373475638481, + "learning_rate": 1.6846418446182262e-06, + "loss": 0.1495, + "step": 10801 + }, + { + "epoch": 0.7389014296463506, + "grad_norm": 1.8911153929738242, + "learning_rate": 1.6838127095331375e-06, + "loss": 0.2779, + "step": 10802 + }, + { + "epoch": 0.7389698337779602, + "grad_norm": 2.676619505623163, + "learning_rate": 1.6829837372225355e-06, + "loss": 0.3125, + "step": 10803 + }, + { + "epoch": 0.7390382379095697, + "grad_norm": 2.58674584169755, + "learning_rate": 1.6821549277271099e-06, + "loss": 0.4827, + "step": 10804 + }, + { + "epoch": 0.7391066420411793, + "grad_norm": 1.9884360052711951, + "learning_rate": 1.6813262810875424e-06, + "loss": 0.2121, + "step": 10805 + }, + { + "epoch": 0.7391750461727888, + "grad_norm": 2.3826106564441427, + "learning_rate": 1.6804977973445063e-06, + "loss": 0.4406, + "step": 10806 + }, + { + "epoch": 0.7392434503043984, + "grad_norm": 1.866088911106301, + "learning_rate": 1.6796694765386668e-06, + "loss": 0.258, + "step": 10807 + }, + { + "epoch": 0.7393118544360079, + "grad_norm": 2.101236154963388, + "learning_rate": 1.6788413187106844e-06, + "loss": 0.2377, + "step": 10808 + }, + { + "epoch": 0.7393802585676175, + "grad_norm": 2.8865762244362156, + "learning_rate": 1.6780133239012075e-06, + "loss": 0.3205, + "step": 10809 + }, + { + "epoch": 0.739448662699227, + "grad_norm": 2.7802009693666143, + "learning_rate": 1.677185492150879e-06, + "loss": 0.2265, + "step": 10810 + }, + { + "epoch": 0.7395170668308366, + "grad_norm": 2.033296269638577, + "learning_rate": 1.676357823500332e-06, + "loss": 0.3459, + "step": 10811 + }, + { + "epoch": 0.7395854709624461, + "grad_norm": 1.612494560905273, + "learning_rate": 1.675530317990191e-06, + "loss": 0.2246, + "step": 10812 + }, + { + "epoch": 0.7396538750940557, + "grad_norm": 2.118380502535304, + "learning_rate": 1.6747029756610778e-06, + "loss": 0.2418, + "step": 10813 + }, + { + "epoch": 0.7397222792256652, + "grad_norm": 2.035909048541825, + "learning_rate": 1.673875796553598e-06, + "loss": 0.3177, + "step": 10814 + }, + { + "epoch": 0.7397906833572748, + "grad_norm": 1.8995311475425694, + "learning_rate": 1.6730487807083578e-06, + "loss": 0.3627, + "step": 10815 + }, + { + "epoch": 0.7398590874888843, + "grad_norm": 1.951408135178555, + "learning_rate": 1.6722219281659486e-06, + "loss": 0.4305, + "step": 10816 + }, + { + "epoch": 0.7399274916204939, + "grad_norm": 2.9890987499728974, + "learning_rate": 1.6713952389669574e-06, + "loss": 0.3178, + "step": 10817 + }, + { + "epoch": 0.7399958957521034, + "grad_norm": 2.4415692389441723, + "learning_rate": 1.6705687131519616e-06, + "loss": 0.5133, + "step": 10818 + }, + { + "epoch": 0.740064299883713, + "grad_norm": 2.425744577029269, + "learning_rate": 1.6697423507615307e-06, + "loss": 0.29, + "step": 10819 + }, + { + "epoch": 0.7401327040153225, + "grad_norm": 2.316927872026968, + "learning_rate": 1.668916151836225e-06, + "loss": 0.3904, + "step": 10820 + }, + { + "epoch": 0.7402011081469321, + "grad_norm": 4.964935529987542, + "learning_rate": 1.6680901164166025e-06, + "loss": 0.1206, + "step": 10821 + }, + { + "epoch": 0.7402695122785417, + "grad_norm": 2.2700469699907204, + "learning_rate": 1.6672642445432048e-06, + "loss": 0.3141, + "step": 10822 + }, + { + "epoch": 0.7403379164101511, + "grad_norm": 2.064812591929618, + "learning_rate": 1.6664385362565733e-06, + "loss": 0.0998, + "step": 10823 + }, + { + "epoch": 0.7404063205417607, + "grad_norm": 4.829346511537598, + "learning_rate": 1.6656129915972357e-06, + "loss": 0.2808, + "step": 10824 + }, + { + "epoch": 0.7404747246733703, + "grad_norm": 2.114452305541134, + "learning_rate": 1.6647876106057142e-06, + "loss": 0.2981, + "step": 10825 + }, + { + "epoch": 0.7405431288049799, + "grad_norm": 2.6107751607507117, + "learning_rate": 1.6639623933225223e-06, + "loss": 0.252, + "step": 10826 + }, + { + "epoch": 0.7406115329365893, + "grad_norm": 1.9190265589785007, + "learning_rate": 1.6631373397881644e-06, + "loss": 0.3011, + "step": 10827 + }, + { + "epoch": 0.7406799370681989, + "grad_norm": 1.9054528899273722, + "learning_rate": 1.6623124500431409e-06, + "loss": 0.2761, + "step": 10828 + }, + { + "epoch": 0.7407483411998085, + "grad_norm": 1.7720526358959854, + "learning_rate": 1.6614877241279397e-06, + "loss": 0.2403, + "step": 10829 + }, + { + "epoch": 0.7408167453314181, + "grad_norm": 2.6282242206255155, + "learning_rate": 1.6606631620830427e-06, + "loss": 0.6993, + "step": 10830 + }, + { + "epoch": 0.7408851494630275, + "grad_norm": 2.1327692733129004, + "learning_rate": 1.6598387639489233e-06, + "loss": 0.2135, + "step": 10831 + }, + { + "epoch": 0.7409535535946371, + "grad_norm": 3.42214731326076, + "learning_rate": 1.6590145297660447e-06, + "loss": 0.3384, + "step": 10832 + }, + { + "epoch": 0.7410219577262467, + "grad_norm": 2.226331692160085, + "learning_rate": 1.6581904595748682e-06, + "loss": 0.3456, + "step": 10833 + }, + { + "epoch": 0.7410903618578563, + "grad_norm": 2.3196674182325285, + "learning_rate": 1.6573665534158401e-06, + "loss": 0.5337, + "step": 10834 + }, + { + "epoch": 0.7411587659894657, + "grad_norm": 1.7586114042828802, + "learning_rate": 1.6565428113294047e-06, + "loss": 0.2556, + "step": 10835 + }, + { + "epoch": 0.7412271701210753, + "grad_norm": 2.1276151049055354, + "learning_rate": 1.655719233355993e-06, + "loss": 0.4751, + "step": 10836 + }, + { + "epoch": 0.7412955742526849, + "grad_norm": 2.4474431831264463, + "learning_rate": 1.654895819536031e-06, + "loss": 0.5211, + "step": 10837 + }, + { + "epoch": 0.7413639783842945, + "grad_norm": 3.6414570000972315, + "learning_rate": 1.6540725699099358e-06, + "loss": 0.5971, + "step": 10838 + }, + { + "epoch": 0.7414323825159039, + "grad_norm": 2.2628938524436224, + "learning_rate": 1.6532494845181157e-06, + "loss": 0.4225, + "step": 10839 + }, + { + "epoch": 0.7415007866475135, + "grad_norm": 2.2351994827286807, + "learning_rate": 1.6524265634009705e-06, + "loss": 0.3956, + "step": 10840 + }, + { + "epoch": 0.7415691907791231, + "grad_norm": 2.256565789264035, + "learning_rate": 1.651603806598897e-06, + "loss": 0.4586, + "step": 10841 + }, + { + "epoch": 0.7416375949107326, + "grad_norm": 2.139214517389838, + "learning_rate": 1.6507812141522762e-06, + "loss": 0.3452, + "step": 10842 + }, + { + "epoch": 0.7417059990423421, + "grad_norm": 3.288721207333548, + "learning_rate": 1.6499587861014875e-06, + "loss": 0.5979, + "step": 10843 + }, + { + "epoch": 0.7417744031739517, + "grad_norm": 1.8401819068199976, + "learning_rate": 1.6491365224868995e-06, + "loss": 0.2713, + "step": 10844 + }, + { + "epoch": 0.7418428073055613, + "grad_norm": 1.8785170629872232, + "learning_rate": 1.6483144233488718e-06, + "loss": 0.2643, + "step": 10845 + }, + { + "epoch": 0.7419112114371708, + "grad_norm": 2.515787119409898, + "learning_rate": 1.6474924887277578e-06, + "loss": 0.7674, + "step": 10846 + }, + { + "epoch": 0.7419796155687803, + "grad_norm": 2.2744162310945777, + "learning_rate": 1.6466707186638993e-06, + "loss": 0.431, + "step": 10847 + }, + { + "epoch": 0.7420480197003899, + "grad_norm": 1.917868174539964, + "learning_rate": 1.6458491131976367e-06, + "loss": 0.3617, + "step": 10848 + }, + { + "epoch": 0.7421164238319995, + "grad_norm": 1.8762803156955818, + "learning_rate": 1.645027672369297e-06, + "loss": 0.1869, + "step": 10849 + }, + { + "epoch": 0.742184827963609, + "grad_norm": 2.2834053308965157, + "learning_rate": 1.6442063962192e-06, + "loss": 0.3847, + "step": 10850 + }, + { + "epoch": 0.7422532320952185, + "grad_norm": 2.107595116651123, + "learning_rate": 1.643385284787658e-06, + "loss": 0.2974, + "step": 10851 + }, + { + "epoch": 0.7423216362268281, + "grad_norm": 2.490213263927059, + "learning_rate": 1.6425643381149737e-06, + "loss": 0.1442, + "step": 10852 + }, + { + "epoch": 0.7423900403584377, + "grad_norm": 1.980435869594804, + "learning_rate": 1.6417435562414458e-06, + "loss": 0.3285, + "step": 10853 + }, + { + "epoch": 0.7424584444900472, + "grad_norm": 2.272892129510964, + "learning_rate": 1.6409229392073595e-06, + "loss": 0.4348, + "step": 10854 + }, + { + "epoch": 0.7425268486216567, + "grad_norm": 1.690314846005454, + "learning_rate": 1.640102487052998e-06, + "loss": 0.2014, + "step": 10855 + }, + { + "epoch": 0.7425952527532663, + "grad_norm": 2.1282411708857016, + "learning_rate": 1.6392821998186315e-06, + "loss": 0.4248, + "step": 10856 + }, + { + "epoch": 0.7426636568848759, + "grad_norm": 2.0324585547098457, + "learning_rate": 1.6384620775445226e-06, + "loss": 0.2888, + "step": 10857 + }, + { + "epoch": 0.7427320610164854, + "grad_norm": 2.3396401611304025, + "learning_rate": 1.637642120270928e-06, + "loss": 0.4273, + "step": 10858 + }, + { + "epoch": 0.742800465148095, + "grad_norm": 2.105373854612997, + "learning_rate": 1.6368223280380951e-06, + "loss": 0.4199, + "step": 10859 + }, + { + "epoch": 0.7428688692797045, + "grad_norm": 1.859844657342012, + "learning_rate": 1.636002700886261e-06, + "loss": 0.366, + "step": 10860 + }, + { + "epoch": 0.742937273411314, + "grad_norm": 2.0589647194089364, + "learning_rate": 1.6351832388556604e-06, + "loss": 0.1761, + "step": 10861 + }, + { + "epoch": 0.7430056775429236, + "grad_norm": 2.6368979612129704, + "learning_rate": 1.6343639419865138e-06, + "loss": 0.5945, + "step": 10862 + }, + { + "epoch": 0.7430740816745331, + "grad_norm": 1.4924511352044634, + "learning_rate": 1.6335448103190388e-06, + "loss": 0.1788, + "step": 10863 + }, + { + "epoch": 0.7431424858061427, + "grad_norm": 2.094246897876505, + "learning_rate": 1.6327258438934407e-06, + "loss": 0.4134, + "step": 10864 + }, + { + "epoch": 0.7432108899377522, + "grad_norm": 2.1304378354802553, + "learning_rate": 1.631907042749919e-06, + "loss": 0.2699, + "step": 10865 + }, + { + "epoch": 0.7432792940693618, + "grad_norm": 2.2513084646869603, + "learning_rate": 1.6310884069286636e-06, + "loss": 0.3979, + "step": 10866 + }, + { + "epoch": 0.7433476982009714, + "grad_norm": 2.0989378134589773, + "learning_rate": 1.6302699364698565e-06, + "loss": 0.3598, + "step": 10867 + }, + { + "epoch": 0.7434161023325809, + "grad_norm": 1.797662363452289, + "learning_rate": 1.6294516314136744e-06, + "loss": 0.2897, + "step": 10868 + }, + { + "epoch": 0.7434845064641904, + "grad_norm": 1.965851345281594, + "learning_rate": 1.6286334918002822e-06, + "loss": 0.4427, + "step": 10869 + }, + { + "epoch": 0.7435529105958, + "grad_norm": 1.822276145029724, + "learning_rate": 1.6278155176698385e-06, + "loss": 0.2934, + "step": 10870 + }, + { + "epoch": 0.7436213147274096, + "grad_norm": 2.035547120089343, + "learning_rate": 1.626997709062494e-06, + "loss": 0.3175, + "step": 10871 + }, + { + "epoch": 0.7436897188590191, + "grad_norm": 1.8025219627872962, + "learning_rate": 1.6261800660183875e-06, + "loss": 0.2486, + "step": 10872 + }, + { + "epoch": 0.7437581229906286, + "grad_norm": 2.117857682690498, + "learning_rate": 1.6253625885776576e-06, + "loss": 0.3012, + "step": 10873 + }, + { + "epoch": 0.7438265271222382, + "grad_norm": 1.874971288222781, + "learning_rate": 1.6245452767804254e-06, + "loss": 0.2056, + "step": 10874 + }, + { + "epoch": 0.7438949312538478, + "grad_norm": 1.7119665317361223, + "learning_rate": 1.6237281306668135e-06, + "loss": 0.2274, + "step": 10875 + }, + { + "epoch": 0.7439633353854573, + "grad_norm": 2.255740835535001, + "learning_rate": 1.6229111502769278e-06, + "loss": 0.4545, + "step": 10876 + }, + { + "epoch": 0.7440317395170668, + "grad_norm": 2.6805014194420016, + "learning_rate": 1.6220943356508718e-06, + "loss": 0.3308, + "step": 10877 + }, + { + "epoch": 0.7441001436486764, + "grad_norm": 2.4409105881123256, + "learning_rate": 1.6212776868287366e-06, + "loss": 0.7775, + "step": 10878 + }, + { + "epoch": 0.744168547780286, + "grad_norm": 1.8227242182008454, + "learning_rate": 1.6204612038506068e-06, + "loss": 0.4259, + "step": 10879 + }, + { + "epoch": 0.7442369519118954, + "grad_norm": 1.9028812095702652, + "learning_rate": 1.6196448867565627e-06, + "loss": 0.3089, + "step": 10880 + }, + { + "epoch": 0.744305356043505, + "grad_norm": 1.5674638341004743, + "learning_rate": 1.6188287355866694e-06, + "loss": 0.2858, + "step": 10881 + }, + { + "epoch": 0.7443737601751146, + "grad_norm": 1.8106443673132695, + "learning_rate": 1.6180127503809906e-06, + "loss": 0.1485, + "step": 10882 + }, + { + "epoch": 0.7444421643067242, + "grad_norm": 1.886838999957929, + "learning_rate": 1.6171969311795777e-06, + "loss": 0.2567, + "step": 10883 + }, + { + "epoch": 0.7445105684383336, + "grad_norm": 2.682709174702005, + "learning_rate": 1.6163812780224747e-06, + "loss": 0.4756, + "step": 10884 + }, + { + "epoch": 0.7445789725699432, + "grad_norm": 2.1703779906902767, + "learning_rate": 1.615565790949718e-06, + "loss": 0.4577, + "step": 10885 + }, + { + "epoch": 0.7446473767015528, + "grad_norm": 2.59454122218071, + "learning_rate": 1.614750470001334e-06, + "loss": 0.2873, + "step": 10886 + }, + { + "epoch": 0.7447157808331624, + "grad_norm": 1.683034626991843, + "learning_rate": 1.6139353152173453e-06, + "loss": 0.2835, + "step": 10887 + }, + { + "epoch": 0.7447841849647718, + "grad_norm": 2.08151251537749, + "learning_rate": 1.6131203266377626e-06, + "loss": 0.4508, + "step": 10888 + }, + { + "epoch": 0.7448525890963814, + "grad_norm": 1.9491006628131862, + "learning_rate": 1.6123055043025899e-06, + "loss": 0.153, + "step": 10889 + }, + { + "epoch": 0.744920993227991, + "grad_norm": 2.0480504155423973, + "learning_rate": 1.6114908482518216e-06, + "loss": 0.2734, + "step": 10890 + }, + { + "epoch": 0.7449893973596006, + "grad_norm": 2.2635074453726487, + "learning_rate": 1.6106763585254436e-06, + "loss": 0.4171, + "step": 10891 + }, + { + "epoch": 0.74505780149121, + "grad_norm": 2.2901370094679443, + "learning_rate": 1.609862035163439e-06, + "loss": 0.4026, + "step": 10892 + }, + { + "epoch": 0.7451262056228196, + "grad_norm": 2.4235523629529774, + "learning_rate": 1.6090478782057744e-06, + "loss": 0.7621, + "step": 10893 + }, + { + "epoch": 0.7451946097544292, + "grad_norm": 2.8057229938419113, + "learning_rate": 1.608233887692417e-06, + "loss": 0.4645, + "step": 10894 + }, + { + "epoch": 0.7452630138860388, + "grad_norm": 2.0278908831008398, + "learning_rate": 1.6074200636633186e-06, + "loss": 0.4231, + "step": 10895 + }, + { + "epoch": 0.7453314180176482, + "grad_norm": 2.6160595698066924, + "learning_rate": 1.6066064061584258e-06, + "loss": 0.5399, + "step": 10896 + }, + { + "epoch": 0.7453998221492578, + "grad_norm": 2.4835091227572583, + "learning_rate": 1.6057929152176777e-06, + "loss": 0.5215, + "step": 10897 + }, + { + "epoch": 0.7454682262808674, + "grad_norm": 4.206196595032891, + "learning_rate": 1.6049795908810034e-06, + "loss": 0.5958, + "step": 10898 + }, + { + "epoch": 0.7455366304124769, + "grad_norm": 2.397056977527036, + "learning_rate": 1.6041664331883233e-06, + "loss": 0.308, + "step": 10899 + }, + { + "epoch": 0.7456050345440864, + "grad_norm": 2.564943915975357, + "learning_rate": 1.603353442179555e-06, + "loss": 0.4002, + "step": 10900 + }, + { + "epoch": 0.745673438675696, + "grad_norm": 2.1340369248627775, + "learning_rate": 1.6025406178946002e-06, + "loss": 0.2042, + "step": 10901 + }, + { + "epoch": 0.7457418428073056, + "grad_norm": 2.888305238433764, + "learning_rate": 1.6017279603733599e-06, + "loss": 0.5713, + "step": 10902 + }, + { + "epoch": 0.7458102469389151, + "grad_norm": 2.0739253567600295, + "learning_rate": 1.6009154696557211e-06, + "loss": 0.3708, + "step": 10903 + }, + { + "epoch": 0.7458786510705246, + "grad_norm": 2.3527463866739007, + "learning_rate": 1.6001031457815651e-06, + "loss": 0.4631, + "step": 10904 + }, + { + "epoch": 0.7459470552021342, + "grad_norm": 2.2585361139523457, + "learning_rate": 1.5992909887907648e-06, + "loss": 0.3993, + "step": 10905 + }, + { + "epoch": 0.7460154593337438, + "grad_norm": 2.334663426404426, + "learning_rate": 1.598478998723183e-06, + "loss": 0.6087, + "step": 10906 + }, + { + "epoch": 0.7460838634653533, + "grad_norm": 2.4354196620778072, + "learning_rate": 1.597667175618679e-06, + "loss": 0.2169, + "step": 10907 + }, + { + "epoch": 0.7461522675969628, + "grad_norm": 1.9271352739465417, + "learning_rate": 1.5968555195170999e-06, + "loss": 0.3711, + "step": 10908 + }, + { + "epoch": 0.7462206717285724, + "grad_norm": 2.0186959241704696, + "learning_rate": 1.5960440304582858e-06, + "loss": 0.3021, + "step": 10909 + }, + { + "epoch": 0.746289075860182, + "grad_norm": 2.776192309721829, + "learning_rate": 1.5952327084820679e-06, + "loss": 0.3626, + "step": 10910 + }, + { + "epoch": 0.7463574799917915, + "grad_norm": 2.7031958135248715, + "learning_rate": 1.5944215536282688e-06, + "loss": 0.4848, + "step": 10911 + }, + { + "epoch": 0.746425884123401, + "grad_norm": 2.1160964477156816, + "learning_rate": 1.5936105659367068e-06, + "loss": 0.6051, + "step": 10912 + }, + { + "epoch": 0.7464942882550106, + "grad_norm": 2.233807765211649, + "learning_rate": 1.5927997454471856e-06, + "loss": 0.337, + "step": 10913 + }, + { + "epoch": 0.7465626923866201, + "grad_norm": 2.002070283020256, + "learning_rate": 1.591989092199508e-06, + "loss": 0.2799, + "step": 10914 + }, + { + "epoch": 0.7466310965182297, + "grad_norm": 2.719656343769254, + "learning_rate": 1.5911786062334628e-06, + "loss": 0.4493, + "step": 10915 + }, + { + "epoch": 0.7466995006498393, + "grad_norm": 1.897858982870575, + "learning_rate": 1.590368287588832e-06, + "loss": 0.1995, + "step": 10916 + }, + { + "epoch": 0.7467679047814488, + "grad_norm": 2.1818890647850098, + "learning_rate": 1.5895581363053913e-06, + "loss": 0.339, + "step": 10917 + }, + { + "epoch": 0.7468363089130583, + "grad_norm": 1.6776353675840843, + "learning_rate": 1.5887481524229065e-06, + "loss": 0.3134, + "step": 10918 + }, + { + "epoch": 0.7469047130446679, + "grad_norm": 2.0525612363380765, + "learning_rate": 1.587938335981133e-06, + "loss": 0.5157, + "step": 10919 + }, + { + "epoch": 0.7469731171762775, + "grad_norm": 2.6847227850473385, + "learning_rate": 1.5871286870198237e-06, + "loss": 0.521, + "step": 10920 + }, + { + "epoch": 0.747041521307887, + "grad_norm": 1.9885788940200884, + "learning_rate": 1.5863192055787186e-06, + "loss": 0.2575, + "step": 10921 + }, + { + "epoch": 0.7471099254394965, + "grad_norm": 2.0013143273903333, + "learning_rate": 1.585509891697552e-06, + "loss": 0.343, + "step": 10922 + }, + { + "epoch": 0.7471783295711061, + "grad_norm": 3.362846816091404, + "learning_rate": 1.584700745416049e-06, + "loss": 0.247, + "step": 10923 + }, + { + "epoch": 0.7472467337027157, + "grad_norm": 3.182434516901295, + "learning_rate": 1.583891766773925e-06, + "loss": 0.4709, + "step": 10924 + }, + { + "epoch": 0.7473151378343252, + "grad_norm": 2.291151395581266, + "learning_rate": 1.58308295581089e-06, + "loss": 0.3244, + "step": 10925 + }, + { + "epoch": 0.7473835419659347, + "grad_norm": 2.2243171211895034, + "learning_rate": 1.5822743125666418e-06, + "loss": 0.3935, + "step": 10926 + }, + { + "epoch": 0.7474519460975443, + "grad_norm": 2.1583096439087615, + "learning_rate": 1.581465837080876e-06, + "loss": 0.4288, + "step": 10927 + }, + { + "epoch": 0.7475203502291539, + "grad_norm": 1.9959183632676984, + "learning_rate": 1.5806575293932746e-06, + "loss": 0.2669, + "step": 10928 + }, + { + "epoch": 0.7475887543607634, + "grad_norm": 2.2212339855447576, + "learning_rate": 1.5798493895435142e-06, + "loss": 0.4062, + "step": 10929 + }, + { + "epoch": 0.7476571584923729, + "grad_norm": 1.60345680045283, + "learning_rate": 1.5790414175712609e-06, + "loss": 0.225, + "step": 10930 + }, + { + "epoch": 0.7477255626239825, + "grad_norm": 2.146026645829162, + "learning_rate": 1.578233613516173e-06, + "loss": 0.3683, + "step": 10931 + }, + { + "epoch": 0.7477939667555921, + "grad_norm": 1.943130801766741, + "learning_rate": 1.577425977417904e-06, + "loss": 0.3771, + "step": 10932 + }, + { + "epoch": 0.7478623708872015, + "grad_norm": 1.9855496241617843, + "learning_rate": 1.5766185093160947e-06, + "loss": 0.3079, + "step": 10933 + }, + { + "epoch": 0.7479307750188111, + "grad_norm": 2.1666317743615733, + "learning_rate": 1.5758112092503814e-06, + "loss": 0.4223, + "step": 10934 + }, + { + "epoch": 0.7479991791504207, + "grad_norm": 2.1465800214566886, + "learning_rate": 1.5750040772603892e-06, + "loss": 0.2742, + "step": 10935 + }, + { + "epoch": 0.7480675832820303, + "grad_norm": 1.7733722204217648, + "learning_rate": 1.5741971133857364e-06, + "loss": 0.2382, + "step": 10936 + }, + { + "epoch": 0.7481359874136397, + "grad_norm": 1.6392756680766372, + "learning_rate": 1.573390317666032e-06, + "loss": 0.1395, + "step": 10937 + }, + { + "epoch": 0.7482043915452493, + "grad_norm": 2.1716033916655415, + "learning_rate": 1.572583690140878e-06, + "loss": 0.4103, + "step": 10938 + }, + { + "epoch": 0.7482727956768589, + "grad_norm": 2.04723971517348, + "learning_rate": 1.5717772308498651e-06, + "loss": 0.2895, + "step": 10939 + }, + { + "epoch": 0.7483411998084685, + "grad_norm": 2.479601526723005, + "learning_rate": 1.5709709398325828e-06, + "loss": 0.3534, + "step": 10940 + }, + { + "epoch": 0.7484096039400779, + "grad_norm": 1.5955791962217107, + "learning_rate": 1.5701648171286033e-06, + "loss": 0.1899, + "step": 10941 + }, + { + "epoch": 0.7484780080716875, + "grad_norm": 1.6290685815220163, + "learning_rate": 1.5693588627774986e-06, + "loss": 0.1991, + "step": 10942 + }, + { + "epoch": 0.7485464122032971, + "grad_norm": 1.9756996585927615, + "learning_rate": 1.5685530768188277e-06, + "loss": 0.347, + "step": 10943 + }, + { + "epoch": 0.7486148163349067, + "grad_norm": 2.0906068567306977, + "learning_rate": 1.567747459292142e-06, + "loss": 0.3143, + "step": 10944 + }, + { + "epoch": 0.7486832204665161, + "grad_norm": 1.6931947102282812, + "learning_rate": 1.5669420102369847e-06, + "loss": 0.302, + "step": 10945 + }, + { + "epoch": 0.7487516245981257, + "grad_norm": 1.705693355698836, + "learning_rate": 1.5661367296928908e-06, + "loss": 0.339, + "step": 10946 + }, + { + "epoch": 0.7488200287297353, + "grad_norm": 1.9096544577877266, + "learning_rate": 1.5653316176993893e-06, + "loss": 0.2356, + "step": 10947 + }, + { + "epoch": 0.7488884328613449, + "grad_norm": 2.103549648849194, + "learning_rate": 1.564526674295998e-06, + "loss": 0.4655, + "step": 10948 + }, + { + "epoch": 0.7489568369929543, + "grad_norm": 1.4845114946350757, + "learning_rate": 1.563721899522227e-06, + "loss": 0.1148, + "step": 10949 + }, + { + "epoch": 0.7490252411245639, + "grad_norm": 2.048337224838165, + "learning_rate": 1.562917293417579e-06, + "loss": 0.1569, + "step": 10950 + }, + { + "epoch": 0.7490936452561735, + "grad_norm": 2.1481831590783353, + "learning_rate": 1.5621128560215459e-06, + "loss": 0.3843, + "step": 10951 + }, + { + "epoch": 0.749162049387783, + "grad_norm": 2.1055512208657725, + "learning_rate": 1.5613085873736172e-06, + "loss": 0.4944, + "step": 10952 + }, + { + "epoch": 0.7492304535193925, + "grad_norm": 2.2191500370287347, + "learning_rate": 1.5605044875132664e-06, + "loss": 0.4287, + "step": 10953 + }, + { + "epoch": 0.7492988576510021, + "grad_norm": 2.0212564238503408, + "learning_rate": 1.5597005564799661e-06, + "loss": 0.3658, + "step": 10954 + }, + { + "epoch": 0.7493672617826117, + "grad_norm": 1.8243788581525069, + "learning_rate": 1.5588967943131754e-06, + "loss": 0.2423, + "step": 10955 + }, + { + "epoch": 0.7494356659142212, + "grad_norm": 2.1374741636354453, + "learning_rate": 1.5580932010523474e-06, + "loss": 0.3834, + "step": 10956 + }, + { + "epoch": 0.7495040700458308, + "grad_norm": 1.5463281975125887, + "learning_rate": 1.5572897767369262e-06, + "loss": 0.2299, + "step": 10957 + }, + { + "epoch": 0.7495724741774403, + "grad_norm": 1.85868206255317, + "learning_rate": 1.5564865214063462e-06, + "loss": 0.282, + "step": 10958 + }, + { + "epoch": 0.7496408783090499, + "grad_norm": 1.7295728176800191, + "learning_rate": 1.5556834351000356e-06, + "loss": 0.3658, + "step": 10959 + }, + { + "epoch": 0.7497092824406594, + "grad_norm": 2.3429867200935037, + "learning_rate": 1.5548805178574162e-06, + "loss": 0.3399, + "step": 10960 + }, + { + "epoch": 0.749777686572269, + "grad_norm": 2.5602134295432064, + "learning_rate": 1.554077769717895e-06, + "loss": 0.59, + "step": 10961 + }, + { + "epoch": 0.7498460907038785, + "grad_norm": 1.9828117475816338, + "learning_rate": 1.5532751907208793e-06, + "loss": 0.2848, + "step": 10962 + }, + { + "epoch": 0.7499144948354881, + "grad_norm": 2.5850081142246535, + "learning_rate": 1.5524727809057604e-06, + "loss": 0.3301, + "step": 10963 + }, + { + "epoch": 0.7499828989670976, + "grad_norm": 1.9780985175073056, + "learning_rate": 1.5516705403119264e-06, + "loss": 0.1314, + "step": 10964 + }, + { + "epoch": 0.7500513030987072, + "grad_norm": 2.4806422495800886, + "learning_rate": 1.5508684689787534e-06, + "loss": 0.4966, + "step": 10965 + }, + { + "epoch": 0.7501197072303167, + "grad_norm": 2.5117277023606768, + "learning_rate": 1.5500665669456105e-06, + "loss": 0.5085, + "step": 10966 + }, + { + "epoch": 0.7501881113619263, + "grad_norm": 2.6574727105276352, + "learning_rate": 1.5492648342518613e-06, + "loss": 0.4164, + "step": 10967 + }, + { + "epoch": 0.7502565154935358, + "grad_norm": 2.499749618151407, + "learning_rate": 1.5484632709368574e-06, + "loss": 0.3357, + "step": 10968 + }, + { + "epoch": 0.7503249196251454, + "grad_norm": 2.4256920053279463, + "learning_rate": 1.5476618770399436e-06, + "loss": 0.4792, + "step": 10969 + }, + { + "epoch": 0.7503933237567549, + "grad_norm": 2.2185322661416658, + "learning_rate": 1.5468606526004565e-06, + "loss": 0.3722, + "step": 10970 + }, + { + "epoch": 0.7504617278883644, + "grad_norm": 1.9329457940941799, + "learning_rate": 1.5460595976577215e-06, + "loss": 0.3374, + "step": 10971 + }, + { + "epoch": 0.750530132019974, + "grad_norm": 3.396629606786625, + "learning_rate": 1.545258712251062e-06, + "loss": 0.3858, + "step": 10972 + }, + { + "epoch": 0.7505985361515836, + "grad_norm": 2.1958131728311474, + "learning_rate": 1.5444579964197858e-06, + "loss": 0.2931, + "step": 10973 + }, + { + "epoch": 0.7506669402831931, + "grad_norm": 1.6783680812590527, + "learning_rate": 1.5436574502031998e-06, + "loss": 0.2057, + "step": 10974 + }, + { + "epoch": 0.7507353444148026, + "grad_norm": 2.2479109801738133, + "learning_rate": 1.5428570736405967e-06, + "loss": 0.4613, + "step": 10975 + }, + { + "epoch": 0.7508037485464122, + "grad_norm": 2.441494334348487, + "learning_rate": 1.5420568667712627e-06, + "loss": 0.3481, + "step": 10976 + }, + { + "epoch": 0.7508721526780218, + "grad_norm": 2.193446428536486, + "learning_rate": 1.541256829634476e-06, + "loss": 0.2157, + "step": 10977 + }, + { + "epoch": 0.7509405568096313, + "grad_norm": 2.1004176371680057, + "learning_rate": 1.5404569622695043e-06, + "loss": 0.3695, + "step": 10978 + }, + { + "epoch": 0.7510089609412408, + "grad_norm": 1.9847099291879975, + "learning_rate": 1.5396572647156126e-06, + "loss": 0.2786, + "step": 10979 + }, + { + "epoch": 0.7510773650728504, + "grad_norm": 2.24346226332213, + "learning_rate": 1.5388577370120506e-06, + "loss": 0.2957, + "step": 10980 + }, + { + "epoch": 0.75114576920446, + "grad_norm": 2.772225553027577, + "learning_rate": 1.5380583791980664e-06, + "loss": 0.5784, + "step": 10981 + }, + { + "epoch": 0.7512141733360695, + "grad_norm": 2.2283360160644103, + "learning_rate": 1.5372591913128941e-06, + "loss": 0.4308, + "step": 10982 + }, + { + "epoch": 0.751282577467679, + "grad_norm": 1.450297702038987, + "learning_rate": 1.5364601733957625e-06, + "loss": 0.2049, + "step": 10983 + }, + { + "epoch": 0.7513509815992886, + "grad_norm": 2.0720438932966982, + "learning_rate": 1.5356613254858905e-06, + "loss": 0.3664, + "step": 10984 + }, + { + "epoch": 0.7514193857308982, + "grad_norm": 2.3361516586422493, + "learning_rate": 1.5348626476224882e-06, + "loss": 0.3728, + "step": 10985 + }, + { + "epoch": 0.7514877898625077, + "grad_norm": 2.680778281413706, + "learning_rate": 1.534064139844762e-06, + "loss": 0.4787, + "step": 10986 + }, + { + "epoch": 0.7515561939941172, + "grad_norm": 2.084613808673395, + "learning_rate": 1.5332658021919045e-06, + "loss": 0.3096, + "step": 10987 + }, + { + "epoch": 0.7516245981257268, + "grad_norm": 2.361292375746741, + "learning_rate": 1.532467634703102e-06, + "loss": 0.4872, + "step": 10988 + }, + { + "epoch": 0.7516930022573364, + "grad_norm": 2.1684077119189986, + "learning_rate": 1.5316696374175328e-06, + "loss": 0.4202, + "step": 10989 + }, + { + "epoch": 0.7517614063889458, + "grad_norm": 1.9149112532322905, + "learning_rate": 1.5308718103743643e-06, + "loss": 0.2838, + "step": 10990 + }, + { + "epoch": 0.7518298105205554, + "grad_norm": 1.584060241443998, + "learning_rate": 1.5300741536127611e-06, + "loss": 0.2139, + "step": 10991 + }, + { + "epoch": 0.751898214652165, + "grad_norm": 2.4526336809641127, + "learning_rate": 1.5292766671718728e-06, + "loss": 0.3141, + "step": 10992 + }, + { + "epoch": 0.7519666187837746, + "grad_norm": 2.5059698055252886, + "learning_rate": 1.528479351090848e-06, + "loss": 0.3099, + "step": 10993 + }, + { + "epoch": 0.752035022915384, + "grad_norm": 1.9844608288323282, + "learning_rate": 1.5276822054088197e-06, + "loss": 0.2795, + "step": 10994 + }, + { + "epoch": 0.7521034270469936, + "grad_norm": 2.271134019041296, + "learning_rate": 1.5268852301649168e-06, + "loss": 0.4941, + "step": 10995 + }, + { + "epoch": 0.7521718311786032, + "grad_norm": 2.58752571268252, + "learning_rate": 1.5260884253982583e-06, + "loss": 0.5866, + "step": 10996 + }, + { + "epoch": 0.7522402353102128, + "grad_norm": 2.4586294098272004, + "learning_rate": 1.5252917911479553e-06, + "loss": 0.2724, + "step": 10997 + }, + { + "epoch": 0.7523086394418222, + "grad_norm": 2.1930211826385126, + "learning_rate": 1.5244953274531083e-06, + "loss": 0.4301, + "step": 10998 + }, + { + "epoch": 0.7523770435734318, + "grad_norm": 2.9710491011557556, + "learning_rate": 1.5236990343528157e-06, + "loss": 0.3338, + "step": 10999 + }, + { + "epoch": 0.7524454477050414, + "grad_norm": 2.34243775825496, + "learning_rate": 1.5229029118861599e-06, + "loss": 0.4617, + "step": 11000 + }, + { + "epoch": 0.752513851836651, + "grad_norm": 1.789611686305046, + "learning_rate": 1.5221069600922212e-06, + "loss": 0.2256, + "step": 11001 + }, + { + "epoch": 0.7525822559682604, + "grad_norm": 2.072210075647784, + "learning_rate": 1.5213111790100676e-06, + "loss": 0.1411, + "step": 11002 + }, + { + "epoch": 0.75265066009987, + "grad_norm": 2.201749592121676, + "learning_rate": 1.52051556867876e-06, + "loss": 0.4056, + "step": 11003 + }, + { + "epoch": 0.7527190642314796, + "grad_norm": 2.4741218887607883, + "learning_rate": 1.5197201291373499e-06, + "loss": 0.6414, + "step": 11004 + }, + { + "epoch": 0.7527874683630892, + "grad_norm": 9.324570788612508, + "learning_rate": 1.5189248604248808e-06, + "loss": 0.4424, + "step": 11005 + }, + { + "epoch": 0.7528558724946987, + "grad_norm": 3.9656538863817152, + "learning_rate": 1.5181297625803909e-06, + "loss": 0.3504, + "step": 11006 + }, + { + "epoch": 0.7529242766263082, + "grad_norm": 2.1841107679353224, + "learning_rate": 1.5173348356429064e-06, + "loss": 0.5279, + "step": 11007 + }, + { + "epoch": 0.7529926807579178, + "grad_norm": 2.7753362898638656, + "learning_rate": 1.5165400796514456e-06, + "loss": 0.1906, + "step": 11008 + }, + { + "epoch": 0.7530610848895273, + "grad_norm": 2.2431627472012976, + "learning_rate": 1.515745494645019e-06, + "loss": 0.2996, + "step": 11009 + }, + { + "epoch": 0.7531294890211369, + "grad_norm": 4.035678102292499, + "learning_rate": 1.5149510806626272e-06, + "loss": 0.2468, + "step": 11010 + }, + { + "epoch": 0.7531978931527464, + "grad_norm": 2.8948314063240956, + "learning_rate": 1.5141568377432675e-06, + "loss": 0.3283, + "step": 11011 + }, + { + "epoch": 0.753266297284356, + "grad_norm": 2.232178648934192, + "learning_rate": 1.5133627659259208e-06, + "loss": 0.3342, + "step": 11012 + }, + { + "epoch": 0.7533347014159655, + "grad_norm": 2.4159651236779953, + "learning_rate": 1.5125688652495685e-06, + "loss": 0.565, + "step": 11013 + }, + { + "epoch": 0.7534031055475751, + "grad_norm": 1.8582484494981233, + "learning_rate": 1.5117751357531768e-06, + "loss": 0.3146, + "step": 11014 + }, + { + "epoch": 0.7534715096791846, + "grad_norm": 2.1717082652577107, + "learning_rate": 1.5109815774757058e-06, + "loss": 0.2945, + "step": 11015 + }, + { + "epoch": 0.7535399138107942, + "grad_norm": 1.9526195253500942, + "learning_rate": 1.5101881904561079e-06, + "loss": 0.2102, + "step": 11016 + }, + { + "epoch": 0.7536083179424037, + "grad_norm": 2.381777698137968, + "learning_rate": 1.5093949747333252e-06, + "loss": 0.5978, + "step": 11017 + }, + { + "epoch": 0.7536767220740133, + "grad_norm": 2.4678834046052276, + "learning_rate": 1.5086019303462913e-06, + "loss": 0.4551, + "step": 11018 + }, + { + "epoch": 0.7537451262056228, + "grad_norm": 2.043343460408993, + "learning_rate": 1.5078090573339365e-06, + "loss": 0.3268, + "step": 11019 + }, + { + "epoch": 0.7538135303372324, + "grad_norm": 2.777545470947959, + "learning_rate": 1.5070163557351747e-06, + "loss": 0.555, + "step": 11020 + }, + { + "epoch": 0.7538819344688419, + "grad_norm": 2.0039176159809644, + "learning_rate": 1.50622382558892e-06, + "loss": 0.4684, + "step": 11021 + }, + { + "epoch": 0.7539503386004515, + "grad_norm": 1.8237867476883642, + "learning_rate": 1.5054314669340703e-06, + "loss": 0.2897, + "step": 11022 + }, + { + "epoch": 0.754018742732061, + "grad_norm": 2.2740737509753854, + "learning_rate": 1.50463927980952e-06, + "loss": 0.485, + "step": 11023 + }, + { + "epoch": 0.7540871468636706, + "grad_norm": 2.1651148389220296, + "learning_rate": 1.5038472642541525e-06, + "loss": 0.531, + "step": 11024 + }, + { + "epoch": 0.7541555509952801, + "grad_norm": 2.0219946323242404, + "learning_rate": 1.503055420306842e-06, + "loss": 0.3479, + "step": 11025 + }, + { + "epoch": 0.7542239551268897, + "grad_norm": 1.8364007852278754, + "learning_rate": 1.5022637480064605e-06, + "loss": 0.145, + "step": 11026 + }, + { + "epoch": 0.7542923592584992, + "grad_norm": 2.26934468328155, + "learning_rate": 1.501472247391864e-06, + "loss": 0.4405, + "step": 11027 + }, + { + "epoch": 0.7543607633901087, + "grad_norm": 2.1167496865491815, + "learning_rate": 1.500680918501904e-06, + "loss": 0.2811, + "step": 11028 + }, + { + "epoch": 0.7544291675217183, + "grad_norm": 2.1424610450212254, + "learning_rate": 1.4998897613754226e-06, + "loss": 0.2605, + "step": 11029 + }, + { + "epoch": 0.7544975716533279, + "grad_norm": 2.676295360544245, + "learning_rate": 1.499098776051252e-06, + "loss": 0.322, + "step": 11030 + }, + { + "epoch": 0.7545659757849374, + "grad_norm": 1.7429307883047194, + "learning_rate": 1.4983079625682206e-06, + "loss": 0.2906, + "step": 11031 + }, + { + "epoch": 0.7546343799165469, + "grad_norm": 2.672645836893206, + "learning_rate": 1.4975173209651422e-06, + "loss": 0.4014, + "step": 11032 + }, + { + "epoch": 0.7547027840481565, + "grad_norm": 1.706663851624672, + "learning_rate": 1.4967268512808292e-06, + "loss": 0.2247, + "step": 11033 + }, + { + "epoch": 0.7547711881797661, + "grad_norm": 1.7744926011166926, + "learning_rate": 1.4959365535540786e-06, + "loss": 0.1418, + "step": 11034 + }, + { + "epoch": 0.7548395923113757, + "grad_norm": 2.2985674727338608, + "learning_rate": 1.495146427823684e-06, + "loss": 0.3056, + "step": 11035 + }, + { + "epoch": 0.7549079964429851, + "grad_norm": 2.4942289439839413, + "learning_rate": 1.4943564741284266e-06, + "loss": 0.2297, + "step": 11036 + }, + { + "epoch": 0.7549764005745947, + "grad_norm": 1.6284137685741606, + "learning_rate": 1.4935666925070825e-06, + "loss": 0.1171, + "step": 11037 + }, + { + "epoch": 0.7550448047062043, + "grad_norm": 1.7396351417280966, + "learning_rate": 1.4927770829984155e-06, + "loss": 0.2758, + "step": 11038 + }, + { + "epoch": 0.7551132088378139, + "grad_norm": 2.7830513253559355, + "learning_rate": 1.4919876456411875e-06, + "loss": 0.5896, + "step": 11039 + }, + { + "epoch": 0.7551816129694233, + "grad_norm": 2.7305472257630883, + "learning_rate": 1.4911983804741437e-06, + "loss": 0.4282, + "step": 11040 + }, + { + "epoch": 0.7552500171010329, + "grad_norm": 2.111052517189333, + "learning_rate": 1.4904092875360294e-06, + "loss": 0.2001, + "step": 11041 + }, + { + "epoch": 0.7553184212326425, + "grad_norm": 2.217412740383963, + "learning_rate": 1.4896203668655746e-06, + "loss": 0.5411, + "step": 11042 + }, + { + "epoch": 0.755386825364252, + "grad_norm": 2.7334098638504924, + "learning_rate": 1.488831618501504e-06, + "loss": 0.4072, + "step": 11043 + }, + { + "epoch": 0.7554552294958615, + "grad_norm": 2.6037833554772636, + "learning_rate": 1.488043042482532e-06, + "loss": 0.3075, + "step": 11044 + }, + { + "epoch": 0.7555236336274711, + "grad_norm": 2.3772812290100136, + "learning_rate": 1.4872546388473657e-06, + "loss": 0.2821, + "step": 11045 + }, + { + "epoch": 0.7555920377590807, + "grad_norm": 2.765710290574267, + "learning_rate": 1.4864664076347062e-06, + "loss": 0.4968, + "step": 11046 + }, + { + "epoch": 0.7556604418906901, + "grad_norm": 2.160997037419076, + "learning_rate": 1.4856783488832415e-06, + "loss": 0.4392, + "step": 11047 + }, + { + "epoch": 0.7557288460222997, + "grad_norm": 2.10763015637028, + "learning_rate": 1.4848904626316546e-06, + "loss": 0.3067, + "step": 11048 + }, + { + "epoch": 0.7557972501539093, + "grad_norm": 2.144167294217386, + "learning_rate": 1.4841027489186177e-06, + "loss": 0.5332, + "step": 11049 + }, + { + "epoch": 0.7558656542855189, + "grad_norm": 2.436437913391299, + "learning_rate": 1.4833152077827944e-06, + "loss": 0.3313, + "step": 11050 + }, + { + "epoch": 0.7559340584171284, + "grad_norm": 1.7355757414490478, + "learning_rate": 1.4825278392628446e-06, + "loss": 0.3188, + "step": 11051 + }, + { + "epoch": 0.7560024625487379, + "grad_norm": 2.1153857387443042, + "learning_rate": 1.4817406433974119e-06, + "loss": 0.588, + "step": 11052 + }, + { + "epoch": 0.7560708666803475, + "grad_norm": 2.328733223953471, + "learning_rate": 1.48095362022514e-06, + "loss": 0.5579, + "step": 11053 + }, + { + "epoch": 0.7561392708119571, + "grad_norm": 2.2709953719839007, + "learning_rate": 1.4801667697846573e-06, + "loss": 0.2682, + "step": 11054 + }, + { + "epoch": 0.7562076749435666, + "grad_norm": 4.844250659294299, + "learning_rate": 1.4793800921145874e-06, + "loss": 0.368, + "step": 11055 + }, + { + "epoch": 0.7562760790751761, + "grad_norm": 2.2696380524446718, + "learning_rate": 1.4785935872535428e-06, + "loss": 0.2508, + "step": 11056 + }, + { + "epoch": 0.7563444832067857, + "grad_norm": 1.9429504539386042, + "learning_rate": 1.4778072552401296e-06, + "loss": 0.3838, + "step": 11057 + }, + { + "epoch": 0.7564128873383953, + "grad_norm": 1.848558878843387, + "learning_rate": 1.4770210961129432e-06, + "loss": 0.2528, + "step": 11058 + }, + { + "epoch": 0.7564812914700048, + "grad_norm": 2.072071220922006, + "learning_rate": 1.476235109910576e-06, + "loss": 0.3369, + "step": 11059 + }, + { + "epoch": 0.7565496956016143, + "grad_norm": 2.0652361587508055, + "learning_rate": 1.4754492966716033e-06, + "loss": 0.1969, + "step": 11060 + }, + { + "epoch": 0.7566180997332239, + "grad_norm": 1.4570061883484062, + "learning_rate": 1.474663656434601e-06, + "loss": 0.1187, + "step": 11061 + }, + { + "epoch": 0.7566865038648334, + "grad_norm": 2.4074645259458007, + "learning_rate": 1.47387818923813e-06, + "loss": 0.3392, + "step": 11062 + }, + { + "epoch": 0.756754907996443, + "grad_norm": 2.254470324879156, + "learning_rate": 1.473092895120745e-06, + "loss": 0.1405, + "step": 11063 + }, + { + "epoch": 0.7568233121280525, + "grad_norm": 1.4823874166753341, + "learning_rate": 1.4723077741209918e-06, + "loss": 0.1939, + "step": 11064 + }, + { + "epoch": 0.7568917162596621, + "grad_norm": 1.696484024711852, + "learning_rate": 1.4715228262774067e-06, + "loss": 0.2993, + "step": 11065 + }, + { + "epoch": 0.7569601203912716, + "grad_norm": 3.0126388569245632, + "learning_rate": 1.470738051628522e-06, + "loss": 0.4785, + "step": 11066 + }, + { + "epoch": 0.7570285245228812, + "grad_norm": 2.489769662617196, + "learning_rate": 1.469953450212856e-06, + "loss": 0.2601, + "step": 11067 + }, + { + "epoch": 0.7570969286544907, + "grad_norm": 2.5518457081515264, + "learning_rate": 1.4691690220689209e-06, + "loss": 0.6227, + "step": 11068 + }, + { + "epoch": 0.7571653327861003, + "grad_norm": 2.422323357703503, + "learning_rate": 1.4683847672352207e-06, + "loss": 0.355, + "step": 11069 + }, + { + "epoch": 0.7572337369177098, + "grad_norm": 1.9100703432700927, + "learning_rate": 1.4676006857502484e-06, + "loss": 0.3543, + "step": 11070 + }, + { + "epoch": 0.7573021410493194, + "grad_norm": 2.3082934214886226, + "learning_rate": 1.466816777652494e-06, + "loss": 0.4211, + "step": 11071 + }, + { + "epoch": 0.757370545180929, + "grad_norm": 1.9488682747147965, + "learning_rate": 1.466033042980432e-06, + "loss": 0.2483, + "step": 11072 + }, + { + "epoch": 0.7574389493125385, + "grad_norm": 2.1481997479754456, + "learning_rate": 1.465249481772535e-06, + "loss": 0.5339, + "step": 11073 + }, + { + "epoch": 0.757507353444148, + "grad_norm": 2.3757607596811177, + "learning_rate": 1.4644660940672628e-06, + "loss": 0.5132, + "step": 11074 + }, + { + "epoch": 0.7575757575757576, + "grad_norm": 2.879790575290854, + "learning_rate": 1.4636828799030679e-06, + "loss": 0.5126, + "step": 11075 + }, + { + "epoch": 0.7576441617073671, + "grad_norm": 2.41925335779575, + "learning_rate": 1.4628998393183935e-06, + "loss": 0.4633, + "step": 11076 + }, + { + "epoch": 0.7577125658389767, + "grad_norm": 2.2236256019230782, + "learning_rate": 1.4621169723516738e-06, + "loss": 0.3771, + "step": 11077 + }, + { + "epoch": 0.7577809699705862, + "grad_norm": 2.621083357629768, + "learning_rate": 1.4613342790413393e-06, + "loss": 0.3557, + "step": 11078 + }, + { + "epoch": 0.7578493741021958, + "grad_norm": 1.9529980274889742, + "learning_rate": 1.4605517594258046e-06, + "loss": 0.3185, + "step": 11079 + }, + { + "epoch": 0.7579177782338054, + "grad_norm": 1.7257503912581413, + "learning_rate": 1.4597694135434831e-06, + "loss": 0.2202, + "step": 11080 + }, + { + "epoch": 0.7579861823654148, + "grad_norm": 2.070778373267814, + "learning_rate": 1.4589872414327745e-06, + "loss": 0.4328, + "step": 11081 + }, + { + "epoch": 0.7580545864970244, + "grad_norm": 2.550083824490965, + "learning_rate": 1.458205243132071e-06, + "loss": 0.668, + "step": 11082 + }, + { + "epoch": 0.758122990628634, + "grad_norm": 2.752432137812584, + "learning_rate": 1.4574234186797575e-06, + "loss": 0.3049, + "step": 11083 + }, + { + "epoch": 0.7581913947602436, + "grad_norm": 5.348312928345199, + "learning_rate": 1.456641768114208e-06, + "loss": 0.2678, + "step": 11084 + }, + { + "epoch": 0.758259798891853, + "grad_norm": 2.132204828573495, + "learning_rate": 1.4558602914737936e-06, + "loss": 0.383, + "step": 11085 + }, + { + "epoch": 0.7583282030234626, + "grad_norm": 2.2480964953450417, + "learning_rate": 1.4550789887968696e-06, + "loss": 0.3459, + "step": 11086 + }, + { + "epoch": 0.7583966071550722, + "grad_norm": 2.1631221139608074, + "learning_rate": 1.454297860121787e-06, + "loss": 0.3219, + "step": 11087 + }, + { + "epoch": 0.7584650112866818, + "grad_norm": 2.2467502065921434, + "learning_rate": 1.453516905486888e-06, + "loss": 0.3307, + "step": 11088 + }, + { + "epoch": 0.7585334154182912, + "grad_norm": 2.263867123784583, + "learning_rate": 1.4527361249305032e-06, + "loss": 0.4399, + "step": 11089 + }, + { + "epoch": 0.7586018195499008, + "grad_norm": 2.449679086551265, + "learning_rate": 1.4519555184909607e-06, + "loss": 0.6568, + "step": 11090 + }, + { + "epoch": 0.7586702236815104, + "grad_norm": 2.7894942379560512, + "learning_rate": 1.4511750862065731e-06, + "loss": 0.2719, + "step": 11091 + }, + { + "epoch": 0.75873862781312, + "grad_norm": 2.0946416664259524, + "learning_rate": 1.4503948281156511e-06, + "loss": 0.2812, + "step": 11092 + }, + { + "epoch": 0.7588070319447294, + "grad_norm": 2.3467665841550955, + "learning_rate": 1.449614744256491e-06, + "loss": 0.3984, + "step": 11093 + }, + { + "epoch": 0.758875436076339, + "grad_norm": 2.696388162672835, + "learning_rate": 1.4488348346673848e-06, + "loss": 0.3664, + "step": 11094 + }, + { + "epoch": 0.7589438402079486, + "grad_norm": 4.226792708242685, + "learning_rate": 1.4480550993866128e-06, + "loss": 0.5325, + "step": 11095 + }, + { + "epoch": 0.7590122443395582, + "grad_norm": 1.7093485591314954, + "learning_rate": 1.4472755384524489e-06, + "loss": 0.2008, + "step": 11096 + }, + { + "epoch": 0.7590806484711676, + "grad_norm": 2.1391521106282196, + "learning_rate": 1.4464961519031556e-06, + "loss": 0.3539, + "step": 11097 + }, + { + "epoch": 0.7591490526027772, + "grad_norm": 3.1438280160373573, + "learning_rate": 1.4457169397769932e-06, + "loss": 0.4579, + "step": 11098 + }, + { + "epoch": 0.7592174567343868, + "grad_norm": 2.6285302968890556, + "learning_rate": 1.4449379021122045e-06, + "loss": 0.3931, + "step": 11099 + }, + { + "epoch": 0.7592858608659963, + "grad_norm": 2.1918410619518194, + "learning_rate": 1.4441590389470327e-06, + "loss": 0.2534, + "step": 11100 + }, + { + "epoch": 0.7593542649976058, + "grad_norm": 1.9619263743066009, + "learning_rate": 1.4433803503197063e-06, + "loss": 0.4267, + "step": 11101 + }, + { + "epoch": 0.7594226691292154, + "grad_norm": 2.063450861426421, + "learning_rate": 1.4426018362684468e-06, + "loss": 0.4359, + "step": 11102 + }, + { + "epoch": 0.759491073260825, + "grad_norm": 2.5715231667538885, + "learning_rate": 1.441823496831468e-06, + "loss": 0.7494, + "step": 11103 + }, + { + "epoch": 0.7595594773924345, + "grad_norm": 2.4468985036352056, + "learning_rate": 1.4410453320469726e-06, + "loss": 0.2571, + "step": 11104 + }, + { + "epoch": 0.759627881524044, + "grad_norm": 2.487352805838547, + "learning_rate": 1.4402673419531598e-06, + "loss": 0.7348, + "step": 11105 + }, + { + "epoch": 0.7596962856556536, + "grad_norm": 2.518208580892869, + "learning_rate": 1.4394895265882153e-06, + "loss": 0.3124, + "step": 11106 + }, + { + "epoch": 0.7597646897872632, + "grad_norm": 3.7043104085219456, + "learning_rate": 1.4387118859903188e-06, + "loss": 0.3382, + "step": 11107 + }, + { + "epoch": 0.7598330939188727, + "grad_norm": 2.50257167867849, + "learning_rate": 1.4379344201976403e-06, + "loss": 0.4555, + "step": 11108 + }, + { + "epoch": 0.7599014980504822, + "grad_norm": 1.9857824470623866, + "learning_rate": 1.4371571292483395e-06, + "loss": 0.4466, + "step": 11109 + }, + { + "epoch": 0.7599699021820918, + "grad_norm": 3.8144949086953206, + "learning_rate": 1.4363800131805734e-06, + "loss": 0.3354, + "step": 11110 + }, + { + "epoch": 0.7600383063137014, + "grad_norm": 2.599675539523194, + "learning_rate": 1.435603072032483e-06, + "loss": 0.6448, + "step": 11111 + }, + { + "epoch": 0.7601067104453109, + "grad_norm": 2.2014537045282836, + "learning_rate": 1.4348263058422079e-06, + "loss": 0.3975, + "step": 11112 + }, + { + "epoch": 0.7601751145769204, + "grad_norm": 2.2140904109167425, + "learning_rate": 1.4340497146478738e-06, + "loss": 0.4018, + "step": 11113 + }, + { + "epoch": 0.76024351870853, + "grad_norm": 3.0732297166753746, + "learning_rate": 1.4332732984875996e-06, + "loss": 0.5618, + "step": 11114 + }, + { + "epoch": 0.7603119228401396, + "grad_norm": 2.234663687909732, + "learning_rate": 1.4324970573994955e-06, + "loss": 0.5619, + "step": 11115 + }, + { + "epoch": 0.7603803269717491, + "grad_norm": 1.779935728383466, + "learning_rate": 1.431720991421663e-06, + "loss": 0.2568, + "step": 11116 + }, + { + "epoch": 0.7604487311033586, + "grad_norm": 2.007082156927445, + "learning_rate": 1.4309451005921938e-06, + "loss": 0.2352, + "step": 11117 + }, + { + "epoch": 0.7605171352349682, + "grad_norm": 2.5949955557820035, + "learning_rate": 1.430169384949175e-06, + "loss": 0.4094, + "step": 11118 + }, + { + "epoch": 0.7605855393665777, + "grad_norm": 2.458524532210458, + "learning_rate": 1.4293938445306798e-06, + "loss": 0.5975, + "step": 11119 + }, + { + "epoch": 0.7606539434981873, + "grad_norm": 2.382852760350035, + "learning_rate": 1.4286184793747792e-06, + "loss": 0.2514, + "step": 11120 + }, + { + "epoch": 0.7607223476297968, + "grad_norm": 2.331519315110301, + "learning_rate": 1.4278432895195288e-06, + "loss": 0.4305, + "step": 11121 + }, + { + "epoch": 0.7607907517614064, + "grad_norm": 2.106478949478531, + "learning_rate": 1.4270682750029801e-06, + "loss": 0.2577, + "step": 11122 + }, + { + "epoch": 0.7608591558930159, + "grad_norm": 2.3341902638048047, + "learning_rate": 1.4262934358631742e-06, + "loss": 0.4629, + "step": 11123 + }, + { + "epoch": 0.7609275600246255, + "grad_norm": 2.19072537375886, + "learning_rate": 1.4255187721381414e-06, + "loss": 0.2595, + "step": 11124 + }, + { + "epoch": 0.760995964156235, + "grad_norm": 1.8074397489984275, + "learning_rate": 1.4247442838659103e-06, + "loss": 0.3942, + "step": 11125 + }, + { + "epoch": 0.7610643682878446, + "grad_norm": 2.280975299129509, + "learning_rate": 1.4239699710844946e-06, + "loss": 0.3777, + "step": 11126 + }, + { + "epoch": 0.7611327724194541, + "grad_norm": 2.412320579517848, + "learning_rate": 1.4231958338319007e-06, + "loss": 0.4157, + "step": 11127 + }, + { + "epoch": 0.7612011765510637, + "grad_norm": 2.332585407913541, + "learning_rate": 1.422421872146128e-06, + "loss": 0.5027, + "step": 11128 + }, + { + "epoch": 0.7612695806826733, + "grad_norm": 2.1873944938284127, + "learning_rate": 1.4216480860651637e-06, + "loss": 0.1995, + "step": 11129 + }, + { + "epoch": 0.7613379848142828, + "grad_norm": 2.3427187494634683, + "learning_rate": 1.4208744756269931e-06, + "loss": 0.4443, + "step": 11130 + }, + { + "epoch": 0.7614063889458923, + "grad_norm": 1.7311654593555716, + "learning_rate": 1.4201010408695847e-06, + "loss": 0.3154, + "step": 11131 + }, + { + "epoch": 0.7614747930775019, + "grad_norm": 2.222683160548558, + "learning_rate": 1.4193277818309065e-06, + "loss": 0.2828, + "step": 11132 + }, + { + "epoch": 0.7615431972091115, + "grad_norm": 2.6009893387396703, + "learning_rate": 1.4185546985489113e-06, + "loss": 0.2387, + "step": 11133 + }, + { + "epoch": 0.761611601340721, + "grad_norm": 2.020193725481459, + "learning_rate": 1.417781791061546e-06, + "loss": 0.4655, + "step": 11134 + }, + { + "epoch": 0.7616800054723305, + "grad_norm": 3.5768084373593028, + "learning_rate": 1.4170090594067487e-06, + "loss": 0.4171, + "step": 11135 + }, + { + "epoch": 0.7617484096039401, + "grad_norm": 7.367918696277473, + "learning_rate": 1.416236503622449e-06, + "loss": 0.278, + "step": 11136 + }, + { + "epoch": 0.7618168137355497, + "grad_norm": 3.6674680682965453, + "learning_rate": 1.415464123746566e-06, + "loss": 0.6727, + "step": 11137 + }, + { + "epoch": 0.7618852178671591, + "grad_norm": 2.232301536914427, + "learning_rate": 1.4146919198170145e-06, + "loss": 0.266, + "step": 11138 + }, + { + "epoch": 0.7619536219987687, + "grad_norm": 2.7926433026829978, + "learning_rate": 1.4139198918716956e-06, + "loss": 0.2128, + "step": 11139 + }, + { + "epoch": 0.7620220261303783, + "grad_norm": 2.7355054480436194, + "learning_rate": 1.413148039948507e-06, + "loss": 0.4885, + "step": 11140 + }, + { + "epoch": 0.7620904302619879, + "grad_norm": 2.8396776615147443, + "learning_rate": 1.4123763640853333e-06, + "loss": 0.5196, + "step": 11141 + }, + { + "epoch": 0.7621588343935973, + "grad_norm": 2.37850387502679, + "learning_rate": 1.4116048643200525e-06, + "loss": 0.3961, + "step": 11142 + }, + { + "epoch": 0.7622272385252069, + "grad_norm": 2.12628926407422, + "learning_rate": 1.4108335406905326e-06, + "loss": 0.3327, + "step": 11143 + }, + { + "epoch": 0.7622956426568165, + "grad_norm": 2.6799477653786328, + "learning_rate": 1.410062393234633e-06, + "loss": 0.2986, + "step": 11144 + }, + { + "epoch": 0.7623640467884261, + "grad_norm": 2.3614352717042655, + "learning_rate": 1.4092914219902087e-06, + "loss": 0.4566, + "step": 11145 + }, + { + "epoch": 0.7624324509200355, + "grad_norm": 2.5424299123362877, + "learning_rate": 1.4085206269951002e-06, + "loss": 0.2207, + "step": 11146 + }, + { + "epoch": 0.7625008550516451, + "grad_norm": 2.712973388112502, + "learning_rate": 1.4077500082871425e-06, + "loss": 0.4586, + "step": 11147 + }, + { + "epoch": 0.7625692591832547, + "grad_norm": 2.1015964278421477, + "learning_rate": 1.4069795659041619e-06, + "loss": 0.227, + "step": 11148 + }, + { + "epoch": 0.7626376633148643, + "grad_norm": 1.9611275949324212, + "learning_rate": 1.4062092998839722e-06, + "loss": 0.369, + "step": 11149 + }, + { + "epoch": 0.7627060674464737, + "grad_norm": 1.9270437872325792, + "learning_rate": 1.4054392102643866e-06, + "loss": 0.3082, + "step": 11150 + }, + { + "epoch": 0.7627744715780833, + "grad_norm": 2.095767728551039, + "learning_rate": 1.4046692970832004e-06, + "loss": 0.3307, + "step": 11151 + }, + { + "epoch": 0.7628428757096929, + "grad_norm": 3.6641691188306647, + "learning_rate": 1.4038995603782086e-06, + "loss": 0.3728, + "step": 11152 + }, + { + "epoch": 0.7629112798413025, + "grad_norm": 1.9765939823101104, + "learning_rate": 1.4031300001871917e-06, + "loss": 0.1863, + "step": 11153 + }, + { + "epoch": 0.7629796839729119, + "grad_norm": 2.4443036091366355, + "learning_rate": 1.4023606165479236e-06, + "loss": 0.3252, + "step": 11154 + }, + { + "epoch": 0.7630480881045215, + "grad_norm": 4.309341822241301, + "learning_rate": 1.4015914094981692e-06, + "loss": 0.3335, + "step": 11155 + }, + { + "epoch": 0.7631164922361311, + "grad_norm": 4.770033701331765, + "learning_rate": 1.4008223790756847e-06, + "loss": 0.5285, + "step": 11156 + }, + { + "epoch": 0.7631848963677406, + "grad_norm": 2.584652791897707, + "learning_rate": 1.4000535253182167e-06, + "loss": 0.4562, + "step": 11157 + }, + { + "epoch": 0.7632533004993501, + "grad_norm": 2.19816526857119, + "learning_rate": 1.3992848482635075e-06, + "loss": 0.2012, + "step": 11158 + }, + { + "epoch": 0.7633217046309597, + "grad_norm": 2.479780783578234, + "learning_rate": 1.3985163479492842e-06, + "loss": 0.5059, + "step": 11159 + }, + { + "epoch": 0.7633901087625693, + "grad_norm": 2.174357196486071, + "learning_rate": 1.3977480244132707e-06, + "loss": 0.5366, + "step": 11160 + }, + { + "epoch": 0.7634585128941788, + "grad_norm": 2.3489586364227124, + "learning_rate": 1.3969798776931799e-06, + "loss": 0.4947, + "step": 11161 + }, + { + "epoch": 0.7635269170257883, + "grad_norm": 2.4374580379688857, + "learning_rate": 1.3962119078267156e-06, + "loss": 0.285, + "step": 11162 + }, + { + "epoch": 0.7635953211573979, + "grad_norm": 4.298729458222112, + "learning_rate": 1.395444114851573e-06, + "loss": 0.5753, + "step": 11163 + }, + { + "epoch": 0.7636637252890075, + "grad_norm": 2.626575925044982, + "learning_rate": 1.394676498805438e-06, + "loss": 0.3102, + "step": 11164 + }, + { + "epoch": 0.763732129420617, + "grad_norm": 2.131385906002982, + "learning_rate": 1.393909059725992e-06, + "loss": 0.3929, + "step": 11165 + }, + { + "epoch": 0.7638005335522265, + "grad_norm": 2.1536789783539563, + "learning_rate": 1.3931417976509026e-06, + "loss": 0.3595, + "step": 11166 + }, + { + "epoch": 0.7638689376838361, + "grad_norm": 2.0377738285943647, + "learning_rate": 1.3923747126178317e-06, + "loss": 0.3514, + "step": 11167 + }, + { + "epoch": 0.7639373418154457, + "grad_norm": 2.778525046261942, + "learning_rate": 1.3916078046644305e-06, + "loss": 0.4609, + "step": 11168 + }, + { + "epoch": 0.7640057459470552, + "grad_norm": 2.1036327828851475, + "learning_rate": 1.3908410738283418e-06, + "loss": 0.3842, + "step": 11169 + }, + { + "epoch": 0.7640741500786647, + "grad_norm": 2.2559355751808283, + "learning_rate": 1.3900745201472026e-06, + "loss": 0.444, + "step": 11170 + }, + { + "epoch": 0.7641425542102743, + "grad_norm": 1.9359237350208498, + "learning_rate": 1.389308143658637e-06, + "loss": 0.1553, + "step": 11171 + }, + { + "epoch": 0.7642109583418838, + "grad_norm": 1.8701708395012226, + "learning_rate": 1.388541944400265e-06, + "loss": 0.2965, + "step": 11172 + }, + { + "epoch": 0.7642793624734934, + "grad_norm": 1.551464276311577, + "learning_rate": 1.3877759224096944e-06, + "loss": 0.2543, + "step": 11173 + }, + { + "epoch": 0.764347766605103, + "grad_norm": 2.0069148250027236, + "learning_rate": 1.3870100777245243e-06, + "loss": 0.4755, + "step": 11174 + }, + { + "epoch": 0.7644161707367125, + "grad_norm": 1.995219317196834, + "learning_rate": 1.3862444103823463e-06, + "loss": 0.2426, + "step": 11175 + }, + { + "epoch": 0.764484574868322, + "grad_norm": 1.923989799406051, + "learning_rate": 1.385478920420742e-06, + "loss": 0.2131, + "step": 11176 + }, + { + "epoch": 0.7645529789999316, + "grad_norm": 2.075461663599717, + "learning_rate": 1.3847136078772883e-06, + "loss": 0.4642, + "step": 11177 + }, + { + "epoch": 0.7646213831315412, + "grad_norm": 2.3860597933871706, + "learning_rate": 1.3839484727895474e-06, + "loss": 0.3959, + "step": 11178 + }, + { + "epoch": 0.7646897872631507, + "grad_norm": 1.7520962744910373, + "learning_rate": 1.3831835151950784e-06, + "loss": 0.2895, + "step": 11179 + }, + { + "epoch": 0.7647581913947602, + "grad_norm": 2.4511004068966713, + "learning_rate": 1.3824187351314277e-06, + "loss": 0.3856, + "step": 11180 + }, + { + "epoch": 0.7648265955263698, + "grad_norm": 2.70465597997113, + "learning_rate": 1.381654132636135e-06, + "loss": 0.4616, + "step": 11181 + }, + { + "epoch": 0.7648949996579794, + "grad_norm": 2.2870815415484533, + "learning_rate": 1.3808897077467303e-06, + "loss": 0.4592, + "step": 11182 + }, + { + "epoch": 0.7649634037895889, + "grad_norm": 1.7115849441329936, + "learning_rate": 1.3801254605007336e-06, + "loss": 0.2703, + "step": 11183 + }, + { + "epoch": 0.7650318079211984, + "grad_norm": 1.6978457856351528, + "learning_rate": 1.3793613909356612e-06, + "loss": 0.1496, + "step": 11184 + }, + { + "epoch": 0.765100212052808, + "grad_norm": 2.9359920364682455, + "learning_rate": 1.378597499089015e-06, + "loss": 0.4007, + "step": 11185 + }, + { + "epoch": 0.7651686161844176, + "grad_norm": 2.3585036860451956, + "learning_rate": 1.377833784998292e-06, + "loss": 0.6044, + "step": 11186 + }, + { + "epoch": 0.7652370203160271, + "grad_norm": 1.9860971967354972, + "learning_rate": 1.3770702487009773e-06, + "loss": 0.3171, + "step": 11187 + }, + { + "epoch": 0.7653054244476366, + "grad_norm": 2.610871177301253, + "learning_rate": 1.3763068902345488e-06, + "loss": 0.7071, + "step": 11188 + }, + { + "epoch": 0.7653738285792462, + "grad_norm": 2.087765010983963, + "learning_rate": 1.3755437096364782e-06, + "loss": 0.0849, + "step": 11189 + }, + { + "epoch": 0.7654422327108558, + "grad_norm": 2.4102446417775187, + "learning_rate": 1.3747807069442226e-06, + "loss": 0.2771, + "step": 11190 + }, + { + "epoch": 0.7655106368424652, + "grad_norm": 2.5009011075137693, + "learning_rate": 1.3740178821952377e-06, + "loss": 0.62, + "step": 11191 + }, + { + "epoch": 0.7655790409740748, + "grad_norm": 1.5812358552126695, + "learning_rate": 1.3732552354269651e-06, + "loss": 0.0939, + "step": 11192 + }, + { + "epoch": 0.7656474451056844, + "grad_norm": 2.0285236181450235, + "learning_rate": 1.372492766676839e-06, + "loss": 0.3071, + "step": 11193 + }, + { + "epoch": 0.765715849237294, + "grad_norm": 2.3156057714778484, + "learning_rate": 1.3717304759822847e-06, + "loss": 0.3301, + "step": 11194 + }, + { + "epoch": 0.7657842533689034, + "grad_norm": 2.3023634135348603, + "learning_rate": 1.370968363380719e-06, + "loss": 0.3812, + "step": 11195 + }, + { + "epoch": 0.765852657500513, + "grad_norm": 2.4390976511678506, + "learning_rate": 1.3702064289095495e-06, + "loss": 0.2327, + "step": 11196 + }, + { + "epoch": 0.7659210616321226, + "grad_norm": 2.1040334551177744, + "learning_rate": 1.3694446726061777e-06, + "loss": 0.2654, + "step": 11197 + }, + { + "epoch": 0.7659894657637322, + "grad_norm": 2.176294502524151, + "learning_rate": 1.3686830945079921e-06, + "loss": 0.5021, + "step": 11198 + }, + { + "epoch": 0.7660578698953416, + "grad_norm": 2.716629769117592, + "learning_rate": 1.3679216946523771e-06, + "loss": 0.536, + "step": 11199 + }, + { + "epoch": 0.7661262740269512, + "grad_norm": 2.7672873420301496, + "learning_rate": 1.367160473076704e-06, + "loss": 0.2554, + "step": 11200 + }, + { + "epoch": 0.7661946781585608, + "grad_norm": 1.7790253282312514, + "learning_rate": 1.3663994298183381e-06, + "loss": 0.2446, + "step": 11201 + }, + { + "epoch": 0.7662630822901704, + "grad_norm": 2.0646040761106432, + "learning_rate": 1.3656385649146342e-06, + "loss": 0.497, + "step": 11202 + }, + { + "epoch": 0.7663314864217798, + "grad_norm": 2.066491184760609, + "learning_rate": 1.3648778784029387e-06, + "loss": 0.295, + "step": 11203 + }, + { + "epoch": 0.7663998905533894, + "grad_norm": 1.84688428192259, + "learning_rate": 1.3641173703205917e-06, + "loss": 0.2752, + "step": 11204 + }, + { + "epoch": 0.766468294684999, + "grad_norm": 1.8048799296581972, + "learning_rate": 1.3633570407049224e-06, + "loss": 0.146, + "step": 11205 + }, + { + "epoch": 0.7665366988166086, + "grad_norm": 1.7899654475265732, + "learning_rate": 1.3625968895932502e-06, + "loss": 0.2751, + "step": 11206 + }, + { + "epoch": 0.766605102948218, + "grad_norm": 2.4561323210503128, + "learning_rate": 1.361836917022888e-06, + "loss": 0.3087, + "step": 11207 + }, + { + "epoch": 0.7666735070798276, + "grad_norm": 2.50843748581904, + "learning_rate": 1.3610771230311364e-06, + "loss": 0.3018, + "step": 11208 + }, + { + "epoch": 0.7667419112114372, + "grad_norm": 2.175952788187908, + "learning_rate": 1.3603175076552933e-06, + "loss": 0.2706, + "step": 11209 + }, + { + "epoch": 0.7668103153430467, + "grad_norm": 2.434271771564657, + "learning_rate": 1.359558070932641e-06, + "loss": 0.5919, + "step": 11210 + }, + { + "epoch": 0.7668787194746562, + "grad_norm": 2.1753705046222116, + "learning_rate": 1.3587988129004603e-06, + "loss": 0.3127, + "step": 11211 + }, + { + "epoch": 0.7669471236062658, + "grad_norm": 1.8264071389492385, + "learning_rate": 1.3580397335960166e-06, + "loss": 0.3051, + "step": 11212 + }, + { + "epoch": 0.7670155277378754, + "grad_norm": 1.8587912556790738, + "learning_rate": 1.35728083305657e-06, + "loss": 0.2229, + "step": 11213 + }, + { + "epoch": 0.7670839318694849, + "grad_norm": 2.1339292013209588, + "learning_rate": 1.3565221113193705e-06, + "loss": 0.5397, + "step": 11214 + }, + { + "epoch": 0.7671523360010944, + "grad_norm": 2.0049968748353684, + "learning_rate": 1.3557635684216598e-06, + "loss": 0.1415, + "step": 11215 + }, + { + "epoch": 0.767220740132704, + "grad_norm": 1.8554232461268545, + "learning_rate": 1.3550052044006694e-06, + "loss": 0.2648, + "step": 11216 + }, + { + "epoch": 0.7672891442643136, + "grad_norm": 2.0527491836667022, + "learning_rate": 1.3542470192936268e-06, + "loss": 0.1927, + "step": 11217 + }, + { + "epoch": 0.7673575483959231, + "grad_norm": 2.4258907434689063, + "learning_rate": 1.3534890131377438e-06, + "loss": 0.3614, + "step": 11218 + }, + { + "epoch": 0.7674259525275327, + "grad_norm": 1.992330762910432, + "learning_rate": 1.3527311859702308e-06, + "loss": 0.2581, + "step": 11219 + }, + { + "epoch": 0.7674943566591422, + "grad_norm": 2.835328175894415, + "learning_rate": 1.3519735378282834e-06, + "loss": 0.2206, + "step": 11220 + }, + { + "epoch": 0.7675627607907518, + "grad_norm": 2.495979879584737, + "learning_rate": 1.351216068749091e-06, + "loss": 0.3516, + "step": 11221 + }, + { + "epoch": 0.7676311649223613, + "grad_norm": 2.0632821608472414, + "learning_rate": 1.3504587787698337e-06, + "loss": 0.4665, + "step": 11222 + }, + { + "epoch": 0.7676995690539709, + "grad_norm": 2.1867792031891726, + "learning_rate": 1.3497016679276815e-06, + "loss": 0.4234, + "step": 11223 + }, + { + "epoch": 0.7677679731855804, + "grad_norm": 2.654793766492584, + "learning_rate": 1.3489447362598002e-06, + "loss": 0.64, + "step": 11224 + }, + { + "epoch": 0.76783637731719, + "grad_norm": 2.652234127155752, + "learning_rate": 1.3481879838033417e-06, + "loss": 0.3316, + "step": 11225 + }, + { + "epoch": 0.7679047814487995, + "grad_norm": 1.9201718979616273, + "learning_rate": 1.3474314105954517e-06, + "loss": 0.3175, + "step": 11226 + }, + { + "epoch": 0.7679731855804091, + "grad_norm": 2.0805273439562715, + "learning_rate": 1.3466750166732662e-06, + "loss": 0.541, + "step": 11227 + }, + { + "epoch": 0.7680415897120186, + "grad_norm": 2.2807606342029705, + "learning_rate": 1.3459188020739105e-06, + "loss": 0.3925, + "step": 11228 + }, + { + "epoch": 0.7681099938436281, + "grad_norm": 2.6132673368668073, + "learning_rate": 1.3451627668345073e-06, + "loss": 0.3531, + "step": 11229 + }, + { + "epoch": 0.7681783979752377, + "grad_norm": 1.9075811323139142, + "learning_rate": 1.3444069109921632e-06, + "loss": 0.2, + "step": 11230 + }, + { + "epoch": 0.7682468021068473, + "grad_norm": 2.2787114611666186, + "learning_rate": 1.3436512345839819e-06, + "loss": 0.2228, + "step": 11231 + }, + { + "epoch": 0.7683152062384568, + "grad_norm": 2.0920857925422114, + "learning_rate": 1.3428957376470542e-06, + "loss": 0.3927, + "step": 11232 + }, + { + "epoch": 0.7683836103700663, + "grad_norm": 2.0382487182243527, + "learning_rate": 1.342140420218464e-06, + "loss": 0.4186, + "step": 11233 + }, + { + "epoch": 0.7684520145016759, + "grad_norm": 2.673499845829009, + "learning_rate": 1.3413852823352852e-06, + "loss": 0.2862, + "step": 11234 + }, + { + "epoch": 0.7685204186332855, + "grad_norm": 2.431886003362017, + "learning_rate": 1.3406303240345848e-06, + "loss": 0.573, + "step": 11235 + }, + { + "epoch": 0.768588822764895, + "grad_norm": 2.2477202919808583, + "learning_rate": 1.3398755453534169e-06, + "loss": 0.2883, + "step": 11236 + }, + { + "epoch": 0.7686572268965045, + "grad_norm": 2.2779120231884855, + "learning_rate": 1.339120946328833e-06, + "loss": 0.3756, + "step": 11237 + }, + { + "epoch": 0.7687256310281141, + "grad_norm": 2.60032977433685, + "learning_rate": 1.3383665269978701e-06, + "loss": 0.3422, + "step": 11238 + }, + { + "epoch": 0.7687940351597237, + "grad_norm": 2.5581315766116486, + "learning_rate": 1.3376122873975616e-06, + "loss": 0.3844, + "step": 11239 + }, + { + "epoch": 0.7688624392913332, + "grad_norm": 2.264656326786856, + "learning_rate": 1.3368582275649272e-06, + "loss": 0.4182, + "step": 11240 + }, + { + "epoch": 0.7689308434229427, + "grad_norm": 1.963879646402121, + "learning_rate": 1.3361043475369801e-06, + "loss": 0.437, + "step": 11241 + }, + { + "epoch": 0.7689992475545523, + "grad_norm": 1.9432464260076796, + "learning_rate": 1.3353506473507243e-06, + "loss": 0.1129, + "step": 11242 + }, + { + "epoch": 0.7690676516861619, + "grad_norm": 2.415184784783771, + "learning_rate": 1.334597127043154e-06, + "loss": 0.5364, + "step": 11243 + }, + { + "epoch": 0.7691360558177714, + "grad_norm": 1.7551486706050041, + "learning_rate": 1.3338437866512583e-06, + "loss": 0.2337, + "step": 11244 + }, + { + "epoch": 0.7692044599493809, + "grad_norm": 1.968505113751283, + "learning_rate": 1.333090626212013e-06, + "loss": 0.3001, + "step": 11245 + }, + { + "epoch": 0.7692728640809905, + "grad_norm": 2.749405753346758, + "learning_rate": 1.332337645762387e-06, + "loss": 0.268, + "step": 11246 + }, + { + "epoch": 0.7693412682126001, + "grad_norm": 4.561124326428267, + "learning_rate": 1.331584845339341e-06, + "loss": 0.3909, + "step": 11247 + }, + { + "epoch": 0.7694096723442095, + "grad_norm": 2.9050725274380222, + "learning_rate": 1.3308322249798234e-06, + "loss": 0.5986, + "step": 11248 + }, + { + "epoch": 0.7694780764758191, + "grad_norm": 2.1356970420778962, + "learning_rate": 1.3300797847207796e-06, + "loss": 0.4668, + "step": 11249 + }, + { + "epoch": 0.7695464806074287, + "grad_norm": 2.1061994774359363, + "learning_rate": 1.3293275245991405e-06, + "loss": 0.324, + "step": 11250 + }, + { + "epoch": 0.7696148847390383, + "grad_norm": 1.7894133927506553, + "learning_rate": 1.3285754446518335e-06, + "loss": 0.3025, + "step": 11251 + }, + { + "epoch": 0.7696832888706477, + "grad_norm": 1.5483428127389074, + "learning_rate": 1.3278235449157734e-06, + "loss": 0.3038, + "step": 11252 + }, + { + "epoch": 0.7697516930022573, + "grad_norm": 2.5103631284288968, + "learning_rate": 1.327071825427866e-06, + "loss": 0.3814, + "step": 11253 + }, + { + "epoch": 0.7698200971338669, + "grad_norm": 2.670230336987154, + "learning_rate": 1.3263202862250097e-06, + "loss": 0.3962, + "step": 11254 + }, + { + "epoch": 0.7698885012654765, + "grad_norm": 2.1368351683205473, + "learning_rate": 1.3255689273440937e-06, + "loss": 0.5518, + "step": 11255 + }, + { + "epoch": 0.7699569053970859, + "grad_norm": 1.8592865826717673, + "learning_rate": 1.324817748821997e-06, + "loss": 0.3613, + "step": 11256 + }, + { + "epoch": 0.7700253095286955, + "grad_norm": 2.37880908911718, + "learning_rate": 1.324066750695594e-06, + "loss": 0.3853, + "step": 11257 + }, + { + "epoch": 0.7700937136603051, + "grad_norm": 1.8390644310071689, + "learning_rate": 1.3233159330017442e-06, + "loss": 0.2467, + "step": 11258 + }, + { + "epoch": 0.7701621177919147, + "grad_norm": 1.5552640258090389, + "learning_rate": 1.3225652957773044e-06, + "loss": 0.1625, + "step": 11259 + }, + { + "epoch": 0.7702305219235241, + "grad_norm": 2.0579269458457135, + "learning_rate": 1.3218148390591184e-06, + "loss": 0.4812, + "step": 11260 + }, + { + "epoch": 0.7702989260551337, + "grad_norm": 1.8635584926876225, + "learning_rate": 1.3210645628840213e-06, + "loss": 0.2766, + "step": 11261 + }, + { + "epoch": 0.7703673301867433, + "grad_norm": 2.384476446540939, + "learning_rate": 1.3203144672888414e-06, + "loss": 0.3287, + "step": 11262 + }, + { + "epoch": 0.7704357343183529, + "grad_norm": 2.0163480437670693, + "learning_rate": 1.3195645523103946e-06, + "loss": 0.2661, + "step": 11263 + }, + { + "epoch": 0.7705041384499624, + "grad_norm": 1.6980553864835495, + "learning_rate": 1.3188148179854937e-06, + "loss": 0.239, + "step": 11264 + }, + { + "epoch": 0.7705725425815719, + "grad_norm": 1.687015618459166, + "learning_rate": 1.3180652643509384e-06, + "loss": 0.3026, + "step": 11265 + }, + { + "epoch": 0.7706409467131815, + "grad_norm": 1.9894670049668903, + "learning_rate": 1.3173158914435192e-06, + "loss": 0.3204, + "step": 11266 + }, + { + "epoch": 0.770709350844791, + "grad_norm": 2.6162894743735894, + "learning_rate": 1.3165666993000197e-06, + "loss": 0.5981, + "step": 11267 + }, + { + "epoch": 0.7707777549764006, + "grad_norm": 2.344696140052224, + "learning_rate": 1.315817687957212e-06, + "loss": 0.5532, + "step": 11268 + }, + { + "epoch": 0.7708461591080101, + "grad_norm": 2.533236104122597, + "learning_rate": 1.3150688574518649e-06, + "loss": 0.4803, + "step": 11269 + }, + { + "epoch": 0.7709145632396197, + "grad_norm": 2.396632468972615, + "learning_rate": 1.3143202078207307e-06, + "loss": 0.5217, + "step": 11270 + }, + { + "epoch": 0.7709829673712292, + "grad_norm": 2.3206119729417924, + "learning_rate": 1.3135717391005603e-06, + "loss": 0.4532, + "step": 11271 + }, + { + "epoch": 0.7710513715028388, + "grad_norm": 2.437659988899266, + "learning_rate": 1.3128234513280908e-06, + "loss": 0.4647, + "step": 11272 + }, + { + "epoch": 0.7711197756344483, + "grad_norm": 2.9241455761995145, + "learning_rate": 1.3120753445400513e-06, + "loss": 0.3332, + "step": 11273 + }, + { + "epoch": 0.7711881797660579, + "grad_norm": 2.2474269393294692, + "learning_rate": 1.3113274187731623e-06, + "loss": 0.3091, + "step": 11274 + }, + { + "epoch": 0.7712565838976674, + "grad_norm": 3.161717447705256, + "learning_rate": 1.3105796740641352e-06, + "loss": 0.442, + "step": 11275 + }, + { + "epoch": 0.771324988029277, + "grad_norm": 2.6020985643863948, + "learning_rate": 1.3098321104496747e-06, + "loss": 0.4422, + "step": 11276 + }, + { + "epoch": 0.7713933921608865, + "grad_norm": 3.04167237782051, + "learning_rate": 1.3090847279664725e-06, + "loss": 0.1854, + "step": 11277 + }, + { + "epoch": 0.7714617962924961, + "grad_norm": 2.266014453694962, + "learning_rate": 1.3083375266512165e-06, + "loss": 0.2484, + "step": 11278 + }, + { + "epoch": 0.7715302004241056, + "grad_norm": 2.3044447819547464, + "learning_rate": 1.307590506540582e-06, + "loss": 0.5336, + "step": 11279 + }, + { + "epoch": 0.7715986045557152, + "grad_norm": 2.0250074526272908, + "learning_rate": 1.3068436676712353e-06, + "loss": 0.3276, + "step": 11280 + }, + { + "epoch": 0.7716670086873247, + "grad_norm": 3.248320136213888, + "learning_rate": 1.3060970100798359e-06, + "loss": 0.2534, + "step": 11281 + }, + { + "epoch": 0.7717354128189343, + "grad_norm": 6.177269695316198, + "learning_rate": 1.3053505338030315e-06, + "loss": 0.551, + "step": 11282 + }, + { + "epoch": 0.7718038169505438, + "grad_norm": 2.4471343812814808, + "learning_rate": 1.3046042388774655e-06, + "loss": 0.3425, + "step": 11283 + }, + { + "epoch": 0.7718722210821534, + "grad_norm": 1.6821682818805124, + "learning_rate": 1.3038581253397682e-06, + "loss": 0.2738, + "step": 11284 + }, + { + "epoch": 0.7719406252137629, + "grad_norm": 2.166363281473191, + "learning_rate": 1.3031121932265628e-06, + "loss": 0.2361, + "step": 11285 + }, + { + "epoch": 0.7720090293453724, + "grad_norm": 2.201002722048257, + "learning_rate": 1.3023664425744624e-06, + "loss": 0.1668, + "step": 11286 + }, + { + "epoch": 0.772077433476982, + "grad_norm": 2.3206976689684047, + "learning_rate": 1.301620873420072e-06, + "loss": 0.5955, + "step": 11287 + }, + { + "epoch": 0.7721458376085916, + "grad_norm": 2.1852977089072447, + "learning_rate": 1.3008754857999894e-06, + "loss": 0.4561, + "step": 11288 + }, + { + "epoch": 0.7722142417402011, + "grad_norm": 2.6312975217736723, + "learning_rate": 1.3001302797507997e-06, + "loss": 0.5955, + "step": 11289 + }, + { + "epoch": 0.7722826458718106, + "grad_norm": 2.203469563993305, + "learning_rate": 1.299385255309084e-06, + "loss": 0.4548, + "step": 11290 + }, + { + "epoch": 0.7723510500034202, + "grad_norm": 2.5567556461720606, + "learning_rate": 1.29864041251141e-06, + "loss": 0.6124, + "step": 11291 + }, + { + "epoch": 0.7724194541350298, + "grad_norm": 2.093661769032754, + "learning_rate": 1.297895751394338e-06, + "loss": 0.5142, + "step": 11292 + }, + { + "epoch": 0.7724878582666393, + "grad_norm": 2.3646616285584283, + "learning_rate": 1.2971512719944196e-06, + "loss": 0.5514, + "step": 11293 + }, + { + "epoch": 0.7725562623982488, + "grad_norm": 2.127422638241903, + "learning_rate": 1.2964069743481978e-06, + "loss": 0.3383, + "step": 11294 + }, + { + "epoch": 0.7726246665298584, + "grad_norm": 2.032657253192704, + "learning_rate": 1.2956628584922048e-06, + "loss": 0.2659, + "step": 11295 + }, + { + "epoch": 0.772693070661468, + "grad_norm": 2.4170229576541202, + "learning_rate": 1.2949189244629684e-06, + "loss": 0.8079, + "step": 11296 + }, + { + "epoch": 0.7727614747930776, + "grad_norm": 2.4171438211983753, + "learning_rate": 1.2941751722970008e-06, + "loss": 0.3219, + "step": 11297 + }, + { + "epoch": 0.772829878924687, + "grad_norm": 2.414252656403884, + "learning_rate": 1.293431602030813e-06, + "loss": 0.409, + "step": 11298 + }, + { + "epoch": 0.7728982830562966, + "grad_norm": 2.5798320320810078, + "learning_rate": 1.2926882137009012e-06, + "loss": 0.633, + "step": 11299 + }, + { + "epoch": 0.7729666871879062, + "grad_norm": 2.595606834399493, + "learning_rate": 1.2919450073437539e-06, + "loss": 0.2347, + "step": 11300 + }, + { + "epoch": 0.7730350913195156, + "grad_norm": 3.1009618528271528, + "learning_rate": 1.2912019829958516e-06, + "loss": 0.3906, + "step": 11301 + }, + { + "epoch": 0.7731034954511252, + "grad_norm": 3.2045467942854233, + "learning_rate": 1.290459140693664e-06, + "loss": 0.466, + "step": 11302 + }, + { + "epoch": 0.7731718995827348, + "grad_norm": 2.5787949818595606, + "learning_rate": 1.2897164804736568e-06, + "loss": 0.403, + "step": 11303 + }, + { + "epoch": 0.7732403037143444, + "grad_norm": 1.1916676312750578, + "learning_rate": 1.2889740023722813e-06, + "loss": 0.1214, + "step": 11304 + }, + { + "epoch": 0.7733087078459538, + "grad_norm": 1.9819989058547363, + "learning_rate": 1.2882317064259824e-06, + "loss": 0.2255, + "step": 11305 + }, + { + "epoch": 0.7733771119775634, + "grad_norm": 2.0228377119744656, + "learning_rate": 1.287489592671195e-06, + "loss": 0.5082, + "step": 11306 + }, + { + "epoch": 0.773445516109173, + "grad_norm": 2.0867659391711, + "learning_rate": 1.2867476611443447e-06, + "loss": 0.2967, + "step": 11307 + }, + { + "epoch": 0.7735139202407826, + "grad_norm": 1.25817376925341, + "learning_rate": 1.2860059118818513e-06, + "loss": 0.1071, + "step": 11308 + }, + { + "epoch": 0.773582324372392, + "grad_norm": 1.335131775892307, + "learning_rate": 1.2852643449201213e-06, + "loss": 0.1522, + "step": 11309 + }, + { + "epoch": 0.7736507285040016, + "grad_norm": 2.1016747540412366, + "learning_rate": 1.2845229602955573e-06, + "loss": 0.2071, + "step": 11310 + }, + { + "epoch": 0.7737191326356112, + "grad_norm": 1.7555535106288225, + "learning_rate": 1.2837817580445478e-06, + "loss": 0.3826, + "step": 11311 + }, + { + "epoch": 0.7737875367672208, + "grad_norm": 2.1711253100418095, + "learning_rate": 1.283040738203475e-06, + "loss": 0.4276, + "step": 11312 + }, + { + "epoch": 0.7738559408988303, + "grad_norm": 1.7041654121220207, + "learning_rate": 1.2822999008087122e-06, + "loss": 0.3229, + "step": 11313 + }, + { + "epoch": 0.7739243450304398, + "grad_norm": 2.444365516084333, + "learning_rate": 1.2815592458966225e-06, + "loss": 0.3245, + "step": 11314 + }, + { + "epoch": 0.7739927491620494, + "grad_norm": 1.9641174088710545, + "learning_rate": 1.28081877350356e-06, + "loss": 0.3186, + "step": 11315 + }, + { + "epoch": 0.774061153293659, + "grad_norm": 2.258407078896102, + "learning_rate": 1.2800784836658732e-06, + "loss": 0.1662, + "step": 11316 + }, + { + "epoch": 0.7741295574252685, + "grad_norm": 1.8584696264458456, + "learning_rate": 1.2793383764198957e-06, + "loss": 0.2048, + "step": 11317 + }, + { + "epoch": 0.774197961556878, + "grad_norm": 2.3734728395544558, + "learning_rate": 1.2785984518019601e-06, + "loss": 0.39, + "step": 11318 + }, + { + "epoch": 0.7742663656884876, + "grad_norm": 2.300205203936592, + "learning_rate": 1.277858709848382e-06, + "loss": 0.5286, + "step": 11319 + }, + { + "epoch": 0.7743347698200971, + "grad_norm": 2.0239278353875094, + "learning_rate": 1.2771191505954727e-06, + "loss": 0.3265, + "step": 11320 + }, + { + "epoch": 0.7744031739517067, + "grad_norm": 2.7060590927765475, + "learning_rate": 1.2763797740795337e-06, + "loss": 0.2725, + "step": 11321 + }, + { + "epoch": 0.7744715780833162, + "grad_norm": 2.175166928528378, + "learning_rate": 1.2756405803368544e-06, + "loss": 0.491, + "step": 11322 + }, + { + "epoch": 0.7745399822149258, + "grad_norm": 1.5263640244463987, + "learning_rate": 1.274901569403722e-06, + "loss": 0.2841, + "step": 11323 + }, + { + "epoch": 0.7746083863465353, + "grad_norm": 2.7346215436686636, + "learning_rate": 1.2741627413164088e-06, + "loss": 0.7171, + "step": 11324 + }, + { + "epoch": 0.7746767904781449, + "grad_norm": 2.348436000770197, + "learning_rate": 1.2734240961111804e-06, + "loss": 0.3898, + "step": 11325 + }, + { + "epoch": 0.7747451946097544, + "grad_norm": 3.001417590033537, + "learning_rate": 1.2726856338242922e-06, + "loss": 0.2876, + "step": 11326 + }, + { + "epoch": 0.774813598741364, + "grad_norm": 2.546481517515702, + "learning_rate": 1.2719473544919913e-06, + "loss": 0.6469, + "step": 11327 + }, + { + "epoch": 0.7748820028729735, + "grad_norm": 2.2433472036715347, + "learning_rate": 1.271209258150518e-06, + "loss": 0.4553, + "step": 11328 + }, + { + "epoch": 0.7749504070045831, + "grad_norm": 2.0300869032622484, + "learning_rate": 1.270471344836099e-06, + "loss": 0.5051, + "step": 11329 + }, + { + "epoch": 0.7750188111361926, + "grad_norm": 1.774850526076905, + "learning_rate": 1.2697336145849576e-06, + "loss": 0.2909, + "step": 11330 + }, + { + "epoch": 0.7750872152678022, + "grad_norm": 2.0047785401577736, + "learning_rate": 1.2689960674333036e-06, + "loss": 0.2686, + "step": 11331 + }, + { + "epoch": 0.7751556193994117, + "grad_norm": 2.4894449396386427, + "learning_rate": 1.2682587034173393e-06, + "loss": 0.3348, + "step": 11332 + }, + { + "epoch": 0.7752240235310213, + "grad_norm": 2.1046894237733245, + "learning_rate": 1.2675215225732584e-06, + "loss": 0.3918, + "step": 11333 + }, + { + "epoch": 0.7752924276626308, + "grad_norm": 2.1589605081287084, + "learning_rate": 1.2667845249372452e-06, + "loss": 0.555, + "step": 11334 + }, + { + "epoch": 0.7753608317942404, + "grad_norm": 1.994961710588724, + "learning_rate": 1.2660477105454732e-06, + "loss": 0.3351, + "step": 11335 + }, + { + "epoch": 0.7754292359258499, + "grad_norm": 7.656660014481645, + "learning_rate": 1.2653110794341123e-06, + "loss": 0.3492, + "step": 11336 + }, + { + "epoch": 0.7754976400574595, + "grad_norm": 1.6200077831631765, + "learning_rate": 1.264574631639317e-06, + "loss": 0.1976, + "step": 11337 + }, + { + "epoch": 0.775566044189069, + "grad_norm": 2.8312093449719966, + "learning_rate": 1.2638383671972382e-06, + "loss": 0.3658, + "step": 11338 + }, + { + "epoch": 0.7756344483206785, + "grad_norm": 1.7896347067852734, + "learning_rate": 1.2631022861440145e-06, + "loss": 0.2592, + "step": 11339 + }, + { + "epoch": 0.7757028524522881, + "grad_norm": 2.1303394218222125, + "learning_rate": 1.2623663885157755e-06, + "loss": 0.5317, + "step": 11340 + }, + { + "epoch": 0.7757712565838977, + "grad_norm": 3.0154816390831978, + "learning_rate": 1.2616306743486428e-06, + "loss": 0.3144, + "step": 11341 + }, + { + "epoch": 0.7758396607155073, + "grad_norm": 1.671617563873675, + "learning_rate": 1.260895143678728e-06, + "loss": 0.2408, + "step": 11342 + }, + { + "epoch": 0.7759080648471167, + "grad_norm": 2.9928960733095993, + "learning_rate": 1.260159796542137e-06, + "loss": 0.6695, + "step": 11343 + }, + { + "epoch": 0.7759764689787263, + "grad_norm": 2.3931875892876846, + "learning_rate": 1.2594246329749622e-06, + "loss": 0.1577, + "step": 11344 + }, + { + "epoch": 0.7760448731103359, + "grad_norm": 2.1959322465253175, + "learning_rate": 1.2586896530132897e-06, + "loss": 0.4673, + "step": 11345 + }, + { + "epoch": 0.7761132772419455, + "grad_norm": 2.1688803328562365, + "learning_rate": 1.2579548566931959e-06, + "loss": 0.3157, + "step": 11346 + }, + { + "epoch": 0.7761816813735549, + "grad_norm": 2.369817267618475, + "learning_rate": 1.257220244050746e-06, + "loss": 0.4956, + "step": 11347 + }, + { + "epoch": 0.7762500855051645, + "grad_norm": 2.230773958345357, + "learning_rate": 1.2564858151220022e-06, + "loss": 0.3396, + "step": 11348 + }, + { + "epoch": 0.7763184896367741, + "grad_norm": 2.162698577572475, + "learning_rate": 1.2557515699430095e-06, + "loss": 0.1297, + "step": 11349 + }, + { + "epoch": 0.7763868937683837, + "grad_norm": 6.183043792231429, + "learning_rate": 1.255017508549813e-06, + "loss": 0.2047, + "step": 11350 + }, + { + "epoch": 0.7764552978999931, + "grad_norm": 1.477176155254975, + "learning_rate": 1.2542836309784407e-06, + "loss": 0.091, + "step": 11351 + }, + { + "epoch": 0.7765237020316027, + "grad_norm": 1.9301847674322028, + "learning_rate": 1.2535499372649163e-06, + "loss": 0.3067, + "step": 11352 + }, + { + "epoch": 0.7765921061632123, + "grad_norm": 1.974353845440114, + "learning_rate": 1.252816427445252e-06, + "loss": 0.1325, + "step": 11353 + }, + { + "epoch": 0.7766605102948219, + "grad_norm": 2.1292652002503965, + "learning_rate": 1.2520831015554525e-06, + "loss": 0.4128, + "step": 11354 + }, + { + "epoch": 0.7767289144264313, + "grad_norm": 2.043000791746986, + "learning_rate": 1.2513499596315116e-06, + "loss": 0.3775, + "step": 11355 + }, + { + "epoch": 0.7767973185580409, + "grad_norm": 2.022797619596804, + "learning_rate": 1.2506170017094183e-06, + "loss": 0.2109, + "step": 11356 + }, + { + "epoch": 0.7768657226896505, + "grad_norm": 2.076039659052859, + "learning_rate": 1.249884227825146e-06, + "loss": 0.4822, + "step": 11357 + }, + { + "epoch": 0.77693412682126, + "grad_norm": 1.9888582198131792, + "learning_rate": 1.2491516380146667e-06, + "loss": 0.3002, + "step": 11358 + }, + { + "epoch": 0.7770025309528695, + "grad_norm": 1.9751289582100278, + "learning_rate": 1.2484192323139382e-06, + "loss": 0.3916, + "step": 11359 + }, + { + "epoch": 0.7770709350844791, + "grad_norm": 1.8889340728571156, + "learning_rate": 1.2476870107589096e-06, + "loss": 0.3747, + "step": 11360 + }, + { + "epoch": 0.7771393392160887, + "grad_norm": 1.8035898173115095, + "learning_rate": 1.2469549733855219e-06, + "loss": 0.3889, + "step": 11361 + }, + { + "epoch": 0.7772077433476982, + "grad_norm": 2.1156007725893455, + "learning_rate": 1.2462231202297065e-06, + "loss": 0.1727, + "step": 11362 + }, + { + "epoch": 0.7772761474793077, + "grad_norm": 2.0170064044095626, + "learning_rate": 1.245491451327388e-06, + "loss": 0.3651, + "step": 11363 + }, + { + "epoch": 0.7773445516109173, + "grad_norm": 2.2691789442265615, + "learning_rate": 1.2447599667144799e-06, + "loss": 0.4165, + "step": 11364 + }, + { + "epoch": 0.7774129557425269, + "grad_norm": 2.1648596627770136, + "learning_rate": 1.2440286664268863e-06, + "loss": 0.5511, + "step": 11365 + }, + { + "epoch": 0.7774813598741364, + "grad_norm": 2.2893694974461902, + "learning_rate": 1.2432975505005035e-06, + "loss": 0.3682, + "step": 11366 + }, + { + "epoch": 0.7775497640057459, + "grad_norm": 1.8358597921605315, + "learning_rate": 1.242566618971216e-06, + "loss": 0.2063, + "step": 11367 + }, + { + "epoch": 0.7776181681373555, + "grad_norm": 2.433257580311756, + "learning_rate": 1.2418358718749047e-06, + "loss": 0.5196, + "step": 11368 + }, + { + "epoch": 0.7776865722689651, + "grad_norm": 2.1726429425503815, + "learning_rate": 1.2411053092474356e-06, + "loss": 0.2149, + "step": 11369 + }, + { + "epoch": 0.7777549764005746, + "grad_norm": 2.9799481758469373, + "learning_rate": 1.240374931124671e-06, + "loss": 0.2135, + "step": 11370 + }, + { + "epoch": 0.7778233805321841, + "grad_norm": 2.628005509716221, + "learning_rate": 1.2396447375424597e-06, + "loss": 0.2254, + "step": 11371 + }, + { + "epoch": 0.7778917846637937, + "grad_norm": 2.3546422517374044, + "learning_rate": 1.2389147285366432e-06, + "loss": 0.187, + "step": 11372 + }, + { + "epoch": 0.7779601887954033, + "grad_norm": 2.0006019982681393, + "learning_rate": 1.2381849041430538e-06, + "loss": 0.4185, + "step": 11373 + }, + { + "epoch": 0.7780285929270128, + "grad_norm": 1.9938045745363964, + "learning_rate": 1.2374552643975136e-06, + "loss": 0.381, + "step": 11374 + }, + { + "epoch": 0.7780969970586223, + "grad_norm": 2.526363554171503, + "learning_rate": 1.2367258093358391e-06, + "loss": 0.5341, + "step": 11375 + }, + { + "epoch": 0.7781654011902319, + "grad_norm": 1.9656626168610336, + "learning_rate": 1.2359965389938354e-06, + "loss": 0.3071, + "step": 11376 + }, + { + "epoch": 0.7782338053218414, + "grad_norm": 3.3463894417928373, + "learning_rate": 1.2352674534072956e-06, + "loss": 0.3984, + "step": 11377 + }, + { + "epoch": 0.778302209453451, + "grad_norm": 2.332705452999476, + "learning_rate": 1.2345385526120108e-06, + "loss": 0.6093, + "step": 11378 + }, + { + "epoch": 0.7783706135850605, + "grad_norm": 2.137720699516803, + "learning_rate": 1.2338098366437574e-06, + "loss": 0.3781, + "step": 11379 + }, + { + "epoch": 0.7784390177166701, + "grad_norm": 2.3325057194017162, + "learning_rate": 1.2330813055383034e-06, + "loss": 0.2272, + "step": 11380 + }, + { + "epoch": 0.7785074218482796, + "grad_norm": 2.464560382109505, + "learning_rate": 1.2323529593314076e-06, + "loss": 0.5684, + "step": 11381 + }, + { + "epoch": 0.7785758259798892, + "grad_norm": 1.8866781107946626, + "learning_rate": 1.2316247980588243e-06, + "loss": 0.2647, + "step": 11382 + }, + { + "epoch": 0.7786442301114987, + "grad_norm": 1.2022965229220361, + "learning_rate": 1.2308968217562928e-06, + "loss": 0.0989, + "step": 11383 + }, + { + "epoch": 0.7787126342431083, + "grad_norm": 1.8438093773930597, + "learning_rate": 1.2301690304595459e-06, + "loss": 0.3792, + "step": 11384 + }, + { + "epoch": 0.7787810383747178, + "grad_norm": 2.0403934899485408, + "learning_rate": 1.229441424204308e-06, + "loss": 0.2328, + "step": 11385 + }, + { + "epoch": 0.7788494425063274, + "grad_norm": 2.267631457849139, + "learning_rate": 1.2287140030262911e-06, + "loss": 0.3284, + "step": 11386 + }, + { + "epoch": 0.778917846637937, + "grad_norm": 2.2980423769857405, + "learning_rate": 1.2279867669612032e-06, + "loss": 0.5794, + "step": 11387 + }, + { + "epoch": 0.7789862507695465, + "grad_norm": 2.1823896729002095, + "learning_rate": 1.227259716044739e-06, + "loss": 0.346, + "step": 11388 + }, + { + "epoch": 0.779054654901156, + "grad_norm": 2.3058255117683757, + "learning_rate": 1.2265328503125873e-06, + "loss": 0.49, + "step": 11389 + }, + { + "epoch": 0.7791230590327656, + "grad_norm": 2.035712512064522, + "learning_rate": 1.2258061698004254e-06, + "loss": 0.2687, + "step": 11390 + }, + { + "epoch": 0.7791914631643752, + "grad_norm": 1.8345126933809917, + "learning_rate": 1.2250796745439219e-06, + "loss": 0.322, + "step": 11391 + }, + { + "epoch": 0.7792598672959847, + "grad_norm": 2.4007523266495903, + "learning_rate": 1.224353364578737e-06, + "loss": 0.3934, + "step": 11392 + }, + { + "epoch": 0.7793282714275942, + "grad_norm": 5.332552024365768, + "learning_rate": 1.223627239940522e-06, + "loss": 0.4231, + "step": 11393 + }, + { + "epoch": 0.7793966755592038, + "grad_norm": 2.417300432374458, + "learning_rate": 1.2229013006649154e-06, + "loss": 0.3056, + "step": 11394 + }, + { + "epoch": 0.7794650796908134, + "grad_norm": 3.0880971087933524, + "learning_rate": 1.2221755467875547e-06, + "loss": 0.4338, + "step": 11395 + }, + { + "epoch": 0.7795334838224228, + "grad_norm": 2.4377623441319276, + "learning_rate": 1.2214499783440592e-06, + "loss": 0.3292, + "step": 11396 + }, + { + "epoch": 0.7796018879540324, + "grad_norm": 2.3340165969986657, + "learning_rate": 1.2207245953700464e-06, + "loss": 0.5662, + "step": 11397 + }, + { + "epoch": 0.779670292085642, + "grad_norm": 1.8997963194625067, + "learning_rate": 1.219999397901121e-06, + "loss": 0.2846, + "step": 11398 + }, + { + "epoch": 0.7797386962172516, + "grad_norm": 2.4264280528829953, + "learning_rate": 1.2192743859728784e-06, + "loss": 0.4989, + "step": 11399 + }, + { + "epoch": 0.779807100348861, + "grad_norm": 4.187452714567666, + "learning_rate": 1.2185495596209058e-06, + "loss": 0.5868, + "step": 11400 + }, + { + "epoch": 0.7798755044804706, + "grad_norm": 1.9832284273190652, + "learning_rate": 1.2178249188807796e-06, + "loss": 0.4006, + "step": 11401 + }, + { + "epoch": 0.7799439086120802, + "grad_norm": 1.7872433137991386, + "learning_rate": 1.2171004637880719e-06, + "loss": 0.3492, + "step": 11402 + }, + { + "epoch": 0.7800123127436898, + "grad_norm": 2.393683035937275, + "learning_rate": 1.2163761943783408e-06, + "loss": 0.4697, + "step": 11403 + }, + { + "epoch": 0.7800807168752992, + "grad_norm": 1.8157102748654081, + "learning_rate": 1.2156521106871373e-06, + "loss": 0.1607, + "step": 11404 + }, + { + "epoch": 0.7801491210069088, + "grad_norm": 2.6300204269718255, + "learning_rate": 1.2149282127500028e-06, + "loss": 0.6968, + "step": 11405 + }, + { + "epoch": 0.7802175251385184, + "grad_norm": 2.1426292003013803, + "learning_rate": 1.2142045006024678e-06, + "loss": 0.4985, + "step": 11406 + }, + { + "epoch": 0.780285929270128, + "grad_norm": 2.429302516255359, + "learning_rate": 1.2134809742800584e-06, + "loss": 0.4052, + "step": 11407 + }, + { + "epoch": 0.7803543334017374, + "grad_norm": 2.1716995580055602, + "learning_rate": 1.2127576338182868e-06, + "loss": 0.394, + "step": 11408 + }, + { + "epoch": 0.780422737533347, + "grad_norm": 4.019288985326033, + "learning_rate": 1.21203447925266e-06, + "loss": 0.5549, + "step": 11409 + }, + { + "epoch": 0.7804911416649566, + "grad_norm": 2.832884624423613, + "learning_rate": 1.2113115106186728e-06, + "loss": 0.4328, + "step": 11410 + }, + { + "epoch": 0.7805595457965662, + "grad_norm": 2.548844272372373, + "learning_rate": 1.2105887279518124e-06, + "loss": 0.4942, + "step": 11411 + }, + { + "epoch": 0.7806279499281756, + "grad_norm": 2.775556587556366, + "learning_rate": 1.2098661312875566e-06, + "loss": 0.6958, + "step": 11412 + }, + { + "epoch": 0.7806963540597852, + "grad_norm": 2.909800391233401, + "learning_rate": 1.2091437206613726e-06, + "loss": 0.3082, + "step": 11413 + }, + { + "epoch": 0.7807647581913948, + "grad_norm": 1.961914358713404, + "learning_rate": 1.2084214961087192e-06, + "loss": 0.274, + "step": 11414 + }, + { + "epoch": 0.7808331623230043, + "grad_norm": 2.140119151748947, + "learning_rate": 1.2076994576650502e-06, + "loss": 0.3017, + "step": 11415 + }, + { + "epoch": 0.7809015664546138, + "grad_norm": 2.4019345873236624, + "learning_rate": 1.2069776053658028e-06, + "loss": 0.3712, + "step": 11416 + }, + { + "epoch": 0.7809699705862234, + "grad_norm": 2.663320723375667, + "learning_rate": 1.2062559392464124e-06, + "loss": 0.585, + "step": 11417 + }, + { + "epoch": 0.781038374717833, + "grad_norm": 2.3296677898798714, + "learning_rate": 1.2055344593423e-06, + "loss": 0.3482, + "step": 11418 + }, + { + "epoch": 0.7811067788494425, + "grad_norm": 2.315149220277461, + "learning_rate": 1.2048131656888801e-06, + "loss": 0.5671, + "step": 11419 + }, + { + "epoch": 0.781175182981052, + "grad_norm": 1.664846942814415, + "learning_rate": 1.2040920583215566e-06, + "loss": 0.2108, + "step": 11420 + }, + { + "epoch": 0.7812435871126616, + "grad_norm": 2.57572392870521, + "learning_rate": 1.2033711372757234e-06, + "loss": 0.3347, + "step": 11421 + }, + { + "epoch": 0.7813119912442712, + "grad_norm": 2.2698132968366225, + "learning_rate": 1.2026504025867702e-06, + "loss": 0.5102, + "step": 11422 + }, + { + "epoch": 0.7813803953758807, + "grad_norm": 2.035326029894943, + "learning_rate": 1.2019298542900725e-06, + "loss": 0.4566, + "step": 11423 + }, + { + "epoch": 0.7814487995074902, + "grad_norm": 2.559849915556697, + "learning_rate": 1.2012094924209982e-06, + "loss": 0.6179, + "step": 11424 + }, + { + "epoch": 0.7815172036390998, + "grad_norm": 1.7008604340525675, + "learning_rate": 1.2004893170149062e-06, + "loss": 0.1414, + "step": 11425 + }, + { + "epoch": 0.7815856077707094, + "grad_norm": 1.6876401489834667, + "learning_rate": 1.1997693281071442e-06, + "loss": 0.3298, + "step": 11426 + }, + { + "epoch": 0.7816540119023189, + "grad_norm": 2.115721994202275, + "learning_rate": 1.199049525733057e-06, + "loss": 0.3648, + "step": 11427 + }, + { + "epoch": 0.7817224160339284, + "grad_norm": 1.893768503079809, + "learning_rate": 1.1983299099279716e-06, + "loss": 0.3214, + "step": 11428 + }, + { + "epoch": 0.781790820165538, + "grad_norm": 2.3693907905199354, + "learning_rate": 1.1976104807272143e-06, + "loss": 0.4924, + "step": 11429 + }, + { + "epoch": 0.7818592242971475, + "grad_norm": 1.6434644469195245, + "learning_rate": 1.1968912381660957e-06, + "loss": 0.2345, + "step": 11430 + }, + { + "epoch": 0.7819276284287571, + "grad_norm": 2.0840264935514057, + "learning_rate": 1.1961721822799204e-06, + "loss": 0.3182, + "step": 11431 + }, + { + "epoch": 0.7819960325603666, + "grad_norm": 1.8742430231797205, + "learning_rate": 1.195453313103983e-06, + "loss": 0.3294, + "step": 11432 + }, + { + "epoch": 0.7820644366919762, + "grad_norm": 1.9614445802007563, + "learning_rate": 1.1947346306735685e-06, + "loss": 0.2316, + "step": 11433 + }, + { + "epoch": 0.7821328408235857, + "grad_norm": 1.7998373574051392, + "learning_rate": 1.1940161350239527e-06, + "loss": 0.2014, + "step": 11434 + }, + { + "epoch": 0.7822012449551953, + "grad_norm": 2.7989216354305873, + "learning_rate": 1.1932978261904054e-06, + "loss": 0.5206, + "step": 11435 + }, + { + "epoch": 0.7822696490868049, + "grad_norm": 2.824966468865225, + "learning_rate": 1.1925797042081815e-06, + "loss": 0.2035, + "step": 11436 + }, + { + "epoch": 0.7823380532184144, + "grad_norm": 2.960912649005083, + "learning_rate": 1.1918617691125333e-06, + "loss": 0.3939, + "step": 11437 + }, + { + "epoch": 0.7824064573500239, + "grad_norm": 2.106009475829786, + "learning_rate": 1.1911440209386981e-06, + "loss": 0.4779, + "step": 11438 + }, + { + "epoch": 0.7824748614816335, + "grad_norm": 2.3375913321659754, + "learning_rate": 1.1904264597219078e-06, + "loss": 0.2284, + "step": 11439 + }, + { + "epoch": 0.782543265613243, + "grad_norm": 2.4055810283290358, + "learning_rate": 1.189709085497383e-06, + "loss": 0.6777, + "step": 11440 + }, + { + "epoch": 0.7826116697448526, + "grad_norm": 2.1240049023751255, + "learning_rate": 1.1889918983003346e-06, + "loss": 0.4547, + "step": 11441 + }, + { + "epoch": 0.7826800738764621, + "grad_norm": 2.88032178097753, + "learning_rate": 1.1882748981659682e-06, + "loss": 0.3019, + "step": 11442 + }, + { + "epoch": 0.7827484780080717, + "grad_norm": 1.891308314211674, + "learning_rate": 1.1875580851294766e-06, + "loss": 0.2498, + "step": 11443 + }, + { + "epoch": 0.7828168821396813, + "grad_norm": 2.102015924995469, + "learning_rate": 1.186841459226044e-06, + "loss": 0.1166, + "step": 11444 + }, + { + "epoch": 0.7828852862712908, + "grad_norm": 2.3263271802021537, + "learning_rate": 1.1861250204908464e-06, + "loss": 0.2467, + "step": 11445 + }, + { + "epoch": 0.7829536904029003, + "grad_norm": 3.647663761408154, + "learning_rate": 1.1854087689590481e-06, + "loss": 0.1567, + "step": 11446 + }, + { + "epoch": 0.7830220945345099, + "grad_norm": 2.19793294987967, + "learning_rate": 1.1846927046658097e-06, + "loss": 0.4098, + "step": 11447 + }, + { + "epoch": 0.7830904986661195, + "grad_norm": 2.5069113327356973, + "learning_rate": 1.1839768276462759e-06, + "loss": 0.3777, + "step": 11448 + }, + { + "epoch": 0.7831589027977289, + "grad_norm": 2.5554842158624806, + "learning_rate": 1.183261137935588e-06, + "loss": 0.3389, + "step": 11449 + }, + { + "epoch": 0.7832273069293385, + "grad_norm": 4.256110076732937, + "learning_rate": 1.182545635568874e-06, + "loss": 0.3283, + "step": 11450 + }, + { + "epoch": 0.7832957110609481, + "grad_norm": 2.67221712967032, + "learning_rate": 1.1818303205812547e-06, + "loss": 0.3635, + "step": 11451 + }, + { + "epoch": 0.7833641151925577, + "grad_norm": 2.457479554292104, + "learning_rate": 1.1811151930078414e-06, + "loss": 0.3184, + "step": 11452 + }, + { + "epoch": 0.7834325193241671, + "grad_norm": 2.4075799276427685, + "learning_rate": 1.1804002528837355e-06, + "loss": 0.2683, + "step": 11453 + }, + { + "epoch": 0.7835009234557767, + "grad_norm": 1.9786104290646396, + "learning_rate": 1.1796855002440283e-06, + "loss": 0.4547, + "step": 11454 + }, + { + "epoch": 0.7835693275873863, + "grad_norm": 2.503938909784807, + "learning_rate": 1.178970935123806e-06, + "loss": 0.2979, + "step": 11455 + }, + { + "epoch": 0.7836377317189959, + "grad_norm": 2.5121991660659493, + "learning_rate": 1.1782565575581406e-06, + "loss": 0.5876, + "step": 11456 + }, + { + "epoch": 0.7837061358506053, + "grad_norm": 3.1295601188954274, + "learning_rate": 1.1775423675821e-06, + "loss": 0.3054, + "step": 11457 + }, + { + "epoch": 0.7837745399822149, + "grad_norm": 2.9259339153964934, + "learning_rate": 1.1768283652307383e-06, + "loss": 0.4984, + "step": 11458 + }, + { + "epoch": 0.7838429441138245, + "grad_norm": 1.9589379895576635, + "learning_rate": 1.1761145505391025e-06, + "loss": 0.2943, + "step": 11459 + }, + { + "epoch": 0.7839113482454341, + "grad_norm": 2.9195130945569585, + "learning_rate": 1.1754009235422293e-06, + "loss": 0.615, + "step": 11460 + }, + { + "epoch": 0.7839797523770435, + "grad_norm": 2.9348250749797384, + "learning_rate": 1.1746874842751467e-06, + "loss": 0.4464, + "step": 11461 + }, + { + "epoch": 0.7840481565086531, + "grad_norm": 1.8102276303537455, + "learning_rate": 1.1739742327728759e-06, + "loss": 0.2332, + "step": 11462 + }, + { + "epoch": 0.7841165606402627, + "grad_norm": 2.551737677592897, + "learning_rate": 1.1732611690704254e-06, + "loss": 0.5973, + "step": 11463 + }, + { + "epoch": 0.7841849647718723, + "grad_norm": 2.799788060852434, + "learning_rate": 1.172548293202796e-06, + "loss": 0.3792, + "step": 11464 + }, + { + "epoch": 0.7842533689034817, + "grad_norm": 2.410122044226278, + "learning_rate": 1.171835605204979e-06, + "loss": 0.4833, + "step": 11465 + }, + { + "epoch": 0.7843217730350913, + "grad_norm": 2.1162526364656484, + "learning_rate": 1.1711231051119548e-06, + "loss": 0.3827, + "step": 11466 + }, + { + "epoch": 0.7843901771667009, + "grad_norm": 2.2166729392474442, + "learning_rate": 1.1704107929586994e-06, + "loss": 0.3773, + "step": 11467 + }, + { + "epoch": 0.7844585812983104, + "grad_norm": 3.4318926766961044, + "learning_rate": 1.1696986687801743e-06, + "loss": 0.4272, + "step": 11468 + }, + { + "epoch": 0.7845269854299199, + "grad_norm": 2.9995563894351926, + "learning_rate": 1.1689867326113358e-06, + "loss": 0.641, + "step": 11469 + }, + { + "epoch": 0.7845953895615295, + "grad_norm": 2.446837964027319, + "learning_rate": 1.1682749844871282e-06, + "loss": 0.397, + "step": 11470 + }, + { + "epoch": 0.7846637936931391, + "grad_norm": 2.751719024209376, + "learning_rate": 1.1675634244424883e-06, + "loss": 0.6445, + "step": 11471 + }, + { + "epoch": 0.7847321978247486, + "grad_norm": 2.2790028301510596, + "learning_rate": 1.1668520525123418e-06, + "loss": 0.5153, + "step": 11472 + }, + { + "epoch": 0.7848006019563581, + "grad_norm": 2.233517135314847, + "learning_rate": 1.166140868731605e-06, + "loss": 0.494, + "step": 11473 + }, + { + "epoch": 0.7848690060879677, + "grad_norm": 2.1997032783637174, + "learning_rate": 1.1654298731351894e-06, + "loss": 0.427, + "step": 11474 + }, + { + "epoch": 0.7849374102195773, + "grad_norm": 2.0834070923781836, + "learning_rate": 1.164719065757992e-06, + "loss": 0.0764, + "step": 11475 + }, + { + "epoch": 0.7850058143511868, + "grad_norm": 2.510782995971779, + "learning_rate": 1.1640084466349028e-06, + "loss": 0.3101, + "step": 11476 + }, + { + "epoch": 0.7850742184827963, + "grad_norm": 2.9057859542520603, + "learning_rate": 1.1632980158008038e-06, + "loss": 0.6309, + "step": 11477 + }, + { + "epoch": 0.7851426226144059, + "grad_norm": 3.815467133251629, + "learning_rate": 1.1625877732905654e-06, + "loss": 0.4721, + "step": 11478 + }, + { + "epoch": 0.7852110267460155, + "grad_norm": 1.7644988663178733, + "learning_rate": 1.1618777191390502e-06, + "loss": 0.2481, + "step": 11479 + }, + { + "epoch": 0.785279430877625, + "grad_norm": 2.4418707565922966, + "learning_rate": 1.1611678533811089e-06, + "loss": 0.6348, + "step": 11480 + }, + { + "epoch": 0.7853478350092346, + "grad_norm": 2.5493893527978053, + "learning_rate": 1.1604581760515882e-06, + "loss": 0.2835, + "step": 11481 + }, + { + "epoch": 0.7854162391408441, + "grad_norm": 1.950260151724573, + "learning_rate": 1.1597486871853209e-06, + "loss": 0.3573, + "step": 11482 + }, + { + "epoch": 0.7854846432724537, + "grad_norm": 1.8549798421715595, + "learning_rate": 1.1590393868171329e-06, + "loss": 0.2654, + "step": 11483 + }, + { + "epoch": 0.7855530474040632, + "grad_norm": 1.9694000790596282, + "learning_rate": 1.1583302749818392e-06, + "loss": 0.1759, + "step": 11484 + }, + { + "epoch": 0.7856214515356728, + "grad_norm": 2.891356452981029, + "learning_rate": 1.157621351714246e-06, + "loss": 0.4296, + "step": 11485 + }, + { + "epoch": 0.7856898556672823, + "grad_norm": 2.109007485480882, + "learning_rate": 1.1569126170491523e-06, + "loss": 0.4101, + "step": 11486 + }, + { + "epoch": 0.7857582597988918, + "grad_norm": 1.9208716820704153, + "learning_rate": 1.1562040710213462e-06, + "loss": 0.357, + "step": 11487 + }, + { + "epoch": 0.7858266639305014, + "grad_norm": 2.4937048092908736, + "learning_rate": 1.1554957136656036e-06, + "loss": 0.3355, + "step": 11488 + }, + { + "epoch": 0.785895068062111, + "grad_norm": 2.1085560939800168, + "learning_rate": 1.1547875450166984e-06, + "loss": 0.2093, + "step": 11489 + }, + { + "epoch": 0.7859634721937205, + "grad_norm": 2.057027014276336, + "learning_rate": 1.1540795651093884e-06, + "loss": 0.4682, + "step": 11490 + }, + { + "epoch": 0.78603187632533, + "grad_norm": 2.277396317622852, + "learning_rate": 1.153371773978425e-06, + "loss": 0.2612, + "step": 11491 + }, + { + "epoch": 0.7861002804569396, + "grad_norm": 2.7640393231909757, + "learning_rate": 1.1526641716585507e-06, + "loss": 0.2114, + "step": 11492 + }, + { + "epoch": 0.7861686845885492, + "grad_norm": 5.72960162736901, + "learning_rate": 1.1519567581844954e-06, + "loss": 0.3947, + "step": 11493 + }, + { + "epoch": 0.7862370887201587, + "grad_norm": 2.1612980396321695, + "learning_rate": 1.1512495335909862e-06, + "loss": 0.3827, + "step": 11494 + }, + { + "epoch": 0.7863054928517682, + "grad_norm": 2.047748597648624, + "learning_rate": 1.1505424979127338e-06, + "loss": 0.26, + "step": 11495 + }, + { + "epoch": 0.7863738969833778, + "grad_norm": 2.2489779058196344, + "learning_rate": 1.1498356511844455e-06, + "loss": 0.2188, + "step": 11496 + }, + { + "epoch": 0.7864423011149874, + "grad_norm": 2.220411249605017, + "learning_rate": 1.1491289934408162e-06, + "loss": 0.4056, + "step": 11497 + }, + { + "epoch": 0.7865107052465969, + "grad_norm": 2.494147041569449, + "learning_rate": 1.1484225247165315e-06, + "loss": 0.5317, + "step": 11498 + }, + { + "epoch": 0.7865791093782064, + "grad_norm": 1.9914192738037926, + "learning_rate": 1.1477162450462681e-06, + "loss": 0.2767, + "step": 11499 + }, + { + "epoch": 0.786647513509816, + "grad_norm": 2.8426969485802185, + "learning_rate": 1.1470101544646928e-06, + "loss": 0.3759, + "step": 11500 + }, + { + "epoch": 0.7867159176414256, + "grad_norm": 1.953091277785052, + "learning_rate": 1.146304253006466e-06, + "loss": 0.4531, + "step": 11501 + }, + { + "epoch": 0.7867843217730351, + "grad_norm": 2.900066154646211, + "learning_rate": 1.1455985407062365e-06, + "loss": 0.2957, + "step": 11502 + }, + { + "epoch": 0.7868527259046446, + "grad_norm": 2.332898181688299, + "learning_rate": 1.1448930175986429e-06, + "loss": 0.2517, + "step": 11503 + }, + { + "epoch": 0.7869211300362542, + "grad_norm": 1.7872087169614461, + "learning_rate": 1.1441876837183158e-06, + "loss": 0.3744, + "step": 11504 + }, + { + "epoch": 0.7869895341678638, + "grad_norm": 3.491353688556089, + "learning_rate": 1.1434825390998756e-06, + "loss": 0.4462, + "step": 11505 + }, + { + "epoch": 0.7870579382994732, + "grad_norm": 2.247230480972924, + "learning_rate": 1.1427775837779364e-06, + "loss": 0.5566, + "step": 11506 + }, + { + "epoch": 0.7871263424310828, + "grad_norm": 1.7818053617888727, + "learning_rate": 1.1420728177870988e-06, + "loss": 0.1571, + "step": 11507 + }, + { + "epoch": 0.7871947465626924, + "grad_norm": 1.7455183228834386, + "learning_rate": 1.1413682411619587e-06, + "loss": 0.3074, + "step": 11508 + }, + { + "epoch": 0.787263150694302, + "grad_norm": 2.625613458794661, + "learning_rate": 1.140663853937098e-06, + "loss": 0.1899, + "step": 11509 + }, + { + "epoch": 0.7873315548259114, + "grad_norm": 2.306812780611384, + "learning_rate": 1.1399596561470916e-06, + "loss": 0.4297, + "step": 11510 + }, + { + "epoch": 0.787399958957521, + "grad_norm": 2.1961554269940415, + "learning_rate": 1.1392556478265055e-06, + "loss": 0.2253, + "step": 11511 + }, + { + "epoch": 0.7874683630891306, + "grad_norm": 2.448984788978647, + "learning_rate": 1.1385518290098957e-06, + "loss": 0.6314, + "step": 11512 + }, + { + "epoch": 0.7875367672207402, + "grad_norm": 3.1050973826108867, + "learning_rate": 1.1378481997318074e-06, + "loss": 0.589, + "step": 11513 + }, + { + "epoch": 0.7876051713523496, + "grad_norm": 2.8043799696611105, + "learning_rate": 1.1371447600267809e-06, + "loss": 0.2558, + "step": 11514 + }, + { + "epoch": 0.7876735754839592, + "grad_norm": 2.3980327227358638, + "learning_rate": 1.1364415099293418e-06, + "loss": 0.4723, + "step": 11515 + }, + { + "epoch": 0.7877419796155688, + "grad_norm": 1.965967092326185, + "learning_rate": 1.135738449474012e-06, + "loss": 0.2056, + "step": 11516 + }, + { + "epoch": 0.7878103837471784, + "grad_norm": 2.2377192672189135, + "learning_rate": 1.1350355786952989e-06, + "loss": 0.6747, + "step": 11517 + }, + { + "epoch": 0.7878787878787878, + "grad_norm": 2.857321403025112, + "learning_rate": 1.1343328976277034e-06, + "loss": 0.2117, + "step": 11518 + }, + { + "epoch": 0.7879471920103974, + "grad_norm": 2.82861946061821, + "learning_rate": 1.1336304063057169e-06, + "loss": 0.4883, + "step": 11519 + }, + { + "epoch": 0.788015596142007, + "grad_norm": 1.725959171847725, + "learning_rate": 1.1329281047638186e-06, + "loss": 0.276, + "step": 11520 + }, + { + "epoch": 0.7880840002736166, + "grad_norm": 3.29047519310233, + "learning_rate": 1.1322259930364848e-06, + "loss": 0.227, + "step": 11521 + }, + { + "epoch": 0.788152404405226, + "grad_norm": 2.284636831346543, + "learning_rate": 1.131524071158176e-06, + "loss": 0.4594, + "step": 11522 + }, + { + "epoch": 0.7882208085368356, + "grad_norm": 2.0742745688303934, + "learning_rate": 1.1308223391633466e-06, + "loss": 0.345, + "step": 11523 + }, + { + "epoch": 0.7882892126684452, + "grad_norm": 2.5336844722985945, + "learning_rate": 1.130120797086441e-06, + "loss": 0.6922, + "step": 11524 + }, + { + "epoch": 0.7883576168000547, + "grad_norm": 2.157814398232301, + "learning_rate": 1.129419444961893e-06, + "loss": 0.3234, + "step": 11525 + }, + { + "epoch": 0.7884260209316643, + "grad_norm": 1.2929404448669621, + "learning_rate": 1.1287182828241306e-06, + "loss": 0.1199, + "step": 11526 + }, + { + "epoch": 0.7884944250632738, + "grad_norm": 1.618344662337258, + "learning_rate": 1.1280173107075676e-06, + "loss": 0.1348, + "step": 11527 + }, + { + "epoch": 0.7885628291948834, + "grad_norm": 1.9019471916439938, + "learning_rate": 1.1273165286466137e-06, + "loss": 0.2684, + "step": 11528 + }, + { + "epoch": 0.7886312333264929, + "grad_norm": 1.8734476179878936, + "learning_rate": 1.1266159366756663e-06, + "loss": 0.2471, + "step": 11529 + }, + { + "epoch": 0.7886996374581025, + "grad_norm": 2.1080439147148047, + "learning_rate": 1.1259155348291124e-06, + "loss": 0.2907, + "step": 11530 + }, + { + "epoch": 0.788768041589712, + "grad_norm": 2.1645191895089653, + "learning_rate": 1.1252153231413321e-06, + "loss": 0.4076, + "step": 11531 + }, + { + "epoch": 0.7888364457213216, + "grad_norm": 2.0757848714467406, + "learning_rate": 1.1245153016466954e-06, + "loss": 0.4253, + "step": 11532 + }, + { + "epoch": 0.7889048498529311, + "grad_norm": 2.0394174624928754, + "learning_rate": 1.12381547037956e-06, + "loss": 0.5263, + "step": 11533 + }, + { + "epoch": 0.7889732539845407, + "grad_norm": 2.268861484678373, + "learning_rate": 1.123115829374281e-06, + "loss": 0.521, + "step": 11534 + }, + { + "epoch": 0.7890416581161502, + "grad_norm": 2.151498801785126, + "learning_rate": 1.122416378665197e-06, + "loss": 0.4025, + "step": 11535 + }, + { + "epoch": 0.7891100622477598, + "grad_norm": 3.115997804473843, + "learning_rate": 1.1217171182866432e-06, + "loss": 0.7605, + "step": 11536 + }, + { + "epoch": 0.7891784663793693, + "grad_norm": 1.8874781286791387, + "learning_rate": 1.121018048272941e-06, + "loss": 0.3484, + "step": 11537 + }, + { + "epoch": 0.7892468705109789, + "grad_norm": 2.003541143518505, + "learning_rate": 1.1203191686584048e-06, + "loss": 0.3687, + "step": 11538 + }, + { + "epoch": 0.7893152746425884, + "grad_norm": 1.6193475004406086, + "learning_rate": 1.1196204794773385e-06, + "loss": 0.2292, + "step": 11539 + }, + { + "epoch": 0.789383678774198, + "grad_norm": 2.3474378150964714, + "learning_rate": 1.1189219807640355e-06, + "loss": 0.4652, + "step": 11540 + }, + { + "epoch": 0.7894520829058075, + "grad_norm": 2.220805478651939, + "learning_rate": 1.118223672552785e-06, + "loss": 0.4656, + "step": 11541 + }, + { + "epoch": 0.7895204870374171, + "grad_norm": 1.6849762446825929, + "learning_rate": 1.1175255548778618e-06, + "loss": 0.2638, + "step": 11542 + }, + { + "epoch": 0.7895888911690266, + "grad_norm": 2.2308684891087243, + "learning_rate": 1.1168276277735319e-06, + "loss": 0.4747, + "step": 11543 + }, + { + "epoch": 0.7896572953006361, + "grad_norm": 2.346326435502451, + "learning_rate": 1.1161298912740542e-06, + "loss": 0.6717, + "step": 11544 + }, + { + "epoch": 0.7897256994322457, + "grad_norm": 2.1539432169634125, + "learning_rate": 1.115432345413675e-06, + "loss": 0.3694, + "step": 11545 + }, + { + "epoch": 0.7897941035638553, + "grad_norm": 2.2730598004870597, + "learning_rate": 1.1147349902266357e-06, + "loss": 0.1794, + "step": 11546 + }, + { + "epoch": 0.7898625076954648, + "grad_norm": 2.212785393417296, + "learning_rate": 1.114037825747164e-06, + "loss": 0.5834, + "step": 11547 + }, + { + "epoch": 0.7899309118270743, + "grad_norm": 2.5934610205839888, + "learning_rate": 1.1133408520094819e-06, + "loss": 0.3056, + "step": 11548 + }, + { + "epoch": 0.7899993159586839, + "grad_norm": 1.320504231046783, + "learning_rate": 1.1126440690477996e-06, + "loss": 0.1484, + "step": 11549 + }, + { + "epoch": 0.7900677200902935, + "grad_norm": 1.4693243116278611, + "learning_rate": 1.111947476896318e-06, + "loss": 0.102, + "step": 11550 + }, + { + "epoch": 0.790136124221903, + "grad_norm": 2.1268578289951936, + "learning_rate": 1.111251075589229e-06, + "loss": 0.3144, + "step": 11551 + }, + { + "epoch": 0.7902045283535125, + "grad_norm": 1.7114914559076662, + "learning_rate": 1.1105548651607167e-06, + "loss": 0.1883, + "step": 11552 + }, + { + "epoch": 0.7902729324851221, + "grad_norm": 2.049720333520929, + "learning_rate": 1.1098588456449517e-06, + "loss": 0.3709, + "step": 11553 + }, + { + "epoch": 0.7903413366167317, + "grad_norm": 2.501126905612238, + "learning_rate": 1.1091630170761014e-06, + "loss": 0.1983, + "step": 11554 + }, + { + "epoch": 0.7904097407483412, + "grad_norm": 2.177946920243763, + "learning_rate": 1.1084673794883172e-06, + "loss": 0.1891, + "step": 11555 + }, + { + "epoch": 0.7904781448799507, + "grad_norm": 1.99379494834625, + "learning_rate": 1.1077719329157477e-06, + "loss": 0.3938, + "step": 11556 + }, + { + "epoch": 0.7905465490115603, + "grad_norm": 2.5448203428893192, + "learning_rate": 1.1070766773925273e-06, + "loss": 0.2284, + "step": 11557 + }, + { + "epoch": 0.7906149531431699, + "grad_norm": 1.8213170312774245, + "learning_rate": 1.1063816129527815e-06, + "loss": 0.1965, + "step": 11558 + }, + { + "epoch": 0.7906833572747793, + "grad_norm": 2.550576144818267, + "learning_rate": 1.1056867396306293e-06, + "loss": 0.1853, + "step": 11559 + }, + { + "epoch": 0.7907517614063889, + "grad_norm": 2.9396317285158275, + "learning_rate": 1.1049920574601753e-06, + "loss": 0.4236, + "step": 11560 + }, + { + "epoch": 0.7908201655379985, + "grad_norm": 2.416742161621179, + "learning_rate": 1.1042975664755211e-06, + "loss": 0.1488, + "step": 11561 + }, + { + "epoch": 0.7908885696696081, + "grad_norm": 2.487554632717934, + "learning_rate": 1.103603266710755e-06, + "loss": 0.4606, + "step": 11562 + }, + { + "epoch": 0.7909569738012175, + "grad_norm": 2.238789405859335, + "learning_rate": 1.1029091581999564e-06, + "loss": 0.5083, + "step": 11563 + }, + { + "epoch": 0.7910253779328271, + "grad_norm": 4.126542612105157, + "learning_rate": 1.1022152409771942e-06, + "loss": 0.3627, + "step": 11564 + }, + { + "epoch": 0.7910937820644367, + "grad_norm": 2.0559849432002433, + "learning_rate": 1.1015215150765296e-06, + "loss": 0.301, + "step": 11565 + }, + { + "epoch": 0.7911621861960463, + "grad_norm": 1.693657220783832, + "learning_rate": 1.1008279805320154e-06, + "loss": 0.3471, + "step": 11566 + }, + { + "epoch": 0.7912305903276557, + "grad_norm": 2.4470341444330264, + "learning_rate": 1.1001346373776916e-06, + "loss": 0.3509, + "step": 11567 + }, + { + "epoch": 0.7912989944592653, + "grad_norm": 1.8284339154719644, + "learning_rate": 1.0994414856475938e-06, + "loss": 0.1641, + "step": 11568 + }, + { + "epoch": 0.7913673985908749, + "grad_norm": 2.374024404119925, + "learning_rate": 1.0987485253757435e-06, + "loss": 0.5277, + "step": 11569 + }, + { + "epoch": 0.7914358027224845, + "grad_norm": 2.382725892290908, + "learning_rate": 1.0980557565961542e-06, + "loss": 0.2229, + "step": 11570 + }, + { + "epoch": 0.791504206854094, + "grad_norm": 2.3388864281486117, + "learning_rate": 1.0973631793428307e-06, + "loss": 0.6448, + "step": 11571 + }, + { + "epoch": 0.7915726109857035, + "grad_norm": 2.2808374952967756, + "learning_rate": 1.0966707936497667e-06, + "loss": 0.4766, + "step": 11572 + }, + { + "epoch": 0.7916410151173131, + "grad_norm": 1.8649275202767792, + "learning_rate": 1.0959785995509504e-06, + "loss": 0.0857, + "step": 11573 + }, + { + "epoch": 0.7917094192489227, + "grad_norm": 1.5635705284650252, + "learning_rate": 1.0952865970803566e-06, + "loss": 0.106, + "step": 11574 + }, + { + "epoch": 0.7917778233805322, + "grad_norm": 2.24279966205326, + "learning_rate": 1.0945947862719514e-06, + "loss": 0.2945, + "step": 11575 + }, + { + "epoch": 0.7918462275121417, + "grad_norm": 2.186657160783983, + "learning_rate": 1.0939031671596938e-06, + "loss": 0.4406, + "step": 11576 + }, + { + "epoch": 0.7919146316437513, + "grad_norm": 4.548052160202163, + "learning_rate": 1.0932117397775317e-06, + "loss": 0.5094, + "step": 11577 + }, + { + "epoch": 0.7919830357753608, + "grad_norm": 2.602016707168793, + "learning_rate": 1.0925205041594029e-06, + "loss": 0.344, + "step": 11578 + }, + { + "epoch": 0.7920514399069704, + "grad_norm": 2.364824872277094, + "learning_rate": 1.0918294603392371e-06, + "loss": 0.3575, + "step": 11579 + }, + { + "epoch": 0.7921198440385799, + "grad_norm": 2.0042701450276206, + "learning_rate": 1.0911386083509517e-06, + "loss": 0.407, + "step": 11580 + }, + { + "epoch": 0.7921882481701895, + "grad_norm": 2.407905974854232, + "learning_rate": 1.0904479482284607e-06, + "loss": 0.5527, + "step": 11581 + }, + { + "epoch": 0.792256652301799, + "grad_norm": 2.3189398690398124, + "learning_rate": 1.0897574800056637e-06, + "loss": 0.5076, + "step": 11582 + }, + { + "epoch": 0.7923250564334086, + "grad_norm": 2.0296238625691165, + "learning_rate": 1.0890672037164518e-06, + "loss": 0.3014, + "step": 11583 + }, + { + "epoch": 0.7923934605650181, + "grad_norm": 2.3056861136390596, + "learning_rate": 1.0883771193947063e-06, + "loss": 0.4426, + "step": 11584 + }, + { + "epoch": 0.7924618646966277, + "grad_norm": 2.3473041352877337, + "learning_rate": 1.0876872270743016e-06, + "loss": 0.4245, + "step": 11585 + }, + { + "epoch": 0.7925302688282372, + "grad_norm": 2.912028669112684, + "learning_rate": 1.0869975267891004e-06, + "loss": 0.5139, + "step": 11586 + }, + { + "epoch": 0.7925986729598468, + "grad_norm": 2.3615882294855797, + "learning_rate": 1.0863080185729546e-06, + "loss": 0.3839, + "step": 11587 + }, + { + "epoch": 0.7926670770914563, + "grad_norm": 1.8864345119238934, + "learning_rate": 1.085618702459712e-06, + "loss": 0.1574, + "step": 11588 + }, + { + "epoch": 0.7927354812230659, + "grad_norm": 1.8299652765059626, + "learning_rate": 1.084929578483206e-06, + "loss": 0.3277, + "step": 11589 + }, + { + "epoch": 0.7928038853546754, + "grad_norm": 2.4385299279572576, + "learning_rate": 1.0842406466772615e-06, + "loss": 0.5104, + "step": 11590 + }, + { + "epoch": 0.792872289486285, + "grad_norm": 2.258822574853403, + "learning_rate": 1.0835519070756955e-06, + "loss": 0.3498, + "step": 11591 + }, + { + "epoch": 0.7929406936178945, + "grad_norm": 2.514979404039023, + "learning_rate": 1.0828633597123128e-06, + "loss": 0.4962, + "step": 11592 + }, + { + "epoch": 0.7930090977495041, + "grad_norm": 2.3366221059431185, + "learning_rate": 1.0821750046209128e-06, + "loss": 0.2765, + "step": 11593 + }, + { + "epoch": 0.7930775018811136, + "grad_norm": 2.7068606511424713, + "learning_rate": 1.081486841835282e-06, + "loss": 0.2343, + "step": 11594 + }, + { + "epoch": 0.7931459060127232, + "grad_norm": 1.9148721033718488, + "learning_rate": 1.0807988713892003e-06, + "loss": 0.1605, + "step": 11595 + }, + { + "epoch": 0.7932143101443327, + "grad_norm": 4.082070543017935, + "learning_rate": 1.0801110933164356e-06, + "loss": 0.3933, + "step": 11596 + }, + { + "epoch": 0.7932827142759422, + "grad_norm": 2.6743876259851853, + "learning_rate": 1.0794235076507475e-06, + "loss": 0.3416, + "step": 11597 + }, + { + "epoch": 0.7933511184075518, + "grad_norm": 3.1275969967639488, + "learning_rate": 1.0787361144258857e-06, + "loss": 0.7697, + "step": 11598 + }, + { + "epoch": 0.7934195225391614, + "grad_norm": 2.1772270946097207, + "learning_rate": 1.07804891367559e-06, + "loss": 0.351, + "step": 11599 + }, + { + "epoch": 0.793487926670771, + "grad_norm": 2.3655367216220333, + "learning_rate": 1.0773619054335937e-06, + "loss": 0.4587, + "step": 11600 + }, + { + "epoch": 0.7935563308023804, + "grad_norm": 1.917483145933174, + "learning_rate": 1.0766750897336175e-06, + "loss": 0.3691, + "step": 11601 + }, + { + "epoch": 0.79362473493399, + "grad_norm": 2.8183681622723102, + "learning_rate": 1.075988466609373e-06, + "loss": 0.4485, + "step": 11602 + }, + { + "epoch": 0.7936931390655996, + "grad_norm": 3.1742518168592393, + "learning_rate": 1.0753020360945637e-06, + "loss": 0.3981, + "step": 11603 + }, + { + "epoch": 0.7937615431972092, + "grad_norm": 2.6457906454252114, + "learning_rate": 1.074615798222881e-06, + "loss": 0.3782, + "step": 11604 + }, + { + "epoch": 0.7938299473288186, + "grad_norm": 2.290766748827768, + "learning_rate": 1.0739297530280125e-06, + "loss": 0.4581, + "step": 11605 + }, + { + "epoch": 0.7938983514604282, + "grad_norm": 3.6784376706713213, + "learning_rate": 1.0732439005436285e-06, + "loss": 0.5522, + "step": 11606 + }, + { + "epoch": 0.7939667555920378, + "grad_norm": 2.7923766355492234, + "learning_rate": 1.0725582408033974e-06, + "loss": 0.3973, + "step": 11607 + }, + { + "epoch": 0.7940351597236474, + "grad_norm": 1.6676613928744959, + "learning_rate": 1.0718727738409734e-06, + "loss": 0.2891, + "step": 11608 + }, + { + "epoch": 0.7941035638552568, + "grad_norm": 2.376669535336451, + "learning_rate": 1.0711874996900024e-06, + "loss": 0.4336, + "step": 11609 + }, + { + "epoch": 0.7941719679868664, + "grad_norm": 2.5292480228375243, + "learning_rate": 1.0705024183841212e-06, + "loss": 0.3756, + "step": 11610 + }, + { + "epoch": 0.794240372118476, + "grad_norm": 2.4085954530643723, + "learning_rate": 1.0698175299569563e-06, + "loss": 0.5956, + "step": 11611 + }, + { + "epoch": 0.7943087762500856, + "grad_norm": 2.545487242262638, + "learning_rate": 1.0691328344421243e-06, + "loss": 0.653, + "step": 11612 + }, + { + "epoch": 0.794377180381695, + "grad_norm": 2.203502778203662, + "learning_rate": 1.0684483318732364e-06, + "loss": 0.3833, + "step": 11613 + }, + { + "epoch": 0.7944455845133046, + "grad_norm": 2.447370290991636, + "learning_rate": 1.0677640222838874e-06, + "loss": 0.6479, + "step": 11614 + }, + { + "epoch": 0.7945139886449142, + "grad_norm": 2.4427082587745725, + "learning_rate": 1.0670799057076714e-06, + "loss": 0.1754, + "step": 11615 + }, + { + "epoch": 0.7945823927765236, + "grad_norm": 2.0968193928015344, + "learning_rate": 1.066395982178165e-06, + "loss": 0.4464, + "step": 11616 + }, + { + "epoch": 0.7946507969081332, + "grad_norm": 2.0480115297750117, + "learning_rate": 1.0657122517289387e-06, + "loss": 0.4171, + "step": 11617 + }, + { + "epoch": 0.7947192010397428, + "grad_norm": 1.5952420613302527, + "learning_rate": 1.065028714393554e-06, + "loss": 0.2741, + "step": 11618 + }, + { + "epoch": 0.7947876051713524, + "grad_norm": 2.7285699094862266, + "learning_rate": 1.06434537020556e-06, + "loss": 0.1404, + "step": 11619 + }, + { + "epoch": 0.7948560093029619, + "grad_norm": 2.45644816286004, + "learning_rate": 1.0636622191985019e-06, + "loss": 0.5692, + "step": 11620 + }, + { + "epoch": 0.7949244134345714, + "grad_norm": 2.386959581976618, + "learning_rate": 1.0629792614059103e-06, + "loss": 0.6086, + "step": 11621 + }, + { + "epoch": 0.794992817566181, + "grad_norm": 1.4336869316065024, + "learning_rate": 1.0622964968613082e-06, + "loss": 0.1455, + "step": 11622 + }, + { + "epoch": 0.7950612216977906, + "grad_norm": 2.272113361998513, + "learning_rate": 1.0616139255982093e-06, + "loss": 0.3094, + "step": 11623 + }, + { + "epoch": 0.7951296258294, + "grad_norm": 1.6752540586083962, + "learning_rate": 1.0609315476501159e-06, + "loss": 0.1989, + "step": 11624 + }, + { + "epoch": 0.7951980299610096, + "grad_norm": 3.3629115629214548, + "learning_rate": 1.0602493630505245e-06, + "loss": 0.2875, + "step": 11625 + }, + { + "epoch": 0.7952664340926192, + "grad_norm": 1.3941229321315591, + "learning_rate": 1.0595673718329185e-06, + "loss": 0.1016, + "step": 11626 + }, + { + "epoch": 0.7953348382242288, + "grad_norm": 1.755762978051579, + "learning_rate": 1.0588855740307752e-06, + "loss": 0.2367, + "step": 11627 + }, + { + "epoch": 0.7954032423558383, + "grad_norm": 2.217043095833347, + "learning_rate": 1.0582039696775592e-06, + "loss": 0.4412, + "step": 11628 + }, + { + "epoch": 0.7954716464874478, + "grad_norm": 2.8032188289252744, + "learning_rate": 1.0575225588067272e-06, + "loss": 0.2221, + "step": 11629 + }, + { + "epoch": 0.7955400506190574, + "grad_norm": 2.751319146212474, + "learning_rate": 1.056841341451726e-06, + "loss": 0.5031, + "step": 11630 + }, + { + "epoch": 0.795608454750667, + "grad_norm": 2.202616544989256, + "learning_rate": 1.0561603176459934e-06, + "loss": 0.5785, + "step": 11631 + }, + { + "epoch": 0.7956768588822765, + "grad_norm": 2.3667346974510077, + "learning_rate": 1.0554794874229552e-06, + "loss": 0.5634, + "step": 11632 + }, + { + "epoch": 0.795745263013886, + "grad_norm": 2.141260950613887, + "learning_rate": 1.0547988508160333e-06, + "loss": 0.4405, + "step": 11633 + }, + { + "epoch": 0.7958136671454956, + "grad_norm": 2.6759597273285127, + "learning_rate": 1.0541184078586337e-06, + "loss": 0.2617, + "step": 11634 + }, + { + "epoch": 0.7958820712771051, + "grad_norm": 2.014516599615638, + "learning_rate": 1.053438158584158e-06, + "loss": 0.2726, + "step": 11635 + }, + { + "epoch": 0.7959504754087147, + "grad_norm": 2.2185854380656287, + "learning_rate": 1.0527581030259953e-06, + "loss": 0.3359, + "step": 11636 + }, + { + "epoch": 0.7960188795403242, + "grad_norm": 2.141123755330442, + "learning_rate": 1.0520782412175261e-06, + "loss": 0.2918, + "step": 11637 + }, + { + "epoch": 0.7960872836719338, + "grad_norm": 2.0797111939786515, + "learning_rate": 1.0513985731921217e-06, + "loss": 0.1264, + "step": 11638 + }, + { + "epoch": 0.7961556878035433, + "grad_norm": 2.2704323639177995, + "learning_rate": 1.0507190989831412e-06, + "loss": 0.2523, + "step": 11639 + }, + { + "epoch": 0.7962240919351529, + "grad_norm": 5.56221573349747, + "learning_rate": 1.0500398186239392e-06, + "loss": 0.3195, + "step": 11640 + }, + { + "epoch": 0.7962924960667624, + "grad_norm": 2.2930534646080996, + "learning_rate": 1.0493607321478571e-06, + "loss": 0.5506, + "step": 11641 + }, + { + "epoch": 0.796360900198372, + "grad_norm": 1.9295480191250185, + "learning_rate": 1.048681839588228e-06, + "loss": 0.278, + "step": 11642 + }, + { + "epoch": 0.7964293043299815, + "grad_norm": 2.2141186033367664, + "learning_rate": 1.0480031409783753e-06, + "loss": 0.3662, + "step": 11643 + }, + { + "epoch": 0.7964977084615911, + "grad_norm": 1.8820597817873812, + "learning_rate": 1.0473246363516105e-06, + "loss": 0.4233, + "step": 11644 + }, + { + "epoch": 0.7965661125932006, + "grad_norm": 1.9073588632538137, + "learning_rate": 1.0466463257412412e-06, + "loss": 0.3171, + "step": 11645 + }, + { + "epoch": 0.7966345167248102, + "grad_norm": 1.7107021047924738, + "learning_rate": 1.0459682091805595e-06, + "loss": 0.1075, + "step": 11646 + }, + { + "epoch": 0.7967029208564197, + "grad_norm": 2.2157544344393147, + "learning_rate": 1.0452902867028536e-06, + "loss": 0.4123, + "step": 11647 + }, + { + "epoch": 0.7967713249880293, + "grad_norm": 2.5151885643282754, + "learning_rate": 1.0446125583413968e-06, + "loss": 0.3997, + "step": 11648 + }, + { + "epoch": 0.7968397291196389, + "grad_norm": 2.0326524001460475, + "learning_rate": 1.0439350241294566e-06, + "loss": 0.3496, + "step": 11649 + }, + { + "epoch": 0.7969081332512484, + "grad_norm": 2.6675556639411258, + "learning_rate": 1.043257684100289e-06, + "loss": 0.5064, + "step": 11650 + }, + { + "epoch": 0.7969765373828579, + "grad_norm": 7.4350409569859615, + "learning_rate": 1.0425805382871413e-06, + "loss": 0.2789, + "step": 11651 + }, + { + "epoch": 0.7970449415144675, + "grad_norm": 1.9960563342624515, + "learning_rate": 1.041903586723249e-06, + "loss": 0.2369, + "step": 11652 + }, + { + "epoch": 0.797113345646077, + "grad_norm": 2.0203527664868677, + "learning_rate": 1.0412268294418443e-06, + "loss": 0.4689, + "step": 11653 + }, + { + "epoch": 0.7971817497776865, + "grad_norm": 2.142425601228326, + "learning_rate": 1.0405502664761414e-06, + "loss": 0.2408, + "step": 11654 + }, + { + "epoch": 0.7972501539092961, + "grad_norm": 2.414916236305398, + "learning_rate": 1.039873897859353e-06, + "loss": 0.3367, + "step": 11655 + }, + { + "epoch": 0.7973185580409057, + "grad_norm": 2.0924103101869305, + "learning_rate": 1.0391977236246775e-06, + "loss": 0.1902, + "step": 11656 + }, + { + "epoch": 0.7973869621725153, + "grad_norm": 2.177048969606, + "learning_rate": 1.0385217438053035e-06, + "loss": 0.403, + "step": 11657 + }, + { + "epoch": 0.7974553663041247, + "grad_norm": 2.2574021781646922, + "learning_rate": 1.0378459584344126e-06, + "loss": 0.289, + "step": 11658 + }, + { + "epoch": 0.7975237704357343, + "grad_norm": 2.1203810335692, + "learning_rate": 1.0371703675451732e-06, + "loss": 0.371, + "step": 11659 + }, + { + "epoch": 0.7975921745673439, + "grad_norm": 1.9512024305240316, + "learning_rate": 1.0364949711707506e-06, + "loss": 0.3285, + "step": 11660 + }, + { + "epoch": 0.7976605786989535, + "grad_norm": 1.7576716938844272, + "learning_rate": 1.035819769344294e-06, + "loss": 0.2748, + "step": 11661 + }, + { + "epoch": 0.7977289828305629, + "grad_norm": 2.1822033107839784, + "learning_rate": 1.0351447620989457e-06, + "loss": 0.2172, + "step": 11662 + }, + { + "epoch": 0.7977973869621725, + "grad_norm": 2.974743925822393, + "learning_rate": 1.034469949467839e-06, + "loss": 0.2934, + "step": 11663 + }, + { + "epoch": 0.7978657910937821, + "grad_norm": 2.2796645076394912, + "learning_rate": 1.0337953314840949e-06, + "loss": 0.2423, + "step": 11664 + }, + { + "epoch": 0.7979341952253917, + "grad_norm": 2.2293675872833494, + "learning_rate": 1.0331209081808296e-06, + "loss": 0.5851, + "step": 11665 + }, + { + "epoch": 0.7980025993570011, + "grad_norm": 2.7444922062469033, + "learning_rate": 1.0324466795911452e-06, + "loss": 0.8613, + "step": 11666 + }, + { + "epoch": 0.7980710034886107, + "grad_norm": 2.031135027639359, + "learning_rate": 1.031772645748138e-06, + "loss": 0.448, + "step": 11667 + }, + { + "epoch": 0.7981394076202203, + "grad_norm": 1.6672434826919567, + "learning_rate": 1.031098806684892e-06, + "loss": 0.2034, + "step": 11668 + }, + { + "epoch": 0.7982078117518299, + "grad_norm": 2.3805744542038547, + "learning_rate": 1.0304251624344819e-06, + "loss": 0.5462, + "step": 11669 + }, + { + "epoch": 0.7982762158834393, + "grad_norm": 4.21439815283498, + "learning_rate": 1.0297517130299739e-06, + "loss": 0.442, + "step": 11670 + }, + { + "epoch": 0.7983446200150489, + "grad_norm": 1.6913865949444604, + "learning_rate": 1.0290784585044234e-06, + "loss": 0.1477, + "step": 11671 + }, + { + "epoch": 0.7984130241466585, + "grad_norm": 2.1885018098286393, + "learning_rate": 1.0284053988908765e-06, + "loss": 0.4179, + "step": 11672 + }, + { + "epoch": 0.798481428278268, + "grad_norm": 1.7448200555645932, + "learning_rate": 1.0277325342223732e-06, + "loss": 0.1212, + "step": 11673 + }, + { + "epoch": 0.7985498324098775, + "grad_norm": 2.123539134969381, + "learning_rate": 1.0270598645319369e-06, + "loss": 0.4825, + "step": 11674 + }, + { + "epoch": 0.7986182365414871, + "grad_norm": 2.517532904598364, + "learning_rate": 1.0263873898525894e-06, + "loss": 0.3489, + "step": 11675 + }, + { + "epoch": 0.7986866406730967, + "grad_norm": 3.043343781656656, + "learning_rate": 1.0257151102173367e-06, + "loss": 0.5408, + "step": 11676 + }, + { + "epoch": 0.7987550448047062, + "grad_norm": 2.7350137019119294, + "learning_rate": 1.0250430256591787e-06, + "loss": 0.289, + "step": 11677 + }, + { + "epoch": 0.7988234489363157, + "grad_norm": 2.4739836313393107, + "learning_rate": 1.024371136211103e-06, + "loss": 0.4713, + "step": 11678 + }, + { + "epoch": 0.7988918530679253, + "grad_norm": 2.191165067622916, + "learning_rate": 1.0236994419060892e-06, + "loss": 0.3421, + "step": 11679 + }, + { + "epoch": 0.7989602571995349, + "grad_norm": 3.2977311527771183, + "learning_rate": 1.0230279427771094e-06, + "loss": 0.633, + "step": 11680 + }, + { + "epoch": 0.7990286613311444, + "grad_norm": 1.9549065056320811, + "learning_rate": 1.0223566388571232e-06, + "loss": 0.2757, + "step": 11681 + }, + { + "epoch": 0.7990970654627539, + "grad_norm": 2.951447923979896, + "learning_rate": 1.0216855301790807e-06, + "loss": 0.444, + "step": 11682 + }, + { + "epoch": 0.7991654695943635, + "grad_norm": 1.9590740772829545, + "learning_rate": 1.0210146167759221e-06, + "loss": 0.34, + "step": 11683 + }, + { + "epoch": 0.7992338737259731, + "grad_norm": 2.347149278484198, + "learning_rate": 1.0203438986805814e-06, + "loss": 0.5045, + "step": 11684 + }, + { + "epoch": 0.7993022778575826, + "grad_norm": 2.4948719240046646, + "learning_rate": 1.0196733759259803e-06, + "loss": 0.2451, + "step": 11685 + }, + { + "epoch": 0.7993706819891921, + "grad_norm": 2.7295540736080177, + "learning_rate": 1.0190030485450292e-06, + "loss": 0.1473, + "step": 11686 + }, + { + "epoch": 0.7994390861208017, + "grad_norm": 2.505591885925238, + "learning_rate": 1.0183329165706334e-06, + "loss": 0.2814, + "step": 11687 + }, + { + "epoch": 0.7995074902524112, + "grad_norm": 1.8002895316910323, + "learning_rate": 1.0176629800356857e-06, + "loss": 0.2513, + "step": 11688 + }, + { + "epoch": 0.7995758943840208, + "grad_norm": 2.2978796610672627, + "learning_rate": 1.0169932389730698e-06, + "loss": 0.5283, + "step": 11689 + }, + { + "epoch": 0.7996442985156303, + "grad_norm": 1.8328380963423556, + "learning_rate": 1.016323693415659e-06, + "loss": 0.2102, + "step": 11690 + }, + { + "epoch": 0.7997127026472399, + "grad_norm": 2.3180004314421385, + "learning_rate": 1.0156543433963173e-06, + "loss": 0.3259, + "step": 11691 + }, + { + "epoch": 0.7997811067788494, + "grad_norm": 2.214861827689104, + "learning_rate": 1.0149851889479017e-06, + "loss": 0.3761, + "step": 11692 + }, + { + "epoch": 0.799849510910459, + "grad_norm": 2.51376502658065, + "learning_rate": 1.014316230103255e-06, + "loss": 0.4316, + "step": 11693 + }, + { + "epoch": 0.7999179150420686, + "grad_norm": 3.151102466889922, + "learning_rate": 1.0136474668952161e-06, + "loss": 0.358, + "step": 11694 + }, + { + "epoch": 0.7999863191736781, + "grad_norm": 2.3127216606614076, + "learning_rate": 1.0129788993566092e-06, + "loss": 0.5417, + "step": 11695 + }, + { + "epoch": 0.8000547233052876, + "grad_norm": 2.0185219144259947, + "learning_rate": 1.012310527520251e-06, + "loss": 0.3574, + "step": 11696 + } + ], + "logging_steps": 1.0, + "max_steps": 14619, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 1462, + "total_flos": 355752990699520.0, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}