diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,14328 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.0, + "eval_steps": 500, + "global_step": 2042, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0009802720254870727, + "grad_norm": 444.2529296875, + "learning_rate": 0.0, + "loss": 10.736241340637207, + "step": 1 + }, + { + "epoch": 0.0019605440509741453, + "grad_norm": 99.79141235351562, + "learning_rate": 1.6129032258064518e-07, + "loss": 10.998506546020508, + "step": 2 + }, + { + "epoch": 0.002940816076461218, + "grad_norm": 94.0584487915039, + "learning_rate": 3.2258064516129035e-07, + "loss": 10.300856590270996, + "step": 3 + }, + { + "epoch": 0.003921088101948291, + "grad_norm": 754.8390502929688, + "learning_rate": 4.838709677419355e-07, + "loss": 10.12170696258545, + "step": 4 + }, + { + "epoch": 0.004901360127435363, + "grad_norm": 110.3061294555664, + "learning_rate": 6.451612903225807e-07, + "loss": 10.923585891723633, + "step": 5 + }, + { + "epoch": 0.005881632152922436, + "grad_norm": 80.18559265136719, + "learning_rate": 8.064516129032258e-07, + "loss": 9.619321823120117, + "step": 6 + }, + { + "epoch": 0.006861904178409509, + "grad_norm": 120.46674346923828, + "learning_rate": 9.67741935483871e-07, + "loss": 8.93388843536377, + "step": 7 + }, + { + "epoch": 0.007842176203896581, + "grad_norm": 155.403564453125, + "learning_rate": 1.1290322580645162e-06, + "loss": 9.105071067810059, + "step": 8 + }, + { + "epoch": 0.008822448229383655, + "grad_norm": 58.01957702636719, + "learning_rate": 1.2903225806451614e-06, + "loss": 7.533741474151611, + "step": 9 + }, + { + "epoch": 0.009802720254870726, + "grad_norm": 31.391273498535156, + "learning_rate": 1.4516129032258066e-06, + "loss": 7.134483337402344, + "step": 10 + }, + { + "epoch": 0.0107829922803578, + "grad_norm": 26.58418083190918, + "learning_rate": 1.6129032258064516e-06, + "loss": 6.369547367095947, + "step": 11 + }, + { + "epoch": 0.011763264305844873, + "grad_norm": 35.25242233276367, + "learning_rate": 1.774193548387097e-06, + "loss": 5.254162311553955, + "step": 12 + }, + { + "epoch": 0.012743536331331944, + "grad_norm": 46.21750259399414, + "learning_rate": 1.935483870967742e-06, + "loss": 4.887372016906738, + "step": 13 + }, + { + "epoch": 0.013723808356819018, + "grad_norm": 11.279609680175781, + "learning_rate": 2.096774193548387e-06, + "loss": 5.321861267089844, + "step": 14 + }, + { + "epoch": 0.01470408038230609, + "grad_norm": 15.513973236083984, + "learning_rate": 2.2580645161290324e-06, + "loss": 4.800952911376953, + "step": 15 + }, + { + "epoch": 0.015684352407793162, + "grad_norm": 9.2393798828125, + "learning_rate": 2.4193548387096776e-06, + "loss": 4.0728983879089355, + "step": 16 + }, + { + "epoch": 0.016664624433280234, + "grad_norm": 5.680278778076172, + "learning_rate": 2.580645161290323e-06, + "loss": 3.8023455142974854, + "step": 17 + }, + { + "epoch": 0.01764489645876731, + "grad_norm": 11.160239219665527, + "learning_rate": 2.7419354838709676e-06, + "loss": 3.5728161334991455, + "step": 18 + }, + { + "epoch": 0.01862516848425438, + "grad_norm": 9.689987182617188, + "learning_rate": 2.903225806451613e-06, + "loss": 3.6294851303100586, + "step": 19 + }, + { + "epoch": 0.019605440509741452, + "grad_norm": 73.86099243164062, + "learning_rate": 3.0645161290322584e-06, + "loss": 3.190464496612549, + "step": 20 + }, + { + "epoch": 0.020585712535228527, + "grad_norm": 5.178689002990723, + "learning_rate": 3.225806451612903e-06, + "loss": 3.054841995239258, + "step": 21 + }, + { + "epoch": 0.0215659845607156, + "grad_norm": 17.172805786132812, + "learning_rate": 3.3870967741935484e-06, + "loss": 2.9779253005981445, + "step": 22 + }, + { + "epoch": 0.02254625658620267, + "grad_norm": 5.4300537109375, + "learning_rate": 3.548387096774194e-06, + "loss": 3.034024238586426, + "step": 23 + }, + { + "epoch": 0.023526528611689745, + "grad_norm": 9.983994483947754, + "learning_rate": 3.7096774193548392e-06, + "loss": 2.7238316535949707, + "step": 24 + }, + { + "epoch": 0.024506800637176817, + "grad_norm": 5.650508403778076, + "learning_rate": 3.870967741935484e-06, + "loss": 2.50161075592041, + "step": 25 + }, + { + "epoch": 0.02548707266266389, + "grad_norm": 55.20005798339844, + "learning_rate": 4.032258064516129e-06, + "loss": 2.6121983528137207, + "step": 26 + }, + { + "epoch": 0.026467344688150964, + "grad_norm": 34.6417121887207, + "learning_rate": 4.193548387096774e-06, + "loss": 2.351276397705078, + "step": 27 + }, + { + "epoch": 0.027447616713638035, + "grad_norm": 4.153648853302002, + "learning_rate": 4.35483870967742e-06, + "loss": 2.1463727951049805, + "step": 28 + }, + { + "epoch": 0.028427888739125107, + "grad_norm": 5.739168167114258, + "learning_rate": 4.516129032258065e-06, + "loss": 2.129469871520996, + "step": 29 + }, + { + "epoch": 0.02940816076461218, + "grad_norm": 13.234610557556152, + "learning_rate": 4.67741935483871e-06, + "loss": 2.023747444152832, + "step": 30 + }, + { + "epoch": 0.030388432790099253, + "grad_norm": 6.241724491119385, + "learning_rate": 4.838709677419355e-06, + "loss": 2.0241587162017822, + "step": 31 + }, + { + "epoch": 0.031368704815586325, + "grad_norm": 2.4811441898345947, + "learning_rate": 5e-06, + "loss": 1.9200693368911743, + "step": 32 + }, + { + "epoch": 0.0323489768410734, + "grad_norm": 18.006078720092773, + "learning_rate": 5.161290322580646e-06, + "loss": 2.048367977142334, + "step": 33 + }, + { + "epoch": 0.03332924886656047, + "grad_norm": 2.888336181640625, + "learning_rate": 5.322580645161291e-06, + "loss": 1.8217453956604004, + "step": 34 + }, + { + "epoch": 0.03430952089204754, + "grad_norm": 4.508528709411621, + "learning_rate": 5.483870967741935e-06, + "loss": 1.9134241342544556, + "step": 35 + }, + { + "epoch": 0.03528979291753462, + "grad_norm": 5.45286226272583, + "learning_rate": 5.645161290322582e-06, + "loss": 1.5955469608306885, + "step": 36 + }, + { + "epoch": 0.036270064943021686, + "grad_norm": 2.6666250228881836, + "learning_rate": 5.806451612903226e-06, + "loss": 1.5563428401947021, + "step": 37 + }, + { + "epoch": 0.03725033696850876, + "grad_norm": 10.96904468536377, + "learning_rate": 5.967741935483872e-06, + "loss": 1.7565765380859375, + "step": 38 + }, + { + "epoch": 0.038230608993995836, + "grad_norm": 8.063737869262695, + "learning_rate": 6.129032258064517e-06, + "loss": 1.559302806854248, + "step": 39 + }, + { + "epoch": 0.039210881019482904, + "grad_norm": 3.7466046810150146, + "learning_rate": 6.290322580645162e-06, + "loss": 1.3841321468353271, + "step": 40 + }, + { + "epoch": 0.04019115304496998, + "grad_norm": 4.077739715576172, + "learning_rate": 6.451612903225806e-06, + "loss": 1.4145723581314087, + "step": 41 + }, + { + "epoch": 0.041171425070457054, + "grad_norm": 2.0939736366271973, + "learning_rate": 6.612903225806452e-06, + "loss": 1.266073226928711, + "step": 42 + }, + { + "epoch": 0.04215169709594412, + "grad_norm": 1.8341953754425049, + "learning_rate": 6.774193548387097e-06, + "loss": 1.4071226119995117, + "step": 43 + }, + { + "epoch": 0.0431319691214312, + "grad_norm": 27.189706802368164, + "learning_rate": 6.935483870967743e-06, + "loss": 1.3583588600158691, + "step": 44 + }, + { + "epoch": 0.04411224114691827, + "grad_norm": 2.82374906539917, + "learning_rate": 7.096774193548388e-06, + "loss": 1.1239447593688965, + "step": 45 + }, + { + "epoch": 0.04509251317240534, + "grad_norm": 3.163405418395996, + "learning_rate": 7.258064516129033e-06, + "loss": 1.2555480003356934, + "step": 46 + }, + { + "epoch": 0.046072785197892416, + "grad_norm": 9.612832069396973, + "learning_rate": 7.4193548387096784e-06, + "loss": 1.2280371189117432, + "step": 47 + }, + { + "epoch": 0.04705305722337949, + "grad_norm": 2.274650812149048, + "learning_rate": 7.580645161290323e-06, + "loss": 1.2739315032958984, + "step": 48 + }, + { + "epoch": 0.04803332924886656, + "grad_norm": 8.578704833984375, + "learning_rate": 7.741935483870968e-06, + "loss": 1.4535212516784668, + "step": 49 + }, + { + "epoch": 0.049013601274353634, + "grad_norm": 33.340606689453125, + "learning_rate": 7.903225806451613e-06, + "loss": 1.4000167846679688, + "step": 50 + }, + { + "epoch": 0.04999387329984071, + "grad_norm": 6.348473072052002, + "learning_rate": 8.064516129032258e-06, + "loss": 1.6060235500335693, + "step": 51 + }, + { + "epoch": 0.05097414532532778, + "grad_norm": 2.6229915618896484, + "learning_rate": 8.225806451612904e-06, + "loss": 1.1890780925750732, + "step": 52 + }, + { + "epoch": 0.05195441735081485, + "grad_norm": 2.0519227981567383, + "learning_rate": 8.387096774193549e-06, + "loss": 1.3458662033081055, + "step": 53 + }, + { + "epoch": 0.05293468937630193, + "grad_norm": 1.8942406177520752, + "learning_rate": 8.548387096774194e-06, + "loss": 1.1449463367462158, + "step": 54 + }, + { + "epoch": 0.053914961401788995, + "grad_norm": 3.5907156467437744, + "learning_rate": 8.70967741935484e-06, + "loss": 1.2168039083480835, + "step": 55 + }, + { + "epoch": 0.05489523342727607, + "grad_norm": 5.486432075500488, + "learning_rate": 8.870967741935484e-06, + "loss": 1.039926528930664, + "step": 56 + }, + { + "epoch": 0.05587550545276314, + "grad_norm": 2.0273101329803467, + "learning_rate": 9.03225806451613e-06, + "loss": 1.225830078125, + "step": 57 + }, + { + "epoch": 0.05685577747825021, + "grad_norm": 3.0957894325256348, + "learning_rate": 9.193548387096775e-06, + "loss": 1.272838830947876, + "step": 58 + }, + { + "epoch": 0.05783604950373729, + "grad_norm": 5.01629114151001, + "learning_rate": 9.35483870967742e-06, + "loss": 1.0743775367736816, + "step": 59 + }, + { + "epoch": 0.05881632152922436, + "grad_norm": 4.781734466552734, + "learning_rate": 9.516129032258065e-06, + "loss": 1.0559556484222412, + "step": 60 + }, + { + "epoch": 0.05979659355471143, + "grad_norm": 2.8775441646575928, + "learning_rate": 9.67741935483871e-06, + "loss": 1.2778871059417725, + "step": 61 + }, + { + "epoch": 0.06077686558019851, + "grad_norm": 1.9450877904891968, + "learning_rate": 9.838709677419356e-06, + "loss": 1.1817667484283447, + "step": 62 + }, + { + "epoch": 0.061757137605685575, + "grad_norm": 8.134469985961914, + "learning_rate": 1e-05, + "loss": 1.0630282163619995, + "step": 63 + }, + { + "epoch": 0.06273740963117265, + "grad_norm": 1.958270788192749, + "learning_rate": 9.999993706252978e-06, + "loss": 1.2392271757125854, + "step": 64 + }, + { + "epoch": 0.06371768165665972, + "grad_norm": 2.5432846546173096, + "learning_rate": 9.999974825027756e-06, + "loss": 1.108931541442871, + "step": 65 + }, + { + "epoch": 0.0646979536821468, + "grad_norm": 2.372438669204712, + "learning_rate": 9.999943356371867e-06, + "loss": 0.9794902801513672, + "step": 66 + }, + { + "epoch": 0.06567822570763387, + "grad_norm": 1.4607375860214233, + "learning_rate": 9.999899300364534e-06, + "loss": 0.8074288368225098, + "step": 67 + }, + { + "epoch": 0.06665849773312094, + "grad_norm": 3.0865185260772705, + "learning_rate": 9.999842657116667e-06, + "loss": 0.9628023505210876, + "step": 68 + }, + { + "epoch": 0.06763876975860801, + "grad_norm": 1.5494884252548218, + "learning_rate": 9.999773426770864e-06, + "loss": 0.9801100492477417, + "step": 69 + }, + { + "epoch": 0.06861904178409509, + "grad_norm": 1.4236968755722046, + "learning_rate": 9.999691609501418e-06, + "loss": 1.0176517963409424, + "step": 70 + }, + { + "epoch": 0.06959931380958216, + "grad_norm": 1.5794986486434937, + "learning_rate": 9.999597205514298e-06, + "loss": 0.970253050327301, + "step": 71 + }, + { + "epoch": 0.07057958583506924, + "grad_norm": 2.677732229232788, + "learning_rate": 9.999490215047167e-06, + "loss": 0.8624630570411682, + "step": 72 + }, + { + "epoch": 0.07155985786055631, + "grad_norm": 1.9615784883499146, + "learning_rate": 9.999370638369377e-06, + "loss": 1.0122735500335693, + "step": 73 + }, + { + "epoch": 0.07254012988604337, + "grad_norm": 3.2134084701538086, + "learning_rate": 9.999238475781957e-06, + "loss": 1.0808191299438477, + "step": 74 + }, + { + "epoch": 0.07352040191153045, + "grad_norm": 1.50729501247406, + "learning_rate": 9.99909372761763e-06, + "loss": 1.0038952827453613, + "step": 75 + }, + { + "epoch": 0.07450067393701752, + "grad_norm": 1.5557780265808105, + "learning_rate": 9.998936394240798e-06, + "loss": 1.1535766124725342, + "step": 76 + }, + { + "epoch": 0.0754809459625046, + "grad_norm": 2.1890480518341064, + "learning_rate": 9.998766476047546e-06, + "loss": 0.917161762714386, + "step": 77 + }, + { + "epoch": 0.07646121798799167, + "grad_norm": 1.307956337928772, + "learning_rate": 9.998583973465647e-06, + "loss": 0.9573119878768921, + "step": 78 + }, + { + "epoch": 0.07744149001347873, + "grad_norm": 1.9047560691833496, + "learning_rate": 9.998388886954546e-06, + "loss": 0.8896398544311523, + "step": 79 + }, + { + "epoch": 0.07842176203896581, + "grad_norm": 3.1031126976013184, + "learning_rate": 9.998181217005377e-06, + "loss": 0.9498454332351685, + "step": 80 + }, + { + "epoch": 0.07940203406445288, + "grad_norm": 2.284217357635498, + "learning_rate": 9.997960964140946e-06, + "loss": 0.9973191022872925, + "step": 81 + }, + { + "epoch": 0.08038230608993996, + "grad_norm": 1.6305043697357178, + "learning_rate": 9.997728128915742e-06, + "loss": 0.935388445854187, + "step": 82 + }, + { + "epoch": 0.08136257811542703, + "grad_norm": 1.501710295677185, + "learning_rate": 9.997482711915926e-06, + "loss": 1.019201397895813, + "step": 83 + }, + { + "epoch": 0.08234285014091411, + "grad_norm": 1.4190490245819092, + "learning_rate": 9.997224713759335e-06, + "loss": 1.099223256111145, + "step": 84 + }, + { + "epoch": 0.08332312216640117, + "grad_norm": 1.886734127998352, + "learning_rate": 9.99695413509548e-06, + "loss": 0.9991997480392456, + "step": 85 + }, + { + "epoch": 0.08430339419188825, + "grad_norm": 1.508866548538208, + "learning_rate": 9.99667097660554e-06, + "loss": 0.9771667122840881, + "step": 86 + }, + { + "epoch": 0.08528366621737532, + "grad_norm": 2.9389193058013916, + "learning_rate": 9.99637523900237e-06, + "loss": 0.9765088558197021, + "step": 87 + }, + { + "epoch": 0.0862639382428624, + "grad_norm": 1.4393686056137085, + "learning_rate": 9.996066923030484e-06, + "loss": 0.9267507791519165, + "step": 88 + }, + { + "epoch": 0.08724421026834947, + "grad_norm": 1.0979642868041992, + "learning_rate": 9.995746029466071e-06, + "loss": 0.8212970495223999, + "step": 89 + }, + { + "epoch": 0.08822448229383655, + "grad_norm": 1.4870240688323975, + "learning_rate": 9.995412559116979e-06, + "loss": 0.8318418264389038, + "step": 90 + }, + { + "epoch": 0.0892047543193236, + "grad_norm": 1.707983136177063, + "learning_rate": 9.99506651282272e-06, + "loss": 0.9875527620315552, + "step": 91 + }, + { + "epoch": 0.09018502634481068, + "grad_norm": 1.1789133548736572, + "learning_rate": 9.994707891454466e-06, + "loss": 0.8298413753509521, + "step": 92 + }, + { + "epoch": 0.09116529837029776, + "grad_norm": 1.3218693733215332, + "learning_rate": 9.994336695915041e-06, + "loss": 1.0132757425308228, + "step": 93 + }, + { + "epoch": 0.09214557039578483, + "grad_norm": 1.0833605527877808, + "learning_rate": 9.993952927138933e-06, + "loss": 1.0068328380584717, + "step": 94 + }, + { + "epoch": 0.0931258424212719, + "grad_norm": 1.2598189115524292, + "learning_rate": 9.993556586092281e-06, + "loss": 0.8056250214576721, + "step": 95 + }, + { + "epoch": 0.09410611444675898, + "grad_norm": 1.4173874855041504, + "learning_rate": 9.993147673772869e-06, + "loss": 1.070406198501587, + "step": 96 + }, + { + "epoch": 0.09508638647224604, + "grad_norm": 1.3954414129257202, + "learning_rate": 9.992726191210139e-06, + "loss": 0.8578159213066101, + "step": 97 + }, + { + "epoch": 0.09606665849773312, + "grad_norm": 1.028037190437317, + "learning_rate": 9.992292139465166e-06, + "loss": 0.8095722794532776, + "step": 98 + }, + { + "epoch": 0.09704693052322019, + "grad_norm": 1.4925620555877686, + "learning_rate": 9.991845519630679e-06, + "loss": 1.0127127170562744, + "step": 99 + }, + { + "epoch": 0.09802720254870727, + "grad_norm": 1.0345618724822998, + "learning_rate": 9.991386332831042e-06, + "loss": 0.8328266143798828, + "step": 100 + }, + { + "epoch": 0.09900747457419434, + "grad_norm": 0.8846564292907715, + "learning_rate": 9.990914580222258e-06, + "loss": 0.8187386989593506, + "step": 101 + }, + { + "epoch": 0.09998774659968142, + "grad_norm": 1.276199460029602, + "learning_rate": 9.990430262991962e-06, + "loss": 1.0163452625274658, + "step": 102 + }, + { + "epoch": 0.10096801862516848, + "grad_norm": 0.9905869364738464, + "learning_rate": 9.989933382359423e-06, + "loss": 0.7760695219039917, + "step": 103 + }, + { + "epoch": 0.10194829065065555, + "grad_norm": 1.4109216928482056, + "learning_rate": 9.989423939575538e-06, + "loss": 0.9511476755142212, + "step": 104 + }, + { + "epoch": 0.10292856267614263, + "grad_norm": 1.2112044095993042, + "learning_rate": 9.988901935922826e-06, + "loss": 0.7780454754829407, + "step": 105 + }, + { + "epoch": 0.1039088347016297, + "grad_norm": 1.1124036312103271, + "learning_rate": 9.988367372715434e-06, + "loss": 0.9685283899307251, + "step": 106 + }, + { + "epoch": 0.10488910672711678, + "grad_norm": 0.8848640322685242, + "learning_rate": 9.987820251299121e-06, + "loss": 0.7924320697784424, + "step": 107 + }, + { + "epoch": 0.10586937875260385, + "grad_norm": 1.4570457935333252, + "learning_rate": 9.987260573051268e-06, + "loss": 0.8897344470024109, + "step": 108 + }, + { + "epoch": 0.10684965077809092, + "grad_norm": 0.8776848316192627, + "learning_rate": 9.986688339380863e-06, + "loss": 0.7704064846038818, + "step": 109 + }, + { + "epoch": 0.10782992280357799, + "grad_norm": 0.9722417593002319, + "learning_rate": 9.9861035517285e-06, + "loss": 0.8362361192703247, + "step": 110 + }, + { + "epoch": 0.10881019482906507, + "grad_norm": 1.0186680555343628, + "learning_rate": 9.985506211566388e-06, + "loss": 0.827409029006958, + "step": 111 + }, + { + "epoch": 0.10979046685455214, + "grad_norm": 0.8622536659240723, + "learning_rate": 9.984896320398326e-06, + "loss": 0.9348365068435669, + "step": 112 + }, + { + "epoch": 0.11077073888003922, + "grad_norm": 0.9350709915161133, + "learning_rate": 9.984273879759713e-06, + "loss": 0.7396246194839478, + "step": 113 + }, + { + "epoch": 0.11175101090552628, + "grad_norm": 1.7236732244491577, + "learning_rate": 9.983638891217544e-06, + "loss": 1.1119165420532227, + "step": 114 + }, + { + "epoch": 0.11273128293101335, + "grad_norm": 0.9292817115783691, + "learning_rate": 9.982991356370404e-06, + "loss": 0.951391339302063, + "step": 115 + }, + { + "epoch": 0.11371155495650043, + "grad_norm": 2.126209259033203, + "learning_rate": 9.982331276848457e-06, + "loss": 1.053914189338684, + "step": 116 + }, + { + "epoch": 0.1146918269819875, + "grad_norm": 0.8694857954978943, + "learning_rate": 9.981658654313458e-06, + "loss": 0.6709041595458984, + "step": 117 + }, + { + "epoch": 0.11567209900747458, + "grad_norm": 1.5078258514404297, + "learning_rate": 9.980973490458728e-06, + "loss": 0.8591851592063904, + "step": 118 + }, + { + "epoch": 0.11665237103296165, + "grad_norm": 1.173314094543457, + "learning_rate": 9.98027578700917e-06, + "loss": 0.9200767278671265, + "step": 119 + }, + { + "epoch": 0.11763264305844871, + "grad_norm": 1.153979778289795, + "learning_rate": 9.979565545721249e-06, + "loss": 0.7959837913513184, + "step": 120 + }, + { + "epoch": 0.11861291508393579, + "grad_norm": 2.0307085514068604, + "learning_rate": 9.978842768382999e-06, + "loss": 0.8494062423706055, + "step": 121 + }, + { + "epoch": 0.11959318710942286, + "grad_norm": 2.860264539718628, + "learning_rate": 9.978107456814007e-06, + "loss": 0.8416390419006348, + "step": 122 + }, + { + "epoch": 0.12057345913490994, + "grad_norm": 1.8330377340316772, + "learning_rate": 9.977359612865424e-06, + "loss": 0.9158598184585571, + "step": 123 + }, + { + "epoch": 0.12155373116039701, + "grad_norm": 1.220954179763794, + "learning_rate": 9.976599238419943e-06, + "loss": 0.8735285997390747, + "step": 124 + }, + { + "epoch": 0.12253400318588409, + "grad_norm": 1.7368972301483154, + "learning_rate": 9.975826335391808e-06, + "loss": 1.2024664878845215, + "step": 125 + }, + { + "epoch": 0.12351427521137115, + "grad_norm": 4.530630111694336, + "learning_rate": 9.975040905726799e-06, + "loss": 0.8800772428512573, + "step": 126 + }, + { + "epoch": 0.12449454723685822, + "grad_norm": 1.2298669815063477, + "learning_rate": 9.974242951402236e-06, + "loss": 0.8384938836097717, + "step": 127 + }, + { + "epoch": 0.1254748192623453, + "grad_norm": 3.0307226181030273, + "learning_rate": 9.973432474426968e-06, + "loss": 0.7943012118339539, + "step": 128 + }, + { + "epoch": 0.12645509128783236, + "grad_norm": 1.2162606716156006, + "learning_rate": 9.972609476841368e-06, + "loss": 0.906460165977478, + "step": 129 + }, + { + "epoch": 0.12743536331331945, + "grad_norm": 1.1699104309082031, + "learning_rate": 9.971773960717333e-06, + "loss": 0.7454752326011658, + "step": 130 + }, + { + "epoch": 0.1284156353388065, + "grad_norm": 0.871055006980896, + "learning_rate": 9.970925928158275e-06, + "loss": 0.7852927446365356, + "step": 131 + }, + { + "epoch": 0.1293959073642936, + "grad_norm": 0.7216456532478333, + "learning_rate": 9.970065381299112e-06, + "loss": 0.7714865207672119, + "step": 132 + }, + { + "epoch": 0.13037617938978066, + "grad_norm": 1.2287877798080444, + "learning_rate": 9.969192322306271e-06, + "loss": 0.8582690954208374, + "step": 133 + }, + { + "epoch": 0.13135645141526775, + "grad_norm": 0.7745174765586853, + "learning_rate": 9.968306753377679e-06, + "loss": 0.8014340400695801, + "step": 134 + }, + { + "epoch": 0.1323367234407548, + "grad_norm": 1.20053231716156, + "learning_rate": 9.96740867674275e-06, + "loss": 1.02126145362854, + "step": 135 + }, + { + "epoch": 0.13331699546624187, + "grad_norm": 1.1115248203277588, + "learning_rate": 9.966498094662398e-06, + "loss": 0.8542506694793701, + "step": 136 + }, + { + "epoch": 0.13429726749172896, + "grad_norm": 0.7779384255409241, + "learning_rate": 9.965575009429006e-06, + "loss": 0.6853510141372681, + "step": 137 + }, + { + "epoch": 0.13527753951721602, + "grad_norm": 0.8397386074066162, + "learning_rate": 9.964639423366442e-06, + "loss": 0.7850790023803711, + "step": 138 + }, + { + "epoch": 0.1362578115427031, + "grad_norm": 3.3384697437286377, + "learning_rate": 9.963691338830045e-06, + "loss": 0.8284836411476135, + "step": 139 + }, + { + "epoch": 0.13723808356819017, + "grad_norm": 1.1726006269454956, + "learning_rate": 9.962730758206612e-06, + "loss": 0.850119948387146, + "step": 140 + }, + { + "epoch": 0.13821835559367723, + "grad_norm": 1.0025110244750977, + "learning_rate": 9.961757683914406e-06, + "loss": 0.9284964203834534, + "step": 141 + }, + { + "epoch": 0.13919862761916432, + "grad_norm": 0.8656743764877319, + "learning_rate": 9.960772118403142e-06, + "loss": 0.7905132174491882, + "step": 142 + }, + { + "epoch": 0.14017889964465138, + "grad_norm": 0.7995918393135071, + "learning_rate": 9.959774064153977e-06, + "loss": 0.6824674606323242, + "step": 143 + }, + { + "epoch": 0.14115917167013847, + "grad_norm": 0.9554956555366516, + "learning_rate": 9.958763523679515e-06, + "loss": 0.8615709543228149, + "step": 144 + }, + { + "epoch": 0.14213944369562553, + "grad_norm": 0.8376150727272034, + "learning_rate": 9.957740499523787e-06, + "loss": 0.782550573348999, + "step": 145 + }, + { + "epoch": 0.14311971572111262, + "grad_norm": 1.0808218717575073, + "learning_rate": 9.956704994262256e-06, + "loss": 0.8757920265197754, + "step": 146 + }, + { + "epoch": 0.14409998774659968, + "grad_norm": 0.8293938040733337, + "learning_rate": 9.955657010501807e-06, + "loss": 0.8046925067901611, + "step": 147 + }, + { + "epoch": 0.14508025977208674, + "grad_norm": 0.9780465960502625, + "learning_rate": 9.954596550880735e-06, + "loss": 0.769169270992279, + "step": 148 + }, + { + "epoch": 0.14606053179757383, + "grad_norm": 0.6602773070335388, + "learning_rate": 9.95352361806875e-06, + "loss": 0.6232231855392456, + "step": 149 + }, + { + "epoch": 0.1470408038230609, + "grad_norm": 1.0579054355621338, + "learning_rate": 9.952438214766956e-06, + "loss": 0.7667106986045837, + "step": 150 + }, + { + "epoch": 0.14802107584854798, + "grad_norm": 8.054926872253418, + "learning_rate": 9.951340343707852e-06, + "loss": 0.9584368467330933, + "step": 151 + }, + { + "epoch": 0.14900134787403504, + "grad_norm": 2.2190775871276855, + "learning_rate": 9.950230007655333e-06, + "loss": 0.6349459886550903, + "step": 152 + }, + { + "epoch": 0.1499816198995221, + "grad_norm": 2.298706531524658, + "learning_rate": 9.949107209404664e-06, + "loss": 0.9834344983100891, + "step": 153 + }, + { + "epoch": 0.1509618919250092, + "grad_norm": 0.9804521799087524, + "learning_rate": 9.94797195178249e-06, + "loss": 0.9633491039276123, + "step": 154 + }, + { + "epoch": 0.15194216395049626, + "grad_norm": 0.8359074592590332, + "learning_rate": 9.946824237646823e-06, + "loss": 0.90494704246521, + "step": 155 + }, + { + "epoch": 0.15292243597598335, + "grad_norm": 1.383854866027832, + "learning_rate": 9.945664069887028e-06, + "loss": 1.0030543804168701, + "step": 156 + }, + { + "epoch": 0.1539027080014704, + "grad_norm": 0.7327127456665039, + "learning_rate": 9.944491451423829e-06, + "loss": 0.7229695916175842, + "step": 157 + }, + { + "epoch": 0.15488298002695747, + "grad_norm": 0.8276205658912659, + "learning_rate": 9.94330638520929e-06, + "loss": 0.7939118146896362, + "step": 158 + }, + { + "epoch": 0.15586325205244456, + "grad_norm": 0.7848310470581055, + "learning_rate": 9.942108874226812e-06, + "loss": 0.7627387046813965, + "step": 159 + }, + { + "epoch": 0.15684352407793162, + "grad_norm": 0.7724266052246094, + "learning_rate": 9.940898921491132e-06, + "loss": 0.7301936149597168, + "step": 160 + }, + { + "epoch": 0.1578237961034187, + "grad_norm": 1.3323100805282593, + "learning_rate": 9.9396765300483e-06, + "loss": 0.683293879032135, + "step": 161 + }, + { + "epoch": 0.15880406812890577, + "grad_norm": 1.311335563659668, + "learning_rate": 9.938441702975689e-06, + "loss": 0.7212080359458923, + "step": 162 + }, + { + "epoch": 0.15978434015439286, + "grad_norm": 0.9447430372238159, + "learning_rate": 9.937194443381972e-06, + "loss": 0.9628721475601196, + "step": 163 + }, + { + "epoch": 0.16076461217987992, + "grad_norm": 0.6951046586036682, + "learning_rate": 9.935934754407125e-06, + "loss": 0.7402358055114746, + "step": 164 + }, + { + "epoch": 0.16174488420536698, + "grad_norm": 0.7210685014724731, + "learning_rate": 9.934662639222412e-06, + "loss": 0.7613071799278259, + "step": 165 + }, + { + "epoch": 0.16272515623085407, + "grad_norm": 0.7304868102073669, + "learning_rate": 9.933378101030382e-06, + "loss": 0.6772556900978088, + "step": 166 + }, + { + "epoch": 0.16370542825634113, + "grad_norm": 0.7971092462539673, + "learning_rate": 9.93208114306486e-06, + "loss": 0.7645301818847656, + "step": 167 + }, + { + "epoch": 0.16468570028182822, + "grad_norm": 0.7262613773345947, + "learning_rate": 9.930771768590934e-06, + "loss": 0.7944133877754211, + "step": 168 + }, + { + "epoch": 0.16566597230731528, + "grad_norm": 0.5857324600219727, + "learning_rate": 9.929449980904952e-06, + "loss": 0.6576533913612366, + "step": 169 + }, + { + "epoch": 0.16664624433280234, + "grad_norm": 0.9272939562797546, + "learning_rate": 9.928115783334518e-06, + "loss": 0.8526591658592224, + "step": 170 + }, + { + "epoch": 0.16762651635828943, + "grad_norm": 0.6524718403816223, + "learning_rate": 9.926769179238467e-06, + "loss": 0.6662750244140625, + "step": 171 + }, + { + "epoch": 0.1686067883837765, + "grad_norm": 1.075812578201294, + "learning_rate": 9.925410172006874e-06, + "loss": 0.7937478423118591, + "step": 172 + }, + { + "epoch": 0.16958706040926358, + "grad_norm": 0.822924017906189, + "learning_rate": 9.924038765061042e-06, + "loss": 0.8538663387298584, + "step": 173 + }, + { + "epoch": 0.17056733243475064, + "grad_norm": 1.1041499376296997, + "learning_rate": 9.922654961853483e-06, + "loss": 0.8885059356689453, + "step": 174 + }, + { + "epoch": 0.17154760446023773, + "grad_norm": 1.4029029607772827, + "learning_rate": 9.921258765867919e-06, + "loss": 0.9263203144073486, + "step": 175 + }, + { + "epoch": 0.1725278764857248, + "grad_norm": 1.096556305885315, + "learning_rate": 9.919850180619274e-06, + "loss": 0.7909505367279053, + "step": 176 + }, + { + "epoch": 0.17350814851121185, + "grad_norm": 0.7070975303649902, + "learning_rate": 9.918429209653662e-06, + "loss": 0.7344837188720703, + "step": 177 + }, + { + "epoch": 0.17448842053669894, + "grad_norm": 1.3108410835266113, + "learning_rate": 9.916995856548371e-06, + "loss": 0.8295502662658691, + "step": 178 + }, + { + "epoch": 0.175468692562186, + "grad_norm": 0.7204867601394653, + "learning_rate": 9.915550124911866e-06, + "loss": 0.8103526830673218, + "step": 179 + }, + { + "epoch": 0.1764489645876731, + "grad_norm": 0.7560566067695618, + "learning_rate": 9.914092018383778e-06, + "loss": 0.7113056778907776, + "step": 180 + }, + { + "epoch": 0.17742923661316015, + "grad_norm": 1.012421727180481, + "learning_rate": 9.912621540634889e-06, + "loss": 0.7535778880119324, + "step": 181 + }, + { + "epoch": 0.1784095086386472, + "grad_norm": 0.8265753388404846, + "learning_rate": 9.91113869536712e-06, + "loss": 0.6618527173995972, + "step": 182 + }, + { + "epoch": 0.1793897806641343, + "grad_norm": 0.7197481989860535, + "learning_rate": 9.909643486313533e-06, + "loss": 0.7319732904434204, + "step": 183 + }, + { + "epoch": 0.18037005268962136, + "grad_norm": 0.8569908738136292, + "learning_rate": 9.908135917238321e-06, + "loss": 0.8834559917449951, + "step": 184 + }, + { + "epoch": 0.18135032471510845, + "grad_norm": 0.7561427354812622, + "learning_rate": 9.906615991936781e-06, + "loss": 0.7185438871383667, + "step": 185 + }, + { + "epoch": 0.1823305967405955, + "grad_norm": 0.8701999187469482, + "learning_rate": 9.905083714235326e-06, + "loss": 0.784651517868042, + "step": 186 + }, + { + "epoch": 0.18331086876608257, + "grad_norm": 0.8645495176315308, + "learning_rate": 9.903539087991462e-06, + "loss": 0.8580794930458069, + "step": 187 + }, + { + "epoch": 0.18429114079156966, + "grad_norm": 0.8038344979286194, + "learning_rate": 9.901982117093786e-06, + "loss": 0.8493238687515259, + "step": 188 + }, + { + "epoch": 0.18527141281705672, + "grad_norm": 0.7347437739372253, + "learning_rate": 9.900412805461968e-06, + "loss": 0.5884836316108704, + "step": 189 + }, + { + "epoch": 0.1862516848425438, + "grad_norm": 0.7285265922546387, + "learning_rate": 9.898831157046748e-06, + "loss": 0.6254444122314453, + "step": 190 + }, + { + "epoch": 0.18723195686803087, + "grad_norm": 0.848617672920227, + "learning_rate": 9.897237175829927e-06, + "loss": 0.747313380241394, + "step": 191 + }, + { + "epoch": 0.18821222889351796, + "grad_norm": 1.6770778894424438, + "learning_rate": 9.895630865824347e-06, + "loss": 0.691787838935852, + "step": 192 + }, + { + "epoch": 0.18919250091900502, + "grad_norm": 1.24128258228302, + "learning_rate": 9.894012231073895e-06, + "loss": 0.7585203647613525, + "step": 193 + }, + { + "epoch": 0.19017277294449209, + "grad_norm": 4.074337005615234, + "learning_rate": 9.89238127565348e-06, + "loss": 0.6597365736961365, + "step": 194 + }, + { + "epoch": 0.19115304496997917, + "grad_norm": 1.5159838199615479, + "learning_rate": 9.890738003669029e-06, + "loss": 0.8449289798736572, + "step": 195 + }, + { + "epoch": 0.19213331699546624, + "grad_norm": 0.9095994830131531, + "learning_rate": 9.88908241925748e-06, + "loss": 0.7729909420013428, + "step": 196 + }, + { + "epoch": 0.19311358902095332, + "grad_norm": 0.8426398038864136, + "learning_rate": 9.887414526586764e-06, + "loss": 0.7727930545806885, + "step": 197 + }, + { + "epoch": 0.19409386104644039, + "grad_norm": 0.6428210735321045, + "learning_rate": 9.885734329855798e-06, + "loss": 0.6707614064216614, + "step": 198 + }, + { + "epoch": 0.19507413307192745, + "grad_norm": 0.9163340926170349, + "learning_rate": 9.884041833294477e-06, + "loss": 0.769806981086731, + "step": 199 + }, + { + "epoch": 0.19605440509741454, + "grad_norm": 0.9019668698310852, + "learning_rate": 9.882337041163656e-06, + "loss": 0.7200514078140259, + "step": 200 + }, + { + "epoch": 0.1970346771229016, + "grad_norm": 0.8410512804985046, + "learning_rate": 9.880619957755151e-06, + "loss": 0.7845771908760071, + "step": 201 + }, + { + "epoch": 0.19801494914838869, + "grad_norm": 0.8458231687545776, + "learning_rate": 9.878890587391715e-06, + "loss": 0.7418243288993835, + "step": 202 + }, + { + "epoch": 0.19899522117387575, + "grad_norm": 0.6429169774055481, + "learning_rate": 9.877148934427037e-06, + "loss": 0.606464684009552, + "step": 203 + }, + { + "epoch": 0.19997549319936284, + "grad_norm": 0.5864877700805664, + "learning_rate": 9.875395003245724e-06, + "loss": 0.6223068833351135, + "step": 204 + }, + { + "epoch": 0.2009557652248499, + "grad_norm": 1.1887778043746948, + "learning_rate": 9.873628798263297e-06, + "loss": 0.8293617963790894, + "step": 205 + }, + { + "epoch": 0.20193603725033696, + "grad_norm": 0.7091985940933228, + "learning_rate": 9.871850323926178e-06, + "loss": 0.7142840027809143, + "step": 206 + }, + { + "epoch": 0.20291630927582405, + "grad_norm": 0.5980514287948608, + "learning_rate": 9.870059584711668e-06, + "loss": 0.6900460720062256, + "step": 207 + }, + { + "epoch": 0.2038965813013111, + "grad_norm": 0.6878236532211304, + "learning_rate": 9.868256585127956e-06, + "loss": 0.6946916580200195, + "step": 208 + }, + { + "epoch": 0.2048768533267982, + "grad_norm": 1.0756114721298218, + "learning_rate": 9.86644132971409e-06, + "loss": 0.6794236898422241, + "step": 209 + }, + { + "epoch": 0.20585712535228526, + "grad_norm": 0.634649932384491, + "learning_rate": 9.864613823039971e-06, + "loss": 0.7387229204177856, + "step": 210 + }, + { + "epoch": 0.20683739737777232, + "grad_norm": 1.0275253057479858, + "learning_rate": 9.862774069706346e-06, + "loss": 0.870427668094635, + "step": 211 + }, + { + "epoch": 0.2078176694032594, + "grad_norm": 0.704164981842041, + "learning_rate": 9.860922074344794e-06, + "loss": 0.7373089790344238, + "step": 212 + }, + { + "epoch": 0.20879794142874647, + "grad_norm": 0.8564486503601074, + "learning_rate": 9.859057841617709e-06, + "loss": 0.5971276164054871, + "step": 213 + }, + { + "epoch": 0.20977821345423356, + "grad_norm": 0.6043328046798706, + "learning_rate": 9.857181376218295e-06, + "loss": 0.6864156126976013, + "step": 214 + }, + { + "epoch": 0.21075848547972062, + "grad_norm": 0.7761573195457458, + "learning_rate": 9.855292682870552e-06, + "loss": 0.6779805421829224, + "step": 215 + }, + { + "epoch": 0.2117387575052077, + "grad_norm": 1.5135776996612549, + "learning_rate": 9.853391766329264e-06, + "loss": 0.6988109350204468, + "step": 216 + }, + { + "epoch": 0.21271902953069477, + "grad_norm": 0.6150910258293152, + "learning_rate": 9.851478631379982e-06, + "loss": 0.6663477420806885, + "step": 217 + }, + { + "epoch": 0.21369930155618183, + "grad_norm": 1.6363004446029663, + "learning_rate": 9.849553282839025e-06, + "loss": 0.6891165971755981, + "step": 218 + }, + { + "epoch": 0.21467957358166892, + "grad_norm": 0.7080655694007874, + "learning_rate": 9.847615725553457e-06, + "loss": 0.6857892870903015, + "step": 219 + }, + { + "epoch": 0.21565984560715598, + "grad_norm": 0.639260470867157, + "learning_rate": 9.845665964401072e-06, + "loss": 0.6488146781921387, + "step": 220 + }, + { + "epoch": 0.21664011763264307, + "grad_norm": 0.5393828749656677, + "learning_rate": 9.843704004290393e-06, + "loss": 0.7031430006027222, + "step": 221 + }, + { + "epoch": 0.21762038965813013, + "grad_norm": 0.6357012391090393, + "learning_rate": 9.841729850160652e-06, + "loss": 0.7031742334365845, + "step": 222 + }, + { + "epoch": 0.2186006616836172, + "grad_norm": 0.7872418165206909, + "learning_rate": 9.839743506981783e-06, + "loss": 0.6754440665245056, + "step": 223 + }, + { + "epoch": 0.21958093370910428, + "grad_norm": 0.6326587796211243, + "learning_rate": 9.837744979754397e-06, + "loss": 0.6297906041145325, + "step": 224 + }, + { + "epoch": 0.22056120573459134, + "grad_norm": 0.6477714776992798, + "learning_rate": 9.835734273509787e-06, + "loss": 0.6901752948760986, + "step": 225 + }, + { + "epoch": 0.22154147776007843, + "grad_norm": 0.7842188477516174, + "learning_rate": 9.833711393309902e-06, + "loss": 0.8501895666122437, + "step": 226 + }, + { + "epoch": 0.2225217497855655, + "grad_norm": 0.7060827016830444, + "learning_rate": 9.831676344247343e-06, + "loss": 0.7715115547180176, + "step": 227 + }, + { + "epoch": 0.22350202181105255, + "grad_norm": 0.7102391123771667, + "learning_rate": 9.829629131445342e-06, + "loss": 0.6443570256233215, + "step": 228 + }, + { + "epoch": 0.22448229383653964, + "grad_norm": 0.5093991756439209, + "learning_rate": 9.827569760057755e-06, + "loss": 0.531701922416687, + "step": 229 + }, + { + "epoch": 0.2254625658620267, + "grad_norm": 0.7141419053077698, + "learning_rate": 9.825498235269046e-06, + "loss": 0.741948127746582, + "step": 230 + }, + { + "epoch": 0.2264428378875138, + "grad_norm": 0.6522760391235352, + "learning_rate": 9.82341456229428e-06, + "loss": 0.6692137122154236, + "step": 231 + }, + { + "epoch": 0.22742310991300085, + "grad_norm": 0.8308090567588806, + "learning_rate": 9.821318746379097e-06, + "loss": 0.7460157871246338, + "step": 232 + }, + { + "epoch": 0.22840338193848794, + "grad_norm": 0.5567009449005127, + "learning_rate": 9.819210792799711e-06, + "loss": 0.6734646558761597, + "step": 233 + }, + { + "epoch": 0.229383653963975, + "grad_norm": 0.5670619606971741, + "learning_rate": 9.817090706862895e-06, + "loss": 0.6297429800033569, + "step": 234 + }, + { + "epoch": 0.23036392598946207, + "grad_norm": 0.6522438526153564, + "learning_rate": 9.814958493905962e-06, + "loss": 0.8140875101089478, + "step": 235 + }, + { + "epoch": 0.23134419801494915, + "grad_norm": 0.8690782189369202, + "learning_rate": 9.812814159296757e-06, + "loss": 0.7535607814788818, + "step": 236 + }, + { + "epoch": 0.23232447004043622, + "grad_norm": 0.7359212636947632, + "learning_rate": 9.810657708433637e-06, + "loss": 0.8144797086715698, + "step": 237 + }, + { + "epoch": 0.2333047420659233, + "grad_norm": 0.6369541883468628, + "learning_rate": 9.808489146745466e-06, + "loss": 0.674457848072052, + "step": 238 + }, + { + "epoch": 0.23428501409141037, + "grad_norm": 0.5617554187774658, + "learning_rate": 9.806308479691595e-06, + "loss": 0.668819785118103, + "step": 239 + }, + { + "epoch": 0.23526528611689743, + "grad_norm": 0.8029347658157349, + "learning_rate": 9.804115712761851e-06, + "loss": 0.9395022392272949, + "step": 240 + }, + { + "epoch": 0.23624555814238452, + "grad_norm": 0.5776801109313965, + "learning_rate": 9.801910851476524e-06, + "loss": 0.776624858379364, + "step": 241 + }, + { + "epoch": 0.23722583016787158, + "grad_norm": 0.6731631755828857, + "learning_rate": 9.799693901386346e-06, + "loss": 0.8745022416114807, + "step": 242 + }, + { + "epoch": 0.23820610219335867, + "grad_norm": 0.6632353067398071, + "learning_rate": 9.797464868072489e-06, + "loss": 0.6243095993995667, + "step": 243 + }, + { + "epoch": 0.23918637421884573, + "grad_norm": 0.5712333917617798, + "learning_rate": 9.79522375714654e-06, + "loss": 0.6474231481552124, + "step": 244 + }, + { + "epoch": 0.24016664624433282, + "grad_norm": 1.6545488834381104, + "learning_rate": 9.792970574250493e-06, + "loss": 0.7908797264099121, + "step": 245 + }, + { + "epoch": 0.24114691826981988, + "grad_norm": 0.5723490715026855, + "learning_rate": 9.790705325056735e-06, + "loss": 0.6788420677185059, + "step": 246 + }, + { + "epoch": 0.24212719029530694, + "grad_norm": 0.5819388628005981, + "learning_rate": 9.788428015268027e-06, + "loss": 0.627650797367096, + "step": 247 + }, + { + "epoch": 0.24310746232079403, + "grad_norm": 1.171949863433838, + "learning_rate": 9.786138650617494e-06, + "loss": 0.8022439479827881, + "step": 248 + }, + { + "epoch": 0.2440877343462811, + "grad_norm": 0.4696350693702698, + "learning_rate": 9.78383723686861e-06, + "loss": 0.5586154460906982, + "step": 249 + }, + { + "epoch": 0.24506800637176818, + "grad_norm": 0.5712862014770508, + "learning_rate": 9.781523779815178e-06, + "loss": 0.6085636615753174, + "step": 250 + }, + { + "epoch": 0.24604827839725524, + "grad_norm": 0.5892388820648193, + "learning_rate": 9.779198285281326e-06, + "loss": 0.6630043983459473, + "step": 251 + }, + { + "epoch": 0.2470285504227423, + "grad_norm": 0.6063647866249084, + "learning_rate": 9.776860759121485e-06, + "loss": 0.7851894497871399, + "step": 252 + }, + { + "epoch": 0.2480088224482294, + "grad_norm": 0.5152275562286377, + "learning_rate": 9.774511207220369e-06, + "loss": 0.6777390241622925, + "step": 253 + }, + { + "epoch": 0.24898909447371645, + "grad_norm": 0.6084205508232117, + "learning_rate": 9.772149635492979e-06, + "loss": 0.6391557455062866, + "step": 254 + }, + { + "epoch": 0.24996936649920354, + "grad_norm": 0.7668172121047974, + "learning_rate": 9.769776049884564e-06, + "loss": 0.662729024887085, + "step": 255 + }, + { + "epoch": 0.2509496385246906, + "grad_norm": 1.0538101196289062, + "learning_rate": 9.767390456370625e-06, + "loss": 0.9381958246231079, + "step": 256 + }, + { + "epoch": 0.25192991055017766, + "grad_norm": 0.5735601186752319, + "learning_rate": 9.76499286095689e-06, + "loss": 0.7146620750427246, + "step": 257 + }, + { + "epoch": 0.2529101825756647, + "grad_norm": 0.6433756351470947, + "learning_rate": 9.762583269679304e-06, + "loss": 0.6317933797836304, + "step": 258 + }, + { + "epoch": 0.25389045460115184, + "grad_norm": 0.852685809135437, + "learning_rate": 9.760161688604008e-06, + "loss": 0.8174331188201904, + "step": 259 + }, + { + "epoch": 0.2548707266266389, + "grad_norm": 0.5660604238510132, + "learning_rate": 9.75772812382733e-06, + "loss": 0.6192631721496582, + "step": 260 + }, + { + "epoch": 0.25585099865212596, + "grad_norm": 0.7691637873649597, + "learning_rate": 9.755282581475769e-06, + "loss": 0.6771009564399719, + "step": 261 + }, + { + "epoch": 0.256831270677613, + "grad_norm": 0.591249942779541, + "learning_rate": 9.75282506770597e-06, + "loss": 0.6050739288330078, + "step": 262 + }, + { + "epoch": 0.2578115427031001, + "grad_norm": 0.8145242929458618, + "learning_rate": 9.750355588704728e-06, + "loss": 0.6268457174301147, + "step": 263 + }, + { + "epoch": 0.2587918147285872, + "grad_norm": 0.5998173952102661, + "learning_rate": 9.747874150688948e-06, + "loss": 0.6692138910293579, + "step": 264 + }, + { + "epoch": 0.25977208675407426, + "grad_norm": 0.5826566219329834, + "learning_rate": 9.745380759905648e-06, + "loss": 0.6950204372406006, + "step": 265 + }, + { + "epoch": 0.2607523587795613, + "grad_norm": 0.5034763813018799, + "learning_rate": 9.742875422631937e-06, + "loss": 0.5742853879928589, + "step": 266 + }, + { + "epoch": 0.2617326308050484, + "grad_norm": 0.6649979948997498, + "learning_rate": 9.740358145174999e-06, + "loss": 0.6680870056152344, + "step": 267 + }, + { + "epoch": 0.2627129028305355, + "grad_norm": 0.5235174894332886, + "learning_rate": 9.737828933872076e-06, + "loss": 0.6588406562805176, + "step": 268 + }, + { + "epoch": 0.26369317485602256, + "grad_norm": 0.5755462646484375, + "learning_rate": 9.735287795090455e-06, + "loss": 0.7380120754241943, + "step": 269 + }, + { + "epoch": 0.2646734468815096, + "grad_norm": 0.7009977698326111, + "learning_rate": 9.73273473522745e-06, + "loss": 0.6192063093185425, + "step": 270 + }, + { + "epoch": 0.2656537189069967, + "grad_norm": 0.612812340259552, + "learning_rate": 9.730169760710385e-06, + "loss": 0.6974408626556396, + "step": 271 + }, + { + "epoch": 0.26663399093248374, + "grad_norm": 0.8031991124153137, + "learning_rate": 9.727592877996585e-06, + "loss": 0.664556086063385, + "step": 272 + }, + { + "epoch": 0.26761426295797086, + "grad_norm": 0.5645917057991028, + "learning_rate": 9.725004093573343e-06, + "loss": 0.7329329252243042, + "step": 273 + }, + { + "epoch": 0.2685945349834579, + "grad_norm": 0.7882102131843567, + "learning_rate": 9.722403413957924e-06, + "loss": 0.6108486652374268, + "step": 274 + }, + { + "epoch": 0.269574807008945, + "grad_norm": 0.5724264979362488, + "learning_rate": 9.719790845697534e-06, + "loss": 0.7101311683654785, + "step": 275 + }, + { + "epoch": 0.27055507903443204, + "grad_norm": 0.5334291458129883, + "learning_rate": 9.717166395369312e-06, + "loss": 0.626849889755249, + "step": 276 + }, + { + "epoch": 0.2715353510599191, + "grad_norm": 0.5758081078529358, + "learning_rate": 9.71453006958031e-06, + "loss": 0.6180967092514038, + "step": 277 + }, + { + "epoch": 0.2725156230854062, + "grad_norm": 0.4749200642108917, + "learning_rate": 9.711881874967471e-06, + "loss": 0.4926038682460785, + "step": 278 + }, + { + "epoch": 0.2734958951108933, + "grad_norm": 0.6426512002944946, + "learning_rate": 9.709221818197626e-06, + "loss": 0.7141193151473999, + "step": 279 + }, + { + "epoch": 0.27447616713638034, + "grad_norm": 0.8100751042366028, + "learning_rate": 9.70654990596746e-06, + "loss": 0.8084619045257568, + "step": 280 + }, + { + "epoch": 0.2754564391618674, + "grad_norm": 0.5176052451133728, + "learning_rate": 9.703866145003512e-06, + "loss": 0.6762614250183105, + "step": 281 + }, + { + "epoch": 0.27643671118735447, + "grad_norm": 0.584479033946991, + "learning_rate": 9.701170542062149e-06, + "loss": 0.6309274435043335, + "step": 282 + }, + { + "epoch": 0.2774169832128416, + "grad_norm": 0.5708397030830383, + "learning_rate": 9.698463103929542e-06, + "loss": 0.7273784875869751, + "step": 283 + }, + { + "epoch": 0.27839725523832864, + "grad_norm": 0.49435490369796753, + "learning_rate": 9.69574383742167e-06, + "loss": 0.5571186542510986, + "step": 284 + }, + { + "epoch": 0.2793775272638157, + "grad_norm": 0.6723142862319946, + "learning_rate": 9.69301274938428e-06, + "loss": 0.7581765651702881, + "step": 285 + }, + { + "epoch": 0.28035779928930277, + "grad_norm": 0.5881115198135376, + "learning_rate": 9.690269846692882e-06, + "loss": 0.6826155185699463, + "step": 286 + }, + { + "epoch": 0.28133807131478983, + "grad_norm": 0.5932793021202087, + "learning_rate": 9.687515136252732e-06, + "loss": 0.6524624824523926, + "step": 287 + }, + { + "epoch": 0.28231834334027694, + "grad_norm": 0.6231981515884399, + "learning_rate": 9.68474862499881e-06, + "loss": 0.6794503927230835, + "step": 288 + }, + { + "epoch": 0.283298615365764, + "grad_norm": 0.582617998123169, + "learning_rate": 9.681970319895804e-06, + "loss": 0.7047141194343567, + "step": 289 + }, + { + "epoch": 0.28427888739125107, + "grad_norm": 1.219525694847107, + "learning_rate": 9.679180227938095e-06, + "loss": 0.6844112873077393, + "step": 290 + }, + { + "epoch": 0.28525915941673813, + "grad_norm": 0.646946132183075, + "learning_rate": 9.676378356149733e-06, + "loss": 0.6921058893203735, + "step": 291 + }, + { + "epoch": 0.28623943144222525, + "grad_norm": 0.6992022395133972, + "learning_rate": 9.67356471158443e-06, + "loss": 0.6100419759750366, + "step": 292 + }, + { + "epoch": 0.2872197034677123, + "grad_norm": 0.623558759689331, + "learning_rate": 9.670739301325534e-06, + "loss": 0.6630300879478455, + "step": 293 + }, + { + "epoch": 0.28819997549319937, + "grad_norm": 0.6203799247741699, + "learning_rate": 9.667902132486009e-06, + "loss": 0.8250691890716553, + "step": 294 + }, + { + "epoch": 0.28918024751868643, + "grad_norm": 0.5172725915908813, + "learning_rate": 9.665053212208426e-06, + "loss": 0.5871583223342896, + "step": 295 + }, + { + "epoch": 0.2901605195441735, + "grad_norm": 0.585059404373169, + "learning_rate": 9.662192547664938e-06, + "loss": 0.6851358413696289, + "step": 296 + }, + { + "epoch": 0.2911407915696606, + "grad_norm": 0.5669979453086853, + "learning_rate": 9.659320146057263e-06, + "loss": 0.6426068544387817, + "step": 297 + }, + { + "epoch": 0.29212106359514767, + "grad_norm": 0.4750717282295227, + "learning_rate": 9.65643601461667e-06, + "loss": 0.5325371026992798, + "step": 298 + }, + { + "epoch": 0.29310133562063473, + "grad_norm": 0.5632511973381042, + "learning_rate": 9.653540160603956e-06, + "loss": 0.6740025877952576, + "step": 299 + }, + { + "epoch": 0.2940816076461218, + "grad_norm": 0.6044143438339233, + "learning_rate": 9.650632591309431e-06, + "loss": 0.7068796753883362, + "step": 300 + }, + { + "epoch": 0.29506187967160885, + "grad_norm": 0.5127572417259216, + "learning_rate": 9.647713314052896e-06, + "loss": 0.5933098196983337, + "step": 301 + }, + { + "epoch": 0.29604215169709597, + "grad_norm": 0.5830118060112, + "learning_rate": 9.64478233618363e-06, + "loss": 0.6408910751342773, + "step": 302 + }, + { + "epoch": 0.29702242372258303, + "grad_norm": 0.5796244740486145, + "learning_rate": 9.641839665080363e-06, + "loss": 0.632135272026062, + "step": 303 + }, + { + "epoch": 0.2980026957480701, + "grad_norm": 0.5338972210884094, + "learning_rate": 9.63888530815127e-06, + "loss": 0.535368800163269, + "step": 304 + }, + { + "epoch": 0.29898296777355715, + "grad_norm": 0.490498423576355, + "learning_rate": 9.635919272833938e-06, + "loss": 0.595739483833313, + "step": 305 + }, + { + "epoch": 0.2999632397990442, + "grad_norm": 0.6124870181083679, + "learning_rate": 9.632941566595357e-06, + "loss": 0.7355586886405945, + "step": 306 + }, + { + "epoch": 0.30094351182453133, + "grad_norm": 0.4968045651912689, + "learning_rate": 9.629952196931902e-06, + "loss": 0.5700988173484802, + "step": 307 + }, + { + "epoch": 0.3019237838500184, + "grad_norm": 0.5857828855514526, + "learning_rate": 9.626951171369306e-06, + "loss": 0.6341098546981812, + "step": 308 + }, + { + "epoch": 0.30290405587550545, + "grad_norm": 0.7570348381996155, + "learning_rate": 9.623938497462647e-06, + "loss": 0.9097845554351807, + "step": 309 + }, + { + "epoch": 0.3038843279009925, + "grad_norm": 0.6131370663642883, + "learning_rate": 9.620914182796327e-06, + "loss": 0.49961164593696594, + "step": 310 + }, + { + "epoch": 0.3048645999264796, + "grad_norm": 0.6986640691757202, + "learning_rate": 9.617878234984056e-06, + "loss": 0.6150896549224854, + "step": 311 + }, + { + "epoch": 0.3058448719519667, + "grad_norm": 0.5393136739730835, + "learning_rate": 9.61483066166883e-06, + "loss": 0.5867908000946045, + "step": 312 + }, + { + "epoch": 0.30682514397745375, + "grad_norm": 0.5819171071052551, + "learning_rate": 9.611771470522908e-06, + "loss": 0.665069580078125, + "step": 313 + }, + { + "epoch": 0.3078054160029408, + "grad_norm": 0.5913625359535217, + "learning_rate": 9.608700669247803e-06, + "loss": 0.6079078316688538, + "step": 314 + }, + { + "epoch": 0.3087856880284279, + "grad_norm": 0.544347882270813, + "learning_rate": 9.60561826557425e-06, + "loss": 0.5789344906806946, + "step": 315 + }, + { + "epoch": 0.30976596005391493, + "grad_norm": 0.6326791644096375, + "learning_rate": 9.602524267262202e-06, + "loss": 0.6234204769134521, + "step": 316 + }, + { + "epoch": 0.31074623207940205, + "grad_norm": 0.6705621480941772, + "learning_rate": 9.599418682100793e-06, + "loss": 0.5959568023681641, + "step": 317 + }, + { + "epoch": 0.3117265041048891, + "grad_norm": 0.5172932744026184, + "learning_rate": 9.596301517908329e-06, + "loss": 0.5890722870826721, + "step": 318 + }, + { + "epoch": 0.3127067761303762, + "grad_norm": 0.9102640151977539, + "learning_rate": 9.59317278253227e-06, + "loss": 0.700379490852356, + "step": 319 + }, + { + "epoch": 0.31368704815586324, + "grad_norm": 0.5571013689041138, + "learning_rate": 9.5900324838492e-06, + "loss": 0.6454604268074036, + "step": 320 + }, + { + "epoch": 0.31466732018135035, + "grad_norm": 0.5500742793083191, + "learning_rate": 9.586880629764817e-06, + "loss": 0.5959937572479248, + "step": 321 + }, + { + "epoch": 0.3156475922068374, + "grad_norm": 0.667962908744812, + "learning_rate": 9.583717228213914e-06, + "loss": 0.6092185974121094, + "step": 322 + }, + { + "epoch": 0.3166278642323245, + "grad_norm": 1.0929455757141113, + "learning_rate": 9.580542287160348e-06, + "loss": 0.6907039284706116, + "step": 323 + }, + { + "epoch": 0.31760813625781154, + "grad_norm": 0.6506457328796387, + "learning_rate": 9.577355814597031e-06, + "loss": 0.7235692739486694, + "step": 324 + }, + { + "epoch": 0.3185884082832986, + "grad_norm": 0.6015403866767883, + "learning_rate": 9.574157818545902e-06, + "loss": 0.5779285430908203, + "step": 325 + }, + { + "epoch": 0.3195686803087857, + "grad_norm": 0.7786325216293335, + "learning_rate": 9.570948307057912e-06, + "loss": 0.7327282428741455, + "step": 326 + }, + { + "epoch": 0.3205489523342728, + "grad_norm": 0.6935060024261475, + "learning_rate": 9.567727288213005e-06, + "loss": 0.5526785850524902, + "step": 327 + }, + { + "epoch": 0.32152922435975984, + "grad_norm": 0.5690432786941528, + "learning_rate": 9.564494770120089e-06, + "loss": 0.5947889089584351, + "step": 328 + }, + { + "epoch": 0.3225094963852469, + "grad_norm": 0.8004881143569946, + "learning_rate": 9.561250760917026e-06, + "loss": 0.6638854742050171, + "step": 329 + }, + { + "epoch": 0.32348976841073396, + "grad_norm": 0.56986004114151, + "learning_rate": 9.557995268770609e-06, + "loss": 0.7149271965026855, + "step": 330 + }, + { + "epoch": 0.3244700404362211, + "grad_norm": 0.5093994736671448, + "learning_rate": 9.554728301876525e-06, + "loss": 0.573759913444519, + "step": 331 + }, + { + "epoch": 0.32545031246170814, + "grad_norm": 0.6659159660339355, + "learning_rate": 9.551449868459371e-06, + "loss": 0.6619836091995239, + "step": 332 + }, + { + "epoch": 0.3264305844871952, + "grad_norm": 0.7009037137031555, + "learning_rate": 9.548159976772593e-06, + "loss": 0.6384791135787964, + "step": 333 + }, + { + "epoch": 0.32741085651268226, + "grad_norm": 0.588800847530365, + "learning_rate": 9.544858635098491e-06, + "loss": 0.5931030511856079, + "step": 334 + }, + { + "epoch": 0.3283911285381693, + "grad_norm": 0.6182215213775635, + "learning_rate": 9.541545851748186e-06, + "loss": 0.7085162401199341, + "step": 335 + }, + { + "epoch": 0.32937140056365644, + "grad_norm": 0.5252450108528137, + "learning_rate": 9.538221635061611e-06, + "loss": 0.5386106967926025, + "step": 336 + }, + { + "epoch": 0.3303516725891435, + "grad_norm": 0.5651358962059021, + "learning_rate": 9.534885993407474e-06, + "loss": 0.6456975936889648, + "step": 337 + }, + { + "epoch": 0.33133194461463056, + "grad_norm": 0.5379696488380432, + "learning_rate": 9.531538935183252e-06, + "loss": 0.6666260361671448, + "step": 338 + }, + { + "epoch": 0.3323122166401176, + "grad_norm": 0.656000554561615, + "learning_rate": 9.528180468815155e-06, + "loss": 0.6818495988845825, + "step": 339 + }, + { + "epoch": 0.3332924886656047, + "grad_norm": 0.8253548741340637, + "learning_rate": 9.524810602758122e-06, + "loss": 0.5551189184188843, + "step": 340 + }, + { + "epoch": 0.3342727606910918, + "grad_norm": 0.8118085265159607, + "learning_rate": 9.521429345495787e-06, + "loss": 0.6034508943557739, + "step": 341 + }, + { + "epoch": 0.33525303271657886, + "grad_norm": 0.8866831660270691, + "learning_rate": 9.518036705540459e-06, + "loss": 0.7978249788284302, + "step": 342 + }, + { + "epoch": 0.3362333047420659, + "grad_norm": 0.5758262276649475, + "learning_rate": 9.514632691433108e-06, + "loss": 0.6506916284561157, + "step": 343 + }, + { + "epoch": 0.337213576767553, + "grad_norm": 0.6339187026023865, + "learning_rate": 9.51121731174333e-06, + "loss": 0.6326726675033569, + "step": 344 + }, + { + "epoch": 0.33819384879304004, + "grad_norm": 0.8769382834434509, + "learning_rate": 9.507790575069347e-06, + "loss": 0.7365003824234009, + "step": 345 + }, + { + "epoch": 0.33917412081852716, + "grad_norm": 0.7927872538566589, + "learning_rate": 9.504352490037957e-06, + "loss": 0.5798900127410889, + "step": 346 + }, + { + "epoch": 0.3401543928440142, + "grad_norm": 0.5708814263343811, + "learning_rate": 9.50090306530454e-06, + "loss": 0.7537117004394531, + "step": 347 + }, + { + "epoch": 0.3411346648695013, + "grad_norm": 0.48839452862739563, + "learning_rate": 9.497442309553017e-06, + "loss": 0.5822838544845581, + "step": 348 + }, + { + "epoch": 0.34211493689498834, + "grad_norm": 0.7066628336906433, + "learning_rate": 9.493970231495836e-06, + "loss": 0.7837392687797546, + "step": 349 + }, + { + "epoch": 0.34309520892047546, + "grad_norm": 0.584050178527832, + "learning_rate": 9.490486839873949e-06, + "loss": 0.7187587022781372, + "step": 350 + }, + { + "epoch": 0.3440754809459625, + "grad_norm": 0.5140157341957092, + "learning_rate": 9.486992143456792e-06, + "loss": 0.6060421466827393, + "step": 351 + }, + { + "epoch": 0.3450557529714496, + "grad_norm": 0.5367881059646606, + "learning_rate": 9.483486151042258e-06, + "loss": 0.5815377235412598, + "step": 352 + }, + { + "epoch": 0.34603602499693664, + "grad_norm": 0.4939191937446594, + "learning_rate": 9.47996887145668e-06, + "loss": 0.5326993465423584, + "step": 353 + }, + { + "epoch": 0.3470162970224237, + "grad_norm": 0.6358892321586609, + "learning_rate": 9.476440313554803e-06, + "loss": 0.6004680395126343, + "step": 354 + }, + { + "epoch": 0.3479965690479108, + "grad_norm": 0.6140431761741638, + "learning_rate": 9.47290048621977e-06, + "loss": 0.6206232309341431, + "step": 355 + }, + { + "epoch": 0.3489768410733979, + "grad_norm": 0.5943885445594788, + "learning_rate": 9.469349398363088e-06, + "loss": 0.5952731966972351, + "step": 356 + }, + { + "epoch": 0.34995711309888494, + "grad_norm": 0.49166974425315857, + "learning_rate": 9.46578705892462e-06, + "loss": 0.6236333847045898, + "step": 357 + }, + { + "epoch": 0.350937385124372, + "grad_norm": 0.6081139445304871, + "learning_rate": 9.46221347687255e-06, + "loss": 0.7176322937011719, + "step": 358 + }, + { + "epoch": 0.35191765714985906, + "grad_norm": 0.5212146639823914, + "learning_rate": 9.458628661203368e-06, + "loss": 0.5600054860115051, + "step": 359 + }, + { + "epoch": 0.3528979291753462, + "grad_norm": 0.5689999461174011, + "learning_rate": 9.45503262094184e-06, + "loss": 0.6379108428955078, + "step": 360 + }, + { + "epoch": 0.35387820120083324, + "grad_norm": 0.5177587866783142, + "learning_rate": 9.451425365140997e-06, + "loss": 0.6243526935577393, + "step": 361 + }, + { + "epoch": 0.3548584732263203, + "grad_norm": 0.4048973619937897, + "learning_rate": 9.447806902882098e-06, + "loss": 0.5460570454597473, + "step": 362 + }, + { + "epoch": 0.35583874525180736, + "grad_norm": 0.5066366195678711, + "learning_rate": 9.444177243274619e-06, + "loss": 0.5914937853813171, + "step": 363 + }, + { + "epoch": 0.3568190172772944, + "grad_norm": 1.1004585027694702, + "learning_rate": 9.440536395456223e-06, + "loss": 0.6613606214523315, + "step": 364 + }, + { + "epoch": 0.35779928930278154, + "grad_norm": 0.4945545196533203, + "learning_rate": 9.43688436859274e-06, + "loss": 0.6563936471939087, + "step": 365 + }, + { + "epoch": 0.3587795613282686, + "grad_norm": 0.46797290444374084, + "learning_rate": 9.433221171878144e-06, + "loss": 0.5227972269058228, + "step": 366 + }, + { + "epoch": 0.35975983335375566, + "grad_norm": 0.6280576586723328, + "learning_rate": 9.429546814534528e-06, + "loss": 0.6633756160736084, + "step": 367 + }, + { + "epoch": 0.3607401053792427, + "grad_norm": 0.6422777771949768, + "learning_rate": 9.425861305812083e-06, + "loss": 0.7747545838356018, + "step": 368 + }, + { + "epoch": 0.3617203774047298, + "grad_norm": 0.6484074592590332, + "learning_rate": 9.422164654989073e-06, + "loss": 0.638740062713623, + "step": 369 + }, + { + "epoch": 0.3627006494302169, + "grad_norm": 0.4812702238559723, + "learning_rate": 9.41845687137181e-06, + "loss": 0.7052801251411438, + "step": 370 + }, + { + "epoch": 0.36368092145570396, + "grad_norm": 0.5138601064682007, + "learning_rate": 9.414737964294636e-06, + "loss": 0.58039790391922, + "step": 371 + }, + { + "epoch": 0.364661193481191, + "grad_norm": 0.7280306220054626, + "learning_rate": 9.411007943119895e-06, + "loss": 0.8962255716323853, + "step": 372 + }, + { + "epoch": 0.3656414655066781, + "grad_norm": 0.5673227906227112, + "learning_rate": 9.40726681723791e-06, + "loss": 0.6811161637306213, + "step": 373 + }, + { + "epoch": 0.36662173753216515, + "grad_norm": 0.5575160980224609, + "learning_rate": 9.403514596066964e-06, + "loss": 0.6203138828277588, + "step": 374 + }, + { + "epoch": 0.36760200955765227, + "grad_norm": 0.5194122195243835, + "learning_rate": 9.399751289053267e-06, + "loss": 0.6002714037895203, + "step": 375 + }, + { + "epoch": 0.3685822815831393, + "grad_norm": 0.476211816072464, + "learning_rate": 9.39597690567094e-06, + "loss": 0.49280551075935364, + "step": 376 + }, + { + "epoch": 0.3695625536086264, + "grad_norm": 0.5374362468719482, + "learning_rate": 9.392191455421989e-06, + "loss": 0.5507866144180298, + "step": 377 + }, + { + "epoch": 0.37054282563411345, + "grad_norm": 0.48763546347618103, + "learning_rate": 9.388394947836278e-06, + "loss": 0.5378220677375793, + "step": 378 + }, + { + "epoch": 0.37152309765960057, + "grad_norm": 0.49194395542144775, + "learning_rate": 9.384587392471516e-06, + "loss": 0.5602411031723022, + "step": 379 + }, + { + "epoch": 0.3725033696850876, + "grad_norm": 0.5846983194351196, + "learning_rate": 9.380768798913213e-06, + "loss": 0.6200522780418396, + "step": 380 + }, + { + "epoch": 0.3734836417105747, + "grad_norm": 0.6853125691413879, + "learning_rate": 9.376939176774678e-06, + "loss": 0.5838781595230103, + "step": 381 + }, + { + "epoch": 0.37446391373606175, + "grad_norm": 0.5014895796775818, + "learning_rate": 9.37309853569698e-06, + "loss": 0.5761265754699707, + "step": 382 + }, + { + "epoch": 0.3754441857615488, + "grad_norm": 0.536831259727478, + "learning_rate": 9.369246885348926e-06, + "loss": 0.6839797496795654, + "step": 383 + }, + { + "epoch": 0.3764244577870359, + "grad_norm": 0.47531557083129883, + "learning_rate": 9.365384235427042e-06, + "loss": 0.6124746799468994, + "step": 384 + }, + { + "epoch": 0.377404729812523, + "grad_norm": 0.5946936011314392, + "learning_rate": 9.361510595655545e-06, + "loss": 0.7137396931648254, + "step": 385 + }, + { + "epoch": 0.37838500183801005, + "grad_norm": 0.502773106098175, + "learning_rate": 9.357625975786318e-06, + "loss": 0.6333768963813782, + "step": 386 + }, + { + "epoch": 0.3793652738634971, + "grad_norm": 0.4880274534225464, + "learning_rate": 9.353730385598887e-06, + "loss": 0.6179925799369812, + "step": 387 + }, + { + "epoch": 0.38034554588898417, + "grad_norm": 0.4644832909107208, + "learning_rate": 9.349823834900396e-06, + "loss": 0.5528720617294312, + "step": 388 + }, + { + "epoch": 0.3813258179144713, + "grad_norm": 0.5169019103050232, + "learning_rate": 9.345906333525582e-06, + "loss": 0.5545703768730164, + "step": 389 + }, + { + "epoch": 0.38230608993995835, + "grad_norm": 0.5275116562843323, + "learning_rate": 9.341977891336748e-06, + "loss": 0.6071966886520386, + "step": 390 + }, + { + "epoch": 0.3832863619654454, + "grad_norm": 0.6112439036369324, + "learning_rate": 9.338038518223746e-06, + "loss": 0.5499929189682007, + "step": 391 + }, + { + "epoch": 0.38426663399093247, + "grad_norm": 0.4731747806072235, + "learning_rate": 9.334088224103942e-06, + "loss": 0.5589090585708618, + "step": 392 + }, + { + "epoch": 0.38524690601641953, + "grad_norm": 0.6080041527748108, + "learning_rate": 9.330127018922195e-06, + "loss": 0.6990543603897095, + "step": 393 + }, + { + "epoch": 0.38622717804190665, + "grad_norm": 0.5127442479133606, + "learning_rate": 9.326154912650834e-06, + "loss": 0.5527446269989014, + "step": 394 + }, + { + "epoch": 0.3872074500673937, + "grad_norm": 0.4985249936580658, + "learning_rate": 9.322171915289635e-06, + "loss": 0.6704634428024292, + "step": 395 + }, + { + "epoch": 0.38818772209288077, + "grad_norm": 0.4670063853263855, + "learning_rate": 9.318178036865786e-06, + "loss": 0.5786021947860718, + "step": 396 + }, + { + "epoch": 0.38916799411836783, + "grad_norm": 0.6596277356147766, + "learning_rate": 9.314173287433874e-06, + "loss": 0.7387778759002686, + "step": 397 + }, + { + "epoch": 0.3901482661438549, + "grad_norm": 0.4874238967895508, + "learning_rate": 9.310157677075847e-06, + "loss": 0.5646405220031738, + "step": 398 + }, + { + "epoch": 0.391128538169342, + "grad_norm": 0.7856294512748718, + "learning_rate": 9.306131215901004e-06, + "loss": 0.641767680644989, + "step": 399 + }, + { + "epoch": 0.39210881019482907, + "grad_norm": 0.3762321174144745, + "learning_rate": 9.302093914045953e-06, + "loss": 0.4375483989715576, + "step": 400 + }, + { + "epoch": 0.39308908222031613, + "grad_norm": 0.5250818133354187, + "learning_rate": 9.298045781674595e-06, + "loss": 0.58973628282547, + "step": 401 + }, + { + "epoch": 0.3940693542458032, + "grad_norm": 0.5813953876495361, + "learning_rate": 9.293986828978106e-06, + "loss": 0.6511713266372681, + "step": 402 + }, + { + "epoch": 0.3950496262712903, + "grad_norm": 0.5056664347648621, + "learning_rate": 9.289917066174887e-06, + "loss": 0.6352890729904175, + "step": 403 + }, + { + "epoch": 0.39602989829677737, + "grad_norm": 0.4536687433719635, + "learning_rate": 9.285836503510562e-06, + "loss": 0.5708507895469666, + "step": 404 + }, + { + "epoch": 0.39701017032226443, + "grad_norm": 0.5224312543869019, + "learning_rate": 9.281745151257946e-06, + "loss": 0.5995397567749023, + "step": 405 + }, + { + "epoch": 0.3979904423477515, + "grad_norm": 0.580417811870575, + "learning_rate": 9.277643019717011e-06, + "loss": 0.7497074604034424, + "step": 406 + }, + { + "epoch": 0.39897071437323856, + "grad_norm": 0.6126673221588135, + "learning_rate": 9.273530119214868e-06, + "loss": 0.5598350763320923, + "step": 407 + }, + { + "epoch": 0.39995098639872567, + "grad_norm": 0.5506340861320496, + "learning_rate": 9.269406460105742e-06, + "loss": 0.5485457181930542, + "step": 408 + }, + { + "epoch": 0.40093125842421273, + "grad_norm": 0.4401313066482544, + "learning_rate": 9.265272052770936e-06, + "loss": 0.5437591075897217, + "step": 409 + }, + { + "epoch": 0.4019115304496998, + "grad_norm": 0.7231860160827637, + "learning_rate": 9.261126907618818e-06, + "loss": 0.5335922837257385, + "step": 410 + }, + { + "epoch": 0.40289180247518686, + "grad_norm": 0.6745095252990723, + "learning_rate": 9.256971035084786e-06, + "loss": 0.6721965074539185, + "step": 411 + }, + { + "epoch": 0.4038720745006739, + "grad_norm": 0.5086560249328613, + "learning_rate": 9.252804445631243e-06, + "loss": 0.5465257167816162, + "step": 412 + }, + { + "epoch": 0.40485234652616103, + "grad_norm": 0.5411233305931091, + "learning_rate": 9.248627149747573e-06, + "loss": 0.76491379737854, + "step": 413 + }, + { + "epoch": 0.4058326185516481, + "grad_norm": 0.5963067412376404, + "learning_rate": 9.244439157950114e-06, + "loss": 0.7780289649963379, + "step": 414 + }, + { + "epoch": 0.40681289057713516, + "grad_norm": 0.5068957805633545, + "learning_rate": 9.24024048078213e-06, + "loss": 0.5478646755218506, + "step": 415 + }, + { + "epoch": 0.4077931626026222, + "grad_norm": 0.45583173632621765, + "learning_rate": 9.236031128813786e-06, + "loss": 0.6128138899803162, + "step": 416 + }, + { + "epoch": 0.4087734346281093, + "grad_norm": 0.5057873129844666, + "learning_rate": 9.231811112642121e-06, + "loss": 0.6155551671981812, + "step": 417 + }, + { + "epoch": 0.4097537066535964, + "grad_norm": 0.44246384501457214, + "learning_rate": 9.227580442891022e-06, + "loss": 0.5776406526565552, + "step": 418 + }, + { + "epoch": 0.41073397867908346, + "grad_norm": 0.5929768681526184, + "learning_rate": 9.223339130211194e-06, + "loss": 0.5738487839698792, + "step": 419 + }, + { + "epoch": 0.4117142507045705, + "grad_norm": 0.5466901659965515, + "learning_rate": 9.219087185280132e-06, + "loss": 0.6888821721076965, + "step": 420 + }, + { + "epoch": 0.4126945227300576, + "grad_norm": 0.5927895903587341, + "learning_rate": 9.214824618802108e-06, + "loss": 0.7219152450561523, + "step": 421 + }, + { + "epoch": 0.41367479475554464, + "grad_norm": 0.45233601331710815, + "learning_rate": 9.210551441508126e-06, + "loss": 0.4950752258300781, + "step": 422 + }, + { + "epoch": 0.41465506678103176, + "grad_norm": 0.4221014082431793, + "learning_rate": 9.206267664155906e-06, + "loss": 0.5129454135894775, + "step": 423 + }, + { + "epoch": 0.4156353388065188, + "grad_norm": 0.517291784286499, + "learning_rate": 9.201973297529852e-06, + "loss": 0.5749497413635254, + "step": 424 + }, + { + "epoch": 0.4166156108320059, + "grad_norm": 0.5062311887741089, + "learning_rate": 9.197668352441025e-06, + "loss": 0.5585308074951172, + "step": 425 + }, + { + "epoch": 0.41759588285749294, + "grad_norm": 0.44994261860847473, + "learning_rate": 9.193352839727122e-06, + "loss": 0.5468809604644775, + "step": 426 + }, + { + "epoch": 0.41857615488298, + "grad_norm": 0.5119527578353882, + "learning_rate": 9.189026770252437e-06, + "loss": 0.581660807132721, + "step": 427 + }, + { + "epoch": 0.4195564269084671, + "grad_norm": 0.5557008981704712, + "learning_rate": 9.18469015490785e-06, + "loss": 0.6926128268241882, + "step": 428 + }, + { + "epoch": 0.4205366989339542, + "grad_norm": 0.4753175377845764, + "learning_rate": 9.18034300461078e-06, + "loss": 0.6371850967407227, + "step": 429 + }, + { + "epoch": 0.42151697095944124, + "grad_norm": 0.6206807494163513, + "learning_rate": 9.175985330305176e-06, + "loss": 0.6325344443321228, + "step": 430 + }, + { + "epoch": 0.4224972429849283, + "grad_norm": 0.49500077962875366, + "learning_rate": 9.171617142961477e-06, + "loss": 0.584123969078064, + "step": 431 + }, + { + "epoch": 0.4234775150104154, + "grad_norm": 0.5801467299461365, + "learning_rate": 9.167238453576589e-06, + "loss": 0.5789539813995361, + "step": 432 + }, + { + "epoch": 0.4244577870359025, + "grad_norm": 0.5026708245277405, + "learning_rate": 9.162849273173857e-06, + "loss": 0.60198974609375, + "step": 433 + }, + { + "epoch": 0.42543805906138954, + "grad_norm": 0.5288900136947632, + "learning_rate": 9.158449612803039e-06, + "loss": 0.7147405743598938, + "step": 434 + }, + { + "epoch": 0.4264183310868766, + "grad_norm": 0.48020800948143005, + "learning_rate": 9.154039483540273e-06, + "loss": 0.6171431541442871, + "step": 435 + }, + { + "epoch": 0.42739860311236366, + "grad_norm": 0.5342614650726318, + "learning_rate": 9.149618896488056e-06, + "loss": 0.5938845872879028, + "step": 436 + }, + { + "epoch": 0.4283788751378508, + "grad_norm": 0.4195495843887329, + "learning_rate": 9.145187862775208e-06, + "loss": 0.44220370054244995, + "step": 437 + }, + { + "epoch": 0.42935914716333784, + "grad_norm": 0.4617707431316376, + "learning_rate": 9.140746393556853e-06, + "loss": 0.6354628801345825, + "step": 438 + }, + { + "epoch": 0.4303394191888249, + "grad_norm": 0.5162332653999329, + "learning_rate": 9.136294500014387e-06, + "loss": 0.6938362717628479, + "step": 439 + }, + { + "epoch": 0.43131969121431196, + "grad_norm": 0.4750451147556305, + "learning_rate": 9.13183219335544e-06, + "loss": 0.6032200455665588, + "step": 440 + }, + { + "epoch": 0.432299963239799, + "grad_norm": 0.5219208002090454, + "learning_rate": 9.12735948481387e-06, + "loss": 0.6199036240577698, + "step": 441 + }, + { + "epoch": 0.43328023526528614, + "grad_norm": 0.4706224203109741, + "learning_rate": 9.122876385649713e-06, + "loss": 0.5312067866325378, + "step": 442 + }, + { + "epoch": 0.4342605072907732, + "grad_norm": 0.4995059669017792, + "learning_rate": 9.118382907149164e-06, + "loss": 0.6517359018325806, + "step": 443 + }, + { + "epoch": 0.43524077931626026, + "grad_norm": 0.5074915885925293, + "learning_rate": 9.113879060624554e-06, + "loss": 0.5837841033935547, + "step": 444 + }, + { + "epoch": 0.4362210513417473, + "grad_norm": 0.5803712606430054, + "learning_rate": 9.109364857414306e-06, + "loss": 0.6096867322921753, + "step": 445 + }, + { + "epoch": 0.4372013233672344, + "grad_norm": 0.7945765852928162, + "learning_rate": 9.104840308882925e-06, + "loss": 0.6078523993492126, + "step": 446 + }, + { + "epoch": 0.4381815953927215, + "grad_norm": 0.5049729347229004, + "learning_rate": 9.100305426420957e-06, + "loss": 0.6547229290008545, + "step": 447 + }, + { + "epoch": 0.43916186741820856, + "grad_norm": 0.4533054530620575, + "learning_rate": 9.09576022144496e-06, + "loss": 0.5432737469673157, + "step": 448 + }, + { + "epoch": 0.4401421394436956, + "grad_norm": 0.4947768747806549, + "learning_rate": 9.091204705397485e-06, + "loss": 0.5290822982788086, + "step": 449 + }, + { + "epoch": 0.4411224114691827, + "grad_norm": 0.5200343728065491, + "learning_rate": 9.086638889747034e-06, + "loss": 0.6132403016090393, + "step": 450 + }, + { + "epoch": 0.44210268349466975, + "grad_norm": 0.4689146876335144, + "learning_rate": 9.08206278598805e-06, + "loss": 0.4153621196746826, + "step": 451 + }, + { + "epoch": 0.44308295552015686, + "grad_norm": 0.5330294370651245, + "learning_rate": 9.077476405640862e-06, + "loss": 0.6979874968528748, + "step": 452 + }, + { + "epoch": 0.4440632275456439, + "grad_norm": 0.4985146224498749, + "learning_rate": 9.07287976025168e-06, + "loss": 0.6385680437088013, + "step": 453 + }, + { + "epoch": 0.445043499571131, + "grad_norm": 0.5191720128059387, + "learning_rate": 9.06827286139255e-06, + "loss": 0.5751811265945435, + "step": 454 + }, + { + "epoch": 0.44602377159661805, + "grad_norm": 0.4784611463546753, + "learning_rate": 9.06365572066134e-06, + "loss": 0.6754035949707031, + "step": 455 + }, + { + "epoch": 0.4470040436221051, + "grad_norm": 0.4339852035045624, + "learning_rate": 9.059028349681693e-06, + "loss": 0.5843990445137024, + "step": 456 + }, + { + "epoch": 0.4479843156475922, + "grad_norm": 0.4005136787891388, + "learning_rate": 9.05439076010301e-06, + "loss": 0.5037546753883362, + "step": 457 + }, + { + "epoch": 0.4489645876730793, + "grad_norm": 0.6022442579269409, + "learning_rate": 9.04974296360042e-06, + "loss": 0.6200989484786987, + "step": 458 + }, + { + "epoch": 0.44994485969856635, + "grad_norm": 0.48745962977409363, + "learning_rate": 9.045084971874738e-06, + "loss": 0.6186924576759338, + "step": 459 + }, + { + "epoch": 0.4509251317240534, + "grad_norm": 0.48209333419799805, + "learning_rate": 9.040416796652459e-06, + "loss": 0.6512715816497803, + "step": 460 + }, + { + "epoch": 0.4519054037495405, + "grad_norm": 0.4796985983848572, + "learning_rate": 9.035738449685707e-06, + "loss": 0.5661883354187012, + "step": 461 + }, + { + "epoch": 0.4528856757750276, + "grad_norm": 0.5448599457740784, + "learning_rate": 9.031049942752215e-06, + "loss": 0.6986958980560303, + "step": 462 + }, + { + "epoch": 0.45386594780051465, + "grad_norm": 0.5704795122146606, + "learning_rate": 9.026351287655294e-06, + "loss": 0.5709372758865356, + "step": 463 + }, + { + "epoch": 0.4548462198260017, + "grad_norm": 0.4772814214229584, + "learning_rate": 9.021642496223802e-06, + "loss": 0.5940738916397095, + "step": 464 + }, + { + "epoch": 0.45582649185148877, + "grad_norm": 0.5291264653205872, + "learning_rate": 9.016923580312114e-06, + "loss": 0.5524354577064514, + "step": 465 + }, + { + "epoch": 0.4568067638769759, + "grad_norm": 0.5132133364677429, + "learning_rate": 9.012194551800098e-06, + "loss": 0.5567898750305176, + "step": 466 + }, + { + "epoch": 0.45778703590246295, + "grad_norm": 0.5544204711914062, + "learning_rate": 9.007455422593077e-06, + "loss": 0.6415997743606567, + "step": 467 + }, + { + "epoch": 0.45876730792795, + "grad_norm": 0.4312683641910553, + "learning_rate": 9.002706204621802e-06, + "loss": 0.5623404383659363, + "step": 468 + }, + { + "epoch": 0.45974757995343707, + "grad_norm": 0.5000928044319153, + "learning_rate": 8.997946909842426e-06, + "loss": 0.5414539575576782, + "step": 469 + }, + { + "epoch": 0.46072785197892413, + "grad_norm": 0.5731768012046814, + "learning_rate": 8.993177550236464e-06, + "loss": 0.5618506073951721, + "step": 470 + }, + { + "epoch": 0.46170812400441125, + "grad_norm": 0.5656113028526306, + "learning_rate": 8.988398137810778e-06, + "loss": 0.5426561832427979, + "step": 471 + }, + { + "epoch": 0.4626883960298983, + "grad_norm": 0.7768886089324951, + "learning_rate": 8.983608684597529e-06, + "loss": 0.6204694509506226, + "step": 472 + }, + { + "epoch": 0.46366866805538537, + "grad_norm": 0.5393010973930359, + "learning_rate": 8.978809202654161e-06, + "loss": 0.6652113199234009, + "step": 473 + }, + { + "epoch": 0.46464894008087243, + "grad_norm": 0.44731763005256653, + "learning_rate": 8.973999704063365e-06, + "loss": 0.595824658870697, + "step": 474 + }, + { + "epoch": 0.4656292121063595, + "grad_norm": 0.48581287264823914, + "learning_rate": 8.969180200933048e-06, + "loss": 0.7073965072631836, + "step": 475 + }, + { + "epoch": 0.4666094841318466, + "grad_norm": 0.447153240442276, + "learning_rate": 8.9643507053963e-06, + "loss": 0.6062104105949402, + "step": 476 + }, + { + "epoch": 0.46758975615733367, + "grad_norm": 0.7264053225517273, + "learning_rate": 8.959511229611377e-06, + "loss": 0.7488131523132324, + "step": 477 + }, + { + "epoch": 0.46857002818282073, + "grad_norm": 0.5424187183380127, + "learning_rate": 8.954661785761648e-06, + "loss": 0.6357473134994507, + "step": 478 + }, + { + "epoch": 0.4695503002083078, + "grad_norm": 0.5364806056022644, + "learning_rate": 8.949802386055582e-06, + "loss": 0.7206878066062927, + "step": 479 + }, + { + "epoch": 0.47053057223379485, + "grad_norm": 0.5717079043388367, + "learning_rate": 8.944933042726715e-06, + "loss": 0.6156141757965088, + "step": 480 + }, + { + "epoch": 0.47151084425928197, + "grad_norm": 0.5376148223876953, + "learning_rate": 8.94005376803361e-06, + "loss": 0.6603047847747803, + "step": 481 + }, + { + "epoch": 0.47249111628476903, + "grad_norm": 0.55712890625, + "learning_rate": 8.935164574259838e-06, + "loss": 0.6593056917190552, + "step": 482 + }, + { + "epoch": 0.4734713883102561, + "grad_norm": 0.5702346563339233, + "learning_rate": 8.930265473713939e-06, + "loss": 0.601768970489502, + "step": 483 + }, + { + "epoch": 0.47445166033574315, + "grad_norm": 0.50825035572052, + "learning_rate": 8.925356478729389e-06, + "loss": 0.5900289416313171, + "step": 484 + }, + { + "epoch": 0.4754319323612302, + "grad_norm": 0.5257771015167236, + "learning_rate": 8.92043760166458e-06, + "loss": 0.6179829835891724, + "step": 485 + }, + { + "epoch": 0.47641220438671733, + "grad_norm": 0.5047194361686707, + "learning_rate": 8.915508854902778e-06, + "loss": 0.5214921236038208, + "step": 486 + }, + { + "epoch": 0.4773924764122044, + "grad_norm": 0.40976911783218384, + "learning_rate": 8.910570250852098e-06, + "loss": 0.4534802734851837, + "step": 487 + }, + { + "epoch": 0.47837274843769145, + "grad_norm": 0.5223997831344604, + "learning_rate": 8.905621801945467e-06, + "loss": 0.5674428939819336, + "step": 488 + }, + { + "epoch": 0.4793530204631785, + "grad_norm": 0.5553629994392395, + "learning_rate": 8.900663520640605e-06, + "loss": 0.6257451772689819, + "step": 489 + }, + { + "epoch": 0.48033329248866563, + "grad_norm": 0.5198466777801514, + "learning_rate": 8.895695419419973e-06, + "loss": 0.8313612937927246, + "step": 490 + }, + { + "epoch": 0.4813135645141527, + "grad_norm": 0.4763185977935791, + "learning_rate": 8.890717510790763e-06, + "loss": 0.5762819051742554, + "step": 491 + }, + { + "epoch": 0.48229383653963975, + "grad_norm": 0.49477419257164, + "learning_rate": 8.885729807284855e-06, + "loss": 0.6347113847732544, + "step": 492 + }, + { + "epoch": 0.4832741085651268, + "grad_norm": 0.5553587079048157, + "learning_rate": 8.880732321458785e-06, + "loss": 0.5741130709648132, + "step": 493 + }, + { + "epoch": 0.4842543805906139, + "grad_norm": 0.4788428843021393, + "learning_rate": 8.875725065893717e-06, + "loss": 0.5938292741775513, + "step": 494 + }, + { + "epoch": 0.485234652616101, + "grad_norm": 0.5414963960647583, + "learning_rate": 8.870708053195414e-06, + "loss": 0.6218856573104858, + "step": 495 + }, + { + "epoch": 0.48621492464158805, + "grad_norm": 0.8729277849197388, + "learning_rate": 8.865681295994195e-06, + "loss": 0.6222516894340515, + "step": 496 + }, + { + "epoch": 0.4871951966670751, + "grad_norm": 0.5980170369148254, + "learning_rate": 8.860644806944917e-06, + "loss": 0.49276524782180786, + "step": 497 + }, + { + "epoch": 0.4881754686925622, + "grad_norm": 0.5419535040855408, + "learning_rate": 8.85559859872694e-06, + "loss": 0.6108378767967224, + "step": 498 + }, + { + "epoch": 0.48915574071804924, + "grad_norm": 0.5596175789833069, + "learning_rate": 8.850542684044078e-06, + "loss": 0.5110671520233154, + "step": 499 + }, + { + "epoch": 0.49013601274353635, + "grad_norm": 0.9812861084938049, + "learning_rate": 8.845477075624598e-06, + "loss": 0.5974424481391907, + "step": 500 + }, + { + "epoch": 0.4911162847690234, + "grad_norm": 0.49719715118408203, + "learning_rate": 8.84040178622116e-06, + "loss": 0.6635510921478271, + "step": 501 + }, + { + "epoch": 0.4920965567945105, + "grad_norm": 0.5727866291999817, + "learning_rate": 8.835316828610801e-06, + "loss": 0.6705462336540222, + "step": 502 + }, + { + "epoch": 0.49307682881999754, + "grad_norm": 0.7227946519851685, + "learning_rate": 8.83022221559489e-06, + "loss": 0.6447646021842957, + "step": 503 + }, + { + "epoch": 0.4940571008454846, + "grad_norm": 0.4858626425266266, + "learning_rate": 8.825117959999117e-06, + "loss": 0.5024073123931885, + "step": 504 + }, + { + "epoch": 0.4950373728709717, + "grad_norm": 0.6126762628555298, + "learning_rate": 8.820004074673433e-06, + "loss": 0.6250107288360596, + "step": 505 + }, + { + "epoch": 0.4960176448964588, + "grad_norm": 0.7225095629692078, + "learning_rate": 8.814880572492043e-06, + "loss": 0.5580613017082214, + "step": 506 + }, + { + "epoch": 0.49699791692194584, + "grad_norm": 0.8311983942985535, + "learning_rate": 8.809747466353356e-06, + "loss": 0.7410059571266174, + "step": 507 + }, + { + "epoch": 0.4979781889474329, + "grad_norm": 0.460429847240448, + "learning_rate": 8.804604769179958e-06, + "loss": 0.46225789189338684, + "step": 508 + }, + { + "epoch": 0.49895846097291996, + "grad_norm": 0.7332952618598938, + "learning_rate": 8.799452493918586e-06, + "loss": 0.7246252298355103, + "step": 509 + }, + { + "epoch": 0.4999387329984071, + "grad_norm": 0.6222991943359375, + "learning_rate": 8.794290653540086e-06, + "loss": 0.7012721300125122, + "step": 510 + }, + { + "epoch": 0.5009190050238941, + "grad_norm": 0.5395691990852356, + "learning_rate": 8.789119261039385e-06, + "loss": 0.47234785556793213, + "step": 511 + }, + { + "epoch": 0.5018992770493812, + "grad_norm": 0.5847768783569336, + "learning_rate": 8.783938329435458e-06, + "loss": 0.5104413032531738, + "step": 512 + }, + { + "epoch": 0.5028795490748683, + "grad_norm": 0.47077733278274536, + "learning_rate": 8.778747871771293e-06, + "loss": 0.5025359392166138, + "step": 513 + }, + { + "epoch": 0.5038598211003553, + "grad_norm": 0.5690563321113586, + "learning_rate": 8.773547901113862e-06, + "loss": 0.526625394821167, + "step": 514 + }, + { + "epoch": 0.5048400931258424, + "grad_norm": 0.5545769333839417, + "learning_rate": 8.768338430554083e-06, + "loss": 0.5559873580932617, + "step": 515 + }, + { + "epoch": 0.5058203651513294, + "grad_norm": 0.5283054113388062, + "learning_rate": 8.763119473206795e-06, + "loss": 0.6010313034057617, + "step": 516 + }, + { + "epoch": 0.5068006371768166, + "grad_norm": 0.5638696551322937, + "learning_rate": 8.757891042210713e-06, + "loss": 0.7054392099380493, + "step": 517 + }, + { + "epoch": 0.5077809092023037, + "grad_norm": 0.5060840845108032, + "learning_rate": 8.752653150728412e-06, + "loss": 0.5484815239906311, + "step": 518 + }, + { + "epoch": 0.5087611812277907, + "grad_norm": 0.6180709600448608, + "learning_rate": 8.747405811946272e-06, + "loss": 0.5184864401817322, + "step": 519 + }, + { + "epoch": 0.5097414532532778, + "grad_norm": 0.4981224834918976, + "learning_rate": 8.742149039074463e-06, + "loss": 0.4991441071033478, + "step": 520 + }, + { + "epoch": 0.5107217252787648, + "grad_norm": 0.6549201011657715, + "learning_rate": 8.736882845346906e-06, + "loss": 0.6914041638374329, + "step": 521 + }, + { + "epoch": 0.5117019973042519, + "grad_norm": 0.559851884841919, + "learning_rate": 8.731607244021236e-06, + "loss": 0.621010422706604, + "step": 522 + }, + { + "epoch": 0.512682269329739, + "grad_norm": 0.6031873226165771, + "learning_rate": 8.726322248378775e-06, + "loss": 0.6384186744689941, + "step": 523 + }, + { + "epoch": 0.513662541355226, + "grad_norm": 0.5719360113143921, + "learning_rate": 8.72102787172449e-06, + "loss": 0.5937843918800354, + "step": 524 + }, + { + "epoch": 0.5146428133807132, + "grad_norm": 0.5336046814918518, + "learning_rate": 8.715724127386971e-06, + "loss": 0.6553807258605957, + "step": 525 + }, + { + "epoch": 0.5156230854062002, + "grad_norm": 0.6381956934928894, + "learning_rate": 8.710411028718388e-06, + "loss": 0.5429552793502808, + "step": 526 + }, + { + "epoch": 0.5166033574316873, + "grad_norm": 0.459621787071228, + "learning_rate": 8.705088589094458e-06, + "loss": 0.48719850182533264, + "step": 527 + }, + { + "epoch": 0.5175836294571744, + "grad_norm": 0.5203189253807068, + "learning_rate": 8.69975682191442e-06, + "loss": 0.633024275302887, + "step": 528 + }, + { + "epoch": 0.5185639014826614, + "grad_norm": 0.5211911797523499, + "learning_rate": 8.69441574060099e-06, + "loss": 0.5052200555801392, + "step": 529 + }, + { + "epoch": 0.5195441735081485, + "grad_norm": 0.4485588073730469, + "learning_rate": 8.68906535860033e-06, + "loss": 0.5522943139076233, + "step": 530 + }, + { + "epoch": 0.5205244455336356, + "grad_norm": 0.5212096571922302, + "learning_rate": 8.683705689382025e-06, + "loss": 0.6452137231826782, + "step": 531 + }, + { + "epoch": 0.5215047175591226, + "grad_norm": 0.54129558801651, + "learning_rate": 8.678336746439036e-06, + "loss": 0.68224036693573, + "step": 532 + }, + { + "epoch": 0.5224849895846098, + "grad_norm": 0.5869094133377075, + "learning_rate": 8.672958543287666e-06, + "loss": 0.5855770111083984, + "step": 533 + }, + { + "epoch": 0.5234652616100968, + "grad_norm": 0.6560388207435608, + "learning_rate": 8.667571093467542e-06, + "loss": 0.5621611475944519, + "step": 534 + }, + { + "epoch": 0.5244455336355839, + "grad_norm": 0.7414459586143494, + "learning_rate": 8.662174410541556e-06, + "loss": 0.6210222244262695, + "step": 535 + }, + { + "epoch": 0.525425805661071, + "grad_norm": 0.5764961242675781, + "learning_rate": 8.656768508095853e-06, + "loss": 0.6334097385406494, + "step": 536 + }, + { + "epoch": 0.526406077686558, + "grad_norm": 0.6460157036781311, + "learning_rate": 8.651353399739787e-06, + "loss": 0.6059669256210327, + "step": 537 + }, + { + "epoch": 0.5273863497120451, + "grad_norm": 0.7143170237541199, + "learning_rate": 8.645929099105886e-06, + "loss": 0.5628725290298462, + "step": 538 + }, + { + "epoch": 0.5283666217375321, + "grad_norm": 0.7480341792106628, + "learning_rate": 8.640495619849821e-06, + "loss": 0.6259042620658875, + "step": 539 + }, + { + "epoch": 0.5293468937630192, + "grad_norm": 0.4735183119773865, + "learning_rate": 8.63505297565037e-06, + "loss": 0.5363203287124634, + "step": 540 + }, + { + "epoch": 0.5303271657885064, + "grad_norm": 0.6985523104667664, + "learning_rate": 8.629601180209382e-06, + "loss": 0.6217838525772095, + "step": 541 + }, + { + "epoch": 0.5313074378139934, + "grad_norm": 0.43614763021469116, + "learning_rate": 8.624140247251745e-06, + "loss": 0.538539469242096, + "step": 542 + }, + { + "epoch": 0.5322877098394805, + "grad_norm": 0.45327121019363403, + "learning_rate": 8.61867019052535e-06, + "loss": 0.5169327259063721, + "step": 543 + }, + { + "epoch": 0.5332679818649675, + "grad_norm": 0.5319138169288635, + "learning_rate": 8.613191023801065e-06, + "loss": 0.5326805710792542, + "step": 544 + }, + { + "epoch": 0.5342482538904546, + "grad_norm": 0.5725491642951965, + "learning_rate": 8.607702760872679e-06, + "loss": 0.5866974592208862, + "step": 545 + }, + { + "epoch": 0.5352285259159417, + "grad_norm": 0.4773092269897461, + "learning_rate": 8.602205415556889e-06, + "loss": 0.5357339382171631, + "step": 546 + }, + { + "epoch": 0.5362087979414287, + "grad_norm": 0.5068699717521667, + "learning_rate": 8.596699001693257e-06, + "loss": 0.5180469751358032, + "step": 547 + }, + { + "epoch": 0.5371890699669158, + "grad_norm": 0.4932785928249359, + "learning_rate": 8.591183533144172e-06, + "loss": 0.5822703838348389, + "step": 548 + }, + { + "epoch": 0.5381693419924028, + "grad_norm": 0.41486823558807373, + "learning_rate": 8.585659023794818e-06, + "loss": 0.5033004283905029, + "step": 549 + }, + { + "epoch": 0.53914961401789, + "grad_norm": 0.48342612385749817, + "learning_rate": 8.580125487553143e-06, + "loss": 0.5849289894104004, + "step": 550 + }, + { + "epoch": 0.5401298860433771, + "grad_norm": 0.41986194252967834, + "learning_rate": 8.574582938349818e-06, + "loss": 0.5392067432403564, + "step": 551 + }, + { + "epoch": 0.5411101580688641, + "grad_norm": 0.5170447826385498, + "learning_rate": 8.569031390138202e-06, + "loss": 0.6853705644607544, + "step": 552 + }, + { + "epoch": 0.5420904300943512, + "grad_norm": 0.560274064540863, + "learning_rate": 8.563470856894316e-06, + "loss": 0.5396522283554077, + "step": 553 + }, + { + "epoch": 0.5430707021198382, + "grad_norm": 0.4939210116863251, + "learning_rate": 8.557901352616788e-06, + "loss": 0.6326029896736145, + "step": 554 + }, + { + "epoch": 0.5440509741453253, + "grad_norm": 0.557123064994812, + "learning_rate": 8.552322891326846e-06, + "loss": 0.5660082101821899, + "step": 555 + }, + { + "epoch": 0.5450312461708124, + "grad_norm": 0.49068284034729004, + "learning_rate": 8.546735487068253e-06, + "loss": 0.6054900884628296, + "step": 556 + }, + { + "epoch": 0.5460115181962995, + "grad_norm": 0.5976768732070923, + "learning_rate": 8.541139153907296e-06, + "loss": 0.558587372303009, + "step": 557 + }, + { + "epoch": 0.5469917902217866, + "grad_norm": 0.5144990086555481, + "learning_rate": 8.535533905932739e-06, + "loss": 0.49542275071144104, + "step": 558 + }, + { + "epoch": 0.5479720622472736, + "grad_norm": 0.4800182282924652, + "learning_rate": 8.529919757255783e-06, + "loss": 0.5426228046417236, + "step": 559 + }, + { + "epoch": 0.5489523342727607, + "grad_norm": 0.47078582644462585, + "learning_rate": 8.524296722010042e-06, + "loss": 0.5946049690246582, + "step": 560 + }, + { + "epoch": 0.5499326062982478, + "grad_norm": 0.5674675703048706, + "learning_rate": 8.518664814351502e-06, + "loss": 0.5650599002838135, + "step": 561 + }, + { + "epoch": 0.5509128783237348, + "grad_norm": 0.44857510924339294, + "learning_rate": 8.513024048458484e-06, + "loss": 0.5294644832611084, + "step": 562 + }, + { + "epoch": 0.5518931503492219, + "grad_norm": 1.0770368576049805, + "learning_rate": 8.507374438531606e-06, + "loss": 0.6094412803649902, + "step": 563 + }, + { + "epoch": 0.5528734223747089, + "grad_norm": 0.5167132019996643, + "learning_rate": 8.501715998793758e-06, + "loss": 0.4995839297771454, + "step": 564 + }, + { + "epoch": 0.553853694400196, + "grad_norm": 0.5282765030860901, + "learning_rate": 8.496048743490053e-06, + "loss": 0.581527590751648, + "step": 565 + }, + { + "epoch": 0.5548339664256832, + "grad_norm": 0.5315512418746948, + "learning_rate": 8.490372686887802e-06, + "loss": 0.58648282289505, + "step": 566 + }, + { + "epoch": 0.5558142384511702, + "grad_norm": 0.6867757439613342, + "learning_rate": 8.48468784327647e-06, + "loss": 0.6812595129013062, + "step": 567 + }, + { + "epoch": 0.5567945104766573, + "grad_norm": 0.5088253617286682, + "learning_rate": 8.478994226967638e-06, + "loss": 0.6054092645645142, + "step": 568 + }, + { + "epoch": 0.5577747825021443, + "grad_norm": 0.5003196597099304, + "learning_rate": 8.473291852294986e-06, + "loss": 0.6233060359954834, + "step": 569 + }, + { + "epoch": 0.5587550545276314, + "grad_norm": 0.553752064704895, + "learning_rate": 8.467580733614233e-06, + "loss": 0.640609085559845, + "step": 570 + }, + { + "epoch": 0.5597353265531185, + "grad_norm": 0.5246297717094421, + "learning_rate": 8.461860885303116e-06, + "loss": 0.5541980266571045, + "step": 571 + }, + { + "epoch": 0.5607155985786055, + "grad_norm": 0.5524768233299255, + "learning_rate": 8.45613232176134e-06, + "loss": 0.5221663117408752, + "step": 572 + }, + { + "epoch": 0.5616958706040927, + "grad_norm": 0.5871309041976929, + "learning_rate": 8.450395057410561e-06, + "loss": 0.600644588470459, + "step": 573 + }, + { + "epoch": 0.5626761426295797, + "grad_norm": 0.5105273127555847, + "learning_rate": 8.444649106694335e-06, + "loss": 0.6499615907669067, + "step": 574 + }, + { + "epoch": 0.5636564146550668, + "grad_norm": 0.46583378314971924, + "learning_rate": 8.438894484078086e-06, + "loss": 0.48685839772224426, + "step": 575 + }, + { + "epoch": 0.5646366866805539, + "grad_norm": 0.6281800270080566, + "learning_rate": 8.433131204049067e-06, + "loss": 0.6305126547813416, + "step": 576 + }, + { + "epoch": 0.5656169587060409, + "grad_norm": 0.4780385196208954, + "learning_rate": 8.427359281116335e-06, + "loss": 0.5220742225646973, + "step": 577 + }, + { + "epoch": 0.566597230731528, + "grad_norm": 0.5597613453865051, + "learning_rate": 8.421578729810693e-06, + "loss": 0.6230953931808472, + "step": 578 + }, + { + "epoch": 0.567577502757015, + "grad_norm": 0.5474322438240051, + "learning_rate": 8.415789564684673e-06, + "loss": 0.7254103422164917, + "step": 579 + }, + { + "epoch": 0.5685577747825021, + "grad_norm": 0.48552948236465454, + "learning_rate": 8.409991800312493e-06, + "loss": 0.5996748805046082, + "step": 580 + }, + { + "epoch": 0.5695380468079893, + "grad_norm": 0.5050139427185059, + "learning_rate": 8.404185451290017e-06, + "loss": 0.6497757434844971, + "step": 581 + }, + { + "epoch": 0.5705183188334763, + "grad_norm": 0.5027022957801819, + "learning_rate": 8.398370532234723e-06, + "loss": 0.7140090465545654, + "step": 582 + }, + { + "epoch": 0.5714985908589634, + "grad_norm": 0.39737898111343384, + "learning_rate": 8.392547057785662e-06, + "loss": 0.42447859048843384, + "step": 583 + }, + { + "epoch": 0.5724788628844505, + "grad_norm": 0.5955871939659119, + "learning_rate": 8.386715042603424e-06, + "loss": 0.5741234421730042, + "step": 584 + }, + { + "epoch": 0.5734591349099375, + "grad_norm": 0.49950310587882996, + "learning_rate": 8.380874501370098e-06, + "loss": 0.5596312284469604, + "step": 585 + }, + { + "epoch": 0.5744394069354246, + "grad_norm": 0.44867679476737976, + "learning_rate": 8.375025448789242e-06, + "loss": 0.49942177534103394, + "step": 586 + }, + { + "epoch": 0.5754196789609116, + "grad_norm": 0.4992743730545044, + "learning_rate": 8.36916789958584e-06, + "loss": 0.5540944337844849, + "step": 587 + }, + { + "epoch": 0.5763999509863987, + "grad_norm": 1.0275784730911255, + "learning_rate": 8.363301868506264e-06, + "loss": 0.6042492985725403, + "step": 588 + }, + { + "epoch": 0.5773802230118859, + "grad_norm": 0.5012497901916504, + "learning_rate": 8.357427370318239e-06, + "loss": 0.545590877532959, + "step": 589 + }, + { + "epoch": 0.5783604950373729, + "grad_norm": 0.44835513830184937, + "learning_rate": 8.351544419810808e-06, + "loss": 0.5180589556694031, + "step": 590 + }, + { + "epoch": 0.57934076706286, + "grad_norm": 0.6875878572463989, + "learning_rate": 8.345653031794292e-06, + "loss": 0.5486396551132202, + "step": 591 + }, + { + "epoch": 0.580321039088347, + "grad_norm": 0.5637878179550171, + "learning_rate": 8.339753221100254e-06, + "loss": 0.5895923376083374, + "step": 592 + }, + { + "epoch": 0.5813013111138341, + "grad_norm": 0.5228107571601868, + "learning_rate": 8.33384500258146e-06, + "loss": 0.7664077877998352, + "step": 593 + }, + { + "epoch": 0.5822815831393212, + "grad_norm": 0.44823041558265686, + "learning_rate": 8.327928391111841e-06, + "loss": 0.5698608160018921, + "step": 594 + }, + { + "epoch": 0.5832618551648082, + "grad_norm": 0.7237053513526917, + "learning_rate": 8.322003401586463e-06, + "loss": 0.6401317119598389, + "step": 595 + }, + { + "epoch": 0.5842421271902953, + "grad_norm": 0.49424564838409424, + "learning_rate": 8.316070048921477e-06, + "loss": 0.5031653642654419, + "step": 596 + }, + { + "epoch": 0.5852223992157823, + "grad_norm": 0.5683320164680481, + "learning_rate": 8.310128348054093e-06, + "loss": 0.4508371353149414, + "step": 597 + }, + { + "epoch": 0.5862026712412695, + "grad_norm": 0.5639766454696655, + "learning_rate": 8.304178313942536e-06, + "loss": 0.5411083698272705, + "step": 598 + }, + { + "epoch": 0.5871829432667566, + "grad_norm": 0.5008068084716797, + "learning_rate": 8.298219961566008e-06, + "loss": 0.5488749146461487, + "step": 599 + }, + { + "epoch": 0.5881632152922436, + "grad_norm": 0.6290292739868164, + "learning_rate": 8.292253305924656e-06, + "loss": 0.5881930589675903, + "step": 600 + }, + { + "epoch": 0.5891434873177307, + "grad_norm": 0.47255659103393555, + "learning_rate": 8.286278362039527e-06, + "loss": 0.6302595138549805, + "step": 601 + }, + { + "epoch": 0.5901237593432177, + "grad_norm": 0.5140688419342041, + "learning_rate": 8.280295144952537e-06, + "loss": 0.6385571956634521, + "step": 602 + }, + { + "epoch": 0.5911040313687048, + "grad_norm": 0.46549689769744873, + "learning_rate": 8.274303669726427e-06, + "loss": 0.5014755129814148, + "step": 603 + }, + { + "epoch": 0.5920843033941919, + "grad_norm": 0.41634997725486755, + "learning_rate": 8.268303951444727e-06, + "loss": 0.495285302400589, + "step": 604 + }, + { + "epoch": 0.5930645754196789, + "grad_norm": 1.8150643110275269, + "learning_rate": 8.262296005211722e-06, + "loss": 0.47511863708496094, + "step": 605 + }, + { + "epoch": 0.5940448474451661, + "grad_norm": 0.5939587950706482, + "learning_rate": 8.25627984615241e-06, + "loss": 0.676226019859314, + "step": 606 + }, + { + "epoch": 0.5950251194706531, + "grad_norm": 0.8281105756759644, + "learning_rate": 8.250255489412464e-06, + "loss": 0.5587377548217773, + "step": 607 + }, + { + "epoch": 0.5960053914961402, + "grad_norm": 0.4769623577594757, + "learning_rate": 8.244222950158194e-06, + "loss": 0.6315554976463318, + "step": 608 + }, + { + "epoch": 0.5969856635216273, + "grad_norm": 0.4458005726337433, + "learning_rate": 8.238182243576512e-06, + "loss": 0.4711849093437195, + "step": 609 + }, + { + "epoch": 0.5979659355471143, + "grad_norm": 0.47836652398109436, + "learning_rate": 8.232133384874887e-06, + "loss": 0.573494553565979, + "step": 610 + }, + { + "epoch": 0.5989462075726014, + "grad_norm": 0.44612884521484375, + "learning_rate": 8.226076389281316e-06, + "loss": 0.6105935573577881, + "step": 611 + }, + { + "epoch": 0.5999264795980884, + "grad_norm": 0.4678818881511688, + "learning_rate": 8.220011272044276e-06, + "loss": 0.5955692529678345, + "step": 612 + }, + { + "epoch": 0.6009067516235755, + "grad_norm": 0.47035977244377136, + "learning_rate": 8.213938048432697e-06, + "loss": 0.5746393799781799, + "step": 613 + }, + { + "epoch": 0.6018870236490627, + "grad_norm": 0.43257585167884827, + "learning_rate": 8.207856733735908e-06, + "loss": 0.5300241112709045, + "step": 614 + }, + { + "epoch": 0.6028672956745497, + "grad_norm": 0.42712169885635376, + "learning_rate": 8.201767343263612e-06, + "loss": 0.5474828481674194, + "step": 615 + }, + { + "epoch": 0.6038475677000368, + "grad_norm": 0.4949519634246826, + "learning_rate": 8.195669892345842e-06, + "loss": 0.539609432220459, + "step": 616 + }, + { + "epoch": 0.6048278397255238, + "grad_norm": 0.466345876455307, + "learning_rate": 8.189564396332927e-06, + "loss": 0.5650939345359802, + "step": 617 + }, + { + "epoch": 0.6058081117510109, + "grad_norm": 0.4587682783603668, + "learning_rate": 8.183450870595443e-06, + "loss": 0.4892288148403168, + "step": 618 + }, + { + "epoch": 0.606788383776498, + "grad_norm": 0.4350122809410095, + "learning_rate": 8.177329330524182e-06, + "loss": 0.5417685508728027, + "step": 619 + }, + { + "epoch": 0.607768655801985, + "grad_norm": 0.44926467537879944, + "learning_rate": 8.171199791530117e-06, + "loss": 0.5792407989501953, + "step": 620 + }, + { + "epoch": 0.6087489278274721, + "grad_norm": 0.48948216438293457, + "learning_rate": 8.165062269044353e-06, + "loss": 0.5384902954101562, + "step": 621 + }, + { + "epoch": 0.6097291998529591, + "grad_norm": 0.4751755893230438, + "learning_rate": 8.158916778518097e-06, + "loss": 0.571563184261322, + "step": 622 + }, + { + "epoch": 0.6107094718784463, + "grad_norm": 0.5100057125091553, + "learning_rate": 8.152763335422612e-06, + "loss": 0.573106050491333, + "step": 623 + }, + { + "epoch": 0.6116897439039334, + "grad_norm": 0.5238609313964844, + "learning_rate": 8.146601955249187e-06, + "loss": 0.5689284801483154, + "step": 624 + }, + { + "epoch": 0.6126700159294204, + "grad_norm": 0.5913124680519104, + "learning_rate": 8.140432653509089e-06, + "loss": 0.5576924085617065, + "step": 625 + }, + { + "epoch": 0.6136502879549075, + "grad_norm": 0.5904691815376282, + "learning_rate": 8.134255445733524e-06, + "loss": 0.6356901526451111, + "step": 626 + }, + { + "epoch": 0.6146305599803945, + "grad_norm": 0.43635961413383484, + "learning_rate": 8.128070347473609e-06, + "loss": 0.5801491737365723, + "step": 627 + }, + { + "epoch": 0.6156108320058816, + "grad_norm": 0.541221022605896, + "learning_rate": 8.121877374300318e-06, + "loss": 0.5040395259857178, + "step": 628 + }, + { + "epoch": 0.6165911040313687, + "grad_norm": 0.4131251573562622, + "learning_rate": 8.115676541804456e-06, + "loss": 0.5251767635345459, + "step": 629 + }, + { + "epoch": 0.6175713760568557, + "grad_norm": 0.45905354619026184, + "learning_rate": 8.109467865596612e-06, + "loss": 0.5374346971511841, + "step": 630 + }, + { + "epoch": 0.6185516480823429, + "grad_norm": 0.4692007005214691, + "learning_rate": 8.10325136130712e-06, + "loss": 0.5571949481964111, + "step": 631 + }, + { + "epoch": 0.6195319201078299, + "grad_norm": 0.5968528985977173, + "learning_rate": 8.097027044586021e-06, + "loss": 0.7638495564460754, + "step": 632 + }, + { + "epoch": 0.620512192133317, + "grad_norm": 0.5194112062454224, + "learning_rate": 8.090794931103026e-06, + "loss": 0.5503364205360413, + "step": 633 + }, + { + "epoch": 0.6214924641588041, + "grad_norm": 0.536645770072937, + "learning_rate": 8.084555036547475e-06, + "loss": 0.6009919047355652, + "step": 634 + }, + { + "epoch": 0.6224727361842911, + "grad_norm": 0.43642154335975647, + "learning_rate": 8.078307376628292e-06, + "loss": 0.4636218547821045, + "step": 635 + }, + { + "epoch": 0.6234530082097782, + "grad_norm": 0.4677816331386566, + "learning_rate": 8.072051967073955e-06, + "loss": 0.5536704063415527, + "step": 636 + }, + { + "epoch": 0.6244332802352652, + "grad_norm": 0.41060787439346313, + "learning_rate": 8.065788823632451e-06, + "loss": 0.41056686639785767, + "step": 637 + }, + { + "epoch": 0.6254135522607523, + "grad_norm": 0.4008098840713501, + "learning_rate": 8.059517962071234e-06, + "loss": 0.4722449779510498, + "step": 638 + }, + { + "epoch": 0.6263938242862395, + "grad_norm": 0.4182589054107666, + "learning_rate": 8.053239398177191e-06, + "loss": 0.5422540307044983, + "step": 639 + }, + { + "epoch": 0.6273740963117265, + "grad_norm": 0.5321866869926453, + "learning_rate": 8.046953147756601e-06, + "loss": 0.5848727226257324, + "step": 640 + }, + { + "epoch": 0.6283543683372136, + "grad_norm": 0.3844715356826782, + "learning_rate": 8.04065922663509e-06, + "loss": 0.4720197319984436, + "step": 641 + }, + { + "epoch": 0.6293346403627007, + "grad_norm": 0.5278245806694031, + "learning_rate": 8.034357650657597e-06, + "loss": 0.779409646987915, + "step": 642 + }, + { + "epoch": 0.6303149123881877, + "grad_norm": 0.4623364806175232, + "learning_rate": 8.028048435688333e-06, + "loss": 0.5392259955406189, + "step": 643 + }, + { + "epoch": 0.6312951844136748, + "grad_norm": 0.5639168620109558, + "learning_rate": 8.02173159761074e-06, + "loss": 0.5717742443084717, + "step": 644 + }, + { + "epoch": 0.6322754564391618, + "grad_norm": 0.46666693687438965, + "learning_rate": 8.015407152327448e-06, + "loss": 0.6617095470428467, + "step": 645 + }, + { + "epoch": 0.633255728464649, + "grad_norm": 0.5085638761520386, + "learning_rate": 8.009075115760243e-06, + "loss": 0.6582084894180298, + "step": 646 + }, + { + "epoch": 0.6342360004901361, + "grad_norm": 0.43695488572120667, + "learning_rate": 8.002735503850016e-06, + "loss": 0.5222762823104858, + "step": 647 + }, + { + "epoch": 0.6352162725156231, + "grad_norm": 0.46036913990974426, + "learning_rate": 7.996388332556735e-06, + "loss": 0.6211761832237244, + "step": 648 + }, + { + "epoch": 0.6361965445411102, + "grad_norm": 0.467732697725296, + "learning_rate": 7.990033617859396e-06, + "loss": 0.4865642488002777, + "step": 649 + }, + { + "epoch": 0.6371768165665972, + "grad_norm": 0.5058535933494568, + "learning_rate": 7.983671375755986e-06, + "loss": 0.5791799426078796, + "step": 650 + }, + { + "epoch": 0.6381570885920843, + "grad_norm": 0.48842498660087585, + "learning_rate": 7.97730162226344e-06, + "loss": 0.6354533433914185, + "step": 651 + }, + { + "epoch": 0.6391373606175714, + "grad_norm": 0.45112404227256775, + "learning_rate": 7.970924373417608e-06, + "loss": 0.6128513813018799, + "step": 652 + }, + { + "epoch": 0.6401176326430584, + "grad_norm": 0.4545324444770813, + "learning_rate": 7.964539645273204e-06, + "loss": 0.5118072032928467, + "step": 653 + }, + { + "epoch": 0.6410979046685455, + "grad_norm": 0.3487434387207031, + "learning_rate": 7.958147453903773e-06, + "loss": 0.4512544870376587, + "step": 654 + }, + { + "epoch": 0.6420781766940326, + "grad_norm": 0.42427530884742737, + "learning_rate": 7.951747815401651e-06, + "loss": 0.5383853912353516, + "step": 655 + }, + { + "epoch": 0.6430584487195197, + "grad_norm": 0.6650370359420776, + "learning_rate": 7.945340745877919e-06, + "loss": 0.559260368347168, + "step": 656 + }, + { + "epoch": 0.6440387207450068, + "grad_norm": 0.49584558606147766, + "learning_rate": 7.938926261462366e-06, + "loss": 0.6616352200508118, + "step": 657 + }, + { + "epoch": 0.6450189927704938, + "grad_norm": 0.4980330765247345, + "learning_rate": 7.932504378303452e-06, + "loss": 0.700371265411377, + "step": 658 + }, + { + "epoch": 0.6459992647959809, + "grad_norm": 0.6228008270263672, + "learning_rate": 7.92607511256826e-06, + "loss": 0.48944735527038574, + "step": 659 + }, + { + "epoch": 0.6469795368214679, + "grad_norm": 0.4910522997379303, + "learning_rate": 7.919638480442453e-06, + "loss": 0.5541126728057861, + "step": 660 + }, + { + "epoch": 0.647959808846955, + "grad_norm": 0.4713754653930664, + "learning_rate": 7.913194498130252e-06, + "loss": 0.4977142810821533, + "step": 661 + }, + { + "epoch": 0.6489400808724421, + "grad_norm": 0.4953506886959076, + "learning_rate": 7.90674318185437e-06, + "loss": 0.6201741099357605, + "step": 662 + }, + { + "epoch": 0.6499203528979292, + "grad_norm": 0.5087093114852905, + "learning_rate": 7.900284547855992e-06, + "loss": 0.6099828481674194, + "step": 663 + }, + { + "epoch": 0.6509006249234163, + "grad_norm": 0.5764956474304199, + "learning_rate": 7.893818612394717e-06, + "loss": 0.5428135395050049, + "step": 664 + }, + { + "epoch": 0.6518808969489033, + "grad_norm": 0.419497549533844, + "learning_rate": 7.887345391748533e-06, + "loss": 0.5326085686683655, + "step": 665 + }, + { + "epoch": 0.6528611689743904, + "grad_norm": 0.7227897047996521, + "learning_rate": 7.880864902213765e-06, + "loss": 0.6780993342399597, + "step": 666 + }, + { + "epoch": 0.6538414409998775, + "grad_norm": 0.5869578719139099, + "learning_rate": 7.874377160105037e-06, + "loss": 0.5142019987106323, + "step": 667 + }, + { + "epoch": 0.6548217130253645, + "grad_norm": 0.49598124623298645, + "learning_rate": 7.86788218175523e-06, + "loss": 0.6338085532188416, + "step": 668 + }, + { + "epoch": 0.6558019850508516, + "grad_norm": 0.44636520743370056, + "learning_rate": 7.861379983515449e-06, + "loss": 0.6195379495620728, + "step": 669 + }, + { + "epoch": 0.6567822570763386, + "grad_norm": 0.4613161087036133, + "learning_rate": 7.854870581754967e-06, + "loss": 0.4916152358055115, + "step": 670 + }, + { + "epoch": 0.6577625291018258, + "grad_norm": 0.4831985831260681, + "learning_rate": 7.848353992861195e-06, + "loss": 0.5819408297538757, + "step": 671 + }, + { + "epoch": 0.6587428011273129, + "grad_norm": 0.3623875677585602, + "learning_rate": 7.841830233239638e-06, + "loss": 0.5120476484298706, + "step": 672 + }, + { + "epoch": 0.6597230731527999, + "grad_norm": 0.40355730056762695, + "learning_rate": 7.835299319313854e-06, + "loss": 0.5201019048690796, + "step": 673 + }, + { + "epoch": 0.660703345178287, + "grad_norm": 0.36736786365509033, + "learning_rate": 7.82876126752541e-06, + "loss": 0.508677065372467, + "step": 674 + }, + { + "epoch": 0.661683617203774, + "grad_norm": 0.3610830307006836, + "learning_rate": 7.822216094333847e-06, + "loss": 0.45365607738494873, + "step": 675 + }, + { + "epoch": 0.6626638892292611, + "grad_norm": 0.4933203160762787, + "learning_rate": 7.815663816216626e-06, + "loss": 0.6202186346054077, + "step": 676 + }, + { + "epoch": 0.6636441612547482, + "grad_norm": 0.47557780146598816, + "learning_rate": 7.8091044496691e-06, + "loss": 0.5591720938682556, + "step": 677 + }, + { + "epoch": 0.6646244332802352, + "grad_norm": 0.4458239674568176, + "learning_rate": 7.80253801120447e-06, + "loss": 0.5407501459121704, + "step": 678 + }, + { + "epoch": 0.6656047053057224, + "grad_norm": 0.3904792070388794, + "learning_rate": 7.795964517353734e-06, + "loss": 0.48473650217056274, + "step": 679 + }, + { + "epoch": 0.6665849773312094, + "grad_norm": 0.4341505169868469, + "learning_rate": 7.789383984665658e-06, + "loss": 0.5052921772003174, + "step": 680 + }, + { + "epoch": 0.6675652493566965, + "grad_norm": 0.40881484746932983, + "learning_rate": 7.782796429706721e-06, + "loss": 0.497772753238678, + "step": 681 + }, + { + "epoch": 0.6685455213821836, + "grad_norm": 0.45834603905677795, + "learning_rate": 7.776201869061086e-06, + "loss": 0.5965415835380554, + "step": 682 + }, + { + "epoch": 0.6695257934076706, + "grad_norm": 0.4278583824634552, + "learning_rate": 7.769600319330553e-06, + "loss": 0.4990220069885254, + "step": 683 + }, + { + "epoch": 0.6705060654331577, + "grad_norm": 0.4458213150501251, + "learning_rate": 7.762991797134513e-06, + "loss": 0.5628777146339417, + "step": 684 + }, + { + "epoch": 0.6714863374586447, + "grad_norm": 0.38521257042884827, + "learning_rate": 7.756376319109917e-06, + "loss": 0.49859195947647095, + "step": 685 + }, + { + "epoch": 0.6724666094841318, + "grad_norm": 0.45445364713668823, + "learning_rate": 7.74975390191122e-06, + "loss": 0.5584304928779602, + "step": 686 + }, + { + "epoch": 0.673446881509619, + "grad_norm": 0.47879061102867126, + "learning_rate": 7.743124562210351e-06, + "loss": 0.6553927659988403, + "step": 687 + }, + { + "epoch": 0.674427153535106, + "grad_norm": 0.41177281737327576, + "learning_rate": 7.736488316696663e-06, + "loss": 0.42918023467063904, + "step": 688 + }, + { + "epoch": 0.6754074255605931, + "grad_norm": 0.569825291633606, + "learning_rate": 7.729845182076896e-06, + "loss": 0.6259717345237732, + "step": 689 + }, + { + "epoch": 0.6763876975860801, + "grad_norm": 0.5575708746910095, + "learning_rate": 7.723195175075136e-06, + "loss": 0.5303859710693359, + "step": 690 + }, + { + "epoch": 0.6773679696115672, + "grad_norm": 0.5137466788291931, + "learning_rate": 7.716538312432767e-06, + "loss": 0.5614829063415527, + "step": 691 + }, + { + "epoch": 0.6783482416370543, + "grad_norm": 0.440349817276001, + "learning_rate": 7.70987461090843e-06, + "loss": 0.5800457000732422, + "step": 692 + }, + { + "epoch": 0.6793285136625413, + "grad_norm": 0.4403285086154938, + "learning_rate": 7.703204087277989e-06, + "loss": 0.5892413854598999, + "step": 693 + }, + { + "epoch": 0.6803087856880284, + "grad_norm": 0.4497002065181732, + "learning_rate": 7.696526758334477e-06, + "loss": 0.562138557434082, + "step": 694 + }, + { + "epoch": 0.6812890577135156, + "grad_norm": 0.4190099239349365, + "learning_rate": 7.689842640888063e-06, + "loss": 0.4345162510871887, + "step": 695 + }, + { + "epoch": 0.6822693297390026, + "grad_norm": 0.5989111661911011, + "learning_rate": 7.683151751766005e-06, + "loss": 0.512845516204834, + "step": 696 + }, + { + "epoch": 0.6832496017644897, + "grad_norm": 0.4910346269607544, + "learning_rate": 7.676454107812608e-06, + "loss": 0.733169674873352, + "step": 697 + }, + { + "epoch": 0.6842298737899767, + "grad_norm": 0.42751452326774597, + "learning_rate": 7.669749725889182e-06, + "loss": 0.5693173408508301, + "step": 698 + }, + { + "epoch": 0.6852101458154638, + "grad_norm": 0.5142582654953003, + "learning_rate": 7.663038622873999e-06, + "loss": 0.5878850221633911, + "step": 699 + }, + { + "epoch": 0.6861904178409509, + "grad_norm": 0.8307844400405884, + "learning_rate": 7.656320815662257e-06, + "loss": 0.6216428279876709, + "step": 700 + }, + { + "epoch": 0.6871706898664379, + "grad_norm": 0.47172537446022034, + "learning_rate": 7.649596321166024e-06, + "loss": 0.549314022064209, + "step": 701 + }, + { + "epoch": 0.688150961891925, + "grad_norm": 0.47968435287475586, + "learning_rate": 7.64286515631421e-06, + "loss": 0.48883455991744995, + "step": 702 + }, + { + "epoch": 0.689131233917412, + "grad_norm": 0.6333498358726501, + "learning_rate": 7.636127338052513e-06, + "loss": 0.6025599241256714, + "step": 703 + }, + { + "epoch": 0.6901115059428992, + "grad_norm": 0.640839159488678, + "learning_rate": 7.629382883343381e-06, + "loss": 0.5830345153808594, + "step": 704 + }, + { + "epoch": 0.6910917779683863, + "grad_norm": 0.39502251148223877, + "learning_rate": 7.622631809165972e-06, + "loss": 0.406398743391037, + "step": 705 + }, + { + "epoch": 0.6920720499938733, + "grad_norm": 0.3786174952983856, + "learning_rate": 7.615874132516108e-06, + "loss": 0.39167964458465576, + "step": 706 + }, + { + "epoch": 0.6930523220193604, + "grad_norm": 0.5710326433181763, + "learning_rate": 7.60910987040623e-06, + "loss": 0.5021277666091919, + "step": 707 + }, + { + "epoch": 0.6940325940448474, + "grad_norm": 0.45050370693206787, + "learning_rate": 7.602339039865362e-06, + "loss": 0.5112322568893433, + "step": 708 + }, + { + "epoch": 0.6950128660703345, + "grad_norm": 0.49357518553733826, + "learning_rate": 7.595561657939061e-06, + "loss": 0.5479626655578613, + "step": 709 + }, + { + "epoch": 0.6959931380958216, + "grad_norm": 0.6366994380950928, + "learning_rate": 7.588777741689376e-06, + "loss": 0.6149463057518005, + "step": 710 + }, + { + "epoch": 0.6969734101213086, + "grad_norm": 0.4910570979118347, + "learning_rate": 7.5819873081948105e-06, + "loss": 0.5862258076667786, + "step": 711 + }, + { + "epoch": 0.6979536821467958, + "grad_norm": 0.41112977266311646, + "learning_rate": 7.575190374550272e-06, + "loss": 0.49828004837036133, + "step": 712 + }, + { + "epoch": 0.6989339541722828, + "grad_norm": 0.5109178423881531, + "learning_rate": 7.568386957867033e-06, + "loss": 0.5555713176727295, + "step": 713 + }, + { + "epoch": 0.6999142261977699, + "grad_norm": 0.46127986907958984, + "learning_rate": 7.561577075272686e-06, + "loss": 0.599381685256958, + "step": 714 + }, + { + "epoch": 0.700894498223257, + "grad_norm": 0.4238424301147461, + "learning_rate": 7.554760743911104e-06, + "loss": 0.4853740930557251, + "step": 715 + }, + { + "epoch": 0.701874770248744, + "grad_norm": 0.5007547736167908, + "learning_rate": 7.5479379809423905e-06, + "loss": 0.5677107572555542, + "step": 716 + }, + { + "epoch": 0.7028550422742311, + "grad_norm": 0.44779297709465027, + "learning_rate": 7.541108803542846e-06, + "loss": 0.5496746301651001, + "step": 717 + }, + { + "epoch": 0.7038353142997181, + "grad_norm": 0.4502071738243103, + "learning_rate": 7.534273228904916e-06, + "loss": 0.5206332206726074, + "step": 718 + }, + { + "epoch": 0.7048155863252052, + "grad_norm": 0.43733149766921997, + "learning_rate": 7.527431274237149e-06, + "loss": 0.4367731809616089, + "step": 719 + }, + { + "epoch": 0.7057958583506924, + "grad_norm": 0.511375367641449, + "learning_rate": 7.52058295676416e-06, + "loss": 0.5530825853347778, + "step": 720 + }, + { + "epoch": 0.7067761303761794, + "grad_norm": 0.5255171656608582, + "learning_rate": 7.5137282937265796e-06, + "loss": 0.5948051810264587, + "step": 721 + }, + { + "epoch": 0.7077564024016665, + "grad_norm": 0.4267750680446625, + "learning_rate": 7.506867302381016e-06, + "loss": 0.506665050983429, + "step": 722 + }, + { + "epoch": 0.7087366744271535, + "grad_norm": 0.40493378043174744, + "learning_rate": 7.500000000000001e-06, + "loss": 0.46698686480522156, + "step": 723 + }, + { + "epoch": 0.7097169464526406, + "grad_norm": 0.4602975845336914, + "learning_rate": 7.493126403871964e-06, + "loss": 0.4653869867324829, + "step": 724 + }, + { + "epoch": 0.7106972184781277, + "grad_norm": 0.5179893970489502, + "learning_rate": 7.486246531301178e-06, + "loss": 0.6415598392486572, + "step": 725 + }, + { + "epoch": 0.7116774905036147, + "grad_norm": 0.4917897582054138, + "learning_rate": 7.479360399607707e-06, + "loss": 0.5062524080276489, + "step": 726 + }, + { + "epoch": 0.7126577625291018, + "grad_norm": 0.42604467272758484, + "learning_rate": 7.472468026127385e-06, + "loss": 0.5530419945716858, + "step": 727 + }, + { + "epoch": 0.7136380345545889, + "grad_norm": 0.48969146609306335, + "learning_rate": 7.465569428211752e-06, + "loss": 0.6023317575454712, + "step": 728 + }, + { + "epoch": 0.714618306580076, + "grad_norm": 0.4161062240600586, + "learning_rate": 7.45866462322802e-06, + "loss": 0.5091547966003418, + "step": 729 + }, + { + "epoch": 0.7155985786055631, + "grad_norm": 0.4053821861743927, + "learning_rate": 7.451753628559027e-06, + "loss": 0.4768165349960327, + "step": 730 + }, + { + "epoch": 0.7165788506310501, + "grad_norm": 0.4812045097351074, + "learning_rate": 7.444836461603195e-06, + "loss": 0.6372095346450806, + "step": 731 + }, + { + "epoch": 0.7175591226565372, + "grad_norm": 0.414230614900589, + "learning_rate": 7.437913139774482e-06, + "loss": 0.4736112058162689, + "step": 732 + }, + { + "epoch": 0.7185393946820242, + "grad_norm": 0.40388473868370056, + "learning_rate": 7.430983680502344e-06, + "loss": 0.452359139919281, + "step": 733 + }, + { + "epoch": 0.7195196667075113, + "grad_norm": 0.4632888436317444, + "learning_rate": 7.424048101231687e-06, + "loss": 0.48538899421691895, + "step": 734 + }, + { + "epoch": 0.7204999387329984, + "grad_norm": 0.5188507437705994, + "learning_rate": 7.4171064194228196e-06, + "loss": 0.49239328503608704, + "step": 735 + }, + { + "epoch": 0.7214802107584855, + "grad_norm": 0.4390290677547455, + "learning_rate": 7.410158652551421e-06, + "loss": 0.5711298584938049, + "step": 736 + }, + { + "epoch": 0.7224604827839726, + "grad_norm": 0.49117282032966614, + "learning_rate": 7.403204818108487e-06, + "loss": 0.6047239899635315, + "step": 737 + }, + { + "epoch": 0.7234407548094596, + "grad_norm": 0.5015243887901306, + "learning_rate": 7.396244933600285e-06, + "loss": 0.6040792465209961, + "step": 738 + }, + { + "epoch": 0.7244210268349467, + "grad_norm": 1.2947595119476318, + "learning_rate": 7.3892790165483164e-06, + "loss": 0.5264707207679749, + "step": 739 + }, + { + "epoch": 0.7254012988604338, + "grad_norm": 0.49317121505737305, + "learning_rate": 7.382307084489269e-06, + "loss": 0.5585997104644775, + "step": 740 + }, + { + "epoch": 0.7263815708859208, + "grad_norm": 0.4826045632362366, + "learning_rate": 7.3753291549749764e-06, + "loss": 0.5586689114570618, + "step": 741 + }, + { + "epoch": 0.7273618429114079, + "grad_norm": 0.4288710057735443, + "learning_rate": 7.368345245572362e-06, + "loss": 0.43690255284309387, + "step": 742 + }, + { + "epoch": 0.7283421149368949, + "grad_norm": 0.4085274934768677, + "learning_rate": 7.361355373863415e-06, + "loss": 0.46534493565559387, + "step": 743 + }, + { + "epoch": 0.729322386962382, + "grad_norm": 0.627498209476471, + "learning_rate": 7.354359557445125e-06, + "loss": 0.5892662405967712, + "step": 744 + }, + { + "epoch": 0.7303026589878692, + "grad_norm": 0.4909018874168396, + "learning_rate": 7.347357813929455e-06, + "loss": 0.6983001232147217, + "step": 745 + }, + { + "epoch": 0.7312829310133562, + "grad_norm": 0.4056415855884552, + "learning_rate": 7.3403501609432835e-06, + "loss": 0.5443574786186218, + "step": 746 + }, + { + "epoch": 0.7322632030388433, + "grad_norm": 0.4906488060951233, + "learning_rate": 7.333336616128369e-06, + "loss": 0.5687349438667297, + "step": 747 + }, + { + "epoch": 0.7332434750643303, + "grad_norm": 0.5490400791168213, + "learning_rate": 7.326317197141304e-06, + "loss": 0.6537408828735352, + "step": 748 + }, + { + "epoch": 0.7342237470898174, + "grad_norm": 0.45163455605506897, + "learning_rate": 7.319291921653464e-06, + "loss": 0.5867169499397278, + "step": 749 + }, + { + "epoch": 0.7352040191153045, + "grad_norm": 0.4942396283149719, + "learning_rate": 7.312260807350975e-06, + "loss": 0.5694727897644043, + "step": 750 + }, + { + "epoch": 0.7361842911407915, + "grad_norm": 0.4633714258670807, + "learning_rate": 7.305223871934657e-06, + "loss": 0.5190877914428711, + "step": 751 + }, + { + "epoch": 0.7371645631662787, + "grad_norm": 0.4772140383720398, + "learning_rate": 7.298181133119987e-06, + "loss": 0.4779459834098816, + "step": 752 + }, + { + "epoch": 0.7381448351917658, + "grad_norm": 0.5711033344268799, + "learning_rate": 7.291132608637053e-06, + "loss": 0.6020298004150391, + "step": 753 + }, + { + "epoch": 0.7391251072172528, + "grad_norm": 0.49676769971847534, + "learning_rate": 7.284078316230504e-06, + "loss": 0.4763784408569336, + "step": 754 + }, + { + "epoch": 0.7401053792427399, + "grad_norm": 0.42109596729278564, + "learning_rate": 7.2770182736595164e-06, + "loss": 0.48202162981033325, + "step": 755 + }, + { + "epoch": 0.7410856512682269, + "grad_norm": 0.6203530430793762, + "learning_rate": 7.269952498697734e-06, + "loss": 0.5142531394958496, + "step": 756 + }, + { + "epoch": 0.742065923293714, + "grad_norm": 0.5919861793518066, + "learning_rate": 7.262881009133242e-06, + "loss": 0.6466728448867798, + "step": 757 + }, + { + "epoch": 0.7430461953192011, + "grad_norm": 0.5874010920524597, + "learning_rate": 7.255803822768504e-06, + "loss": 0.48682892322540283, + "step": 758 + }, + { + "epoch": 0.7440264673446881, + "grad_norm": 0.505060613155365, + "learning_rate": 7.24872095742033e-06, + "loss": 0.5157826542854309, + "step": 759 + }, + { + "epoch": 0.7450067393701753, + "grad_norm": 0.672005295753479, + "learning_rate": 7.241632430919823e-06, + "loss": 0.5602332949638367, + "step": 760 + }, + { + "epoch": 0.7459870113956623, + "grad_norm": 0.42120110988616943, + "learning_rate": 7.234538261112342e-06, + "loss": 0.42714083194732666, + "step": 761 + }, + { + "epoch": 0.7469672834211494, + "grad_norm": 0.7334301471710205, + "learning_rate": 7.227438465857448e-06, + "loss": 0.6078978776931763, + "step": 762 + }, + { + "epoch": 0.7479475554466365, + "grad_norm": 0.6107139587402344, + "learning_rate": 7.2203330630288714e-06, + "loss": 0.4394034445285797, + "step": 763 + }, + { + "epoch": 0.7489278274721235, + "grad_norm": 0.6603996753692627, + "learning_rate": 7.213222070514453e-06, + "loss": 0.7087306976318359, + "step": 764 + }, + { + "epoch": 0.7499080994976106, + "grad_norm": 0.5546720623970032, + "learning_rate": 7.206105506216107e-06, + "loss": 0.4940693974494934, + "step": 765 + }, + { + "epoch": 0.7508883715230976, + "grad_norm": 0.8178077340126038, + "learning_rate": 7.198983388049779e-06, + "loss": 0.5339226126670837, + "step": 766 + }, + { + "epoch": 0.7518686435485847, + "grad_norm": 0.5073369145393372, + "learning_rate": 7.191855733945388e-06, + "loss": 0.5095759034156799, + "step": 767 + }, + { + "epoch": 0.7528489155740719, + "grad_norm": 0.39711958169937134, + "learning_rate": 7.1847225618467975e-06, + "loss": 0.41847747564315796, + "step": 768 + }, + { + "epoch": 0.7538291875995589, + "grad_norm": 0.5229822397232056, + "learning_rate": 7.177583889711763e-06, + "loss": 0.5046806335449219, + "step": 769 + }, + { + "epoch": 0.754809459625046, + "grad_norm": 0.4823722839355469, + "learning_rate": 7.170439735511876e-06, + "loss": 0.5876001119613647, + "step": 770 + }, + { + "epoch": 0.755789731650533, + "grad_norm": 0.7995231747627258, + "learning_rate": 7.163290117232542e-06, + "loss": 0.5199307799339294, + "step": 771 + }, + { + "epoch": 0.7567700036760201, + "grad_norm": 0.5239847898483276, + "learning_rate": 7.1561350528729145e-06, + "loss": 0.5662480592727661, + "step": 772 + }, + { + "epoch": 0.7577502757015072, + "grad_norm": 0.4220186173915863, + "learning_rate": 7.148974560445859e-06, + "loss": 0.409650593996048, + "step": 773 + }, + { + "epoch": 0.7587305477269942, + "grad_norm": 0.40589994192123413, + "learning_rate": 7.1418086579779075e-06, + "loss": 0.482654869556427, + "step": 774 + }, + { + "epoch": 0.7597108197524813, + "grad_norm": 0.4408826529979706, + "learning_rate": 7.1346373635092095e-06, + "loss": 0.5564194917678833, + "step": 775 + }, + { + "epoch": 0.7606910917779683, + "grad_norm": 0.5282425284385681, + "learning_rate": 7.127460695093493e-06, + "loss": 0.6081736087799072, + "step": 776 + }, + { + "epoch": 0.7616713638034555, + "grad_norm": 3.734576940536499, + "learning_rate": 7.12027867079801e-06, + "loss": 0.5723667144775391, + "step": 777 + }, + { + "epoch": 0.7626516358289426, + "grad_norm": 0.494312584400177, + "learning_rate": 7.113091308703498e-06, + "loss": 0.5435426831245422, + "step": 778 + }, + { + "epoch": 0.7636319078544296, + "grad_norm": 0.42854106426239014, + "learning_rate": 7.105898626904134e-06, + "loss": 0.5378055572509766, + "step": 779 + }, + { + "epoch": 0.7646121798799167, + "grad_norm": 0.48089906573295593, + "learning_rate": 7.098700643507485e-06, + "loss": 0.591333270072937, + "step": 780 + }, + { + "epoch": 0.7655924519054037, + "grad_norm": 0.3623630106449127, + "learning_rate": 7.0914973766344645e-06, + "loss": 0.4385789632797241, + "step": 781 + }, + { + "epoch": 0.7665727239308908, + "grad_norm": 0.6053583025932312, + "learning_rate": 7.0842888444192905e-06, + "loss": 0.6852695941925049, + "step": 782 + }, + { + "epoch": 0.7675529959563779, + "grad_norm": 0.43562591075897217, + "learning_rate": 7.0770750650094335e-06, + "loss": 0.5775709748268127, + "step": 783 + }, + { + "epoch": 0.7685332679818649, + "grad_norm": 0.39110779762268066, + "learning_rate": 7.069856056565573e-06, + "loss": 0.48135459423065186, + "step": 784 + }, + { + "epoch": 0.7695135400073521, + "grad_norm": 0.41838300228118896, + "learning_rate": 7.062631837261556e-06, + "loss": 0.5081921815872192, + "step": 785 + }, + { + "epoch": 0.7704938120328391, + "grad_norm": 0.37609559297561646, + "learning_rate": 7.055402425284346e-06, + "loss": 0.4609271287918091, + "step": 786 + }, + { + "epoch": 0.7714740840583262, + "grad_norm": 0.4856472909450531, + "learning_rate": 7.048167838833977e-06, + "loss": 0.5818092823028564, + "step": 787 + }, + { + "epoch": 0.7724543560838133, + "grad_norm": 0.691068708896637, + "learning_rate": 7.040928096123516e-06, + "loss": 0.5119889974594116, + "step": 788 + }, + { + "epoch": 0.7734346281093003, + "grad_norm": 0.5112810730934143, + "learning_rate": 7.033683215379002e-06, + "loss": 0.5305988788604736, + "step": 789 + }, + { + "epoch": 0.7744149001347874, + "grad_norm": 0.4650510549545288, + "learning_rate": 7.026433214839416e-06, + "loss": 0.5277286171913147, + "step": 790 + }, + { + "epoch": 0.7753951721602744, + "grad_norm": 0.40126878023147583, + "learning_rate": 7.019178112756625e-06, + "loss": 0.46395719051361084, + "step": 791 + }, + { + "epoch": 0.7763754441857615, + "grad_norm": 0.4711439609527588, + "learning_rate": 7.011917927395341e-06, + "loss": 0.49734482169151306, + "step": 792 + }, + { + "epoch": 0.7773557162112487, + "grad_norm": 0.4137747585773468, + "learning_rate": 7.004652677033069e-06, + "loss": 0.5272625684738159, + "step": 793 + }, + { + "epoch": 0.7783359882367357, + "grad_norm": 0.44795504212379456, + "learning_rate": 6.997382379960072e-06, + "loss": 0.5505017042160034, + "step": 794 + }, + { + "epoch": 0.7793162602622228, + "grad_norm": 0.43027734756469727, + "learning_rate": 6.990107054479313e-06, + "loss": 0.5101114511489868, + "step": 795 + }, + { + "epoch": 0.7802965322877098, + "grad_norm": 0.35909900069236755, + "learning_rate": 6.9828267189064125e-06, + "loss": 0.52973473072052, + "step": 796 + }, + { + "epoch": 0.7812768043131969, + "grad_norm": 0.45199069380760193, + "learning_rate": 6.9755413915696105e-06, + "loss": 0.597319483757019, + "step": 797 + }, + { + "epoch": 0.782257076338684, + "grad_norm": 0.3661534786224365, + "learning_rate": 6.968251090809708e-06, + "loss": 0.4127325713634491, + "step": 798 + }, + { + "epoch": 0.783237348364171, + "grad_norm": 0.41393721103668213, + "learning_rate": 6.960955834980028e-06, + "loss": 0.4885827302932739, + "step": 799 + }, + { + "epoch": 0.7842176203896581, + "grad_norm": 0.406561017036438, + "learning_rate": 6.953655642446368e-06, + "loss": 0.5904073715209961, + "step": 800 + }, + { + "epoch": 0.7851978924151451, + "grad_norm": 0.4422285556793213, + "learning_rate": 6.946350531586959e-06, + "loss": 0.47113174200057983, + "step": 801 + }, + { + "epoch": 0.7861781644406323, + "grad_norm": 0.40606558322906494, + "learning_rate": 6.9390405207924014e-06, + "loss": 0.536541223526001, + "step": 802 + }, + { + "epoch": 0.7871584364661194, + "grad_norm": 0.4594433307647705, + "learning_rate": 6.931725628465643e-06, + "loss": 0.47606784105300903, + "step": 803 + }, + { + "epoch": 0.7881387084916064, + "grad_norm": 0.47338026762008667, + "learning_rate": 6.9244058730219176e-06, + "loss": 0.5939470529556274, + "step": 804 + }, + { + "epoch": 0.7891189805170935, + "grad_norm": 0.3534870445728302, + "learning_rate": 6.917081272888697e-06, + "loss": 0.3604448437690735, + "step": 805 + }, + { + "epoch": 0.7900992525425806, + "grad_norm": 0.44684967398643494, + "learning_rate": 6.909751846505657e-06, + "loss": 0.4442647695541382, + "step": 806 + }, + { + "epoch": 0.7910795245680676, + "grad_norm": 0.4265282154083252, + "learning_rate": 6.902417612324615e-06, + "loss": 0.5449322462081909, + "step": 807 + }, + { + "epoch": 0.7920597965935547, + "grad_norm": 0.37107881903648376, + "learning_rate": 6.895078588809503e-06, + "loss": 0.48995405435562134, + "step": 808 + }, + { + "epoch": 0.7930400686190417, + "grad_norm": 0.3696930706501007, + "learning_rate": 6.887734794436301e-06, + "loss": 0.40083059668540955, + "step": 809 + }, + { + "epoch": 0.7940203406445289, + "grad_norm": 0.4195983111858368, + "learning_rate": 6.880386247692999e-06, + "loss": 0.4219232499599457, + "step": 810 + }, + { + "epoch": 0.795000612670016, + "grad_norm": 0.3787069022655487, + "learning_rate": 6.873032967079562e-06, + "loss": 0.48830705881118774, + "step": 811 + }, + { + "epoch": 0.795980884695503, + "grad_norm": 0.4739780128002167, + "learning_rate": 6.865674971107859e-06, + "loss": 0.5899959802627563, + "step": 812 + }, + { + "epoch": 0.7969611567209901, + "grad_norm": 0.40056106448173523, + "learning_rate": 6.858312278301638e-06, + "loss": 0.5551238059997559, + "step": 813 + }, + { + "epoch": 0.7979414287464771, + "grad_norm": 0.3908003568649292, + "learning_rate": 6.850944907196472e-06, + "loss": 0.48024556040763855, + "step": 814 + }, + { + "epoch": 0.7989217007719642, + "grad_norm": 0.3721950948238373, + "learning_rate": 6.8435728763397045e-06, + "loss": 0.454034686088562, + "step": 815 + }, + { + "epoch": 0.7999019727974513, + "grad_norm": 0.458629846572876, + "learning_rate": 6.836196204290417e-06, + "loss": 0.5407589673995972, + "step": 816 + }, + { + "epoch": 0.8008822448229383, + "grad_norm": 0.3916063904762268, + "learning_rate": 6.828814909619374e-06, + "loss": 0.5556098222732544, + "step": 817 + }, + { + "epoch": 0.8018625168484255, + "grad_norm": 0.37506920099258423, + "learning_rate": 6.821429010908972e-06, + "loss": 0.4908115565776825, + "step": 818 + }, + { + "epoch": 0.8028427888739125, + "grad_norm": 0.4200496971607208, + "learning_rate": 6.814038526753205e-06, + "loss": 0.5301434397697449, + "step": 819 + }, + { + "epoch": 0.8038230608993996, + "grad_norm": 0.466868132352829, + "learning_rate": 6.806643475757608e-06, + "loss": 0.46871984004974365, + "step": 820 + }, + { + "epoch": 0.8048033329248867, + "grad_norm": 0.4289228320121765, + "learning_rate": 6.799243876539213e-06, + "loss": 0.5200282335281372, + "step": 821 + }, + { + "epoch": 0.8057836049503737, + "grad_norm": 0.43029043078422546, + "learning_rate": 6.7918397477265e-06, + "loss": 0.6185582876205444, + "step": 822 + }, + { + "epoch": 0.8067638769758608, + "grad_norm": 0.4174940288066864, + "learning_rate": 6.78443110795936e-06, + "loss": 0.49757710099220276, + "step": 823 + }, + { + "epoch": 0.8077441490013478, + "grad_norm": 0.3718063533306122, + "learning_rate": 6.77701797588903e-06, + "loss": 0.48157191276550293, + "step": 824 + }, + { + "epoch": 0.808724421026835, + "grad_norm": 0.6138043999671936, + "learning_rate": 6.76960037017806e-06, + "loss": 0.44370216131210327, + "step": 825 + }, + { + "epoch": 0.8097046930523221, + "grad_norm": 1.0954099893569946, + "learning_rate": 6.762178309500267e-06, + "loss": 0.5945616364479065, + "step": 826 + }, + { + "epoch": 0.8106849650778091, + "grad_norm": 0.4381861388683319, + "learning_rate": 6.75475181254068e-06, + "loss": 0.5567284822463989, + "step": 827 + }, + { + "epoch": 0.8116652371032962, + "grad_norm": 0.4149746596813202, + "learning_rate": 6.747320897995493e-06, + "loss": 0.5557035207748413, + "step": 828 + }, + { + "epoch": 0.8126455091287832, + "grad_norm": 0.427548885345459, + "learning_rate": 6.739885584572026e-06, + "loss": 0.48417359590530396, + "step": 829 + }, + { + "epoch": 0.8136257811542703, + "grad_norm": 0.5222792625427246, + "learning_rate": 6.732445890988675e-06, + "loss": 0.6526938676834106, + "step": 830 + }, + { + "epoch": 0.8146060531797574, + "grad_norm": 0.5752202272415161, + "learning_rate": 6.725001835974854e-06, + "loss": 0.5249187350273132, + "step": 831 + }, + { + "epoch": 0.8155863252052444, + "grad_norm": 0.46711474657058716, + "learning_rate": 6.717553438270966e-06, + "loss": 0.452995240688324, + "step": 832 + }, + { + "epoch": 0.8165665972307316, + "grad_norm": 0.47589167952537537, + "learning_rate": 6.710100716628345e-06, + "loss": 0.532433271408081, + "step": 833 + }, + { + "epoch": 0.8175468692562186, + "grad_norm": 0.3826428949832916, + "learning_rate": 6.702643689809205e-06, + "loss": 0.40372392535209656, + "step": 834 + }, + { + "epoch": 0.8185271412817057, + "grad_norm": 0.47821974754333496, + "learning_rate": 6.695182376586603e-06, + "loss": 0.4737083315849304, + "step": 835 + }, + { + "epoch": 0.8195074133071928, + "grad_norm": 3.390791177749634, + "learning_rate": 6.68771679574439e-06, + "loss": 0.5816224217414856, + "step": 836 + }, + { + "epoch": 0.8204876853326798, + "grad_norm": 0.4647277593612671, + "learning_rate": 6.680246966077151e-06, + "loss": 0.5226106643676758, + "step": 837 + }, + { + "epoch": 0.8214679573581669, + "grad_norm": 0.4741794466972351, + "learning_rate": 6.672772906390177e-06, + "loss": 0.4869573414325714, + "step": 838 + }, + { + "epoch": 0.8224482293836539, + "grad_norm": 0.45704689621925354, + "learning_rate": 6.665294635499404e-06, + "loss": 0.43767136335372925, + "step": 839 + }, + { + "epoch": 0.823428501409141, + "grad_norm": 0.4436359703540802, + "learning_rate": 6.65781217223137e-06, + "loss": 0.5588634014129639, + "step": 840 + }, + { + "epoch": 0.8244087734346282, + "grad_norm": 0.4628417193889618, + "learning_rate": 6.650325535423166e-06, + "loss": 0.6288622617721558, + "step": 841 + }, + { + "epoch": 0.8253890454601152, + "grad_norm": 0.4618736207485199, + "learning_rate": 6.642834743922393e-06, + "loss": 0.6996235847473145, + "step": 842 + }, + { + "epoch": 0.8263693174856023, + "grad_norm": 0.4331154227256775, + "learning_rate": 6.635339816587109e-06, + "loss": 0.5010419487953186, + "step": 843 + }, + { + "epoch": 0.8273495895110893, + "grad_norm": 0.48043107986450195, + "learning_rate": 6.627840772285784e-06, + "loss": 0.5202208161354065, + "step": 844 + }, + { + "epoch": 0.8283298615365764, + "grad_norm": 0.4572037160396576, + "learning_rate": 6.6203376298972535e-06, + "loss": 0.5409632921218872, + "step": 845 + }, + { + "epoch": 0.8293101335620635, + "grad_norm": 0.5208296775817871, + "learning_rate": 6.612830408310671e-06, + "loss": 0.5930708050727844, + "step": 846 + }, + { + "epoch": 0.8302904055875505, + "grad_norm": 0.4406164884567261, + "learning_rate": 6.605319126425455e-06, + "loss": 0.5349196195602417, + "step": 847 + }, + { + "epoch": 0.8312706776130376, + "grad_norm": 0.418474406003952, + "learning_rate": 6.5978038031512505e-06, + "loss": 0.41959160566329956, + "step": 848 + }, + { + "epoch": 0.8322509496385246, + "grad_norm": 0.5490645170211792, + "learning_rate": 6.590284457407876e-06, + "loss": 0.6460529565811157, + "step": 849 + }, + { + "epoch": 0.8332312216640118, + "grad_norm": 0.44365769624710083, + "learning_rate": 6.582761108125272e-06, + "loss": 0.46053674817085266, + "step": 850 + }, + { + "epoch": 0.8342114936894989, + "grad_norm": 0.43043237924575806, + "learning_rate": 6.5752337742434644e-06, + "loss": 0.41906946897506714, + "step": 851 + }, + { + "epoch": 0.8351917657149859, + "grad_norm": 0.5311118364334106, + "learning_rate": 6.567702474712507e-06, + "loss": 0.5444743037223816, + "step": 852 + }, + { + "epoch": 0.836172037740473, + "grad_norm": 0.6719686985015869, + "learning_rate": 6.560167228492436e-06, + "loss": 0.488853394985199, + "step": 853 + }, + { + "epoch": 0.83715230976596, + "grad_norm": 0.41844865679740906, + "learning_rate": 6.552628054553225e-06, + "loss": 0.49421727657318115, + "step": 854 + }, + { + "epoch": 0.8381325817914471, + "grad_norm": 0.4129052460193634, + "learning_rate": 6.545084971874738e-06, + "loss": 0.5810033082962036, + "step": 855 + }, + { + "epoch": 0.8391128538169342, + "grad_norm": 0.44456323981285095, + "learning_rate": 6.537537999446673e-06, + "loss": 0.5115293860435486, + "step": 856 + }, + { + "epoch": 0.8400931258424212, + "grad_norm": 0.4097927212715149, + "learning_rate": 6.529987156268527e-06, + "loss": 0.4882569909095764, + "step": 857 + }, + { + "epoch": 0.8410733978679084, + "grad_norm": 0.44071686267852783, + "learning_rate": 6.522432461349536e-06, + "loss": 0.5030308961868286, + "step": 858 + }, + { + "epoch": 0.8420536698933954, + "grad_norm": 0.5421448349952698, + "learning_rate": 6.514873933708637e-06, + "loss": 0.6325263977050781, + "step": 859 + }, + { + "epoch": 0.8430339419188825, + "grad_norm": 0.5930724143981934, + "learning_rate": 6.507311592374415e-06, + "loss": 0.5336626768112183, + "step": 860 + }, + { + "epoch": 0.8440142139443696, + "grad_norm": 0.4966267943382263, + "learning_rate": 6.499745456385054e-06, + "loss": 0.5958417654037476, + "step": 861 + }, + { + "epoch": 0.8449944859698566, + "grad_norm": 0.4763844311237335, + "learning_rate": 6.4921755447882926e-06, + "loss": 0.5004958510398865, + "step": 862 + }, + { + "epoch": 0.8459747579953437, + "grad_norm": 0.5007293224334717, + "learning_rate": 6.484601876641375e-06, + "loss": 0.5528030395507812, + "step": 863 + }, + { + "epoch": 0.8469550300208308, + "grad_norm": 0.41702571511268616, + "learning_rate": 6.477024471011001e-06, + "loss": 0.49212419986724854, + "step": 864 + }, + { + "epoch": 0.8479353020463178, + "grad_norm": 0.4272618889808655, + "learning_rate": 6.469443346973281e-06, + "loss": 0.5188068151473999, + "step": 865 + }, + { + "epoch": 0.848915574071805, + "grad_norm": 0.5673028230667114, + "learning_rate": 6.461858523613684e-06, + "loss": 0.5160360932350159, + "step": 866 + }, + { + "epoch": 0.849895846097292, + "grad_norm": 0.4834825396537781, + "learning_rate": 6.454270020026996e-06, + "loss": 0.5802666544914246, + "step": 867 + }, + { + "epoch": 0.8508761181227791, + "grad_norm": 0.48169195652008057, + "learning_rate": 6.446677855317265e-06, + "loss": 0.4749450087547302, + "step": 868 + }, + { + "epoch": 0.8518563901482662, + "grad_norm": 0.5354800820350647, + "learning_rate": 6.439082048597755e-06, + "loss": 0.49622076749801636, + "step": 869 + }, + { + "epoch": 0.8528366621737532, + "grad_norm": 0.42235618829727173, + "learning_rate": 6.431482618990902e-06, + "loss": 0.46224337816238403, + "step": 870 + }, + { + "epoch": 0.8538169341992403, + "grad_norm": 0.5284849405288696, + "learning_rate": 6.423879585628262e-06, + "loss": 0.6243587136268616, + "step": 871 + }, + { + "epoch": 0.8547972062247273, + "grad_norm": 0.6278819441795349, + "learning_rate": 6.41627296765046e-06, + "loss": 0.6148986220359802, + "step": 872 + }, + { + "epoch": 0.8557774782502144, + "grad_norm": 0.45493778586387634, + "learning_rate": 6.408662784207149e-06, + "loss": 0.5227400660514832, + "step": 873 + }, + { + "epoch": 0.8567577502757016, + "grad_norm": 0.5376487374305725, + "learning_rate": 6.401049054456957e-06, + "loss": 0.6814272403717041, + "step": 874 + }, + { + "epoch": 0.8577380223011886, + "grad_norm": 0.3643544316291809, + "learning_rate": 6.39343179756744e-06, + "loss": 0.44438761472702026, + "step": 875 + }, + { + "epoch": 0.8587182943266757, + "grad_norm": 0.4614802598953247, + "learning_rate": 6.385811032715031e-06, + "loss": 0.4618123769760132, + "step": 876 + }, + { + "epoch": 0.8596985663521627, + "grad_norm": 0.5620284676551819, + "learning_rate": 6.378186779084996e-06, + "loss": 0.5348123908042908, + "step": 877 + }, + { + "epoch": 0.8606788383776498, + "grad_norm": 0.6737039089202881, + "learning_rate": 6.370559055871389e-06, + "loss": 0.5795971751213074, + "step": 878 + }, + { + "epoch": 0.8616591104031369, + "grad_norm": 0.46896591782569885, + "learning_rate": 6.362927882276991e-06, + "loss": 0.7106213569641113, + "step": 879 + }, + { + "epoch": 0.8626393824286239, + "grad_norm": 0.4269348680973053, + "learning_rate": 6.355293277513271e-06, + "loss": 0.4726841449737549, + "step": 880 + }, + { + "epoch": 0.863619654454111, + "grad_norm": 0.41544944047927856, + "learning_rate": 6.34765526080034e-06, + "loss": 0.4410395920276642, + "step": 881 + }, + { + "epoch": 0.864599926479598, + "grad_norm": 0.4709013104438782, + "learning_rate": 6.3400138513668955e-06, + "loss": 0.4410967230796814, + "step": 882 + }, + { + "epoch": 0.8655801985050852, + "grad_norm": 0.4805549681186676, + "learning_rate": 6.332369068450175e-06, + "loss": 0.5561596155166626, + "step": 883 + }, + { + "epoch": 0.8665604705305723, + "grad_norm": 0.35023778676986694, + "learning_rate": 6.324720931295913e-06, + "loss": 0.3837100863456726, + "step": 884 + }, + { + "epoch": 0.8675407425560593, + "grad_norm": 0.9413726925849915, + "learning_rate": 6.317069459158284e-06, + "loss": 0.5233207941055298, + "step": 885 + }, + { + "epoch": 0.8685210145815464, + "grad_norm": 0.43868952989578247, + "learning_rate": 6.3094146712998595e-06, + "loss": 0.5409136414527893, + "step": 886 + }, + { + "epoch": 0.8695012866070334, + "grad_norm": 0.4314418435096741, + "learning_rate": 6.301756586991561e-06, + "loss": 0.5296670794487, + "step": 887 + }, + { + "epoch": 0.8704815586325205, + "grad_norm": 0.4177858829498291, + "learning_rate": 6.294095225512604e-06, + "loss": 0.4831322729587555, + "step": 888 + }, + { + "epoch": 0.8714618306580076, + "grad_norm": 0.4337747395038605, + "learning_rate": 6.286430606150458e-06, + "loss": 0.5630544424057007, + "step": 889 + }, + { + "epoch": 0.8724421026834946, + "grad_norm": 0.4146502614021301, + "learning_rate": 6.278762748200795e-06, + "loss": 0.4446233808994293, + "step": 890 + }, + { + "epoch": 0.8734223747089818, + "grad_norm": 0.4662277102470398, + "learning_rate": 6.271091670967437e-06, + "loss": 0.5407575964927673, + "step": 891 + }, + { + "epoch": 0.8744026467344688, + "grad_norm": 0.48252469301223755, + "learning_rate": 6.2634173937623075e-06, + "loss": 0.46057388186454773, + "step": 892 + }, + { + "epoch": 0.8753829187599559, + "grad_norm": 0.4287121593952179, + "learning_rate": 6.255739935905396e-06, + "loss": 0.4808887243270874, + "step": 893 + }, + { + "epoch": 0.876363190785443, + "grad_norm": 0.42432844638824463, + "learning_rate": 6.248059316724693e-06, + "loss": 0.4453814625740051, + "step": 894 + }, + { + "epoch": 0.87734346281093, + "grad_norm": 0.42549434304237366, + "learning_rate": 6.240375555556145e-06, + "loss": 0.5829179883003235, + "step": 895 + }, + { + "epoch": 0.8783237348364171, + "grad_norm": 0.4437717795372009, + "learning_rate": 6.232688671743613e-06, + "loss": 0.4659510850906372, + "step": 896 + }, + { + "epoch": 0.8793040068619041, + "grad_norm": 0.5647913813591003, + "learning_rate": 6.22499868463882e-06, + "loss": 0.5178962349891663, + "step": 897 + }, + { + "epoch": 0.8802842788873912, + "grad_norm": 0.6277410984039307, + "learning_rate": 6.217305613601296e-06, + "loss": 0.6390823721885681, + "step": 898 + }, + { + "epoch": 0.8812645509128784, + "grad_norm": 0.43650683760643005, + "learning_rate": 6.209609477998339e-06, + "loss": 0.5823169946670532, + "step": 899 + }, + { + "epoch": 0.8822448229383654, + "grad_norm": 0.43029022216796875, + "learning_rate": 6.201910297204963e-06, + "loss": 0.4697658121585846, + "step": 900 + }, + { + "epoch": 0.8832250949638525, + "grad_norm": 0.4142000675201416, + "learning_rate": 6.194208090603845e-06, + "loss": 0.49903035163879395, + "step": 901 + }, + { + "epoch": 0.8842053669893395, + "grad_norm": 0.990412712097168, + "learning_rate": 6.186502877585281e-06, + "loss": 0.4805365800857544, + "step": 902 + }, + { + "epoch": 0.8851856390148266, + "grad_norm": 1.582311987876892, + "learning_rate": 6.178794677547138e-06, + "loss": 0.5169810056686401, + "step": 903 + }, + { + "epoch": 0.8861659110403137, + "grad_norm": 0.4250990152359009, + "learning_rate": 6.171083509894795e-06, + "loss": 0.4501458406448364, + "step": 904 + }, + { + "epoch": 0.8871461830658007, + "grad_norm": 0.42763155698776245, + "learning_rate": 6.163369394041112e-06, + "loss": 0.4777667820453644, + "step": 905 + }, + { + "epoch": 0.8881264550912878, + "grad_norm": 0.5017222166061401, + "learning_rate": 6.155652349406366e-06, + "loss": 0.4914003610610962, + "step": 906 + }, + { + "epoch": 0.8891067271167749, + "grad_norm": 0.44423478841781616, + "learning_rate": 6.1479323954182055e-06, + "loss": 0.48054519295692444, + "step": 907 + }, + { + "epoch": 0.890086999142262, + "grad_norm": 0.6064445376396179, + "learning_rate": 6.140209551511609e-06, + "loss": 0.5427241921424866, + "step": 908 + }, + { + "epoch": 0.8910672711677491, + "grad_norm": 0.45574769377708435, + "learning_rate": 6.132483837128823e-06, + "loss": 0.46265387535095215, + "step": 909 + }, + { + "epoch": 0.8920475431932361, + "grad_norm": 0.5986409187316895, + "learning_rate": 6.124755271719326e-06, + "loss": 0.7392776608467102, + "step": 910 + }, + { + "epoch": 0.8930278152187232, + "grad_norm": 0.5339890718460083, + "learning_rate": 6.1170238747397715e-06, + "loss": 0.5947340726852417, + "step": 911 + }, + { + "epoch": 0.8940080872442102, + "grad_norm": 0.5864500403404236, + "learning_rate": 6.109289665653944e-06, + "loss": 0.48198479413986206, + "step": 912 + }, + { + "epoch": 0.8949883592696973, + "grad_norm": 0.4087126553058624, + "learning_rate": 6.101552663932704e-06, + "loss": 0.516301155090332, + "step": 913 + }, + { + "epoch": 0.8959686312951844, + "grad_norm": 0.4832591116428375, + "learning_rate": 6.093812889053945e-06, + "loss": 0.5776762962341309, + "step": 914 + }, + { + "epoch": 0.8969489033206715, + "grad_norm": 0.4292212724685669, + "learning_rate": 6.08607036050254e-06, + "loss": 0.48135313391685486, + "step": 915 + }, + { + "epoch": 0.8979291753461586, + "grad_norm": 0.4345339834690094, + "learning_rate": 6.078325097770296e-06, + "loss": 0.5092774033546448, + "step": 916 + }, + { + "epoch": 0.8989094473716457, + "grad_norm": 0.5339592099189758, + "learning_rate": 6.070577120355903e-06, + "loss": 0.6810686588287354, + "step": 917 + }, + { + "epoch": 0.8998897193971327, + "grad_norm": 0.40882620215415955, + "learning_rate": 6.062826447764883e-06, + "loss": 0.49107372760772705, + "step": 918 + }, + { + "epoch": 0.9008699914226198, + "grad_norm": 0.475019246339798, + "learning_rate": 6.055073099509549e-06, + "loss": 0.5392287969589233, + "step": 919 + }, + { + "epoch": 0.9018502634481068, + "grad_norm": 0.45199888944625854, + "learning_rate": 6.047317095108943e-06, + "loss": 0.5883407592773438, + "step": 920 + }, + { + "epoch": 0.9028305354735939, + "grad_norm": 0.3889693319797516, + "learning_rate": 6.039558454088796e-06, + "loss": 0.40601295232772827, + "step": 921 + }, + { + "epoch": 0.903810807499081, + "grad_norm": 0.4438765048980713, + "learning_rate": 6.031797195981481e-06, + "loss": 0.5861967206001282, + "step": 922 + }, + { + "epoch": 0.904791079524568, + "grad_norm": 0.4316740036010742, + "learning_rate": 6.024033340325954e-06, + "loss": 0.46414318680763245, + "step": 923 + }, + { + "epoch": 0.9057713515500552, + "grad_norm": 0.38185253739356995, + "learning_rate": 6.016266906667711e-06, + "loss": 0.48858797550201416, + "step": 924 + }, + { + "epoch": 0.9067516235755422, + "grad_norm": 0.41384172439575195, + "learning_rate": 6.0084979145587444e-06, + "loss": 0.4836603105068207, + "step": 925 + }, + { + "epoch": 0.9077318956010293, + "grad_norm": 0.4198357164859772, + "learning_rate": 6.000726383557476e-06, + "loss": 0.4855630695819855, + "step": 926 + }, + { + "epoch": 0.9087121676265164, + "grad_norm": 0.4103989899158478, + "learning_rate": 5.9929523332287275e-06, + "loss": 0.4840914011001587, + "step": 927 + }, + { + "epoch": 0.9096924396520034, + "grad_norm": 0.3962758779525757, + "learning_rate": 5.985175783143666e-06, + "loss": 0.5013476610183716, + "step": 928 + }, + { + "epoch": 0.9106727116774905, + "grad_norm": 0.4539540410041809, + "learning_rate": 5.977396752879742e-06, + "loss": 0.5926103591918945, + "step": 929 + }, + { + "epoch": 0.9116529837029775, + "grad_norm": 0.37731069326400757, + "learning_rate": 5.9696152620206575e-06, + "loss": 0.4235096871852875, + "step": 930 + }, + { + "epoch": 0.9126332557284647, + "grad_norm": 0.4731782078742981, + "learning_rate": 5.961831330156306e-06, + "loss": 0.6326535940170288, + "step": 931 + }, + { + "epoch": 0.9136135277539518, + "grad_norm": 0.4542580246925354, + "learning_rate": 5.954044976882725e-06, + "loss": 0.558417558670044, + "step": 932 + }, + { + "epoch": 0.9145937997794388, + "grad_norm": 0.4298331141471863, + "learning_rate": 5.946256221802052e-06, + "loss": 0.5137132406234741, + "step": 933 + }, + { + "epoch": 0.9155740718049259, + "grad_norm": 0.47129830718040466, + "learning_rate": 5.938465084522468e-06, + "loss": 0.5925625562667847, + "step": 934 + }, + { + "epoch": 0.9165543438304129, + "grad_norm": 0.42570555210113525, + "learning_rate": 5.930671584658151e-06, + "loss": 0.47181135416030884, + "step": 935 + }, + { + "epoch": 0.9175346158559, + "grad_norm": 0.4356832802295685, + "learning_rate": 5.922875741829227e-06, + "loss": 0.5676063299179077, + "step": 936 + }, + { + "epoch": 0.9185148878813871, + "grad_norm": 0.5222892761230469, + "learning_rate": 5.915077575661723e-06, + "loss": 0.6410059928894043, + "step": 937 + }, + { + "epoch": 0.9194951599068741, + "grad_norm": 0.5957534313201904, + "learning_rate": 5.907277105787513e-06, + "loss": 0.4146493077278137, + "step": 938 + }, + { + "epoch": 0.9204754319323613, + "grad_norm": 0.37144264578819275, + "learning_rate": 5.89947435184427e-06, + "loss": 0.3888629674911499, + "step": 939 + }, + { + "epoch": 0.9214557039578483, + "grad_norm": 0.4696313738822937, + "learning_rate": 5.891669333475416e-06, + "loss": 0.6653344631195068, + "step": 940 + }, + { + "epoch": 0.9224359759833354, + "grad_norm": 0.6026146411895752, + "learning_rate": 5.883862070330079e-06, + "loss": 0.5807779431343079, + "step": 941 + }, + { + "epoch": 0.9234162480088225, + "grad_norm": 0.41022008657455444, + "learning_rate": 5.876052582063031e-06, + "loss": 0.5313568711280823, + "step": 942 + }, + { + "epoch": 0.9243965200343095, + "grad_norm": 0.3925977945327759, + "learning_rate": 5.8682408883346535e-06, + "loss": 0.4375201463699341, + "step": 943 + }, + { + "epoch": 0.9253767920597966, + "grad_norm": 0.9332260489463806, + "learning_rate": 5.860427008810872e-06, + "loss": 0.4921548366546631, + "step": 944 + }, + { + "epoch": 0.9263570640852836, + "grad_norm": 0.5160321593284607, + "learning_rate": 5.85261096316312e-06, + "loss": 0.5410822033882141, + "step": 945 + }, + { + "epoch": 0.9273373361107707, + "grad_norm": 0.4472995102405548, + "learning_rate": 5.844792771068284e-06, + "loss": 0.5569736957550049, + "step": 946 + }, + { + "epoch": 0.9283176081362579, + "grad_norm": 0.34775489568710327, + "learning_rate": 5.8369724522086545e-06, + "loss": 0.48673319816589355, + "step": 947 + }, + { + "epoch": 0.9292978801617449, + "grad_norm": 0.46031758189201355, + "learning_rate": 5.829150026271871e-06, + "loss": 0.6265984177589417, + "step": 948 + }, + { + "epoch": 0.930278152187232, + "grad_norm": 0.3856378197669983, + "learning_rate": 5.821325512950886e-06, + "loss": 0.49659961462020874, + "step": 949 + }, + { + "epoch": 0.931258424212719, + "grad_norm": 0.7397359013557434, + "learning_rate": 5.8134989319439004e-06, + "loss": 0.5615105628967285, + "step": 950 + }, + { + "epoch": 0.9322386962382061, + "grad_norm": 0.4330701529979706, + "learning_rate": 5.805670302954322e-06, + "loss": 0.5041952133178711, + "step": 951 + }, + { + "epoch": 0.9332189682636932, + "grad_norm": 0.4708372950553894, + "learning_rate": 5.797839645690715e-06, + "loss": 0.49568355083465576, + "step": 952 + }, + { + "epoch": 0.9341992402891802, + "grad_norm": 0.4601927399635315, + "learning_rate": 5.79000697986675e-06, + "loss": 0.5010634064674377, + "step": 953 + }, + { + "epoch": 0.9351795123146673, + "grad_norm": 0.4409112334251404, + "learning_rate": 5.782172325201155e-06, + "loss": 0.49562788009643555, + "step": 954 + }, + { + "epoch": 0.9361597843401543, + "grad_norm": 0.4923012852668762, + "learning_rate": 5.774335701417662e-06, + "loss": 0.4498739540576935, + "step": 955 + }, + { + "epoch": 0.9371400563656415, + "grad_norm": 0.40143096446990967, + "learning_rate": 5.766497128244963e-06, + "loss": 0.551247239112854, + "step": 956 + }, + { + "epoch": 0.9381203283911286, + "grad_norm": 0.4007181227207184, + "learning_rate": 5.758656625416659e-06, + "loss": 0.49534258246421814, + "step": 957 + }, + { + "epoch": 0.9391006004166156, + "grad_norm": 0.5263123512268066, + "learning_rate": 5.750814212671202e-06, + "loss": 0.5723960399627686, + "step": 958 + }, + { + "epoch": 0.9400808724421027, + "grad_norm": 0.4463786482810974, + "learning_rate": 5.7429699097518585e-06, + "loss": 0.531633734703064, + "step": 959 + }, + { + "epoch": 0.9410611444675897, + "grad_norm": 0.43471935391426086, + "learning_rate": 5.735123736406655e-06, + "loss": 0.500960111618042, + "step": 960 + }, + { + "epoch": 0.9420414164930768, + "grad_norm": 0.4278755784034729, + "learning_rate": 5.727275712388318e-06, + "loss": 0.4983411431312561, + "step": 961 + }, + { + "epoch": 0.9430216885185639, + "grad_norm": 0.446235328912735, + "learning_rate": 5.719425857454242e-06, + "loss": 0.7012414932250977, + "step": 962 + }, + { + "epoch": 0.9440019605440509, + "grad_norm": 0.3571961522102356, + "learning_rate": 5.711574191366427e-06, + "loss": 0.4507150650024414, + "step": 963 + }, + { + "epoch": 0.9449822325695381, + "grad_norm": 0.43753328919410706, + "learning_rate": 5.7037207338914326e-06, + "loss": 0.5032480955123901, + "step": 964 + }, + { + "epoch": 0.9459625045950251, + "grad_norm": 0.4057547152042389, + "learning_rate": 5.695865504800328e-06, + "loss": 0.5925788879394531, + "step": 965 + }, + { + "epoch": 0.9469427766205122, + "grad_norm": 0.4289734959602356, + "learning_rate": 5.688008523868646e-06, + "loss": 0.5180743336677551, + "step": 966 + }, + { + "epoch": 0.9479230486459993, + "grad_norm": 0.45726263523101807, + "learning_rate": 5.680149810876322e-06, + "loss": 0.6067906618118286, + "step": 967 + }, + { + "epoch": 0.9489033206714863, + "grad_norm": 0.4400206506252289, + "learning_rate": 5.6722893856076596e-06, + "loss": 0.4790126085281372, + "step": 968 + }, + { + "epoch": 0.9498835926969734, + "grad_norm": 0.4761674702167511, + "learning_rate": 5.664427267851271e-06, + "loss": 0.4958120584487915, + "step": 969 + }, + { + "epoch": 0.9508638647224604, + "grad_norm": 0.392704039812088, + "learning_rate": 5.656563477400026e-06, + "loss": 0.5374599099159241, + "step": 970 + }, + { + "epoch": 0.9518441367479475, + "grad_norm": 0.4220191538333893, + "learning_rate": 5.648698034051009e-06, + "loss": 0.5424162149429321, + "step": 971 + }, + { + "epoch": 0.9528244087734347, + "grad_norm": 0.40179651975631714, + "learning_rate": 5.640830957605466e-06, + "loss": 0.4912230670452118, + "step": 972 + }, + { + "epoch": 0.9538046807989217, + "grad_norm": 0.379812628030777, + "learning_rate": 5.632962267868747e-06, + "loss": 0.4763560891151428, + "step": 973 + }, + { + "epoch": 0.9547849528244088, + "grad_norm": 0.447950154542923, + "learning_rate": 5.625091984650273e-06, + "loss": 0.5547003746032715, + "step": 974 + }, + { + "epoch": 0.9557652248498959, + "grad_norm": 0.3814934492111206, + "learning_rate": 5.617220127763474e-06, + "loss": 0.44596758484840393, + "step": 975 + }, + { + "epoch": 0.9567454968753829, + "grad_norm": 0.40731292963027954, + "learning_rate": 5.609346717025738e-06, + "loss": 0.4661918878555298, + "step": 976 + }, + { + "epoch": 0.95772576890087, + "grad_norm": 0.36259421706199646, + "learning_rate": 5.601471772258368e-06, + "loss": 0.5000268220901489, + "step": 977 + }, + { + "epoch": 0.958706040926357, + "grad_norm": 0.36221638321876526, + "learning_rate": 5.593595313286526e-06, + "loss": 0.40143662691116333, + "step": 978 + }, + { + "epoch": 0.9596863129518441, + "grad_norm": 0.4417479634284973, + "learning_rate": 5.585717359939192e-06, + "loss": 0.45912468433380127, + "step": 979 + }, + { + "epoch": 0.9606665849773313, + "grad_norm": 0.3343220353126526, + "learning_rate": 5.5778379320491e-06, + "loss": 0.41304337978363037, + "step": 980 + }, + { + "epoch": 0.9616468570028183, + "grad_norm": 0.4492647647857666, + "learning_rate": 5.569957049452703e-06, + "loss": 0.6581159830093384, + "step": 981 + }, + { + "epoch": 0.9626271290283054, + "grad_norm": 0.43616801500320435, + "learning_rate": 5.562074731990115e-06, + "loss": 0.4837437868118286, + "step": 982 + }, + { + "epoch": 0.9636074010537924, + "grad_norm": 0.3660007417201996, + "learning_rate": 5.5541909995050554e-06, + "loss": 0.45983701944351196, + "step": 983 + }, + { + "epoch": 0.9645876730792795, + "grad_norm": 0.3804008364677429, + "learning_rate": 5.546305871844816e-06, + "loss": 0.38527461886405945, + "step": 984 + }, + { + "epoch": 0.9655679451047666, + "grad_norm": 0.3764747083187103, + "learning_rate": 5.538419368860196e-06, + "loss": 0.43098461627960205, + "step": 985 + }, + { + "epoch": 0.9665482171302536, + "grad_norm": 0.38172152638435364, + "learning_rate": 5.530531510405454e-06, + "loss": 0.43058720231056213, + "step": 986 + }, + { + "epoch": 0.9675284891557407, + "grad_norm": 0.4629911482334137, + "learning_rate": 5.522642316338268e-06, + "loss": 0.4460248351097107, + "step": 987 + }, + { + "epoch": 0.9685087611812278, + "grad_norm": 0.4303341507911682, + "learning_rate": 5.514751806519673e-06, + "loss": 0.5343886613845825, + "step": 988 + }, + { + "epoch": 0.9694890332067149, + "grad_norm": 0.3839262127876282, + "learning_rate": 5.506860000814017e-06, + "loss": 0.4435023367404938, + "step": 989 + }, + { + "epoch": 0.970469305232202, + "grad_norm": 0.3626423478126526, + "learning_rate": 5.498966919088914e-06, + "loss": 0.5118464827537537, + "step": 990 + }, + { + "epoch": 0.971449577257689, + "grad_norm": 0.37724119424819946, + "learning_rate": 5.491072581215186e-06, + "loss": 0.5193969011306763, + "step": 991 + }, + { + "epoch": 0.9724298492831761, + "grad_norm": 0.3465311825275421, + "learning_rate": 5.48317700706682e-06, + "loss": 0.4697530269622803, + "step": 992 + }, + { + "epoch": 0.9734101213086631, + "grad_norm": 0.3911851644515991, + "learning_rate": 5.475280216520913e-06, + "loss": 0.44076526165008545, + "step": 993 + }, + { + "epoch": 0.9743903933341502, + "grad_norm": 0.32924512028694153, + "learning_rate": 5.467382229457629e-06, + "loss": 0.48638254404067993, + "step": 994 + }, + { + "epoch": 0.9753706653596373, + "grad_norm": 0.343056857585907, + "learning_rate": 5.459483065760138e-06, + "loss": 0.48689454793930054, + "step": 995 + }, + { + "epoch": 0.9763509373851244, + "grad_norm": 0.3778912425041199, + "learning_rate": 5.451582745314576e-06, + "loss": 0.5066269636154175, + "step": 996 + }, + { + "epoch": 0.9773312094106115, + "grad_norm": 0.4240163266658783, + "learning_rate": 5.443681288009991e-06, + "loss": 0.5170713067054749, + "step": 997 + }, + { + "epoch": 0.9783114814360985, + "grad_norm": 0.42373648285865784, + "learning_rate": 5.435778713738292e-06, + "loss": 0.4839274287223816, + "step": 998 + }, + { + "epoch": 0.9792917534615856, + "grad_norm": 0.4663996398448944, + "learning_rate": 5.4278750423942e-06, + "loss": 0.5407215356826782, + "step": 999 + }, + { + "epoch": 0.9802720254870727, + "grad_norm": 0.35067281126976013, + "learning_rate": 5.419970293875197e-06, + "loss": 0.42240002751350403, + "step": 1000 + }, + { + "epoch": 0.9812522975125597, + "grad_norm": 0.3952957093715668, + "learning_rate": 5.412064488081482e-06, + "loss": 0.4958457350730896, + "step": 1001 + }, + { + "epoch": 0.9822325695380468, + "grad_norm": 0.3625335097312927, + "learning_rate": 5.404157644915907e-06, + "loss": 0.4905568063259125, + "step": 1002 + }, + { + "epoch": 0.9832128415635338, + "grad_norm": 0.4454256594181061, + "learning_rate": 5.396249784283943e-06, + "loss": 0.46749985218048096, + "step": 1003 + }, + { + "epoch": 0.984193113589021, + "grad_norm": 0.37890392541885376, + "learning_rate": 5.388340926093619e-06, + "loss": 0.4817125201225281, + "step": 1004 + }, + { + "epoch": 0.9851733856145081, + "grad_norm": 0.40377727150917053, + "learning_rate": 5.380431090255475e-06, + "loss": 0.46670079231262207, + "step": 1005 + }, + { + "epoch": 0.9861536576399951, + "grad_norm": 0.4605295658111572, + "learning_rate": 5.372520296682516e-06, + "loss": 0.626600980758667, + "step": 1006 + }, + { + "epoch": 0.9871339296654822, + "grad_norm": 0.37546271085739136, + "learning_rate": 5.364608565290154e-06, + "loss": 0.43172627687454224, + "step": 1007 + }, + { + "epoch": 0.9881142016909692, + "grad_norm": 0.39535439014434814, + "learning_rate": 5.356695915996162e-06, + "loss": 0.4631704092025757, + "step": 1008 + }, + { + "epoch": 0.9890944737164563, + "grad_norm": 0.3779025673866272, + "learning_rate": 5.348782368720627e-06, + "loss": 0.5013574361801147, + "step": 1009 + }, + { + "epoch": 0.9900747457419434, + "grad_norm": 0.37166279554367065, + "learning_rate": 5.340867943385895e-06, + "loss": 0.4716484546661377, + "step": 1010 + }, + { + "epoch": 0.9910550177674304, + "grad_norm": 0.4055781066417694, + "learning_rate": 5.33295265991652e-06, + "loss": 0.5293822288513184, + "step": 1011 + }, + { + "epoch": 0.9920352897929176, + "grad_norm": 0.3550724387168884, + "learning_rate": 5.325036538239221e-06, + "loss": 0.45156705379486084, + "step": 1012 + }, + { + "epoch": 0.9930155618184046, + "grad_norm": 0.3812902867794037, + "learning_rate": 5.317119598282823e-06, + "loss": 0.4375635087490082, + "step": 1013 + }, + { + "epoch": 0.9939958338438917, + "grad_norm": 0.447942852973938, + "learning_rate": 5.3092018599782155e-06, + "loss": 0.542604923248291, + "step": 1014 + }, + { + "epoch": 0.9949761058693788, + "grad_norm": 0.44398513436317444, + "learning_rate": 5.301283343258293e-06, + "loss": 0.547914981842041, + "step": 1015 + }, + { + "epoch": 0.9959563778948658, + "grad_norm": 0.42243313789367676, + "learning_rate": 5.293364068057912e-06, + "loss": 0.5050883293151855, + "step": 1016 + }, + { + "epoch": 0.9969366499203529, + "grad_norm": 0.34492605924606323, + "learning_rate": 5.285444054313841e-06, + "loss": 0.36160433292388916, + "step": 1017 + }, + { + "epoch": 0.9979169219458399, + "grad_norm": 0.4846872091293335, + "learning_rate": 5.277523321964701e-06, + "loss": 0.5536131262779236, + "step": 1018 + }, + { + "epoch": 0.998897193971327, + "grad_norm": 0.34226441383361816, + "learning_rate": 5.26960189095093e-06, + "loss": 0.4014680087566376, + "step": 1019 + }, + { + "epoch": 0.9998774659968142, + "grad_norm": 0.3548482358455658, + "learning_rate": 5.2616797812147205e-06, + "loss": 0.4133622348308563, + "step": 1020 + }, + { + "epoch": 1.0, + "grad_norm": 1.1780668497085571, + "learning_rate": 5.253757012699972e-06, + "loss": 0.6116985082626343, + "step": 1021 + }, + { + "epoch": 1.000980272025487, + "grad_norm": 0.3691287338733673, + "learning_rate": 5.245833605352246e-06, + "loss": 0.45967456698417664, + "step": 1022 + }, + { + "epoch": 1.0019605440509742, + "grad_norm": 0.37535277009010315, + "learning_rate": 5.237909579118713e-06, + "loss": 0.4363602101802826, + "step": 1023 + }, + { + "epoch": 1.0029408160764612, + "grad_norm": 0.4292650520801544, + "learning_rate": 5.2299849539480965e-06, + "loss": 0.4564576745033264, + "step": 1024 + }, + { + "epoch": 1.0039210881019482, + "grad_norm": 0.3447128236293793, + "learning_rate": 5.2220597497906315e-06, + "loss": 0.3897055387496948, + "step": 1025 + }, + { + "epoch": 1.0049013601274355, + "grad_norm": 0.3557998538017273, + "learning_rate": 5.214133986598014e-06, + "loss": 0.3619903028011322, + "step": 1026 + }, + { + "epoch": 1.0058816321529225, + "grad_norm": 0.39072269201278687, + "learning_rate": 5.206207684323337e-06, + "loss": 0.34538593888282776, + "step": 1027 + }, + { + "epoch": 1.0068619041784095, + "grad_norm": 0.4686689078807831, + "learning_rate": 5.198280862921062e-06, + "loss": 0.4392380118370056, + "step": 1028 + }, + { + "epoch": 1.0078421762038965, + "grad_norm": 0.3751749098300934, + "learning_rate": 5.190353542346951e-06, + "loss": 0.3452165126800537, + "step": 1029 + }, + { + "epoch": 1.0088224482293837, + "grad_norm": 0.4477664828300476, + "learning_rate": 5.182425742558024e-06, + "loss": 0.4791143536567688, + "step": 1030 + }, + { + "epoch": 1.0098027202548707, + "grad_norm": 0.41647619009017944, + "learning_rate": 5.174497483512506e-06, + "loss": 0.4747411906719208, + "step": 1031 + }, + { + "epoch": 1.0107829922803577, + "grad_norm": 0.3926742970943451, + "learning_rate": 5.166568785169781e-06, + "loss": 0.4350982904434204, + "step": 1032 + }, + { + "epoch": 1.011763264305845, + "grad_norm": 0.3539671301841736, + "learning_rate": 5.15863966749034e-06, + "loss": 0.37327489256858826, + "step": 1033 + }, + { + "epoch": 1.012743536331332, + "grad_norm": 0.44685983657836914, + "learning_rate": 5.1507101504357225e-06, + "loss": 0.45712411403656006, + "step": 1034 + }, + { + "epoch": 1.013723808356819, + "grad_norm": 0.3385825455188751, + "learning_rate": 5.142780253968481e-06, + "loss": 0.3279542326927185, + "step": 1035 + }, + { + "epoch": 1.0147040803823062, + "grad_norm": 0.3867848217487335, + "learning_rate": 5.134849998052123e-06, + "loss": 0.4891187846660614, + "step": 1036 + }, + { + "epoch": 1.0156843524077932, + "grad_norm": 0.39323440194129944, + "learning_rate": 5.126919402651053e-06, + "loss": 0.43106579780578613, + "step": 1037 + }, + { + "epoch": 1.0166646244332802, + "grad_norm": 0.39069345593452454, + "learning_rate": 5.118988487730537e-06, + "loss": 0.49026423692703247, + "step": 1038 + }, + { + "epoch": 1.0176448964587672, + "grad_norm": 0.4325862228870392, + "learning_rate": 5.111057273256648e-06, + "loss": 0.48113250732421875, + "step": 1039 + }, + { + "epoch": 1.0186251684842544, + "grad_norm": 0.3621155023574829, + "learning_rate": 5.103125779196204e-06, + "loss": 0.4307592511177063, + "step": 1040 + }, + { + "epoch": 1.0196054405097414, + "grad_norm": 0.3972395956516266, + "learning_rate": 5.095194025516733e-06, + "loss": 0.4181487560272217, + "step": 1041 + }, + { + "epoch": 1.0205857125352285, + "grad_norm": 0.4484376311302185, + "learning_rate": 5.087262032186418e-06, + "loss": 0.49258244037628174, + "step": 1042 + }, + { + "epoch": 1.0215659845607157, + "grad_norm": 0.43103304505348206, + "learning_rate": 5.07932981917404e-06, + "loss": 0.4145815074443817, + "step": 1043 + }, + { + "epoch": 1.0225462565862027, + "grad_norm": 0.37276118993759155, + "learning_rate": 5.071397406448937e-06, + "loss": 0.41032058000564575, + "step": 1044 + }, + { + "epoch": 1.0235265286116897, + "grad_norm": 0.39688998460769653, + "learning_rate": 5.063464813980948e-06, + "loss": 0.4077647030353546, + "step": 1045 + }, + { + "epoch": 1.024506800637177, + "grad_norm": 0.3730168640613556, + "learning_rate": 5.055532061740367e-06, + "loss": 0.3442533612251282, + "step": 1046 + }, + { + "epoch": 1.025487072662664, + "grad_norm": 0.40458160638809204, + "learning_rate": 5.0475991696978844e-06, + "loss": 0.42443233728408813, + "step": 1047 + }, + { + "epoch": 1.026467344688151, + "grad_norm": 0.38275137543678284, + "learning_rate": 5.039666157824549e-06, + "loss": 0.4137828052043915, + "step": 1048 + }, + { + "epoch": 1.027447616713638, + "grad_norm": 0.38976171612739563, + "learning_rate": 5.03173304609171e-06, + "loss": 0.39616531133651733, + "step": 1049 + }, + { + "epoch": 1.0284278887391252, + "grad_norm": 0.3985825181007385, + "learning_rate": 5.023799854470963e-06, + "loss": 0.45660901069641113, + "step": 1050 + }, + { + "epoch": 1.0294081607646122, + "grad_norm": 0.3173740804195404, + "learning_rate": 5.015866602934112e-06, + "loss": 0.3417515158653259, + "step": 1051 + }, + { + "epoch": 1.0303884327900992, + "grad_norm": 0.48825255036354065, + "learning_rate": 5.007933311453108e-06, + "loss": 0.5318125486373901, + "step": 1052 + }, + { + "epoch": 1.0313687048155864, + "grad_norm": 0.4049169719219208, + "learning_rate": 5e-06, + "loss": 0.43396246433258057, + "step": 1053 + }, + { + "epoch": 1.0323489768410734, + "grad_norm": 0.3945798873901367, + "learning_rate": 4.992066688546895e-06, + "loss": 0.4253767132759094, + "step": 1054 + }, + { + "epoch": 1.0333292488665604, + "grad_norm": 0.4103344976902008, + "learning_rate": 4.984133397065889e-06, + "loss": 0.40531277656555176, + "step": 1055 + }, + { + "epoch": 1.0343095208920476, + "grad_norm": 0.3595251441001892, + "learning_rate": 4.976200145529039e-06, + "loss": 0.37361353635787964, + "step": 1056 + }, + { + "epoch": 1.0352897929175346, + "grad_norm": 0.3990113437175751, + "learning_rate": 4.9682669539082914e-06, + "loss": 0.4884231686592102, + "step": 1057 + }, + { + "epoch": 1.0362700649430217, + "grad_norm": 0.31987398862838745, + "learning_rate": 4.960333842175453e-06, + "loss": 0.34926849603652954, + "step": 1058 + }, + { + "epoch": 1.0372503369685087, + "grad_norm": 0.46583428978919983, + "learning_rate": 4.952400830302117e-06, + "loss": 0.47854527831077576, + "step": 1059 + }, + { + "epoch": 1.0382306089939959, + "grad_norm": 0.3727846145629883, + "learning_rate": 4.944467938259636e-06, + "loss": 0.315160870552063, + "step": 1060 + }, + { + "epoch": 1.039210881019483, + "grad_norm": 0.3596564531326294, + "learning_rate": 4.936535186019053e-06, + "loss": 0.36465609073638916, + "step": 1061 + }, + { + "epoch": 1.04019115304497, + "grad_norm": 0.46648547053337097, + "learning_rate": 4.928602593551065e-06, + "loss": 0.47904151678085327, + "step": 1062 + }, + { + "epoch": 1.0411714250704571, + "grad_norm": 0.39216434955596924, + "learning_rate": 4.9206701808259605e-06, + "loss": 0.3930017948150635, + "step": 1063 + }, + { + "epoch": 1.0421516970959441, + "grad_norm": 0.4563177824020386, + "learning_rate": 4.9127379678135825e-06, + "loss": 0.4542958736419678, + "step": 1064 + }, + { + "epoch": 1.0431319691214311, + "grad_norm": 0.44150593876838684, + "learning_rate": 4.904805974483267e-06, + "loss": 0.46942785382270813, + "step": 1065 + }, + { + "epoch": 1.0441122411469184, + "grad_norm": 0.3670433759689331, + "learning_rate": 4.8968742208037976e-06, + "loss": 0.3405688405036926, + "step": 1066 + }, + { + "epoch": 1.0450925131724054, + "grad_norm": 0.30620449781417847, + "learning_rate": 4.888942726743353e-06, + "loss": 0.31899744272232056, + "step": 1067 + }, + { + "epoch": 1.0460727851978924, + "grad_norm": 0.49040576815605164, + "learning_rate": 4.881011512269464e-06, + "loss": 0.6267503499984741, + "step": 1068 + }, + { + "epoch": 1.0470530572233794, + "grad_norm": 0.3658970296382904, + "learning_rate": 4.873080597348948e-06, + "loss": 0.36484646797180176, + "step": 1069 + }, + { + "epoch": 1.0480333292488666, + "grad_norm": 0.3933736979961395, + "learning_rate": 4.86515000194788e-06, + "loss": 0.427493691444397, + "step": 1070 + }, + { + "epoch": 1.0490136012743536, + "grad_norm": 0.4736839532852173, + "learning_rate": 4.85721974603152e-06, + "loss": 0.5376136302947998, + "step": 1071 + }, + { + "epoch": 1.0499938732998406, + "grad_norm": 0.38101696968078613, + "learning_rate": 4.849289849564278e-06, + "loss": 0.39923804998397827, + "step": 1072 + }, + { + "epoch": 1.0509741453253278, + "grad_norm": 0.42541632056236267, + "learning_rate": 4.841360332509663e-06, + "loss": 0.36699429154396057, + "step": 1073 + }, + { + "epoch": 1.0519544173508149, + "grad_norm": 0.42379432916641235, + "learning_rate": 4.83343121483022e-06, + "loss": 0.4543285667896271, + "step": 1074 + }, + { + "epoch": 1.0529346893763019, + "grad_norm": 0.5372243523597717, + "learning_rate": 4.825502516487497e-06, + "loss": 0.4559112787246704, + "step": 1075 + }, + { + "epoch": 1.053914961401789, + "grad_norm": 0.4589729905128479, + "learning_rate": 4.817574257441978e-06, + "loss": 0.4061373472213745, + "step": 1076 + }, + { + "epoch": 1.054895233427276, + "grad_norm": 0.3967716097831726, + "learning_rate": 4.809646457653051e-06, + "loss": 0.3934286832809448, + "step": 1077 + }, + { + "epoch": 1.055875505452763, + "grad_norm": 0.35180771350860596, + "learning_rate": 4.801719137078939e-06, + "loss": 0.3477855920791626, + "step": 1078 + }, + { + "epoch": 1.0568557774782503, + "grad_norm": 0.4047290086746216, + "learning_rate": 4.793792315676665e-06, + "loss": 0.4579012393951416, + "step": 1079 + }, + { + "epoch": 1.0578360495037373, + "grad_norm": 0.39317500591278076, + "learning_rate": 4.785866013401989e-06, + "loss": 0.40592679381370544, + "step": 1080 + }, + { + "epoch": 1.0588163215292243, + "grad_norm": 0.3951299786567688, + "learning_rate": 4.777940250209369e-06, + "loss": 0.4006436765193939, + "step": 1081 + }, + { + "epoch": 1.0597965935547113, + "grad_norm": 0.4077007472515106, + "learning_rate": 4.770015046051904e-06, + "loss": 0.45638835430145264, + "step": 1082 + }, + { + "epoch": 1.0607768655801986, + "grad_norm": 0.34083035588264465, + "learning_rate": 4.762090420881289e-06, + "loss": 0.39168232679367065, + "step": 1083 + }, + { + "epoch": 1.0617571376056856, + "grad_norm": 0.3833742141723633, + "learning_rate": 4.754166394647755e-06, + "loss": 0.3773995637893677, + "step": 1084 + }, + { + "epoch": 1.0627374096311726, + "grad_norm": 0.3877207040786743, + "learning_rate": 4.74624298730003e-06, + "loss": 0.41565343737602234, + "step": 1085 + }, + { + "epoch": 1.0637176816566598, + "grad_norm": 0.44384750723838806, + "learning_rate": 4.738320218785281e-06, + "loss": 0.46749821305274963, + "step": 1086 + }, + { + "epoch": 1.0646979536821468, + "grad_norm": 0.3572728931903839, + "learning_rate": 4.7303981090490715e-06, + "loss": 0.4305860102176666, + "step": 1087 + }, + { + "epoch": 1.0656782257076338, + "grad_norm": 0.367789089679718, + "learning_rate": 4.7224766780353005e-06, + "loss": 0.37522345781326294, + "step": 1088 + }, + { + "epoch": 1.0666584977331208, + "grad_norm": 0.379123717546463, + "learning_rate": 4.71455594568616e-06, + "loss": 0.4396986961364746, + "step": 1089 + }, + { + "epoch": 1.067638769758608, + "grad_norm": 0.3493232727050781, + "learning_rate": 4.706635931942089e-06, + "loss": 0.3614601492881775, + "step": 1090 + }, + { + "epoch": 1.068619041784095, + "grad_norm": 0.3911818265914917, + "learning_rate": 4.6987166567417085e-06, + "loss": 0.43483102321624756, + "step": 1091 + }, + { + "epoch": 1.069599313809582, + "grad_norm": 0.4138401746749878, + "learning_rate": 4.690798140021787e-06, + "loss": 0.3496001362800598, + "step": 1092 + }, + { + "epoch": 1.0705795858350693, + "grad_norm": 0.40637072920799255, + "learning_rate": 4.682880401717178e-06, + "loss": 0.4135543704032898, + "step": 1093 + }, + { + "epoch": 1.0715598578605563, + "grad_norm": 0.36067408323287964, + "learning_rate": 4.674963461760782e-06, + "loss": 0.3597530722618103, + "step": 1094 + }, + { + "epoch": 1.0725401298860433, + "grad_norm": 0.3889940679073334, + "learning_rate": 4.667047340083481e-06, + "loss": 0.37421876192092896, + "step": 1095 + }, + { + "epoch": 1.0735204019115305, + "grad_norm": 0.44483500719070435, + "learning_rate": 4.659132056614108e-06, + "loss": 0.46453529596328735, + "step": 1096 + }, + { + "epoch": 1.0745006739370175, + "grad_norm": 0.3886990547180176, + "learning_rate": 4.651217631279374e-06, + "loss": 0.4254445433616638, + "step": 1097 + }, + { + "epoch": 1.0754809459625045, + "grad_norm": 0.31164583563804626, + "learning_rate": 4.643304084003839e-06, + "loss": 0.2881056070327759, + "step": 1098 + }, + { + "epoch": 1.0764612179879918, + "grad_norm": 0.8180178999900818, + "learning_rate": 4.635391434709847e-06, + "loss": 0.4005112051963806, + "step": 1099 + }, + { + "epoch": 1.0774414900134788, + "grad_norm": 0.4170231819152832, + "learning_rate": 4.627479703317485e-06, + "loss": 0.4223352372646332, + "step": 1100 + }, + { + "epoch": 1.0784217620389658, + "grad_norm": 0.3896201252937317, + "learning_rate": 4.619568909744524e-06, + "loss": 0.38746267557144165, + "step": 1101 + }, + { + "epoch": 1.0794020340644528, + "grad_norm": 0.37979328632354736, + "learning_rate": 4.611659073906382e-06, + "loss": 0.4118471145629883, + "step": 1102 + }, + { + "epoch": 1.08038230608994, + "grad_norm": 0.38320398330688477, + "learning_rate": 4.603750215716057e-06, + "loss": 0.3374844789505005, + "step": 1103 + }, + { + "epoch": 1.081362578115427, + "grad_norm": 0.36406081914901733, + "learning_rate": 4.595842355084094e-06, + "loss": 0.39088350534439087, + "step": 1104 + }, + { + "epoch": 1.082342850140914, + "grad_norm": 0.39569252729415894, + "learning_rate": 4.587935511918521e-06, + "loss": 0.44276106357574463, + "step": 1105 + }, + { + "epoch": 1.0833231221664013, + "grad_norm": 0.3960108458995819, + "learning_rate": 4.580029706124804e-06, + "loss": 0.40090763568878174, + "step": 1106 + }, + { + "epoch": 1.0843033941918883, + "grad_norm": 0.4160557687282562, + "learning_rate": 4.572124957605803e-06, + "loss": 0.40512001514434814, + "step": 1107 + }, + { + "epoch": 1.0852836662173753, + "grad_norm": 0.35392165184020996, + "learning_rate": 4.564221286261709e-06, + "loss": 0.36655810475349426, + "step": 1108 + }, + { + "epoch": 1.0862639382428625, + "grad_norm": 0.36522161960601807, + "learning_rate": 4.55631871199001e-06, + "loss": 0.3843870759010315, + "step": 1109 + }, + { + "epoch": 1.0872442102683495, + "grad_norm": 0.43062299489974976, + "learning_rate": 4.548417254685425e-06, + "loss": 0.5818829536437988, + "step": 1110 + }, + { + "epoch": 1.0882244822938365, + "grad_norm": 0.39051923155784607, + "learning_rate": 4.5405169342398634e-06, + "loss": 0.3421304225921631, + "step": 1111 + }, + { + "epoch": 1.0892047543193235, + "grad_norm": 0.3609357476234436, + "learning_rate": 4.5326177705423725e-06, + "loss": 0.35731297731399536, + "step": 1112 + }, + { + "epoch": 1.0901850263448107, + "grad_norm": 0.3642447292804718, + "learning_rate": 4.524719783479088e-06, + "loss": 0.43409985303878784, + "step": 1113 + }, + { + "epoch": 1.0911652983702977, + "grad_norm": 0.4051709771156311, + "learning_rate": 4.5168229929331805e-06, + "loss": 0.427818238735199, + "step": 1114 + }, + { + "epoch": 1.0921455703957847, + "grad_norm": 0.37739044427871704, + "learning_rate": 4.5089274187848144e-06, + "loss": 0.3848983645439148, + "step": 1115 + }, + { + "epoch": 1.093125842421272, + "grad_norm": 0.42955872416496277, + "learning_rate": 4.501033080911087e-06, + "loss": 0.49801453948020935, + "step": 1116 + }, + { + "epoch": 1.094106114446759, + "grad_norm": 0.43234995007514954, + "learning_rate": 4.493139999185984e-06, + "loss": 0.3880156874656677, + "step": 1117 + }, + { + "epoch": 1.095086386472246, + "grad_norm": 0.40927213430404663, + "learning_rate": 4.485248193480328e-06, + "loss": 0.4579135775566101, + "step": 1118 + }, + { + "epoch": 1.0960666584977332, + "grad_norm": 0.4949118494987488, + "learning_rate": 4.477357683661734e-06, + "loss": 0.5130608081817627, + "step": 1119 + }, + { + "epoch": 1.0970469305232202, + "grad_norm": 0.40647363662719727, + "learning_rate": 4.469468489594546e-06, + "loss": 0.3638862073421478, + "step": 1120 + }, + { + "epoch": 1.0980272025487072, + "grad_norm": 0.3945309519767761, + "learning_rate": 4.461580631139806e-06, + "loss": 0.40762943029403687, + "step": 1121 + }, + { + "epoch": 1.0990074745741945, + "grad_norm": 0.3408013582229614, + "learning_rate": 4.453694128155187e-06, + "loss": 0.37056824564933777, + "step": 1122 + }, + { + "epoch": 1.0999877465996815, + "grad_norm": 0.37208259105682373, + "learning_rate": 4.445809000494945e-06, + "loss": 0.4105289876461029, + "step": 1123 + }, + { + "epoch": 1.1009680186251685, + "grad_norm": 0.37705376744270325, + "learning_rate": 4.4379252680098875e-06, + "loss": 0.4101429581642151, + "step": 1124 + }, + { + "epoch": 1.1019482906506555, + "grad_norm": 0.405033677816391, + "learning_rate": 4.430042950547298e-06, + "loss": 0.43087536096572876, + "step": 1125 + }, + { + "epoch": 1.1029285626761427, + "grad_norm": 0.3684760630130768, + "learning_rate": 4.4221620679509016e-06, + "loss": 0.3694723844528198, + "step": 1126 + }, + { + "epoch": 1.1039088347016297, + "grad_norm": 0.34156399965286255, + "learning_rate": 4.414282640060809e-06, + "loss": 0.3670724630355835, + "step": 1127 + }, + { + "epoch": 1.1048891067271167, + "grad_norm": 0.35699546337127686, + "learning_rate": 4.4064046867134755e-06, + "loss": 0.3098832368850708, + "step": 1128 + }, + { + "epoch": 1.105869378752604, + "grad_norm": 0.34710371494293213, + "learning_rate": 4.398528227741634e-06, + "loss": 0.35312020778656006, + "step": 1129 + }, + { + "epoch": 1.106849650778091, + "grad_norm": 0.344389408826828, + "learning_rate": 4.390653282974264e-06, + "loss": 0.3646032214164734, + "step": 1130 + }, + { + "epoch": 1.107829922803578, + "grad_norm": 0.40715762972831726, + "learning_rate": 4.382779872236527e-06, + "loss": 0.4308096766471863, + "step": 1131 + }, + { + "epoch": 1.108810194829065, + "grad_norm": 0.38126593828201294, + "learning_rate": 4.374908015349728e-06, + "loss": 0.41913601756095886, + "step": 1132 + }, + { + "epoch": 1.1097904668545522, + "grad_norm": 0.3499723970890045, + "learning_rate": 4.367037732131254e-06, + "loss": 0.3470945358276367, + "step": 1133 + }, + { + "epoch": 1.1107707388800392, + "grad_norm": 0.33410918712615967, + "learning_rate": 4.359169042394537e-06, + "loss": 0.37115544080734253, + "step": 1134 + }, + { + "epoch": 1.1117510109055262, + "grad_norm": 0.4114793539047241, + "learning_rate": 4.3513019659489916e-06, + "loss": 0.44457772374153137, + "step": 1135 + }, + { + "epoch": 1.1127312829310134, + "grad_norm": 0.3794882595539093, + "learning_rate": 4.343436522599975e-06, + "loss": 0.4111524224281311, + "step": 1136 + }, + { + "epoch": 1.1137115549565004, + "grad_norm": 0.38420233130455017, + "learning_rate": 4.33557273214873e-06, + "loss": 0.41278114914894104, + "step": 1137 + }, + { + "epoch": 1.1146918269819874, + "grad_norm": 0.39566460251808167, + "learning_rate": 4.327710614392341e-06, + "loss": 0.37005168199539185, + "step": 1138 + }, + { + "epoch": 1.1156720990074747, + "grad_norm": 0.3738879859447479, + "learning_rate": 4.319850189123681e-06, + "loss": 0.3486557602882385, + "step": 1139 + }, + { + "epoch": 1.1166523710329617, + "grad_norm": 0.3536122143268585, + "learning_rate": 4.3119914761313566e-06, + "loss": 0.3847700357437134, + "step": 1140 + }, + { + "epoch": 1.1176326430584487, + "grad_norm": 0.33755171298980713, + "learning_rate": 4.304134495199675e-06, + "loss": 0.3240794837474823, + "step": 1141 + }, + { + "epoch": 1.118612915083936, + "grad_norm": 0.3523722290992737, + "learning_rate": 4.296279266108569e-06, + "loss": 0.4163820445537567, + "step": 1142 + }, + { + "epoch": 1.119593187109423, + "grad_norm": 0.37797531485557556, + "learning_rate": 4.2884258086335755e-06, + "loss": 0.39504697918891907, + "step": 1143 + }, + { + "epoch": 1.12057345913491, + "grad_norm": 0.3944871425628662, + "learning_rate": 4.28057414254576e-06, + "loss": 0.47559815645217896, + "step": 1144 + }, + { + "epoch": 1.121553731160397, + "grad_norm": 0.4183574914932251, + "learning_rate": 4.272724287611684e-06, + "loss": 0.4353412091732025, + "step": 1145 + }, + { + "epoch": 1.1225340031858841, + "grad_norm": 0.40189674496650696, + "learning_rate": 4.264876263593347e-06, + "loss": 0.403361976146698, + "step": 1146 + }, + { + "epoch": 1.1235142752113711, + "grad_norm": 0.3977178633213043, + "learning_rate": 4.257030090248142e-06, + "loss": 0.4541288912296295, + "step": 1147 + }, + { + "epoch": 1.1244945472368582, + "grad_norm": 0.39345574378967285, + "learning_rate": 4.249185787328798e-06, + "loss": 0.347665011882782, + "step": 1148 + }, + { + "epoch": 1.1254748192623454, + "grad_norm": 0.40523380041122437, + "learning_rate": 4.241343374583343e-06, + "loss": 0.4029330611228943, + "step": 1149 + }, + { + "epoch": 1.1264550912878324, + "grad_norm": 0.4071088433265686, + "learning_rate": 4.233502871755036e-06, + "loss": 0.4457220733165741, + "step": 1150 + }, + { + "epoch": 1.1274353633133194, + "grad_norm": 0.35133352875709534, + "learning_rate": 4.225664298582339e-06, + "loss": 0.37571749091148376, + "step": 1151 + }, + { + "epoch": 1.1284156353388064, + "grad_norm": 0.3751051425933838, + "learning_rate": 4.217827674798845e-06, + "loss": 0.4688645601272583, + "step": 1152 + }, + { + "epoch": 1.1293959073642936, + "grad_norm": 0.4077433943748474, + "learning_rate": 4.209993020133251e-06, + "loss": 0.42999720573425293, + "step": 1153 + }, + { + "epoch": 1.1303761793897806, + "grad_norm": 0.40395140647888184, + "learning_rate": 4.202160354309286e-06, + "loss": 0.42764806747436523, + "step": 1154 + }, + { + "epoch": 1.1313564514152676, + "grad_norm": 0.38254719972610474, + "learning_rate": 4.194329697045681e-06, + "loss": 0.3954124450683594, + "step": 1155 + }, + { + "epoch": 1.1323367234407549, + "grad_norm": 0.40836068987846375, + "learning_rate": 4.186501068056102e-06, + "loss": 0.4754643440246582, + "step": 1156 + }, + { + "epoch": 1.1333169954662419, + "grad_norm": 0.3315666615962982, + "learning_rate": 4.178674487049116e-06, + "loss": 0.38767218589782715, + "step": 1157 + }, + { + "epoch": 1.1342972674917289, + "grad_norm": 0.3553229570388794, + "learning_rate": 4.1708499737281305e-06, + "loss": 0.3861744701862335, + "step": 1158 + }, + { + "epoch": 1.135277539517216, + "grad_norm": 0.3296109139919281, + "learning_rate": 4.163027547791347e-06, + "loss": 0.3621108829975128, + "step": 1159 + }, + { + "epoch": 1.1362578115427031, + "grad_norm": 0.3422757685184479, + "learning_rate": 4.1552072289317185e-06, + "loss": 0.3197747468948364, + "step": 1160 + }, + { + "epoch": 1.1372380835681901, + "grad_norm": 0.3334837257862091, + "learning_rate": 4.147389036836881e-06, + "loss": 0.34771430492401123, + "step": 1161 + }, + { + "epoch": 1.1382183555936773, + "grad_norm": 0.4155479967594147, + "learning_rate": 4.139572991189131e-06, + "loss": 0.4543231725692749, + "step": 1162 + }, + { + "epoch": 1.1391986276191644, + "grad_norm": 0.37592974305152893, + "learning_rate": 4.131759111665349e-06, + "loss": 0.42315226793289185, + "step": 1163 + }, + { + "epoch": 1.1401788996446514, + "grad_norm": 0.4641445279121399, + "learning_rate": 4.12394741793697e-06, + "loss": 0.5089558362960815, + "step": 1164 + }, + { + "epoch": 1.1411591716701386, + "grad_norm": 0.39589712023735046, + "learning_rate": 4.116137929669921e-06, + "loss": 0.416471004486084, + "step": 1165 + }, + { + "epoch": 1.1421394436956256, + "grad_norm": 0.3632883131504059, + "learning_rate": 4.1083306665245855e-06, + "loss": 0.4456322491168976, + "step": 1166 + }, + { + "epoch": 1.1431197157211126, + "grad_norm": 0.3645889461040497, + "learning_rate": 4.100525648155731e-06, + "loss": 0.38634130358695984, + "step": 1167 + }, + { + "epoch": 1.1440999877465996, + "grad_norm": 0.533083438873291, + "learning_rate": 4.092722894212488e-06, + "loss": 0.5102528929710388, + "step": 1168 + }, + { + "epoch": 1.1450802597720868, + "grad_norm": 0.3875017762184143, + "learning_rate": 4.084922424338277e-06, + "loss": 0.4213959574699402, + "step": 1169 + }, + { + "epoch": 1.1460605317975738, + "grad_norm": 0.34545421600341797, + "learning_rate": 4.077124258170774e-06, + "loss": 0.38439345359802246, + "step": 1170 + }, + { + "epoch": 1.1470408038230608, + "grad_norm": 0.36993974447250366, + "learning_rate": 4.06932841534185e-06, + "loss": 0.34796881675720215, + "step": 1171 + }, + { + "epoch": 1.148021075848548, + "grad_norm": 0.5000345706939697, + "learning_rate": 4.061534915477533e-06, + "loss": 0.33618277311325073, + "step": 1172 + }, + { + "epoch": 1.149001347874035, + "grad_norm": 0.3687799274921417, + "learning_rate": 4.053743778197951e-06, + "loss": 0.38591301441192627, + "step": 1173 + }, + { + "epoch": 1.149981619899522, + "grad_norm": 0.4422888159751892, + "learning_rate": 4.045955023117276e-06, + "loss": 0.42434555292129517, + "step": 1174 + }, + { + "epoch": 1.150961891925009, + "grad_norm": 0.3722866475582123, + "learning_rate": 4.038168669843698e-06, + "loss": 0.4512460231781006, + "step": 1175 + }, + { + "epoch": 1.1519421639504963, + "grad_norm": 0.4258495569229126, + "learning_rate": 4.030384737979345e-06, + "loss": 0.43189001083374023, + "step": 1176 + }, + { + "epoch": 1.1529224359759833, + "grad_norm": 0.4360964596271515, + "learning_rate": 4.02260324712026e-06, + "loss": 0.38588055968284607, + "step": 1177 + }, + { + "epoch": 1.1539027080014703, + "grad_norm": 0.4099330008029938, + "learning_rate": 4.014824216856336e-06, + "loss": 0.4455612599849701, + "step": 1178 + }, + { + "epoch": 1.1548829800269576, + "grad_norm": 0.3808075189590454, + "learning_rate": 4.007047666771274e-06, + "loss": 0.41428738832473755, + "step": 1179 + }, + { + "epoch": 1.1558632520524446, + "grad_norm": 0.3736558258533478, + "learning_rate": 3.999273616442526e-06, + "loss": 0.3947542905807495, + "step": 1180 + }, + { + "epoch": 1.1568435240779316, + "grad_norm": 0.4050402343273163, + "learning_rate": 3.991502085441259e-06, + "loss": 0.3685680329799652, + "step": 1181 + }, + { + "epoch": 1.1578237961034188, + "grad_norm": 0.42905285954475403, + "learning_rate": 3.9837330933322895e-06, + "loss": 0.4300174117088318, + "step": 1182 + }, + { + "epoch": 1.1588040681289058, + "grad_norm": 0.3575042188167572, + "learning_rate": 3.975966659674048e-06, + "loss": 0.36900150775909424, + "step": 1183 + }, + { + "epoch": 1.1597843401543928, + "grad_norm": 0.3584974706172943, + "learning_rate": 3.968202804018519e-06, + "loss": 0.41339266300201416, + "step": 1184 + }, + { + "epoch": 1.16076461217988, + "grad_norm": 0.3421943485736847, + "learning_rate": 3.960441545911205e-06, + "loss": 0.36757180094718933, + "step": 1185 + }, + { + "epoch": 1.161744884205367, + "grad_norm": 0.36353015899658203, + "learning_rate": 3.952682904891058e-06, + "loss": 0.4084877073764801, + "step": 1186 + }, + { + "epoch": 1.162725156230854, + "grad_norm": 0.39983996748924255, + "learning_rate": 3.944926900490452e-06, + "loss": 0.4480135440826416, + "step": 1187 + }, + { + "epoch": 1.163705428256341, + "grad_norm": 0.45189017057418823, + "learning_rate": 3.937173552235117e-06, + "loss": 0.452054500579834, + "step": 1188 + }, + { + "epoch": 1.1646857002818283, + "grad_norm": 0.4069003462791443, + "learning_rate": 3.929422879644099e-06, + "loss": 0.4110977053642273, + "step": 1189 + }, + { + "epoch": 1.1656659723073153, + "grad_norm": 0.34311729669570923, + "learning_rate": 3.921674902229706e-06, + "loss": 0.34948086738586426, + "step": 1190 + }, + { + "epoch": 1.1666462443328023, + "grad_norm": 0.38811400532722473, + "learning_rate": 3.913929639497462e-06, + "loss": 0.36766964197158813, + "step": 1191 + }, + { + "epoch": 1.1676265163582895, + "grad_norm": 0.3650139570236206, + "learning_rate": 3.906187110946058e-06, + "loss": 0.3843083381652832, + "step": 1192 + }, + { + "epoch": 1.1686067883837765, + "grad_norm": 0.3309456706047058, + "learning_rate": 3.898447336067297e-06, + "loss": 0.34224626421928406, + "step": 1193 + }, + { + "epoch": 1.1695870604092635, + "grad_norm": 0.4625893831253052, + "learning_rate": 3.890710334346058e-06, + "loss": 0.47564953565597534, + "step": 1194 + }, + { + "epoch": 1.1705673324347505, + "grad_norm": 0.36542174220085144, + "learning_rate": 3.882976125260229e-06, + "loss": 0.3954586982727051, + "step": 1195 + }, + { + "epoch": 1.1715476044602378, + "grad_norm": 0.3621373474597931, + "learning_rate": 3.875244728280676e-06, + "loss": 0.4008944034576416, + "step": 1196 + }, + { + "epoch": 1.1725278764857248, + "grad_norm": 0.3834877610206604, + "learning_rate": 3.867516162871177e-06, + "loss": 0.38867902755737305, + "step": 1197 + }, + { + "epoch": 1.1735081485112118, + "grad_norm": 0.40423545241355896, + "learning_rate": 3.859790448488394e-06, + "loss": 0.4608898162841797, + "step": 1198 + }, + { + "epoch": 1.174488420536699, + "grad_norm": 0.39354994893074036, + "learning_rate": 3.8520676045817945e-06, + "loss": 0.4029826521873474, + "step": 1199 + }, + { + "epoch": 1.175468692562186, + "grad_norm": 0.45826849341392517, + "learning_rate": 3.844347650593635e-06, + "loss": 0.448000431060791, + "step": 1200 + }, + { + "epoch": 1.176448964587673, + "grad_norm": 0.3757564425468445, + "learning_rate": 3.8366306059588885e-06, + "loss": 0.36338678002357483, + "step": 1201 + }, + { + "epoch": 1.1774292366131602, + "grad_norm": 0.4462909996509552, + "learning_rate": 3.828916490105206e-06, + "loss": 0.4846561849117279, + "step": 1202 + }, + { + "epoch": 1.1784095086386472, + "grad_norm": 0.4407897889614105, + "learning_rate": 3.821205322452863e-06, + "loss": 0.4513360559940338, + "step": 1203 + }, + { + "epoch": 1.1793897806641342, + "grad_norm": 0.3902183473110199, + "learning_rate": 3.8134971224147204e-06, + "loss": 0.39645785093307495, + "step": 1204 + }, + { + "epoch": 1.1803700526896215, + "grad_norm": 0.573232114315033, + "learning_rate": 3.8057919093961554e-06, + "loss": 0.4645434021949768, + "step": 1205 + }, + { + "epoch": 1.1813503247151085, + "grad_norm": 0.3610806465148926, + "learning_rate": 3.798089702795038e-06, + "loss": 0.4019550085067749, + "step": 1206 + }, + { + "epoch": 1.1823305967405955, + "grad_norm": 0.6123276352882385, + "learning_rate": 3.790390522001662e-06, + "loss": 0.4320734739303589, + "step": 1207 + }, + { + "epoch": 1.1833108687660825, + "grad_norm": 0.36507129669189453, + "learning_rate": 3.782694386398706e-06, + "loss": 0.3751339316368103, + "step": 1208 + }, + { + "epoch": 1.1842911407915697, + "grad_norm": 0.4385579526424408, + "learning_rate": 3.775001315361183e-06, + "loss": 0.4628278613090515, + "step": 1209 + }, + { + "epoch": 1.1852714128170567, + "grad_norm": 0.3933520019054413, + "learning_rate": 3.7673113282563883e-06, + "loss": 0.4253647029399872, + "step": 1210 + }, + { + "epoch": 1.1862516848425437, + "grad_norm": 0.38493654131889343, + "learning_rate": 3.7596244444438577e-06, + "loss": 0.3740760087966919, + "step": 1211 + }, + { + "epoch": 1.187231956868031, + "grad_norm": 0.4316898584365845, + "learning_rate": 3.7519406832753085e-06, + "loss": 0.39512258768081665, + "step": 1212 + }, + { + "epoch": 1.188212228893518, + "grad_norm": 0.360851526260376, + "learning_rate": 3.7442600640946045e-06, + "loss": 0.3304922878742218, + "step": 1213 + }, + { + "epoch": 1.189192500919005, + "grad_norm": 0.36605045199394226, + "learning_rate": 3.7365826062376933e-06, + "loss": 0.3601117730140686, + "step": 1214 + }, + { + "epoch": 1.190172772944492, + "grad_norm": 0.3873301148414612, + "learning_rate": 3.7289083290325668e-06, + "loss": 0.4049808084964752, + "step": 1215 + }, + { + "epoch": 1.1911530449699792, + "grad_norm": 0.41103193163871765, + "learning_rate": 3.721237251799206e-06, + "loss": 0.43388375639915466, + "step": 1216 + }, + { + "epoch": 1.1921333169954662, + "grad_norm": 0.44179049134254456, + "learning_rate": 3.7135693938495433e-06, + "loss": 0.4327327311038971, + "step": 1217 + }, + { + "epoch": 1.1931135890209532, + "grad_norm": 0.3925487995147705, + "learning_rate": 3.705904774487396e-06, + "loss": 0.36374127864837646, + "step": 1218 + }, + { + "epoch": 1.1940938610464404, + "grad_norm": 0.3964214324951172, + "learning_rate": 3.69824341300844e-06, + "loss": 0.43925851583480835, + "step": 1219 + }, + { + "epoch": 1.1950741330719274, + "grad_norm": 0.42010238766670227, + "learning_rate": 3.6905853287001405e-06, + "loss": 0.4596717953681946, + "step": 1220 + }, + { + "epoch": 1.1960544050974145, + "grad_norm": 0.3509490489959717, + "learning_rate": 3.682930540841717e-06, + "loss": 0.35538655519485474, + "step": 1221 + }, + { + "epoch": 1.1970346771229017, + "grad_norm": 0.4000403881072998, + "learning_rate": 3.675279068704087e-06, + "loss": 0.43485313653945923, + "step": 1222 + }, + { + "epoch": 1.1980149491483887, + "grad_norm": 0.4635853171348572, + "learning_rate": 3.667630931549826e-06, + "loss": 0.40438783168792725, + "step": 1223 + }, + { + "epoch": 1.1989952211738757, + "grad_norm": 0.45477378368377686, + "learning_rate": 3.6599861486331074e-06, + "loss": 0.4430294334888458, + "step": 1224 + }, + { + "epoch": 1.199975493199363, + "grad_norm": 0.38992637395858765, + "learning_rate": 3.6523447391996613e-06, + "loss": 0.418027400970459, + "step": 1225 + }, + { + "epoch": 1.20095576522485, + "grad_norm": 0.37911441922187805, + "learning_rate": 3.6447067224867306e-06, + "loss": 0.37032103538513184, + "step": 1226 + }, + { + "epoch": 1.201936037250337, + "grad_norm": 0.4202483892440796, + "learning_rate": 3.637072117723012e-06, + "loss": 0.45127421617507935, + "step": 1227 + }, + { + "epoch": 1.2029163092758242, + "grad_norm": 0.4258260428905487, + "learning_rate": 3.6294409441286133e-06, + "loss": 0.47339314222335815, + "step": 1228 + }, + { + "epoch": 1.2038965813013112, + "grad_norm": 0.32870087027549744, + "learning_rate": 3.6218132209150047e-06, + "loss": 0.3060763478279114, + "step": 1229 + }, + { + "epoch": 1.2048768533267982, + "grad_norm": 0.3770845830440521, + "learning_rate": 3.6141889672849728e-06, + "loss": 0.42122548818588257, + "step": 1230 + }, + { + "epoch": 1.2058571253522852, + "grad_norm": 0.34395742416381836, + "learning_rate": 3.606568202432562e-06, + "loss": 0.3398696780204773, + "step": 1231 + }, + { + "epoch": 1.2068373973777724, + "grad_norm": 0.39930200576782227, + "learning_rate": 3.598950945543045e-06, + "loss": 0.4024962782859802, + "step": 1232 + }, + { + "epoch": 1.2078176694032594, + "grad_norm": 0.3859228491783142, + "learning_rate": 3.5913372157928515e-06, + "loss": 0.37723875045776367, + "step": 1233 + }, + { + "epoch": 1.2087979414287464, + "grad_norm": 0.43469709157943726, + "learning_rate": 3.5837270323495416e-06, + "loss": 0.4358142912387848, + "step": 1234 + }, + { + "epoch": 1.2097782134542336, + "grad_norm": 0.3631856143474579, + "learning_rate": 3.5761204143717387e-06, + "loss": 0.39562028646469116, + "step": 1235 + }, + { + "epoch": 1.2107584854797206, + "grad_norm": 0.41554635763168335, + "learning_rate": 3.568517381009099e-06, + "loss": 0.4201498031616211, + "step": 1236 + }, + { + "epoch": 1.2117387575052077, + "grad_norm": 0.35036760568618774, + "learning_rate": 3.560917951402245e-06, + "loss": 0.3424341082572937, + "step": 1237 + }, + { + "epoch": 1.2127190295306947, + "grad_norm": 0.4890834391117096, + "learning_rate": 3.553322144682737e-06, + "loss": 0.39817193150520325, + "step": 1238 + }, + { + "epoch": 1.2136993015561819, + "grad_norm": 0.37727776169776917, + "learning_rate": 3.5457299799730047e-06, + "loss": 0.376686692237854, + "step": 1239 + }, + { + "epoch": 1.214679573581669, + "grad_norm": 0.3657433092594147, + "learning_rate": 3.538141476386317e-06, + "loss": 0.40161749720573425, + "step": 1240 + }, + { + "epoch": 1.215659845607156, + "grad_norm": 0.3864190876483917, + "learning_rate": 3.5305566530267217e-06, + "loss": 0.4222371578216553, + "step": 1241 + }, + { + "epoch": 1.2166401176326431, + "grad_norm": 0.44883379340171814, + "learning_rate": 3.5229755289890005e-06, + "loss": 0.45256662368774414, + "step": 1242 + }, + { + "epoch": 1.2176203896581301, + "grad_norm": 0.4351153075695038, + "learning_rate": 3.5153981233586277e-06, + "loss": 0.40705186128616333, + "step": 1243 + }, + { + "epoch": 1.2186006616836171, + "grad_norm": 0.3470980226993561, + "learning_rate": 3.507824455211708e-06, + "loss": 0.3777850568294525, + "step": 1244 + }, + { + "epoch": 1.2195809337091044, + "grad_norm": 0.30768781900405884, + "learning_rate": 3.5002545436149478e-06, + "loss": 0.3007776439189911, + "step": 1245 + }, + { + "epoch": 1.2205612057345914, + "grad_norm": 0.39595335721969604, + "learning_rate": 3.4926884076255864e-06, + "loss": 0.37620678544044495, + "step": 1246 + }, + { + "epoch": 1.2215414777600784, + "grad_norm": 0.35394197702407837, + "learning_rate": 3.4851260662913643e-06, + "loss": 0.38582348823547363, + "step": 1247 + }, + { + "epoch": 1.2225217497855656, + "grad_norm": 0.4044612944126129, + "learning_rate": 3.477567538650466e-06, + "loss": 0.4379409849643707, + "step": 1248 + }, + { + "epoch": 1.2235020218110526, + "grad_norm": 0.45545586943626404, + "learning_rate": 3.470012843731476e-06, + "loss": 0.5053396821022034, + "step": 1249 + }, + { + "epoch": 1.2244822938365396, + "grad_norm": 0.30247926712036133, + "learning_rate": 3.4624620005533273e-06, + "loss": 0.2948305308818817, + "step": 1250 + }, + { + "epoch": 1.2254625658620266, + "grad_norm": 0.4061148762702942, + "learning_rate": 3.4549150281252635e-06, + "loss": 0.37848591804504395, + "step": 1251 + }, + { + "epoch": 1.2264428378875138, + "grad_norm": 0.35463598370552063, + "learning_rate": 3.4473719454467747e-06, + "loss": 0.3596639037132263, + "step": 1252 + }, + { + "epoch": 1.2274231099130009, + "grad_norm": 0.34395158290863037, + "learning_rate": 3.439832771507565e-06, + "loss": 0.3485071659088135, + "step": 1253 + }, + { + "epoch": 1.2284033819384879, + "grad_norm": 0.41445282101631165, + "learning_rate": 3.432297525287495e-06, + "loss": 0.43462860584259033, + "step": 1254 + }, + { + "epoch": 1.229383653963975, + "grad_norm": 0.3992398977279663, + "learning_rate": 3.4247662257565372e-06, + "loss": 0.41641998291015625, + "step": 1255 + }, + { + "epoch": 1.230363925989462, + "grad_norm": 0.3352434039115906, + "learning_rate": 3.4172388918747278e-06, + "loss": 0.3666713833808899, + "step": 1256 + }, + { + "epoch": 1.231344198014949, + "grad_norm": 0.35100722312927246, + "learning_rate": 3.4097155425921256e-06, + "loss": 0.383060097694397, + "step": 1257 + }, + { + "epoch": 1.232324470040436, + "grad_norm": 0.3855835795402527, + "learning_rate": 3.402196196848751e-06, + "loss": 0.4248864948749542, + "step": 1258 + }, + { + "epoch": 1.2333047420659233, + "grad_norm": 0.3790622353553772, + "learning_rate": 3.394680873574546e-06, + "loss": 0.3974880576133728, + "step": 1259 + }, + { + "epoch": 1.2342850140914103, + "grad_norm": 0.3722037374973297, + "learning_rate": 3.3871695916893314e-06, + "loss": 0.3850044012069702, + "step": 1260 + }, + { + "epoch": 1.2352652861168973, + "grad_norm": 0.42391470074653625, + "learning_rate": 3.3796623701027477e-06, + "loss": 0.3658878803253174, + "step": 1261 + }, + { + "epoch": 1.2362455581423846, + "grad_norm": 0.34987837076187134, + "learning_rate": 3.372159227714218e-06, + "loss": 0.4010200798511505, + "step": 1262 + }, + { + "epoch": 1.2372258301678716, + "grad_norm": 0.3850593864917755, + "learning_rate": 3.3646601834128924e-06, + "loss": 0.4039258360862732, + "step": 1263 + }, + { + "epoch": 1.2382061021933586, + "grad_norm": 0.3262411057949066, + "learning_rate": 3.357165256077608e-06, + "loss": 0.33648356795310974, + "step": 1264 + }, + { + "epoch": 1.2391863742188458, + "grad_norm": 0.3948027491569519, + "learning_rate": 3.349674464576834e-06, + "loss": 0.39538174867630005, + "step": 1265 + }, + { + "epoch": 1.2401666462443328, + "grad_norm": 0.3772190809249878, + "learning_rate": 3.3421878277686315e-06, + "loss": 0.40817028284072876, + "step": 1266 + }, + { + "epoch": 1.2411469182698198, + "grad_norm": 0.4274192452430725, + "learning_rate": 3.3347053645005965e-06, + "loss": 0.47100597620010376, + "step": 1267 + }, + { + "epoch": 1.242127190295307, + "grad_norm": 0.36291229724884033, + "learning_rate": 3.3272270936098243e-06, + "loss": 0.37679433822631836, + "step": 1268 + }, + { + "epoch": 1.243107462320794, + "grad_norm": 0.4662453234195709, + "learning_rate": 3.319753033922849e-06, + "loss": 0.41882896423339844, + "step": 1269 + }, + { + "epoch": 1.244087734346281, + "grad_norm": 0.3911527097225189, + "learning_rate": 3.3122832042556118e-06, + "loss": 0.40091946721076965, + "step": 1270 + }, + { + "epoch": 1.2450680063717683, + "grad_norm": 0.39950352907180786, + "learning_rate": 3.3048176234133967e-06, + "loss": 0.3761706054210663, + "step": 1271 + }, + { + "epoch": 1.2460482783972553, + "grad_norm": 0.392782062292099, + "learning_rate": 3.297356310190797e-06, + "loss": 0.38706642389297485, + "step": 1272 + }, + { + "epoch": 1.2470285504227423, + "grad_norm": 0.3568131625652313, + "learning_rate": 3.289899283371657e-06, + "loss": 0.38436463475227356, + "step": 1273 + }, + { + "epoch": 1.2480088224482293, + "grad_norm": 0.7199851274490356, + "learning_rate": 3.282446561729035e-06, + "loss": 0.3979560136795044, + "step": 1274 + }, + { + "epoch": 1.2489890944737165, + "grad_norm": 0.353189080953598, + "learning_rate": 3.274998164025148e-06, + "loss": 0.35999202728271484, + "step": 1275 + }, + { + "epoch": 1.2499693664992035, + "grad_norm": 0.4450392723083496, + "learning_rate": 3.267554109011327e-06, + "loss": 0.40954309701919556, + "step": 1276 + }, + { + "epoch": 1.2509496385246905, + "grad_norm": 0.3501782715320587, + "learning_rate": 3.260114415427975e-06, + "loss": 0.33778876066207886, + "step": 1277 + }, + { + "epoch": 1.2519299105501775, + "grad_norm": 0.3724333345890045, + "learning_rate": 3.252679102004509e-06, + "loss": 0.3930673599243164, + "step": 1278 + }, + { + "epoch": 1.2529101825756648, + "grad_norm": 0.47041675448417664, + "learning_rate": 3.2452481874593234e-06, + "loss": 0.44294336438179016, + "step": 1279 + }, + { + "epoch": 1.2538904546011518, + "grad_norm": 0.3746916651725769, + "learning_rate": 3.2378216904997345e-06, + "loss": 0.42315858602523804, + "step": 1280 + }, + { + "epoch": 1.2548707266266388, + "grad_norm": 0.6910121440887451, + "learning_rate": 3.230399629821942e-06, + "loss": 0.401492714881897, + "step": 1281 + }, + { + "epoch": 1.255850998652126, + "grad_norm": 0.4114987254142761, + "learning_rate": 3.2229820241109722e-06, + "loss": 0.4236416220664978, + "step": 1282 + }, + { + "epoch": 1.256831270677613, + "grad_norm": 0.3747204840183258, + "learning_rate": 3.2155688920406415e-06, + "loss": 0.3477349579334259, + "step": 1283 + }, + { + "epoch": 1.2578115427031, + "grad_norm": 0.4492729604244232, + "learning_rate": 3.2081602522734987e-06, + "loss": 0.45774856209754944, + "step": 1284 + }, + { + "epoch": 1.2587918147285873, + "grad_norm": 0.3943130075931549, + "learning_rate": 3.200756123460788e-06, + "loss": 0.43856897950172424, + "step": 1285 + }, + { + "epoch": 1.2597720867540743, + "grad_norm": 0.33616903424263, + "learning_rate": 3.1933565242423925e-06, + "loss": 0.3544430136680603, + "step": 1286 + }, + { + "epoch": 1.2607523587795613, + "grad_norm": 0.39423948526382446, + "learning_rate": 3.1859614732467957e-06, + "loss": 0.4907015264034271, + "step": 1287 + }, + { + "epoch": 1.2617326308050485, + "grad_norm": 0.44739484786987305, + "learning_rate": 3.178570989091028e-06, + "loss": 0.4467267394065857, + "step": 1288 + }, + { + "epoch": 1.2627129028305355, + "grad_norm": 0.3911536633968353, + "learning_rate": 3.171185090380628e-06, + "loss": 0.37290918827056885, + "step": 1289 + }, + { + "epoch": 1.2636931748560225, + "grad_norm": 0.4767346680164337, + "learning_rate": 3.1638037957095836e-06, + "loss": 0.38968420028686523, + "step": 1290 + }, + { + "epoch": 1.2646734468815097, + "grad_norm": 0.400234192609787, + "learning_rate": 3.156427123660297e-06, + "loss": 0.42811959981918335, + "step": 1291 + }, + { + "epoch": 1.2656537189069967, + "grad_norm": 0.40222156047821045, + "learning_rate": 3.1490550928035312e-06, + "loss": 0.3919750452041626, + "step": 1292 + }, + { + "epoch": 1.2666339909324837, + "grad_norm": 0.4121929705142975, + "learning_rate": 3.141687721698363e-06, + "loss": 0.3865625262260437, + "step": 1293 + }, + { + "epoch": 1.267614262957971, + "grad_norm": 0.3889275789260864, + "learning_rate": 3.1343250288921438e-06, + "loss": 0.3191291093826294, + "step": 1294 + }, + { + "epoch": 1.268594534983458, + "grad_norm": 0.5170485973358154, + "learning_rate": 3.12696703292044e-06, + "loss": 0.4491143524646759, + "step": 1295 + }, + { + "epoch": 1.269574807008945, + "grad_norm": 0.40000584721565247, + "learning_rate": 3.119613752307002e-06, + "loss": 0.41072505712509155, + "step": 1296 + }, + { + "epoch": 1.270555079034432, + "grad_norm": 0.4792482256889343, + "learning_rate": 3.1122652055637014e-06, + "loss": 0.40112292766571045, + "step": 1297 + }, + { + "epoch": 1.271535351059919, + "grad_norm": 0.45774105191230774, + "learning_rate": 3.104921411190499e-06, + "loss": 0.2819810211658478, + "step": 1298 + }, + { + "epoch": 1.2725156230854062, + "grad_norm": 0.3816893994808197, + "learning_rate": 3.097582387675385e-06, + "loss": 0.3803802728652954, + "step": 1299 + }, + { + "epoch": 1.2734958951108932, + "grad_norm": 0.3787786066532135, + "learning_rate": 3.0902481534943464e-06, + "loss": 0.39059048891067505, + "step": 1300 + }, + { + "epoch": 1.2744761671363802, + "grad_norm": 0.5583904385566711, + "learning_rate": 3.0829187271113035e-06, + "loss": 0.419122576713562, + "step": 1301 + }, + { + "epoch": 1.2754564391618675, + "grad_norm": 0.4614335000514984, + "learning_rate": 3.075594126978084e-06, + "loss": 0.4940609633922577, + "step": 1302 + }, + { + "epoch": 1.2764367111873545, + "grad_norm": 0.3942849040031433, + "learning_rate": 3.0682743715343565e-06, + "loss": 0.3276735544204712, + "step": 1303 + }, + { + "epoch": 1.2774169832128415, + "grad_norm": 0.39449965953826904, + "learning_rate": 3.0609594792076e-06, + "loss": 0.34671539068222046, + "step": 1304 + }, + { + "epoch": 1.2783972552383287, + "grad_norm": 0.3523140251636505, + "learning_rate": 3.053649468413043e-06, + "loss": 0.3847375512123108, + "step": 1305 + }, + { + "epoch": 1.2793775272638157, + "grad_norm": 0.5079805850982666, + "learning_rate": 3.0463443575536324e-06, + "loss": 0.45407336950302124, + "step": 1306 + }, + { + "epoch": 1.2803577992893027, + "grad_norm": 0.3784370422363281, + "learning_rate": 3.0390441650199727e-06, + "loss": 0.33036309480667114, + "step": 1307 + }, + { + "epoch": 1.28133807131479, + "grad_norm": 0.4176805913448334, + "learning_rate": 3.0317489091902936e-06, + "loss": 0.38364288210868835, + "step": 1308 + }, + { + "epoch": 1.282318343340277, + "grad_norm": 0.4283663034439087, + "learning_rate": 3.0244586084303908e-06, + "loss": 0.39900028705596924, + "step": 1309 + }, + { + "epoch": 1.283298615365764, + "grad_norm": 0.4573337435722351, + "learning_rate": 3.0171732810935887e-06, + "loss": 0.3670453727245331, + "step": 1310 + }, + { + "epoch": 1.2842788873912512, + "grad_norm": 0.40415993332862854, + "learning_rate": 3.0098929455206905e-06, + "loss": 0.3328857421875, + "step": 1311 + }, + { + "epoch": 1.2852591594167382, + "grad_norm": 0.42343538999557495, + "learning_rate": 3.00261762003993e-06, + "loss": 0.38732367753982544, + "step": 1312 + }, + { + "epoch": 1.2862394314422252, + "grad_norm": 0.42217183113098145, + "learning_rate": 2.995347322966933e-06, + "loss": 0.4529569745063782, + "step": 1313 + }, + { + "epoch": 1.2872197034677124, + "grad_norm": 0.6133620142936707, + "learning_rate": 2.9880820726046613e-06, + "loss": 0.36799007654190063, + "step": 1314 + }, + { + "epoch": 1.2881999754931994, + "grad_norm": 0.40401691198349, + "learning_rate": 2.980821887243377e-06, + "loss": 0.3898249864578247, + "step": 1315 + }, + { + "epoch": 1.2891802475186864, + "grad_norm": 0.36774176359176636, + "learning_rate": 2.9735667851605855e-06, + "loss": 0.3981729745864868, + "step": 1316 + }, + { + "epoch": 1.2901605195441734, + "grad_norm": 0.4208289682865143, + "learning_rate": 2.966316784621e-06, + "loss": 0.3395769000053406, + "step": 1317 + }, + { + "epoch": 1.2911407915696607, + "grad_norm": 0.4332103729248047, + "learning_rate": 2.959071903876486e-06, + "loss": 0.3669860363006592, + "step": 1318 + }, + { + "epoch": 1.2921210635951477, + "grad_norm": 0.3393762707710266, + "learning_rate": 2.951832161166024e-06, + "loss": 0.31675875186920166, + "step": 1319 + }, + { + "epoch": 1.2931013356206347, + "grad_norm": 0.5967795252799988, + "learning_rate": 2.9445975747156548e-06, + "loss": 0.468588650226593, + "step": 1320 + }, + { + "epoch": 1.2940816076461217, + "grad_norm": 0.38714200258255005, + "learning_rate": 2.937368162738445e-06, + "loss": 0.35217729210853577, + "step": 1321 + }, + { + "epoch": 1.295061879671609, + "grad_norm": 0.4572497308254242, + "learning_rate": 2.9301439434344275e-06, + "loss": 0.4509837031364441, + "step": 1322 + }, + { + "epoch": 1.296042151697096, + "grad_norm": 0.34165865182876587, + "learning_rate": 2.9229249349905686e-06, + "loss": 0.28333014249801636, + "step": 1323 + }, + { + "epoch": 1.297022423722583, + "grad_norm": 0.4069069027900696, + "learning_rate": 2.9157111555807103e-06, + "loss": 0.33724260330200195, + "step": 1324 + }, + { + "epoch": 1.2980026957480701, + "grad_norm": 0.6604207158088684, + "learning_rate": 2.9085026233655367e-06, + "loss": 0.3322385549545288, + "step": 1325 + }, + { + "epoch": 1.2989829677735572, + "grad_norm": 0.37773096561431885, + "learning_rate": 2.901299356492516e-06, + "loss": 0.37516629695892334, + "step": 1326 + }, + { + "epoch": 1.2999632397990442, + "grad_norm": 0.3801353871822357, + "learning_rate": 2.8941013730958674e-06, + "loss": 0.3868027627468109, + "step": 1327 + }, + { + "epoch": 1.3009435118245314, + "grad_norm": 0.36233529448509216, + "learning_rate": 2.886908691296504e-06, + "loss": 0.3630043566226959, + "step": 1328 + }, + { + "epoch": 1.3019237838500184, + "grad_norm": 0.3876889944076538, + "learning_rate": 2.8797213292019927e-06, + "loss": 0.36869075894355774, + "step": 1329 + }, + { + "epoch": 1.3029040558755054, + "grad_norm": 0.41262805461883545, + "learning_rate": 2.8725393049065098e-06, + "loss": 0.3855483829975128, + "step": 1330 + }, + { + "epoch": 1.3038843279009926, + "grad_norm": 0.3631833493709564, + "learning_rate": 2.8653626364907918e-06, + "loss": 0.38113847374916077, + "step": 1331 + }, + { + "epoch": 1.3048645999264796, + "grad_norm": 0.40543341636657715, + "learning_rate": 2.8581913420220954e-06, + "loss": 0.335469126701355, + "step": 1332 + }, + { + "epoch": 1.3058448719519666, + "grad_norm": 0.3715052008628845, + "learning_rate": 2.851025439554142e-06, + "loss": 0.40020814538002014, + "step": 1333 + }, + { + "epoch": 1.3068251439774539, + "grad_norm": 0.35445529222488403, + "learning_rate": 2.8438649471270876e-06, + "loss": 0.331275999546051, + "step": 1334 + }, + { + "epoch": 1.3078054160029409, + "grad_norm": 0.3255250155925751, + "learning_rate": 2.8367098827674575e-06, + "loss": 0.3754723072052002, + "step": 1335 + }, + { + "epoch": 1.3087856880284279, + "grad_norm": 0.3292457163333893, + "learning_rate": 2.829560264488126e-06, + "loss": 0.3226877450942993, + "step": 1336 + }, + { + "epoch": 1.3097659600539149, + "grad_norm": 0.4526783227920532, + "learning_rate": 2.82241611028824e-06, + "loss": 0.4872207045555115, + "step": 1337 + }, + { + "epoch": 1.310746232079402, + "grad_norm": 0.4097343385219574, + "learning_rate": 2.8152774381532033e-06, + "loss": 0.4654613137245178, + "step": 1338 + }, + { + "epoch": 1.3117265041048891, + "grad_norm": 0.34368985891342163, + "learning_rate": 2.8081442660546126e-06, + "loss": 0.3075721263885498, + "step": 1339 + }, + { + "epoch": 1.3127067761303761, + "grad_norm": 0.4095923900604248, + "learning_rate": 2.8010166119502238e-06, + "loss": 0.37104469537734985, + "step": 1340 + }, + { + "epoch": 1.3136870481558631, + "grad_norm": 0.3320257365703583, + "learning_rate": 2.7938944937838924e-06, + "loss": 0.30803996324539185, + "step": 1341 + }, + { + "epoch": 1.3146673201813504, + "grad_norm": 5.715288162231445, + "learning_rate": 2.7867779294855502e-06, + "loss": 0.3657503128051758, + "step": 1342 + }, + { + "epoch": 1.3156475922068374, + "grad_norm": 0.459286630153656, + "learning_rate": 2.7796669369711294e-06, + "loss": 0.4311758875846863, + "step": 1343 + }, + { + "epoch": 1.3166278642323244, + "grad_norm": 0.41369494795799255, + "learning_rate": 2.7725615341425526e-06, + "loss": 0.4160696268081665, + "step": 1344 + }, + { + "epoch": 1.3176081362578116, + "grad_norm": 0.3941971957683563, + "learning_rate": 2.7654617388876612e-06, + "loss": 0.45054078102111816, + "step": 1345 + }, + { + "epoch": 1.3185884082832986, + "grad_norm": 0.7816408276557922, + "learning_rate": 2.758367569080178e-06, + "loss": 0.40603122115135193, + "step": 1346 + }, + { + "epoch": 1.3195686803087856, + "grad_norm": 0.40323975682258606, + "learning_rate": 2.751279042579672e-06, + "loss": 0.4330912232398987, + "step": 1347 + }, + { + "epoch": 1.3205489523342728, + "grad_norm": 0.3292030692100525, + "learning_rate": 2.744196177231498e-06, + "loss": 0.2993565797805786, + "step": 1348 + }, + { + "epoch": 1.3215292243597598, + "grad_norm": 0.3749717175960541, + "learning_rate": 2.7371189908667604e-06, + "loss": 0.3840376138687134, + "step": 1349 + }, + { + "epoch": 1.3225094963852468, + "grad_norm": 0.3833976686000824, + "learning_rate": 2.7300475013022666e-06, + "loss": 0.36495548486709595, + "step": 1350 + }, + { + "epoch": 1.323489768410734, + "grad_norm": 0.402906596660614, + "learning_rate": 2.722981726340487e-06, + "loss": 0.39486148953437805, + "step": 1351 + }, + { + "epoch": 1.324470040436221, + "grad_norm": 0.5622661113739014, + "learning_rate": 2.715921683769496e-06, + "loss": 0.3580387830734253, + "step": 1352 + }, + { + "epoch": 1.325450312461708, + "grad_norm": 0.4067818522453308, + "learning_rate": 2.708867391362948e-06, + "loss": 0.358877956867218, + "step": 1353 + }, + { + "epoch": 1.3264305844871953, + "grad_norm": 0.40848594903945923, + "learning_rate": 2.7018188668800123e-06, + "loss": 0.4004408121109009, + "step": 1354 + }, + { + "epoch": 1.3274108565126823, + "grad_norm": 0.39419838786125183, + "learning_rate": 2.694776128065345e-06, + "loss": 0.33877044916152954, + "step": 1355 + }, + { + "epoch": 1.3283911285381693, + "grad_norm": 0.403003066778183, + "learning_rate": 2.687739192649026e-06, + "loss": 0.34709805250167847, + "step": 1356 + }, + { + "epoch": 1.3293714005636565, + "grad_norm": 0.35368582606315613, + "learning_rate": 2.6807080783465376e-06, + "loss": 0.33329468965530396, + "step": 1357 + }, + { + "epoch": 1.3303516725891436, + "grad_norm": 0.3720478117465973, + "learning_rate": 2.673682802858697e-06, + "loss": 0.3533756136894226, + "step": 1358 + }, + { + "epoch": 1.3313319446146306, + "grad_norm": 0.3895004391670227, + "learning_rate": 2.6666633838716317e-06, + "loss": 0.40257835388183594, + "step": 1359 + }, + { + "epoch": 1.3323122166401176, + "grad_norm": 0.3760802447795868, + "learning_rate": 2.659649839056716e-06, + "loss": 0.36246705055236816, + "step": 1360 + }, + { + "epoch": 1.3332924886656046, + "grad_norm": 0.376716673374176, + "learning_rate": 2.6526421860705474e-06, + "loss": 0.3696845769882202, + "step": 1361 + }, + { + "epoch": 1.3342727606910918, + "grad_norm": 0.34749388694763184, + "learning_rate": 2.645640442554878e-06, + "loss": 0.319829523563385, + "step": 1362 + }, + { + "epoch": 1.3352530327165788, + "grad_norm": 0.3609011769294739, + "learning_rate": 2.6386446261365874e-06, + "loss": 0.3443223834037781, + "step": 1363 + }, + { + "epoch": 1.3362333047420658, + "grad_norm": 0.3637268543243408, + "learning_rate": 2.63165475442764e-06, + "loss": 0.3592514395713806, + "step": 1364 + }, + { + "epoch": 1.337213576767553, + "grad_norm": 0.38370901346206665, + "learning_rate": 2.6246708450250256e-06, + "loss": 0.3944026231765747, + "step": 1365 + }, + { + "epoch": 1.33819384879304, + "grad_norm": 0.33777371048927307, + "learning_rate": 2.617692915510732e-06, + "loss": 0.3304780125617981, + "step": 1366 + }, + { + "epoch": 1.339174120818527, + "grad_norm": 0.38178232312202454, + "learning_rate": 2.6107209834516857e-06, + "loss": 0.37031328678131104, + "step": 1367 + }, + { + "epoch": 1.3401543928440143, + "grad_norm": 0.37555423378944397, + "learning_rate": 2.603755066399718e-06, + "loss": 0.3902512490749359, + "step": 1368 + }, + { + "epoch": 1.3411346648695013, + "grad_norm": 0.3644770681858063, + "learning_rate": 2.5967951818915137e-06, + "loss": 0.3631979823112488, + "step": 1369 + }, + { + "epoch": 1.3421149368949883, + "grad_norm": 0.40092551708221436, + "learning_rate": 2.5898413474485797e-06, + "loss": 0.3804783225059509, + "step": 1370 + }, + { + "epoch": 1.3430952089204755, + "grad_norm": 0.47878721356391907, + "learning_rate": 2.5828935805771804e-06, + "loss": 0.4761326313018799, + "step": 1371 + }, + { + "epoch": 1.3440754809459625, + "grad_norm": 0.3924635350704193, + "learning_rate": 2.5759518987683154e-06, + "loss": 0.34059667587280273, + "step": 1372 + }, + { + "epoch": 1.3450557529714495, + "grad_norm": 0.37239521741867065, + "learning_rate": 2.5690163194976576e-06, + "loss": 0.3563910126686096, + "step": 1373 + }, + { + "epoch": 1.3460360249969368, + "grad_norm": 0.41610437631607056, + "learning_rate": 2.56208686022552e-06, + "loss": 0.3744328022003174, + "step": 1374 + }, + { + "epoch": 1.3470162970224238, + "grad_norm": 0.3674508035182953, + "learning_rate": 2.5551635383968063e-06, + "loss": 0.3201541602611542, + "step": 1375 + }, + { + "epoch": 1.3479965690479108, + "grad_norm": 0.32904183864593506, + "learning_rate": 2.5482463714409744e-06, + "loss": 0.32733088731765747, + "step": 1376 + }, + { + "epoch": 1.348976841073398, + "grad_norm": 0.4079500436782837, + "learning_rate": 2.5413353767719805e-06, + "loss": 0.41550686955451965, + "step": 1377 + }, + { + "epoch": 1.349957113098885, + "grad_norm": 0.37391263246536255, + "learning_rate": 2.5344305717882487e-06, + "loss": 0.3100280165672302, + "step": 1378 + }, + { + "epoch": 1.350937385124372, + "grad_norm": 0.3452690839767456, + "learning_rate": 2.527531973872617e-06, + "loss": 0.3291139006614685, + "step": 1379 + }, + { + "epoch": 1.351917657149859, + "grad_norm": 0.4535479247570038, + "learning_rate": 2.520639600392295e-06, + "loss": 0.4195869565010071, + "step": 1380 + }, + { + "epoch": 1.3528979291753462, + "grad_norm": 0.3643161654472351, + "learning_rate": 2.5137534686988265e-06, + "loss": 0.431568443775177, + "step": 1381 + }, + { + "epoch": 1.3538782012008332, + "grad_norm": 0.3103673458099365, + "learning_rate": 2.5068735961280367e-06, + "loss": 0.29704877734184265, + "step": 1382 + }, + { + "epoch": 1.3548584732263202, + "grad_norm": 0.4200555086135864, + "learning_rate": 2.5000000000000015e-06, + "loss": 0.40711110830307007, + "step": 1383 + }, + { + "epoch": 1.3558387452518073, + "grad_norm": 0.34306541085243225, + "learning_rate": 2.493132697618986e-06, + "loss": 0.30961981415748596, + "step": 1384 + }, + { + "epoch": 1.3568190172772945, + "grad_norm": 0.4073183536529541, + "learning_rate": 2.486271706273421e-06, + "loss": 0.31677743792533875, + "step": 1385 + }, + { + "epoch": 1.3577992893027815, + "grad_norm": 0.3668540418148041, + "learning_rate": 2.4794170432358415e-06, + "loss": 0.3492218852043152, + "step": 1386 + }, + { + "epoch": 1.3587795613282685, + "grad_norm": 0.3954210579395294, + "learning_rate": 2.4725687257628533e-06, + "loss": 0.4060995578765869, + "step": 1387 + }, + { + "epoch": 1.3597598333537557, + "grad_norm": 0.39518460631370544, + "learning_rate": 2.465726771095086e-06, + "loss": 0.3773648738861084, + "step": 1388 + }, + { + "epoch": 1.3607401053792427, + "grad_norm": 0.37454745173454285, + "learning_rate": 2.4588911964571557e-06, + "loss": 0.36403530836105347, + "step": 1389 + }, + { + "epoch": 1.3617203774047297, + "grad_norm": 0.473150372505188, + "learning_rate": 2.4520620190576094e-06, + "loss": 0.42615050077438354, + "step": 1390 + }, + { + "epoch": 1.362700649430217, + "grad_norm": 0.3816130459308624, + "learning_rate": 2.445239256088898e-06, + "loss": 0.3653256297111511, + "step": 1391 + }, + { + "epoch": 1.363680921455704, + "grad_norm": 0.3447451591491699, + "learning_rate": 2.4384229247273154e-06, + "loss": 0.3344596028327942, + "step": 1392 + }, + { + "epoch": 1.364661193481191, + "grad_norm": 0.41519880294799805, + "learning_rate": 2.4316130421329696e-06, + "loss": 0.4016348123550415, + "step": 1393 + }, + { + "epoch": 1.3656414655066782, + "grad_norm": 0.3167952001094818, + "learning_rate": 2.424809625449729e-06, + "loss": 0.3495238423347473, + "step": 1394 + }, + { + "epoch": 1.3666217375321652, + "grad_norm": 0.39838922023773193, + "learning_rate": 2.418012691805191e-06, + "loss": 0.4294274151325226, + "step": 1395 + }, + { + "epoch": 1.3676020095576522, + "grad_norm": 0.4710291028022766, + "learning_rate": 2.411222258310626e-06, + "loss": 0.3705732822418213, + "step": 1396 + }, + { + "epoch": 1.3685822815831394, + "grad_norm": 0.44181695580482483, + "learning_rate": 2.404438342060941e-06, + "loss": 0.49499452114105225, + "step": 1397 + }, + { + "epoch": 1.3695625536086264, + "grad_norm": 0.38113605976104736, + "learning_rate": 2.3976609601346395e-06, + "loss": 0.3740078806877136, + "step": 1398 + }, + { + "epoch": 1.3705428256341134, + "grad_norm": 0.3366130590438843, + "learning_rate": 2.3908901295937713e-06, + "loss": 0.41622573137283325, + "step": 1399 + }, + { + "epoch": 1.3715230976596007, + "grad_norm": 0.3616192936897278, + "learning_rate": 2.3841258674838948e-06, + "loss": 0.38352787494659424, + "step": 1400 + }, + { + "epoch": 1.3725033696850877, + "grad_norm": 0.4455365836620331, + "learning_rate": 2.3773681908340284e-06, + "loss": 0.5421079397201538, + "step": 1401 + }, + { + "epoch": 1.3734836417105747, + "grad_norm": 0.3265042006969452, + "learning_rate": 2.3706171166566204e-06, + "loss": 0.35201209783554077, + "step": 1402 + }, + { + "epoch": 1.3744639137360617, + "grad_norm": 0.36978763341903687, + "learning_rate": 2.363872661947488e-06, + "loss": 0.4175504446029663, + "step": 1403 + }, + { + "epoch": 1.3754441857615487, + "grad_norm": 0.32741472125053406, + "learning_rate": 2.3571348436857906e-06, + "loss": 0.32883599400520325, + "step": 1404 + }, + { + "epoch": 1.376424457787036, + "grad_norm": 0.3596395254135132, + "learning_rate": 2.3504036788339763e-06, + "loss": 0.42533740401268005, + "step": 1405 + }, + { + "epoch": 1.377404729812523, + "grad_norm": 0.3268555998802185, + "learning_rate": 2.343679184337745e-06, + "loss": 0.3815903067588806, + "step": 1406 + }, + { + "epoch": 1.37838500183801, + "grad_norm": 0.40195927023887634, + "learning_rate": 2.3369613771260006e-06, + "loss": 0.4699534773826599, + "step": 1407 + }, + { + "epoch": 1.3793652738634972, + "grad_norm": 0.33817511796951294, + "learning_rate": 2.33025027411082e-06, + "loss": 0.34417587518692017, + "step": 1408 + }, + { + "epoch": 1.3803455458889842, + "grad_norm": 0.32887962460517883, + "learning_rate": 2.323545892187393e-06, + "loss": 0.3174667954444885, + "step": 1409 + }, + { + "epoch": 1.3813258179144712, + "grad_norm": 0.502831757068634, + "learning_rate": 2.3168482482339955e-06, + "loss": 0.4483807682991028, + "step": 1410 + }, + { + "epoch": 1.3823060899399584, + "grad_norm": 0.3988838791847229, + "learning_rate": 2.310157359111938e-06, + "loss": 0.38515228033065796, + "step": 1411 + }, + { + "epoch": 1.3832863619654454, + "grad_norm": 0.40593546628952026, + "learning_rate": 2.3034732416655247e-06, + "loss": 0.42112022638320923, + "step": 1412 + }, + { + "epoch": 1.3842666339909324, + "grad_norm": 0.37779366970062256, + "learning_rate": 2.296795912722014e-06, + "loss": 0.41379955410957336, + "step": 1413 + }, + { + "epoch": 1.3852469060164196, + "grad_norm": 0.3651103973388672, + "learning_rate": 2.2901253890915714e-06, + "loss": 0.3805370330810547, + "step": 1414 + }, + { + "epoch": 1.3862271780419066, + "grad_norm": 0.3557569086551666, + "learning_rate": 2.2834616875672362e-06, + "loss": 0.40311068296432495, + "step": 1415 + }, + { + "epoch": 1.3872074500673937, + "grad_norm": 0.3499234616756439, + "learning_rate": 2.2768048249248648e-06, + "loss": 0.34748905897140503, + "step": 1416 + }, + { + "epoch": 1.3881877220928809, + "grad_norm": 0.3618040978908539, + "learning_rate": 2.2701548179231048e-06, + "loss": 0.3492986559867859, + "step": 1417 + }, + { + "epoch": 1.3891679941183679, + "grad_norm": 0.41700074076652527, + "learning_rate": 2.263511683303339e-06, + "loss": 0.5027569532394409, + "step": 1418 + }, + { + "epoch": 1.390148266143855, + "grad_norm": 0.33532679080963135, + "learning_rate": 2.2568754377896516e-06, + "loss": 0.33285945653915405, + "step": 1419 + }, + { + "epoch": 1.3911285381693421, + "grad_norm": 0.4351573586463928, + "learning_rate": 2.2502460980887807e-06, + "loss": 0.4646742343902588, + "step": 1420 + }, + { + "epoch": 1.3921088101948291, + "grad_norm": 0.37762102484703064, + "learning_rate": 2.2436236808900846e-06, + "loss": 0.40934765338897705, + "step": 1421 + }, + { + "epoch": 1.3930890822203161, + "grad_norm": 0.3822619616985321, + "learning_rate": 2.2370082028654867e-06, + "loss": 0.4055747985839844, + "step": 1422 + }, + { + "epoch": 1.3940693542458031, + "grad_norm": 0.430327445268631, + "learning_rate": 2.230399680669449e-06, + "loss": 0.4070087671279907, + "step": 1423 + }, + { + "epoch": 1.3950496262712904, + "grad_norm": 0.40656566619873047, + "learning_rate": 2.223798130938916e-06, + "loss": 0.4123728275299072, + "step": 1424 + }, + { + "epoch": 1.3960298982967774, + "grad_norm": 0.4164937138557434, + "learning_rate": 2.2172035702932828e-06, + "loss": 0.4464244246482849, + "step": 1425 + }, + { + "epoch": 1.3970101703222644, + "grad_norm": 0.37057921290397644, + "learning_rate": 2.210616015334344e-06, + "loss": 0.38867858052253723, + "step": 1426 + }, + { + "epoch": 1.3979904423477514, + "grad_norm": 0.39764076471328735, + "learning_rate": 2.204035482646267e-06, + "loss": 0.3937212824821472, + "step": 1427 + }, + { + "epoch": 1.3989707143732386, + "grad_norm": 0.3571273386478424, + "learning_rate": 2.1974619887955294e-06, + "loss": 0.3573153614997864, + "step": 1428 + }, + { + "epoch": 1.3999509863987256, + "grad_norm": 0.3540017604827881, + "learning_rate": 2.190895550330899e-06, + "loss": 0.3838018774986267, + "step": 1429 + }, + { + "epoch": 1.4009312584242126, + "grad_norm": 0.42102551460266113, + "learning_rate": 2.1843361837833755e-06, + "loss": 0.39180850982666016, + "step": 1430 + }, + { + "epoch": 1.4019115304496998, + "grad_norm": 0.380798876285553, + "learning_rate": 2.1777839056661555e-06, + "loss": 0.40284842252731323, + "step": 1431 + }, + { + "epoch": 1.4028918024751869, + "grad_norm": 0.4725682735443115, + "learning_rate": 2.171238732474591e-06, + "loss": 0.5058016777038574, + "step": 1432 + }, + { + "epoch": 1.4038720745006739, + "grad_norm": 0.3676036298274994, + "learning_rate": 2.1647006806861472e-06, + "loss": 0.3790314495563507, + "step": 1433 + }, + { + "epoch": 1.404852346526161, + "grad_norm": 0.36220112442970276, + "learning_rate": 2.1581697667603634e-06, + "loss": 0.40718066692352295, + "step": 1434 + }, + { + "epoch": 1.405832618551648, + "grad_norm": 0.4111344516277313, + "learning_rate": 2.1516460071388062e-06, + "loss": 0.3712926506996155, + "step": 1435 + }, + { + "epoch": 1.406812890577135, + "grad_norm": 0.34678778052330017, + "learning_rate": 2.145129418245035e-06, + "loss": 0.39137139916419983, + "step": 1436 + }, + { + "epoch": 1.4077931626026223, + "grad_norm": 0.4294041097164154, + "learning_rate": 2.1386200164845527e-06, + "loss": 0.414194792509079, + "step": 1437 + }, + { + "epoch": 1.4087734346281093, + "grad_norm": 0.3771069049835205, + "learning_rate": 2.132117818244771e-06, + "loss": 0.41370606422424316, + "step": 1438 + }, + { + "epoch": 1.4097537066535963, + "grad_norm": 0.4165838658809662, + "learning_rate": 2.125622839894964e-06, + "loss": 0.3896535634994507, + "step": 1439 + }, + { + "epoch": 1.4107339786790836, + "grad_norm": 0.36013364791870117, + "learning_rate": 2.1191350977862362e-06, + "loss": 0.31934142112731934, + "step": 1440 + }, + { + "epoch": 1.4117142507045706, + "grad_norm": 0.3193138837814331, + "learning_rate": 2.1126546082514665e-06, + "loss": 0.32281625270843506, + "step": 1441 + }, + { + "epoch": 1.4126945227300576, + "grad_norm": 0.4018670618534088, + "learning_rate": 2.106181387605283e-06, + "loss": 0.4774879217147827, + "step": 1442 + }, + { + "epoch": 1.4136747947555446, + "grad_norm": 0.4102632403373718, + "learning_rate": 2.09971545214401e-06, + "loss": 0.40976428985595703, + "step": 1443 + }, + { + "epoch": 1.4146550667810318, + "grad_norm": 0.34307852387428284, + "learning_rate": 2.0932568181456314e-06, + "loss": 0.3434346616268158, + "step": 1444 + }, + { + "epoch": 1.4156353388065188, + "grad_norm": 0.3488242030143738, + "learning_rate": 2.086805501869749e-06, + "loss": 0.3222504258155823, + "step": 1445 + }, + { + "epoch": 1.4166156108320058, + "grad_norm": 0.3165755271911621, + "learning_rate": 2.080361519557548e-06, + "loss": 0.3054387867450714, + "step": 1446 + }, + { + "epoch": 1.4175958828574928, + "grad_norm": 0.3666515052318573, + "learning_rate": 2.073924887431744e-06, + "loss": 0.3628386855125427, + "step": 1447 + }, + { + "epoch": 1.41857615488298, + "grad_norm": 0.3163360059261322, + "learning_rate": 2.0674956216965484e-06, + "loss": 0.2675761878490448, + "step": 1448 + }, + { + "epoch": 1.419556426908467, + "grad_norm": 0.3186602294445038, + "learning_rate": 2.061073738537635e-06, + "loss": 0.34987667202949524, + "step": 1449 + }, + { + "epoch": 1.420536698933954, + "grad_norm": 0.38075146079063416, + "learning_rate": 2.054659254122084e-06, + "loss": 0.407789021730423, + "step": 1450 + }, + { + "epoch": 1.4215169709594413, + "grad_norm": 0.4103749096393585, + "learning_rate": 2.0482521845983522e-06, + "loss": 0.40918007493019104, + "step": 1451 + }, + { + "epoch": 1.4224972429849283, + "grad_norm": 0.32040518522262573, + "learning_rate": 2.0418525460962283e-06, + "loss": 0.28277337551116943, + "step": 1452 + }, + { + "epoch": 1.4234775150104153, + "grad_norm": 0.39736682176589966, + "learning_rate": 2.0354603547267985e-06, + "loss": 0.38686496019363403, + "step": 1453 + }, + { + "epoch": 1.4244577870359025, + "grad_norm": 0.37901315093040466, + "learning_rate": 2.029075626582393e-06, + "loss": 0.34147050976753235, + "step": 1454 + }, + { + "epoch": 1.4254380590613895, + "grad_norm": 0.4597582221031189, + "learning_rate": 2.0226983777365604e-06, + "loss": 0.4084995985031128, + "step": 1455 + }, + { + "epoch": 1.4264183310868765, + "grad_norm": 0.4245735704898834, + "learning_rate": 2.0163286242440155e-06, + "loss": 0.4759935140609741, + "step": 1456 + }, + { + "epoch": 1.4273986031123638, + "grad_norm": 0.2954086661338806, + "learning_rate": 2.009966382140606e-06, + "loss": 0.2945442497730255, + "step": 1457 + }, + { + "epoch": 1.4283788751378508, + "grad_norm": 0.40143081545829773, + "learning_rate": 2.0036116674432653e-06, + "loss": 0.43226438760757446, + "step": 1458 + }, + { + "epoch": 1.4293591471633378, + "grad_norm": 0.34295380115509033, + "learning_rate": 1.9972644961499853e-06, + "loss": 0.36782434582710266, + "step": 1459 + }, + { + "epoch": 1.430339419188825, + "grad_norm": 0.3968486785888672, + "learning_rate": 1.990924884239758e-06, + "loss": 0.4365782141685486, + "step": 1460 + }, + { + "epoch": 1.431319691214312, + "grad_norm": 0.34507662057876587, + "learning_rate": 1.9845928476725522e-06, + "loss": 0.3597962260246277, + "step": 1461 + }, + { + "epoch": 1.432299963239799, + "grad_norm": 0.47816431522369385, + "learning_rate": 1.9782684023892613e-06, + "loss": 0.3986159563064575, + "step": 1462 + }, + { + "epoch": 1.4332802352652863, + "grad_norm": 0.3591419756412506, + "learning_rate": 1.971951564311668e-06, + "loss": 0.34434974193573, + "step": 1463 + }, + { + "epoch": 1.4342605072907733, + "grad_norm": 0.3620143234729767, + "learning_rate": 1.9656423493424047e-06, + "loss": 0.33680909872055054, + "step": 1464 + }, + { + "epoch": 1.4352407793162603, + "grad_norm": 0.4480261206626892, + "learning_rate": 1.959340773364911e-06, + "loss": 0.4787750244140625, + "step": 1465 + }, + { + "epoch": 1.4362210513417473, + "grad_norm": 0.38664156198501587, + "learning_rate": 1.953046852243401e-06, + "loss": 0.3196617364883423, + "step": 1466 + }, + { + "epoch": 1.4372013233672343, + "grad_norm": 0.5035872459411621, + "learning_rate": 1.946760601822809e-06, + "loss": 0.3775079548358917, + "step": 1467 + }, + { + "epoch": 1.4381815953927215, + "grad_norm": 0.3946012556552887, + "learning_rate": 1.9404820379287677e-06, + "loss": 0.3474588990211487, + "step": 1468 + }, + { + "epoch": 1.4391618674182085, + "grad_norm": 0.38673529028892517, + "learning_rate": 1.9342111763675512e-06, + "loss": 0.3392299711704254, + "step": 1469 + }, + { + "epoch": 1.4401421394436955, + "grad_norm": 0.3585011661052704, + "learning_rate": 1.927948032926047e-06, + "loss": 0.36674612760543823, + "step": 1470 + }, + { + "epoch": 1.4411224114691827, + "grad_norm": 0.36866655945777893, + "learning_rate": 1.9216926233717087e-06, + "loss": 0.3938637971878052, + "step": 1471 + }, + { + "epoch": 1.4421026834946697, + "grad_norm": 0.42264842987060547, + "learning_rate": 1.915444963452527e-06, + "loss": 0.370591402053833, + "step": 1472 + }, + { + "epoch": 1.4430829555201568, + "grad_norm": 0.328696072101593, + "learning_rate": 1.9092050688969736e-06, + "loss": 0.34347325563430786, + "step": 1473 + }, + { + "epoch": 1.444063227545644, + "grad_norm": 0.3298667371273041, + "learning_rate": 1.9029729554139798e-06, + "loss": 0.32899415493011475, + "step": 1474 + }, + { + "epoch": 1.445043499571131, + "grad_norm": 0.4168875515460968, + "learning_rate": 1.8967486386928819e-06, + "loss": 0.4729446470737457, + "step": 1475 + }, + { + "epoch": 1.446023771596618, + "grad_norm": 0.35200831294059753, + "learning_rate": 1.89053213440339e-06, + "loss": 0.41604435443878174, + "step": 1476 + }, + { + "epoch": 1.4470040436221052, + "grad_norm": 0.3220856785774231, + "learning_rate": 1.8843234581955444e-06, + "loss": 0.3407359719276428, + "step": 1477 + }, + { + "epoch": 1.4479843156475922, + "grad_norm": 0.33618050813674927, + "learning_rate": 1.8781226256996838e-06, + "loss": 0.36941471695899963, + "step": 1478 + }, + { + "epoch": 1.4489645876730792, + "grad_norm": 0.33431369066238403, + "learning_rate": 1.8719296525263925e-06, + "loss": 0.35194844007492065, + "step": 1479 + }, + { + "epoch": 1.4499448596985665, + "grad_norm": 0.3651709258556366, + "learning_rate": 1.865744554266477e-06, + "loss": 0.37440216541290283, + "step": 1480 + }, + { + "epoch": 1.4509251317240535, + "grad_norm": 0.48525023460388184, + "learning_rate": 1.859567346490913e-06, + "loss": 0.4449937343597412, + "step": 1481 + }, + { + "epoch": 1.4519054037495405, + "grad_norm": 0.4373418986797333, + "learning_rate": 1.8533980447508138e-06, + "loss": 0.3450752794742584, + "step": 1482 + }, + { + "epoch": 1.4528856757750277, + "grad_norm": 0.39088544249534607, + "learning_rate": 1.8472366645773892e-06, + "loss": 0.32666605710983276, + "step": 1483 + }, + { + "epoch": 1.4538659478005147, + "grad_norm": 0.32912006974220276, + "learning_rate": 1.8410832214819046e-06, + "loss": 0.30559319257736206, + "step": 1484 + }, + { + "epoch": 1.4548462198260017, + "grad_norm": 0.3647792339324951, + "learning_rate": 1.8349377309556487e-06, + "loss": 0.3280768394470215, + "step": 1485 + }, + { + "epoch": 1.4558264918514887, + "grad_norm": 0.36307135224342346, + "learning_rate": 1.8288002084698842e-06, + "loss": 0.34019583463668823, + "step": 1486 + }, + { + "epoch": 1.456806763876976, + "grad_norm": 0.3741042912006378, + "learning_rate": 1.8226706694758194e-06, + "loss": 0.35917502641677856, + "step": 1487 + }, + { + "epoch": 1.457787035902463, + "grad_norm": 0.3453916907310486, + "learning_rate": 1.8165491294045596e-06, + "loss": 0.39770618081092834, + "step": 1488 + }, + { + "epoch": 1.45876730792795, + "grad_norm": 0.41760051250457764, + "learning_rate": 1.810435603667075e-06, + "loss": 0.42916232347488403, + "step": 1489 + }, + { + "epoch": 1.459747579953437, + "grad_norm": 0.4573180675506592, + "learning_rate": 1.8043301076541575e-06, + "loss": 0.47303882241249084, + "step": 1490 + }, + { + "epoch": 1.4607278519789242, + "grad_norm": 0.40351033210754395, + "learning_rate": 1.798232656736389e-06, + "loss": 0.48586228489875793, + "step": 1491 + }, + { + "epoch": 1.4617081240044112, + "grad_norm": 0.4049628973007202, + "learning_rate": 1.792143266264092e-06, + "loss": 0.3597436547279358, + "step": 1492 + }, + { + "epoch": 1.4626883960298982, + "grad_norm": 0.40396711230278015, + "learning_rate": 1.7860619515673034e-06, + "loss": 0.38415706157684326, + "step": 1493 + }, + { + "epoch": 1.4636686680553854, + "grad_norm": 0.3359701335430145, + "learning_rate": 1.7799887279557238e-06, + "loss": 0.3412218689918518, + "step": 1494 + }, + { + "epoch": 1.4646489400808724, + "grad_norm": 0.38843947649002075, + "learning_rate": 1.7739236107186858e-06, + "loss": 0.40858131647109985, + "step": 1495 + }, + { + "epoch": 1.4656292121063594, + "grad_norm": 0.3327353596687317, + "learning_rate": 1.767866615125114e-06, + "loss": 0.31050920486450195, + "step": 1496 + }, + { + "epoch": 1.4666094841318467, + "grad_norm": 0.4126180112361908, + "learning_rate": 1.7618177564234907e-06, + "loss": 0.454841673374176, + "step": 1497 + }, + { + "epoch": 1.4675897561573337, + "grad_norm": 0.3566969335079193, + "learning_rate": 1.7557770498418087e-06, + "loss": 0.4169977307319641, + "step": 1498 + }, + { + "epoch": 1.4685700281828207, + "grad_norm": 0.334134966135025, + "learning_rate": 1.7497445105875377e-06, + "loss": 0.3161314129829407, + "step": 1499 + }, + { + "epoch": 1.469550300208308, + "grad_norm": 0.3449893593788147, + "learning_rate": 1.7437201538475916e-06, + "loss": 0.35423362255096436, + "step": 1500 + }, + { + "epoch": 1.470530572233795, + "grad_norm": 0.3642461895942688, + "learning_rate": 1.7377039947882802e-06, + "loss": 0.3608074188232422, + "step": 1501 + }, + { + "epoch": 1.471510844259282, + "grad_norm": 0.3763211667537689, + "learning_rate": 1.7316960485552758e-06, + "loss": 0.34844836592674255, + "step": 1502 + }, + { + "epoch": 1.4724911162847691, + "grad_norm": 0.3820318579673767, + "learning_rate": 1.7256963302735752e-06, + "loss": 0.36789318919181824, + "step": 1503 + }, + { + "epoch": 1.4734713883102561, + "grad_norm": 0.31845200061798096, + "learning_rate": 1.7197048550474643e-06, + "loss": 0.3099425733089447, + "step": 1504 + }, + { + "epoch": 1.4744516603357432, + "grad_norm": 0.38080859184265137, + "learning_rate": 1.7137216379604727e-06, + "loss": 0.33930373191833496, + "step": 1505 + }, + { + "epoch": 1.4754319323612302, + "grad_norm": 0.4188162386417389, + "learning_rate": 1.7077466940753446e-06, + "loss": 0.352202832698822, + "step": 1506 + }, + { + "epoch": 1.4764122043867174, + "grad_norm": 0.3287113606929779, + "learning_rate": 1.7017800384339928e-06, + "loss": 0.37279272079467773, + "step": 1507 + }, + { + "epoch": 1.4773924764122044, + "grad_norm": 0.32032501697540283, + "learning_rate": 1.6958216860574661e-06, + "loss": 0.29563525319099426, + "step": 1508 + }, + { + "epoch": 1.4783727484376914, + "grad_norm": 0.4033612012863159, + "learning_rate": 1.6898716519459074e-06, + "loss": 0.36490416526794434, + "step": 1509 + }, + { + "epoch": 1.4793530204631784, + "grad_norm": 0.37555792927742004, + "learning_rate": 1.683929951078524e-06, + "loss": 0.39824461936950684, + "step": 1510 + }, + { + "epoch": 1.4803332924886656, + "grad_norm": 0.38234370946884155, + "learning_rate": 1.6779965984135376e-06, + "loss": 0.41880670189857483, + "step": 1511 + }, + { + "epoch": 1.4813135645141526, + "grad_norm": 0.41006413102149963, + "learning_rate": 1.6720716088881595e-06, + "loss": 0.3978181779384613, + "step": 1512 + }, + { + "epoch": 1.4822938365396396, + "grad_norm": 0.35753580927848816, + "learning_rate": 1.6661549974185426e-06, + "loss": 0.297444611787796, + "step": 1513 + }, + { + "epoch": 1.4832741085651269, + "grad_norm": 0.36359643936157227, + "learning_rate": 1.6602467788997483e-06, + "loss": 0.36547040939331055, + "step": 1514 + }, + { + "epoch": 1.4842543805906139, + "grad_norm": 0.4175422787666321, + "learning_rate": 1.6543469682057105e-06, + "loss": 0.3530169725418091, + "step": 1515 + }, + { + "epoch": 1.4852346526161009, + "grad_norm": 0.432218998670578, + "learning_rate": 1.648455580189194e-06, + "loss": 0.41731953620910645, + "step": 1516 + }, + { + "epoch": 1.486214924641588, + "grad_norm": 0.4131220877170563, + "learning_rate": 1.6425726296817634e-06, + "loss": 0.3449380695819855, + "step": 1517 + }, + { + "epoch": 1.4871951966670751, + "grad_norm": 0.34036311507225037, + "learning_rate": 1.6366981314937374e-06, + "loss": 0.3346409797668457, + "step": 1518 + }, + { + "epoch": 1.4881754686925621, + "grad_norm": 0.3888697028160095, + "learning_rate": 1.6308321004141609e-06, + "loss": 0.3721480369567871, + "step": 1519 + }, + { + "epoch": 1.4891557407180493, + "grad_norm": 0.30210623145103455, + "learning_rate": 1.624974551210759e-06, + "loss": 0.2844368815422058, + "step": 1520 + }, + { + "epoch": 1.4901360127435364, + "grad_norm": 0.3465906083583832, + "learning_rate": 1.6191254986299044e-06, + "loss": 0.3504944443702698, + "step": 1521 + }, + { + "epoch": 1.4911162847690234, + "grad_norm": 0.327955961227417, + "learning_rate": 1.6132849573965781e-06, + "loss": 0.3071931004524231, + "step": 1522 + }, + { + "epoch": 1.4920965567945106, + "grad_norm": 0.42220282554626465, + "learning_rate": 1.6074529422143398e-06, + "loss": 0.4126828908920288, + "step": 1523 + }, + { + "epoch": 1.4930768288199976, + "grad_norm": 0.42190858721733093, + "learning_rate": 1.6016294677652772e-06, + "loss": 0.45040827989578247, + "step": 1524 + }, + { + "epoch": 1.4940571008454846, + "grad_norm": 0.44068682193756104, + "learning_rate": 1.5958145487099829e-06, + "loss": 0.37440741062164307, + "step": 1525 + }, + { + "epoch": 1.4950373728709718, + "grad_norm": 0.31783613562583923, + "learning_rate": 1.5900081996875083e-06, + "loss": 0.3338105082511902, + "step": 1526 + }, + { + "epoch": 1.4960176448964588, + "grad_norm": 0.5140895843505859, + "learning_rate": 1.5842104353153286e-06, + "loss": 0.4193746745586395, + "step": 1527 + }, + { + "epoch": 1.4969979169219458, + "grad_norm": 0.30907872319221497, + "learning_rate": 1.5784212701893088e-06, + "loss": 0.3128858208656311, + "step": 1528 + }, + { + "epoch": 1.4979781889474328, + "grad_norm": 0.34407368302345276, + "learning_rate": 1.5726407188836672e-06, + "loss": 0.3326481878757477, + "step": 1529 + }, + { + "epoch": 1.4989584609729198, + "grad_norm": 0.3159351944923401, + "learning_rate": 1.566868795950932e-06, + "loss": 0.31238776445388794, + "step": 1530 + }, + { + "epoch": 1.499938732998407, + "grad_norm": 0.39459702372550964, + "learning_rate": 1.561105515921915e-06, + "loss": 0.42257100343704224, + "step": 1531 + }, + { + "epoch": 1.500919005023894, + "grad_norm": 0.36173015832901, + "learning_rate": 1.5553508933056665e-06, + "loss": 0.33956852555274963, + "step": 1532 + }, + { + "epoch": 1.501899277049381, + "grad_norm": 0.36193543672561646, + "learning_rate": 1.549604942589441e-06, + "loss": 0.36098289489746094, + "step": 1533 + }, + { + "epoch": 1.5028795490748683, + "grad_norm": 0.4178013801574707, + "learning_rate": 1.543867678238663e-06, + "loss": 0.4477638602256775, + "step": 1534 + }, + { + "epoch": 1.5038598211003553, + "grad_norm": 0.40881428122520447, + "learning_rate": 1.5381391146968866e-06, + "loss": 0.42927151918411255, + "step": 1535 + }, + { + "epoch": 1.5048400931258423, + "grad_norm": 0.3852333426475525, + "learning_rate": 1.5324192663857673e-06, + "loss": 0.3718698024749756, + "step": 1536 + }, + { + "epoch": 1.5058203651513296, + "grad_norm": 0.36849144101142883, + "learning_rate": 1.5267081477050132e-06, + "loss": 0.31609517335891724, + "step": 1537 + }, + { + "epoch": 1.5068006371768166, + "grad_norm": 0.4196474850177765, + "learning_rate": 1.521005773032362e-06, + "loss": 0.460585355758667, + "step": 1538 + }, + { + "epoch": 1.5077809092023036, + "grad_norm": 0.42934414744377136, + "learning_rate": 1.5153121567235334e-06, + "loss": 0.4124494194984436, + "step": 1539 + }, + { + "epoch": 1.5087611812277908, + "grad_norm": 0.3763748109340668, + "learning_rate": 1.5096273131121997e-06, + "loss": 0.3496323227882385, + "step": 1540 + }, + { + "epoch": 1.5097414532532778, + "grad_norm": 0.35510021448135376, + "learning_rate": 1.5039512565099468e-06, + "loss": 0.3460981845855713, + "step": 1541 + }, + { + "epoch": 1.5107217252787648, + "grad_norm": 0.3260575532913208, + "learning_rate": 1.4982840012062428e-06, + "loss": 0.31388479471206665, + "step": 1542 + }, + { + "epoch": 1.511701997304252, + "grad_norm": 0.39579617977142334, + "learning_rate": 1.4926255614683931e-06, + "loss": 0.37090277671813965, + "step": 1543 + }, + { + "epoch": 1.512682269329739, + "grad_norm": 0.4031487703323364, + "learning_rate": 1.4869759515415167e-06, + "loss": 0.39675334095954895, + "step": 1544 + }, + { + "epoch": 1.513662541355226, + "grad_norm": 0.3760387897491455, + "learning_rate": 1.4813351856484981e-06, + "loss": 0.33782583475112915, + "step": 1545 + }, + { + "epoch": 1.5146428133807133, + "grad_norm": 0.3884948492050171, + "learning_rate": 1.4757032779899588e-06, + "loss": 0.4089457392692566, + "step": 1546 + }, + { + "epoch": 1.5156230854062, + "grad_norm": 0.4042520225048065, + "learning_rate": 1.470080242744218e-06, + "loss": 0.32891637086868286, + "step": 1547 + }, + { + "epoch": 1.5166033574316873, + "grad_norm": 0.40428227186203003, + "learning_rate": 1.4644660940672628e-06, + "loss": 0.345123827457428, + "step": 1548 + }, + { + "epoch": 1.5175836294571745, + "grad_norm": 0.3905646800994873, + "learning_rate": 1.458860846092705e-06, + "loss": 0.37207287549972534, + "step": 1549 + }, + { + "epoch": 1.5185639014826613, + "grad_norm": 0.42539429664611816, + "learning_rate": 1.453264512931748e-06, + "loss": 0.37235546112060547, + "step": 1550 + }, + { + "epoch": 1.5195441735081485, + "grad_norm": 0.41843000054359436, + "learning_rate": 1.4476771086731567e-06, + "loss": 0.4087207317352295, + "step": 1551 + }, + { + "epoch": 1.5205244455336357, + "grad_norm": 0.39144694805145264, + "learning_rate": 1.4420986473832127e-06, + "loss": 0.3397385776042938, + "step": 1552 + }, + { + "epoch": 1.5215047175591225, + "grad_norm": 0.360812246799469, + "learning_rate": 1.4365291431056871e-06, + "loss": 0.3448691666126251, + "step": 1553 + }, + { + "epoch": 1.5224849895846098, + "grad_norm": 0.4787864685058594, + "learning_rate": 1.4309686098617975e-06, + "loss": 0.45358023047447205, + "step": 1554 + }, + { + "epoch": 1.5234652616100968, + "grad_norm": 0.3954147398471832, + "learning_rate": 1.4254170616501828e-06, + "loss": 0.3767503499984741, + "step": 1555 + }, + { + "epoch": 1.5244455336355838, + "grad_norm": 0.3362419605255127, + "learning_rate": 1.4198745124468565e-06, + "loss": 0.30626094341278076, + "step": 1556 + }, + { + "epoch": 1.525425805661071, + "grad_norm": 0.3179536759853363, + "learning_rate": 1.4143409762051829e-06, + "loss": 0.31327712535858154, + "step": 1557 + }, + { + "epoch": 1.526406077686558, + "grad_norm": 0.41927048563957214, + "learning_rate": 1.4088164668558302e-06, + "loss": 0.40531328320503235, + "step": 1558 + }, + { + "epoch": 1.527386349712045, + "grad_norm": 0.35107892751693726, + "learning_rate": 1.4033009983067454e-06, + "loss": 0.3446226716041565, + "step": 1559 + }, + { + "epoch": 1.5283666217375322, + "grad_norm": 0.3600156009197235, + "learning_rate": 1.3977945844431118e-06, + "loss": 0.30024707317352295, + "step": 1560 + }, + { + "epoch": 1.5293468937630192, + "grad_norm": 0.3746243417263031, + "learning_rate": 1.3922972391273226e-06, + "loss": 0.38197118043899536, + "step": 1561 + }, + { + "epoch": 1.5303271657885062, + "grad_norm": 0.3007654845714569, + "learning_rate": 1.3868089761989356e-06, + "loss": 0.3054050803184509, + "step": 1562 + }, + { + "epoch": 1.5313074378139935, + "grad_norm": 0.3238379657268524, + "learning_rate": 1.3813298094746491e-06, + "loss": 0.33077818155288696, + "step": 1563 + }, + { + "epoch": 1.5322877098394805, + "grad_norm": 0.4369322657585144, + "learning_rate": 1.375859752748257e-06, + "loss": 0.49619340896606445, + "step": 1564 + }, + { + "epoch": 1.5332679818649675, + "grad_norm": 0.41659650206565857, + "learning_rate": 1.3703988197906209e-06, + "loss": 0.4149579405784607, + "step": 1565 + }, + { + "epoch": 1.5342482538904547, + "grad_norm": 0.36634567379951477, + "learning_rate": 1.3649470243496327e-06, + "loss": 0.4047231078147888, + "step": 1566 + }, + { + "epoch": 1.5352285259159417, + "grad_norm": 0.35472002625465393, + "learning_rate": 1.3595043801501794e-06, + "loss": 0.323494017124176, + "step": 1567 + }, + { + "epoch": 1.5362087979414287, + "grad_norm": 0.3257138729095459, + "learning_rate": 1.354070900894115e-06, + "loss": 0.32642731070518494, + "step": 1568 + }, + { + "epoch": 1.537189069966916, + "grad_norm": 0.36110514402389526, + "learning_rate": 1.3486466002602133e-06, + "loss": 0.35283994674682617, + "step": 1569 + }, + { + "epoch": 1.5381693419924027, + "grad_norm": 0.34080949425697327, + "learning_rate": 1.3432314919041478e-06, + "loss": 0.3704628646373749, + "step": 1570 + }, + { + "epoch": 1.53914961401789, + "grad_norm": 0.4577089548110962, + "learning_rate": 1.3378255894584463e-06, + "loss": 0.5338248610496521, + "step": 1571 + }, + { + "epoch": 1.5401298860433772, + "grad_norm": 0.33302542567253113, + "learning_rate": 1.3324289065324608e-06, + "loss": 0.3055313229560852, + "step": 1572 + }, + { + "epoch": 1.541110158068864, + "grad_norm": 0.4024427831172943, + "learning_rate": 1.3270414567123342e-06, + "loss": 0.4424835741519928, + "step": 1573 + }, + { + "epoch": 1.5420904300943512, + "grad_norm": 0.3416128158569336, + "learning_rate": 1.3216632535609664e-06, + "loss": 0.30003440380096436, + "step": 1574 + }, + { + "epoch": 1.5430707021198382, + "grad_norm": 0.43086281418800354, + "learning_rate": 1.3162943106179748e-06, + "loss": 0.4463260769844055, + "step": 1575 + }, + { + "epoch": 1.5440509741453252, + "grad_norm": 0.319501131772995, + "learning_rate": 1.310934641399671e-06, + "loss": 0.28336817026138306, + "step": 1576 + }, + { + "epoch": 1.5450312461708124, + "grad_norm": 0.41629907488822937, + "learning_rate": 1.305584259399013e-06, + "loss": 0.33188754320144653, + "step": 1577 + }, + { + "epoch": 1.5460115181962995, + "grad_norm": 0.40745699405670166, + "learning_rate": 1.3002431780855817e-06, + "loss": 0.4487993121147156, + "step": 1578 + }, + { + "epoch": 1.5469917902217865, + "grad_norm": 0.4176565408706665, + "learning_rate": 1.2949114109055417e-06, + "loss": 0.39183542132377625, + "step": 1579 + }, + { + "epoch": 1.5479720622472737, + "grad_norm": 0.416147917509079, + "learning_rate": 1.2895889712816129e-06, + "loss": 0.4164677858352661, + "step": 1580 + }, + { + "epoch": 1.5489523342727607, + "grad_norm": 0.3067798316478729, + "learning_rate": 1.2842758726130283e-06, + "loss": 0.28684866428375244, + "step": 1581 + }, + { + "epoch": 1.5499326062982477, + "grad_norm": 0.41653069853782654, + "learning_rate": 1.2789721282755102e-06, + "loss": 0.4671644866466522, + "step": 1582 + }, + { + "epoch": 1.550912878323735, + "grad_norm": 0.3526851236820221, + "learning_rate": 1.2736777516212267e-06, + "loss": 0.35661229491233826, + "step": 1583 + }, + { + "epoch": 1.551893150349222, + "grad_norm": 0.3794231712818146, + "learning_rate": 1.2683927559787657e-06, + "loss": 0.3517894148826599, + "step": 1584 + }, + { + "epoch": 1.552873422374709, + "grad_norm": 0.36633118987083435, + "learning_rate": 1.263117154653097e-06, + "loss": 0.3292240798473358, + "step": 1585 + }, + { + "epoch": 1.5538536944001962, + "grad_norm": 0.37773072719573975, + "learning_rate": 1.2578509609255385e-06, + "loss": 0.3643125295639038, + "step": 1586 + }, + { + "epoch": 1.5548339664256832, + "grad_norm": 0.36037200689315796, + "learning_rate": 1.2525941880537307e-06, + "loss": 0.36885929107666016, + "step": 1587 + }, + { + "epoch": 1.5558142384511702, + "grad_norm": 0.40058663487434387, + "learning_rate": 1.2473468492715896e-06, + "loss": 0.3856750726699829, + "step": 1588 + }, + { + "epoch": 1.5567945104766574, + "grad_norm": 0.3196178078651428, + "learning_rate": 1.242108957789287e-06, + "loss": 0.3264007568359375, + "step": 1589 + }, + { + "epoch": 1.5577747825021442, + "grad_norm": 0.34384649991989136, + "learning_rate": 1.2368805267932072e-06, + "loss": 0.3553543984889984, + "step": 1590 + }, + { + "epoch": 1.5587550545276314, + "grad_norm": 0.35931405425071716, + "learning_rate": 1.2316615694459188e-06, + "loss": 0.33355772495269775, + "step": 1591 + }, + { + "epoch": 1.5597353265531186, + "grad_norm": 0.45766180753707886, + "learning_rate": 1.22645209888614e-06, + "loss": 0.40842118859291077, + "step": 1592 + }, + { + "epoch": 1.5607155985786054, + "grad_norm": 0.36616262793540955, + "learning_rate": 1.2212521282287093e-06, + "loss": 0.3450744152069092, + "step": 1593 + }, + { + "epoch": 1.5616958706040927, + "grad_norm": 0.39555516839027405, + "learning_rate": 1.2160616705645434e-06, + "loss": 0.3462858200073242, + "step": 1594 + }, + { + "epoch": 1.5626761426295797, + "grad_norm": 0.38610339164733887, + "learning_rate": 1.210880738960616e-06, + "loss": 0.37060654163360596, + "step": 1595 + }, + { + "epoch": 1.5636564146550667, + "grad_norm": 0.3594928979873657, + "learning_rate": 1.2057093464599156e-06, + "loss": 0.3372204899787903, + "step": 1596 + }, + { + "epoch": 1.564636686680554, + "grad_norm": 0.38657864928245544, + "learning_rate": 1.200547506081416e-06, + "loss": 0.412594735622406, + "step": 1597 + }, + { + "epoch": 1.565616958706041, + "grad_norm": 0.41203126311302185, + "learning_rate": 1.1953952308200428e-06, + "loss": 0.3953709304332733, + "step": 1598 + }, + { + "epoch": 1.566597230731528, + "grad_norm": 0.3372426927089691, + "learning_rate": 1.1902525336466465e-06, + "loss": 0.33884385228157043, + "step": 1599 + }, + { + "epoch": 1.5675775027570151, + "grad_norm": 0.3971172273159027, + "learning_rate": 1.1851194275079586e-06, + "loss": 0.37486666440963745, + "step": 1600 + }, + { + "epoch": 1.5685577747825021, + "grad_norm": 0.4146364629268646, + "learning_rate": 1.1799959253265668e-06, + "loss": 0.3716042637825012, + "step": 1601 + }, + { + "epoch": 1.5695380468079891, + "grad_norm": 0.4184497892856598, + "learning_rate": 1.1748820400008843e-06, + "loss": 0.4815278649330139, + "step": 1602 + }, + { + "epoch": 1.5705183188334764, + "grad_norm": 0.3731510043144226, + "learning_rate": 1.1697777844051105e-06, + "loss": 0.3586379289627075, + "step": 1603 + }, + { + "epoch": 1.5714985908589634, + "grad_norm": 0.42198994755744934, + "learning_rate": 1.164683171389202e-06, + "loss": 0.3905531167984009, + "step": 1604 + }, + { + "epoch": 1.5724788628844504, + "grad_norm": 0.4600583612918854, + "learning_rate": 1.1595982137788403e-06, + "loss": 0.4425617456436157, + "step": 1605 + }, + { + "epoch": 1.5734591349099376, + "grad_norm": 0.38403743505477905, + "learning_rate": 1.154522924375403e-06, + "loss": 0.37622302770614624, + "step": 1606 + }, + { + "epoch": 1.5744394069354246, + "grad_norm": 0.37568604946136475, + "learning_rate": 1.1494573159559214e-06, + "loss": 0.3701554536819458, + "step": 1607 + }, + { + "epoch": 1.5754196789609116, + "grad_norm": 0.3850537836551666, + "learning_rate": 1.144401401273062e-06, + "loss": 0.36834436655044556, + "step": 1608 + }, + { + "epoch": 1.5763999509863988, + "grad_norm": 0.41071683168411255, + "learning_rate": 1.1393551930550828e-06, + "loss": 0.39382147789001465, + "step": 1609 + }, + { + "epoch": 1.5773802230118859, + "grad_norm": 0.3978777229785919, + "learning_rate": 1.1343187040058068e-06, + "loss": 0.3603208661079407, + "step": 1610 + }, + { + "epoch": 1.5783604950373729, + "grad_norm": 0.37316370010375977, + "learning_rate": 1.1292919468045876e-06, + "loss": 0.33115750551223755, + "step": 1611 + }, + { + "epoch": 1.57934076706286, + "grad_norm": 0.38263967633247375, + "learning_rate": 1.1242749341062837e-06, + "loss": 0.42162957787513733, + "step": 1612 + }, + { + "epoch": 1.5803210390883469, + "grad_norm": 0.41667816042900085, + "learning_rate": 1.1192676785412154e-06, + "loss": 0.4899442195892334, + "step": 1613 + }, + { + "epoch": 1.581301311113834, + "grad_norm": 0.3998047113418579, + "learning_rate": 1.1142701927151456e-06, + "loss": 0.3733862638473511, + "step": 1614 + }, + { + "epoch": 1.5822815831393213, + "grad_norm": 0.37669438123703003, + "learning_rate": 1.1092824892092375e-06, + "loss": 0.4294346570968628, + "step": 1615 + }, + { + "epoch": 1.583261855164808, + "grad_norm": 0.32034826278686523, + "learning_rate": 1.1043045805800285e-06, + "loss": 0.275056928396225, + "step": 1616 + }, + { + "epoch": 1.5842421271902953, + "grad_norm": 0.3887910842895508, + "learning_rate": 1.099336479359398e-06, + "loss": 0.35782015323638916, + "step": 1617 + }, + { + "epoch": 1.5852223992157823, + "grad_norm": 0.4010225832462311, + "learning_rate": 1.0943781980545332e-06, + "loss": 0.3785402178764343, + "step": 1618 + }, + { + "epoch": 1.5862026712412693, + "grad_norm": 0.3740882873535156, + "learning_rate": 1.0894297491479044e-06, + "loss": 0.3544098138809204, + "step": 1619 + }, + { + "epoch": 1.5871829432667566, + "grad_norm": 0.3949381113052368, + "learning_rate": 1.0844911450972228e-06, + "loss": 0.39230555295944214, + "step": 1620 + }, + { + "epoch": 1.5881632152922436, + "grad_norm": 0.4530141055583954, + "learning_rate": 1.0795623983354214e-06, + "loss": 0.3948533535003662, + "step": 1621 + }, + { + "epoch": 1.5891434873177306, + "grad_norm": 0.400316059589386, + "learning_rate": 1.0746435212706125e-06, + "loss": 0.40657317638397217, + "step": 1622 + }, + { + "epoch": 1.5901237593432178, + "grad_norm": 0.3851020038127899, + "learning_rate": 1.0697345262860638e-06, + "loss": 0.43135738372802734, + "step": 1623 + }, + { + "epoch": 1.5911040313687048, + "grad_norm": 0.4000897705554962, + "learning_rate": 1.0648354257401622e-06, + "loss": 0.4508344531059265, + "step": 1624 + }, + { + "epoch": 1.5920843033941918, + "grad_norm": 0.4060913920402527, + "learning_rate": 1.0599462319663906e-06, + "loss": 0.3771827220916748, + "step": 1625 + }, + { + "epoch": 1.593064575419679, + "grad_norm": 0.340829074382782, + "learning_rate": 1.0550669572732862e-06, + "loss": 0.3534057140350342, + "step": 1626 + }, + { + "epoch": 1.594044847445166, + "grad_norm": 0.7374085187911987, + "learning_rate": 1.0501976139444191e-06, + "loss": 0.37338921427726746, + "step": 1627 + }, + { + "epoch": 1.595025119470653, + "grad_norm": 0.3648485243320465, + "learning_rate": 1.0453382142383545e-06, + "loss": 0.29381653666496277, + "step": 1628 + }, + { + "epoch": 1.5960053914961403, + "grad_norm": 0.3882230818271637, + "learning_rate": 1.0404887703886252e-06, + "loss": 0.366402804851532, + "step": 1629 + }, + { + "epoch": 1.5969856635216273, + "grad_norm": 0.4072689712047577, + "learning_rate": 1.0356492946036994e-06, + "loss": 0.39244699478149414, + "step": 1630 + }, + { + "epoch": 1.5979659355471143, + "grad_norm": 0.3501201868057251, + "learning_rate": 1.0308197990669538e-06, + "loss": 0.3464279770851135, + "step": 1631 + }, + { + "epoch": 1.5989462075726015, + "grad_norm": 0.41540011763572693, + "learning_rate": 1.026000295936635e-06, + "loss": 0.43793538212776184, + "step": 1632 + }, + { + "epoch": 1.5999264795980883, + "grad_norm": 0.36005133390426636, + "learning_rate": 1.0211907973458391e-06, + "loss": 0.4485785663127899, + "step": 1633 + }, + { + "epoch": 1.6009067516235755, + "grad_norm": 0.42961329221725464, + "learning_rate": 1.0163913154024724e-06, + "loss": 0.33980482816696167, + "step": 1634 + }, + { + "epoch": 1.6018870236490628, + "grad_norm": 0.33254656195640564, + "learning_rate": 1.0116018621892237e-06, + "loss": 0.34293556213378906, + "step": 1635 + }, + { + "epoch": 1.6028672956745496, + "grad_norm": 0.4023721218109131, + "learning_rate": 1.006822449763537e-06, + "loss": 0.31392702460289, + "step": 1636 + }, + { + "epoch": 1.6038475677000368, + "grad_norm": 0.3301834464073181, + "learning_rate": 1.0020530901575754e-06, + "loss": 0.32767170667648315, + "step": 1637 + }, + { + "epoch": 1.6048278397255238, + "grad_norm": 0.41354766488075256, + "learning_rate": 9.972937953781985e-07, + "loss": 0.3896903097629547, + "step": 1638 + }, + { + "epoch": 1.6058081117510108, + "grad_norm": 0.3764152526855469, + "learning_rate": 9.925445774069232e-07, + "loss": 0.3687092661857605, + "step": 1639 + }, + { + "epoch": 1.606788383776498, + "grad_norm": 0.39385542273521423, + "learning_rate": 9.878054481999027e-07, + "loss": 0.3807629942893982, + "step": 1640 + }, + { + "epoch": 1.607768655801985, + "grad_norm": 0.36973652243614197, + "learning_rate": 9.830764196878872e-07, + "loss": 0.38120055198669434, + "step": 1641 + }, + { + "epoch": 1.608748927827472, + "grad_norm": 0.36880338191986084, + "learning_rate": 9.783575037762006e-07, + "loss": 0.32165300846099854, + "step": 1642 + }, + { + "epoch": 1.6097291998529593, + "grad_norm": 0.3916257917881012, + "learning_rate": 9.73648712344707e-07, + "loss": 0.34062325954437256, + "step": 1643 + }, + { + "epoch": 1.6107094718784463, + "grad_norm": 0.34860825538635254, + "learning_rate": 9.689500572477857e-07, + "loss": 0.34618276357650757, + "step": 1644 + }, + { + "epoch": 1.6116897439039333, + "grad_norm": 0.3608398139476776, + "learning_rate": 9.642615503142927e-07, + "loss": 0.38314056396484375, + "step": 1645 + }, + { + "epoch": 1.6126700159294205, + "grad_norm": 0.3841942250728607, + "learning_rate": 9.595832033475417e-07, + "loss": 0.3677881360054016, + "step": 1646 + }, + { + "epoch": 1.6136502879549075, + "grad_norm": 0.4171530306339264, + "learning_rate": 9.549150281252633e-07, + "loss": 0.39177200198173523, + "step": 1647 + }, + { + "epoch": 1.6146305599803945, + "grad_norm": 0.40549641847610474, + "learning_rate": 9.502570363995839e-07, + "loss": 0.3616636395454407, + "step": 1648 + }, + { + "epoch": 1.6156108320058817, + "grad_norm": 0.403617799282074, + "learning_rate": 9.456092398969902e-07, + "loss": 0.3500201106071472, + "step": 1649 + }, + { + "epoch": 1.6165911040313687, + "grad_norm": 0.4164355993270874, + "learning_rate": 9.409716503183075e-07, + "loss": 0.3984196186065674, + "step": 1650 + }, + { + "epoch": 1.6175713760568557, + "grad_norm": 0.4306594729423523, + "learning_rate": 9.363442793386606e-07, + "loss": 0.4459274411201477, + "step": 1651 + }, + { + "epoch": 1.618551648082343, + "grad_norm": 0.40200403332710266, + "learning_rate": 9.317271386074489e-07, + "loss": 0.37599679827690125, + "step": 1652 + }, + { + "epoch": 1.6195319201078298, + "grad_norm": 0.3381712734699249, + "learning_rate": 9.271202397483214e-07, + "loss": 0.3198702037334442, + "step": 1653 + }, + { + "epoch": 1.620512192133317, + "grad_norm": 0.3760193884372711, + "learning_rate": 9.225235943591398e-07, + "loss": 0.34943050146102905, + "step": 1654 + }, + { + "epoch": 1.6214924641588042, + "grad_norm": 0.36681824922561646, + "learning_rate": 9.179372140119524e-07, + "loss": 0.3538042902946472, + "step": 1655 + }, + { + "epoch": 1.622472736184291, + "grad_norm": 0.39741751551628113, + "learning_rate": 9.133611102529655e-07, + "loss": 0.3460010886192322, + "step": 1656 + }, + { + "epoch": 1.6234530082097782, + "grad_norm": 0.35628822445869446, + "learning_rate": 9.087952946025175e-07, + "loss": 0.3123657703399658, + "step": 1657 + }, + { + "epoch": 1.6244332802352652, + "grad_norm": 0.39191627502441406, + "learning_rate": 9.042397785550405e-07, + "loss": 0.3519492745399475, + "step": 1658 + }, + { + "epoch": 1.6254135522607522, + "grad_norm": 0.409349650144577, + "learning_rate": 8.996945735790447e-07, + "loss": 0.4180533289909363, + "step": 1659 + }, + { + "epoch": 1.6263938242862395, + "grad_norm": 0.43744170665740967, + "learning_rate": 8.951596911170756e-07, + "loss": 0.3942815661430359, + "step": 1660 + }, + { + "epoch": 1.6273740963117265, + "grad_norm": 0.3615512549877167, + "learning_rate": 8.906351425856952e-07, + "loss": 0.3662254810333252, + "step": 1661 + }, + { + "epoch": 1.6283543683372135, + "grad_norm": 0.39092549681663513, + "learning_rate": 8.861209393754478e-07, + "loss": 0.37831389904022217, + "step": 1662 + }, + { + "epoch": 1.6293346403627007, + "grad_norm": 0.38348859548568726, + "learning_rate": 8.816170928508367e-07, + "loss": 0.3815270662307739, + "step": 1663 + }, + { + "epoch": 1.6303149123881877, + "grad_norm": 0.5721145272254944, + "learning_rate": 8.771236143502876e-07, + "loss": 0.3905642628669739, + "step": 1664 + }, + { + "epoch": 1.6312951844136747, + "grad_norm": 0.4526739716529846, + "learning_rate": 8.7264051518613e-07, + "loss": 0.40796419978141785, + "step": 1665 + }, + { + "epoch": 1.632275456439162, + "grad_norm": 0.42541760206222534, + "learning_rate": 8.681678066445598e-07, + "loss": 0.40424951910972595, + "step": 1666 + }, + { + "epoch": 1.633255728464649, + "grad_norm": 0.37035465240478516, + "learning_rate": 8.637054999856148e-07, + "loss": 0.3495844006538391, + "step": 1667 + }, + { + "epoch": 1.634236000490136, + "grad_norm": 0.39394092559814453, + "learning_rate": 8.592536064431467e-07, + "loss": 0.384368896484375, + "step": 1668 + }, + { + "epoch": 1.6352162725156232, + "grad_norm": 0.3680286109447479, + "learning_rate": 8.54812137224792e-07, + "loss": 0.3417609930038452, + "step": 1669 + }, + { + "epoch": 1.6361965445411102, + "grad_norm": 0.3347630798816681, + "learning_rate": 8.503811035119452e-07, + "loss": 0.30828940868377686, + "step": 1670 + }, + { + "epoch": 1.6371768165665972, + "grad_norm": 0.3998127579689026, + "learning_rate": 8.459605164597268e-07, + "loss": 0.4099857211112976, + "step": 1671 + }, + { + "epoch": 1.6381570885920844, + "grad_norm": 0.35439401865005493, + "learning_rate": 8.415503871969616e-07, + "loss": 0.2773662209510803, + "step": 1672 + }, + { + "epoch": 1.6391373606175714, + "grad_norm": 0.3643379807472229, + "learning_rate": 8.371507268261436e-07, + "loss": 0.34642860293388367, + "step": 1673 + }, + { + "epoch": 1.6401176326430584, + "grad_norm": 0.41094130277633667, + "learning_rate": 8.327615464234129e-07, + "loss": 0.4250739514827728, + "step": 1674 + }, + { + "epoch": 1.6410979046685457, + "grad_norm": 0.3596060276031494, + "learning_rate": 8.283828570385239e-07, + "loss": 0.3752918839454651, + "step": 1675 + }, + { + "epoch": 1.6420781766940324, + "grad_norm": 0.47694459557533264, + "learning_rate": 8.240146696948254e-07, + "loss": 0.43831154704093933, + "step": 1676 + }, + { + "epoch": 1.6430584487195197, + "grad_norm": 0.4135114252567291, + "learning_rate": 8.196569953892202e-07, + "loss": 0.4069673717021942, + "step": 1677 + }, + { + "epoch": 1.644038720745007, + "grad_norm": 0.4220380485057831, + "learning_rate": 8.153098450921514e-07, + "loss": 0.44206881523132324, + "step": 1678 + }, + { + "epoch": 1.6450189927704937, + "grad_norm": 0.37465569376945496, + "learning_rate": 8.109732297475637e-07, + "loss": 0.36452120542526245, + "step": 1679 + }, + { + "epoch": 1.645999264795981, + "grad_norm": 0.3630717396736145, + "learning_rate": 8.066471602728804e-07, + "loss": 0.33272457122802734, + "step": 1680 + }, + { + "epoch": 1.646979536821468, + "grad_norm": 0.3615228235721588, + "learning_rate": 8.023316475589754e-07, + "loss": 0.3282840847969055, + "step": 1681 + }, + { + "epoch": 1.647959808846955, + "grad_norm": 0.3980245292186737, + "learning_rate": 7.980267024701494e-07, + "loss": 0.40692663192749023, + "step": 1682 + }, + { + "epoch": 1.6489400808724421, + "grad_norm": 0.41923287510871887, + "learning_rate": 7.937323358440935e-07, + "loss": 0.38563722372055054, + "step": 1683 + }, + { + "epoch": 1.6499203528979292, + "grad_norm": 0.4210887849330902, + "learning_rate": 7.894485584918737e-07, + "loss": 0.41898012161254883, + "step": 1684 + }, + { + "epoch": 1.6509006249234162, + "grad_norm": 0.3525547981262207, + "learning_rate": 7.851753811978924e-07, + "loss": 0.31187230348587036, + "step": 1685 + }, + { + "epoch": 1.6518808969489034, + "grad_norm": 0.33336856961250305, + "learning_rate": 7.809128147198692e-07, + "loss": 0.289673775434494, + "step": 1686 + }, + { + "epoch": 1.6528611689743904, + "grad_norm": 0.4095363914966583, + "learning_rate": 7.766608697888095e-07, + "loss": 0.3870829939842224, + "step": 1687 + }, + { + "epoch": 1.6538414409998774, + "grad_norm": 0.4303877651691437, + "learning_rate": 7.724195571089787e-07, + "loss": 0.4495899975299835, + "step": 1688 + }, + { + "epoch": 1.6548217130253646, + "grad_norm": 0.43693169951438904, + "learning_rate": 7.681888873578786e-07, + "loss": 0.4429781436920166, + "step": 1689 + }, + { + "epoch": 1.6558019850508516, + "grad_norm": 0.4128706455230713, + "learning_rate": 7.639688711862131e-07, + "loss": 0.387198269367218, + "step": 1690 + }, + { + "epoch": 1.6567822570763386, + "grad_norm": 0.4205910265445709, + "learning_rate": 7.597595192178702e-07, + "loss": 0.41632279753685, + "step": 1691 + }, + { + "epoch": 1.6577625291018259, + "grad_norm": 0.3953818678855896, + "learning_rate": 7.555608420498872e-07, + "loss": 0.3593927323818207, + "step": 1692 + }, + { + "epoch": 1.6587428011273129, + "grad_norm": 0.3857731819152832, + "learning_rate": 7.513728502524286e-07, + "loss": 0.3423689603805542, + "step": 1693 + }, + { + "epoch": 1.6597230731527999, + "grad_norm": 0.3715440332889557, + "learning_rate": 7.471955543687586e-07, + "loss": 0.36759787797927856, + "step": 1694 + }, + { + "epoch": 1.660703345178287, + "grad_norm": 0.3684738576412201, + "learning_rate": 7.430289649152156e-07, + "loss": 0.27273011207580566, + "step": 1695 + }, + { + "epoch": 1.6616836172037739, + "grad_norm": 0.3769178092479706, + "learning_rate": 7.388730923811827e-07, + "loss": 0.36157768964767456, + "step": 1696 + }, + { + "epoch": 1.6626638892292611, + "grad_norm": 0.36349016427993774, + "learning_rate": 7.347279472290647e-07, + "loss": 0.35967859625816345, + "step": 1697 + }, + { + "epoch": 1.6636441612547483, + "grad_norm": 0.37429091334342957, + "learning_rate": 7.305935398942598e-07, + "loss": 0.3740808367729187, + "step": 1698 + }, + { + "epoch": 1.6646244332802351, + "grad_norm": 0.4239107668399811, + "learning_rate": 7.264698807851328e-07, + "loss": 0.40116116404533386, + "step": 1699 + }, + { + "epoch": 1.6656047053057224, + "grad_norm": 0.35234925150871277, + "learning_rate": 7.2235698028299e-07, + "loss": 0.33047643303871155, + "step": 1700 + }, + { + "epoch": 1.6665849773312094, + "grad_norm": 0.36228421330451965, + "learning_rate": 7.182548487420555e-07, + "loss": 0.33354827761650085, + "step": 1701 + }, + { + "epoch": 1.6675652493566964, + "grad_norm": 0.3616648018360138, + "learning_rate": 7.141634964894389e-07, + "loss": 0.2902921736240387, + "step": 1702 + }, + { + "epoch": 1.6685455213821836, + "grad_norm": 0.4077574908733368, + "learning_rate": 7.100829338251147e-07, + "loss": 0.43251097202301025, + "step": 1703 + }, + { + "epoch": 1.6695257934076706, + "grad_norm": 0.4163186252117157, + "learning_rate": 7.060131710218959e-07, + "loss": 0.4361751079559326, + "step": 1704 + }, + { + "epoch": 1.6705060654331576, + "grad_norm": 0.35189637541770935, + "learning_rate": 7.019542183254047e-07, + "loss": 0.33032262325286865, + "step": 1705 + }, + { + "epoch": 1.6714863374586448, + "grad_norm": 0.3183583915233612, + "learning_rate": 6.979060859540493e-07, + "loss": 0.26121529936790466, + "step": 1706 + }, + { + "epoch": 1.6724666094841318, + "grad_norm": 0.3675546944141388, + "learning_rate": 6.938687840989972e-07, + "loss": 0.3178236484527588, + "step": 1707 + }, + { + "epoch": 1.6734468815096188, + "grad_norm": 0.37069079279899597, + "learning_rate": 6.898423229241535e-07, + "loss": 0.3701722025871277, + "step": 1708 + }, + { + "epoch": 1.674427153535106, + "grad_norm": 0.37248364090919495, + "learning_rate": 6.858267125661272e-07, + "loss": 0.39604297280311584, + "step": 1709 + }, + { + "epoch": 1.675407425560593, + "grad_norm": 0.3417600393295288, + "learning_rate": 6.818219631342149e-07, + "loss": 0.3022978901863098, + "step": 1710 + }, + { + "epoch": 1.67638769758608, + "grad_norm": 0.33110755681991577, + "learning_rate": 6.778280847103668e-07, + "loss": 0.3298816680908203, + "step": 1711 + }, + { + "epoch": 1.6773679696115673, + "grad_norm": 0.3624923825263977, + "learning_rate": 6.738450873491675e-07, + "loss": 0.33920934796333313, + "step": 1712 + }, + { + "epoch": 1.6783482416370543, + "grad_norm": 0.34578874707221985, + "learning_rate": 6.698729810778065e-07, + "loss": 0.25365781784057617, + "step": 1713 + }, + { + "epoch": 1.6793285136625413, + "grad_norm": 0.43730032444000244, + "learning_rate": 6.659117758960598e-07, + "loss": 0.4115995168685913, + "step": 1714 + }, + { + "epoch": 1.6803087856880285, + "grad_norm": 0.49691611528396606, + "learning_rate": 6.619614817762537e-07, + "loss": 0.3700525760650635, + "step": 1715 + }, + { + "epoch": 1.6812890577135156, + "grad_norm": 0.412494421005249, + "learning_rate": 6.580221086632516e-07, + "loss": 0.34133976697921753, + "step": 1716 + }, + { + "epoch": 1.6822693297390026, + "grad_norm": 0.4409703016281128, + "learning_rate": 6.540936664744197e-07, + "loss": 0.4596427083015442, + "step": 1717 + }, + { + "epoch": 1.6832496017644898, + "grad_norm": 0.4154723584651947, + "learning_rate": 6.501761650996052e-07, + "loss": 0.3950229287147522, + "step": 1718 + }, + { + "epoch": 1.6842298737899766, + "grad_norm": 0.3930273950099945, + "learning_rate": 6.462696144011149e-07, + "loss": 0.3304884135723114, + "step": 1719 + }, + { + "epoch": 1.6852101458154638, + "grad_norm": 0.3548828661441803, + "learning_rate": 6.423740242136833e-07, + "loss": 0.34451717138290405, + "step": 1720 + }, + { + "epoch": 1.686190417840951, + "grad_norm": 0.33905333280563354, + "learning_rate": 6.384894043444568e-07, + "loss": 0.28703922033309937, + "step": 1721 + }, + { + "epoch": 1.6871706898664378, + "grad_norm": 0.4317936301231384, + "learning_rate": 6.34615764572959e-07, + "loss": 0.36868083477020264, + "step": 1722 + }, + { + "epoch": 1.688150961891925, + "grad_norm": 0.4037177562713623, + "learning_rate": 6.307531146510754e-07, + "loss": 0.38313159346580505, + "step": 1723 + }, + { + "epoch": 1.689131233917412, + "grad_norm": 0.3752223551273346, + "learning_rate": 6.269014643030214e-07, + "loss": 0.3738197684288025, + "step": 1724 + }, + { + "epoch": 1.690111505942899, + "grad_norm": 0.39177384972572327, + "learning_rate": 6.230608232253227e-07, + "loss": 0.3778371810913086, + "step": 1725 + }, + { + "epoch": 1.6910917779683863, + "grad_norm": 0.40136051177978516, + "learning_rate": 6.192312010867868e-07, + "loss": 0.31901055574417114, + "step": 1726 + }, + { + "epoch": 1.6920720499938733, + "grad_norm": 0.44372859597206116, + "learning_rate": 6.154126075284855e-07, + "loss": 0.38163936138153076, + "step": 1727 + }, + { + "epoch": 1.6930523220193603, + "grad_norm": 0.4267496168613434, + "learning_rate": 6.116050521637218e-07, + "loss": 0.43253737688064575, + "step": 1728 + }, + { + "epoch": 1.6940325940448475, + "grad_norm": 0.6490470767021179, + "learning_rate": 6.07808544578013e-07, + "loss": 0.5197738409042358, + "step": 1729 + }, + { + "epoch": 1.6950128660703345, + "grad_norm": 0.3373274803161621, + "learning_rate": 6.04023094329062e-07, + "loss": 0.3314768373966217, + "step": 1730 + }, + { + "epoch": 1.6959931380958215, + "grad_norm": 0.3126501142978668, + "learning_rate": 6.002487109467347e-07, + "loss": 0.30093029141426086, + "step": 1731 + }, + { + "epoch": 1.6969734101213088, + "grad_norm": 0.3094346225261688, + "learning_rate": 5.964854039330365e-07, + "loss": 0.28220677375793457, + "step": 1732 + }, + { + "epoch": 1.6979536821467958, + "grad_norm": 0.39775609970092773, + "learning_rate": 5.927331827620902e-07, + "loss": 0.43980672955513, + "step": 1733 + }, + { + "epoch": 1.6989339541722828, + "grad_norm": 0.4178506135940552, + "learning_rate": 5.889920568801055e-07, + "loss": 0.4097081124782562, + "step": 1734 + }, + { + "epoch": 1.69991422619777, + "grad_norm": 0.3653453290462494, + "learning_rate": 5.852620357053651e-07, + "loss": 0.36697205901145935, + "step": 1735 + }, + { + "epoch": 1.700894498223257, + "grad_norm": 0.4167642295360565, + "learning_rate": 5.815431286281925e-07, + "loss": 0.4138021469116211, + "step": 1736 + }, + { + "epoch": 1.701874770248744, + "grad_norm": 0.3523234724998474, + "learning_rate": 5.778353450109286e-07, + "loss": 0.3384416401386261, + "step": 1737 + }, + { + "epoch": 1.7028550422742312, + "grad_norm": 0.3431588113307953, + "learning_rate": 5.741386941879179e-07, + "loss": 0.34060513973236084, + "step": 1738 + }, + { + "epoch": 1.703835314299718, + "grad_norm": 0.3320436477661133, + "learning_rate": 5.704531854654721e-07, + "loss": 0.28444021940231323, + "step": 1739 + }, + { + "epoch": 1.7048155863252052, + "grad_norm": 0.36861515045166016, + "learning_rate": 5.667788281218567e-07, + "loss": 0.36222732067108154, + "step": 1740 + }, + { + "epoch": 1.7057958583506925, + "grad_norm": 0.4572397470474243, + "learning_rate": 5.631156314072605e-07, + "loss": 0.4219418466091156, + "step": 1741 + }, + { + "epoch": 1.7067761303761793, + "grad_norm": 0.3636431396007538, + "learning_rate": 5.594636045437796e-07, + "loss": 0.3135250210762024, + "step": 1742 + }, + { + "epoch": 1.7077564024016665, + "grad_norm": 0.3627729117870331, + "learning_rate": 5.558227567253832e-07, + "loss": 0.35144442319869995, + "step": 1743 + }, + { + "epoch": 1.7087366744271535, + "grad_norm": 0.43457746505737305, + "learning_rate": 5.521930971179041e-07, + "loss": 0.4118567705154419, + "step": 1744 + }, + { + "epoch": 1.7097169464526405, + "grad_norm": 0.4134824872016907, + "learning_rate": 5.485746348590048e-07, + "loss": 0.40434730052948, + "step": 1745 + }, + { + "epoch": 1.7106972184781277, + "grad_norm": 0.3714609146118164, + "learning_rate": 5.449673790581611e-07, + "loss": 0.35350048542022705, + "step": 1746 + }, + { + "epoch": 1.7116774905036147, + "grad_norm": 0.3943200409412384, + "learning_rate": 5.413713387966329e-07, + "loss": 0.37914228439331055, + "step": 1747 + }, + { + "epoch": 1.7126577625291017, + "grad_norm": 0.37800419330596924, + "learning_rate": 5.377865231274504e-07, + "loss": 0.3656231760978699, + "step": 1748 + }, + { + "epoch": 1.713638034554589, + "grad_norm": 0.4477095305919647, + "learning_rate": 5.34212941075381e-07, + "loss": 0.45934489369392395, + "step": 1749 + }, + { + "epoch": 1.714618306580076, + "grad_norm": 0.39313989877700806, + "learning_rate": 5.306506016369134e-07, + "loss": 0.437400758266449, + "step": 1750 + }, + { + "epoch": 1.715598578605563, + "grad_norm": 0.4008079469203949, + "learning_rate": 5.270995137802315e-07, + "loss": 0.3860788941383362, + "step": 1751 + }, + { + "epoch": 1.7165788506310502, + "grad_norm": 0.30773061513900757, + "learning_rate": 5.235596864451975e-07, + "loss": 0.2639247179031372, + "step": 1752 + }, + { + "epoch": 1.7175591226565372, + "grad_norm": 0.39925551414489746, + "learning_rate": 5.200311285433213e-07, + "loss": 0.40698951482772827, + "step": 1753 + }, + { + "epoch": 1.7185393946820242, + "grad_norm": 0.4305100739002228, + "learning_rate": 5.165138489577415e-07, + "loss": 0.42061877250671387, + "step": 1754 + }, + { + "epoch": 1.7195196667075114, + "grad_norm": 0.39501580595970154, + "learning_rate": 5.130078565432089e-07, + "loss": 0.30541402101516724, + "step": 1755 + }, + { + "epoch": 1.7204999387329984, + "grad_norm": 0.5031176209449768, + "learning_rate": 5.095131601260517e-07, + "loss": 0.37350988388061523, + "step": 1756 + }, + { + "epoch": 1.7214802107584855, + "grad_norm": 0.3525550961494446, + "learning_rate": 5.06029768504166e-07, + "loss": 0.3409478962421417, + "step": 1757 + }, + { + "epoch": 1.7224604827839727, + "grad_norm": 0.2970249652862549, + "learning_rate": 5.025576904469842e-07, + "loss": 0.334276020526886, + "step": 1758 + }, + { + "epoch": 1.7234407548094595, + "grad_norm": 0.35102379322052, + "learning_rate": 4.990969346954611e-07, + "loss": 0.32869595289230347, + "step": 1759 + }, + { + "epoch": 1.7244210268349467, + "grad_norm": 0.3201766908168793, + "learning_rate": 4.956475099620433e-07, + "loss": 0.3035515546798706, + "step": 1760 + }, + { + "epoch": 1.725401298860434, + "grad_norm": 0.3414289951324463, + "learning_rate": 4.922094249306559e-07, + "loss": 0.34790509939193726, + "step": 1761 + }, + { + "epoch": 1.7263815708859207, + "grad_norm": 0.3776668906211853, + "learning_rate": 4.887826882566699e-07, + "loss": 0.37836602330207825, + "step": 1762 + }, + { + "epoch": 1.727361842911408, + "grad_norm": 0.37328770756721497, + "learning_rate": 4.853673085668947e-07, + "loss": 0.35866159200668335, + "step": 1763 + }, + { + "epoch": 1.728342114936895, + "grad_norm": 0.4130055606365204, + "learning_rate": 4.819632944595415e-07, + "loss": 0.4106901288032532, + "step": 1764 + }, + { + "epoch": 1.729322386962382, + "grad_norm": 0.3384333848953247, + "learning_rate": 4.785706545042141e-07, + "loss": 0.3396153748035431, + "step": 1765 + }, + { + "epoch": 1.7303026589878692, + "grad_norm": 0.4002649486064911, + "learning_rate": 4.751893972418781e-07, + "loss": 0.3552248775959015, + "step": 1766 + }, + { + "epoch": 1.7312829310133562, + "grad_norm": 0.38660842180252075, + "learning_rate": 4.7181953118484556e-07, + "loss": 0.4244087338447571, + "step": 1767 + }, + { + "epoch": 1.7322632030388432, + "grad_norm": 0.33861714601516724, + "learning_rate": 4.6846106481675035e-07, + "loss": 0.30651751160621643, + "step": 1768 + }, + { + "epoch": 1.7332434750643304, + "grad_norm": 0.4157235026359558, + "learning_rate": 4.651140065925269e-07, + "loss": 0.33949941396713257, + "step": 1769 + }, + { + "epoch": 1.7342237470898174, + "grad_norm": 0.401938259601593, + "learning_rate": 4.617783649383906e-07, + "loss": 0.35739219188690186, + "step": 1770 + }, + { + "epoch": 1.7352040191153044, + "grad_norm": 0.4284442663192749, + "learning_rate": 4.58454148251814e-07, + "loss": 0.3342616856098175, + "step": 1771 + }, + { + "epoch": 1.7361842911407916, + "grad_norm": 0.34812623262405396, + "learning_rate": 4.551413649015113e-07, + "loss": 0.2948799133300781, + "step": 1772 + }, + { + "epoch": 1.7371645631662787, + "grad_norm": 0.45614632964134216, + "learning_rate": 4.5184002322740784e-07, + "loss": 0.3512307405471802, + "step": 1773 + }, + { + "epoch": 1.7381448351917657, + "grad_norm": 0.4031081199645996, + "learning_rate": 4.48550131540631e-07, + "loss": 0.36050859093666077, + "step": 1774 + }, + { + "epoch": 1.7391251072172529, + "grad_norm": 0.4034765958786011, + "learning_rate": 4.4527169812347446e-07, + "loss": 0.375060498714447, + "step": 1775 + }, + { + "epoch": 1.74010537924274, + "grad_norm": 0.4362258017063141, + "learning_rate": 4.420047312293946e-07, + "loss": 0.41729527711868286, + "step": 1776 + }, + { + "epoch": 1.741085651268227, + "grad_norm": 0.4252602159976959, + "learning_rate": 4.387492390829734e-07, + "loss": 0.3769991993904114, + "step": 1777 + }, + { + "epoch": 1.7420659232937141, + "grad_norm": 0.36566048860549927, + "learning_rate": 4.355052298799112e-07, + "loss": 0.34430134296417236, + "step": 1778 + }, + { + "epoch": 1.7430461953192011, + "grad_norm": 0.3364507853984833, + "learning_rate": 4.322727117869951e-07, + "loss": 0.3499795198440552, + "step": 1779 + }, + { + "epoch": 1.7440264673446881, + "grad_norm": 0.37793198227882385, + "learning_rate": 4.2905169294208846e-07, + "loss": 0.3552755117416382, + "step": 1780 + }, + { + "epoch": 1.7450067393701754, + "grad_norm": 0.3905639052391052, + "learning_rate": 4.2584218145409916e-07, + "loss": 0.3236846923828125, + "step": 1781 + }, + { + "epoch": 1.7459870113956621, + "grad_norm": 0.3534347712993622, + "learning_rate": 4.2264418540297e-07, + "loss": 0.36665186285972595, + "step": 1782 + }, + { + "epoch": 1.7469672834211494, + "grad_norm": 0.36155980825424194, + "learning_rate": 4.194577128396521e-07, + "loss": 0.32624632120132446, + "step": 1783 + }, + { + "epoch": 1.7479475554466366, + "grad_norm": 0.5735606551170349, + "learning_rate": 4.1628277178608677e-07, + "loss": 0.4880862236022949, + "step": 1784 + }, + { + "epoch": 1.7489278274721234, + "grad_norm": 0.34808674454689026, + "learning_rate": 4.131193702351827e-07, + "loss": 0.4228994846343994, + "step": 1785 + }, + { + "epoch": 1.7499080994976106, + "grad_norm": 0.35863596200942993, + "learning_rate": 4.099675161508021e-07, + "loss": 0.350189208984375, + "step": 1786 + }, + { + "epoch": 1.7508883715230976, + "grad_norm": 0.3939903676509857, + "learning_rate": 4.0682721746773346e-07, + "loss": 0.3640008568763733, + "step": 1787 + }, + { + "epoch": 1.7518686435485846, + "grad_norm": 0.48331713676452637, + "learning_rate": 4.036984820916723e-07, + "loss": 0.39425891637802124, + "step": 1788 + }, + { + "epoch": 1.7528489155740719, + "grad_norm": 0.40826287865638733, + "learning_rate": 4.005813178992091e-07, + "loss": 0.36711686849594116, + "step": 1789 + }, + { + "epoch": 1.7538291875995589, + "grad_norm": 0.41581106185913086, + "learning_rate": 3.9747573273779816e-07, + "loss": 0.45201385021209717, + "step": 1790 + }, + { + "epoch": 1.7548094596250459, + "grad_norm": 0.35795292258262634, + "learning_rate": 3.9438173442575e-07, + "loss": 0.35533955693244934, + "step": 1791 + }, + { + "epoch": 1.755789731650533, + "grad_norm": 0.38203519582748413, + "learning_rate": 3.9129933075219837e-07, + "loss": 0.3351580500602722, + "step": 1792 + }, + { + "epoch": 1.75677000367602, + "grad_norm": 0.3370845913887024, + "learning_rate": 3.882285294770938e-07, + "loss": 0.3960869312286377, + "step": 1793 + }, + { + "epoch": 1.757750275701507, + "grad_norm": 0.37114661931991577, + "learning_rate": 3.851693383311722e-07, + "loss": 0.3129011392593384, + "step": 1794 + }, + { + "epoch": 1.7587305477269943, + "grad_norm": 0.4207909405231476, + "learning_rate": 3.821217650159453e-07, + "loss": 0.4319097399711609, + "step": 1795 + }, + { + "epoch": 1.7597108197524813, + "grad_norm": 0.3931514322757721, + "learning_rate": 3.790858172036738e-07, + "loss": 0.37323957681655884, + "step": 1796 + }, + { + "epoch": 1.7606910917779683, + "grad_norm": 0.46052196621894836, + "learning_rate": 3.760615025373543e-07, + "loss": 0.37961477041244507, + "step": 1797 + }, + { + "epoch": 1.7616713638034556, + "grad_norm": 0.42848390340805054, + "learning_rate": 3.730488286306944e-07, + "loss": 0.39329323172569275, + "step": 1798 + }, + { + "epoch": 1.7626516358289426, + "grad_norm": 0.35349076986312866, + "learning_rate": 3.7004780306809873e-07, + "loss": 0.3262864351272583, + "step": 1799 + }, + { + "epoch": 1.7636319078544296, + "grad_norm": 0.4072122573852539, + "learning_rate": 3.6705843340464287e-07, + "loss": 0.4095666706562042, + "step": 1800 + }, + { + "epoch": 1.7646121798799168, + "grad_norm": 0.3660013675689697, + "learning_rate": 3.6408072716606346e-07, + "loss": 0.3483973741531372, + "step": 1801 + }, + { + "epoch": 1.7655924519054036, + "grad_norm": 0.4415965676307678, + "learning_rate": 3.611146918487307e-07, + "loss": 0.3515743315219879, + "step": 1802 + }, + { + "epoch": 1.7665727239308908, + "grad_norm": 0.36835986375808716, + "learning_rate": 3.581603349196372e-07, + "loss": 0.34636688232421875, + "step": 1803 + }, + { + "epoch": 1.767552995956378, + "grad_norm": 0.3201569616794586, + "learning_rate": 3.552176638163707e-07, + "loss": 0.2817646265029907, + "step": 1804 + }, + { + "epoch": 1.7685332679818648, + "grad_norm": 0.4631844162940979, + "learning_rate": 3.522866859471047e-07, + "loss": 0.42721739411354065, + "step": 1805 + }, + { + "epoch": 1.769513540007352, + "grad_norm": 0.3755486011505127, + "learning_rate": 3.4936740869057075e-07, + "loss": 0.3944542407989502, + "step": 1806 + }, + { + "epoch": 1.770493812032839, + "grad_norm": 0.3666553795337677, + "learning_rate": 3.46459839396045e-07, + "loss": 0.33703193068504333, + "step": 1807 + }, + { + "epoch": 1.771474084058326, + "grad_norm": 0.36981818079948425, + "learning_rate": 3.435639853833317e-07, + "loss": 0.33699771761894226, + "step": 1808 + }, + { + "epoch": 1.7724543560838133, + "grad_norm": 0.39931976795196533, + "learning_rate": 3.406798539427386e-07, + "loss": 0.3801427185535431, + "step": 1809 + }, + { + "epoch": 1.7734346281093003, + "grad_norm": 0.6619271636009216, + "learning_rate": 3.3780745233506384e-07, + "loss": 0.3411650061607361, + "step": 1810 + }, + { + "epoch": 1.7744149001347873, + "grad_norm": 0.3984590470790863, + "learning_rate": 3.3494678779157464e-07, + "loss": 0.3892936706542969, + "step": 1811 + }, + { + "epoch": 1.7753951721602745, + "grad_norm": 0.45333677530288696, + "learning_rate": 3.320978675139919e-07, + "loss": 0.37370234727859497, + "step": 1812 + }, + { + "epoch": 1.7763754441857615, + "grad_norm": 0.36892426013946533, + "learning_rate": 3.2926069867446673e-07, + "loss": 0.37117406725883484, + "step": 1813 + }, + { + "epoch": 1.7773557162112485, + "grad_norm": 0.36695677042007446, + "learning_rate": 3.2643528841556973e-07, + "loss": 0.3996949791908264, + "step": 1814 + }, + { + "epoch": 1.7783359882367358, + "grad_norm": 0.3981246054172516, + "learning_rate": 3.2362164385026704e-07, + "loss": 0.41761890053749084, + "step": 1815 + }, + { + "epoch": 1.7793162602622228, + "grad_norm": 0.4921976923942566, + "learning_rate": 3.2081977206190715e-07, + "loss": 0.5487469434738159, + "step": 1816 + }, + { + "epoch": 1.7802965322877098, + "grad_norm": 0.3893972933292389, + "learning_rate": 3.180296801041971e-07, + "loss": 0.33446502685546875, + "step": 1817 + }, + { + "epoch": 1.781276804313197, + "grad_norm": 0.47832003235816956, + "learning_rate": 3.1525137500119207e-07, + "loss": 0.44355836510658264, + "step": 1818 + }, + { + "epoch": 1.782257076338684, + "grad_norm": 0.32840055227279663, + "learning_rate": 3.1248486374726884e-07, + "loss": 0.2868937849998474, + "step": 1819 + }, + { + "epoch": 1.783237348364171, + "grad_norm": 0.4017927944660187, + "learning_rate": 3.09730153307119e-07, + "loss": 0.411729097366333, + "step": 1820 + }, + { + "epoch": 1.7842176203896583, + "grad_norm": 0.347160667181015, + "learning_rate": 3.069872506157212e-07, + "loss": 0.29913264513015747, + "step": 1821 + }, + { + "epoch": 1.785197892415145, + "grad_norm": 0.3558223843574524, + "learning_rate": 3.042561625783308e-07, + "loss": 0.2906627953052521, + "step": 1822 + }, + { + "epoch": 1.7861781644406323, + "grad_norm": 0.37851837277412415, + "learning_rate": 3.015368960704584e-07, + "loss": 0.3845628499984741, + "step": 1823 + }, + { + "epoch": 1.7871584364661195, + "grad_norm": 0.38107749819755554, + "learning_rate": 2.988294579378537e-07, + "loss": 0.331478476524353, + "step": 1824 + }, + { + "epoch": 1.7881387084916063, + "grad_norm": 0.3775060176849365, + "learning_rate": 2.9613385499648926e-07, + "loss": 0.3944990038871765, + "step": 1825 + }, + { + "epoch": 1.7891189805170935, + "grad_norm": 0.5023965835571289, + "learning_rate": 2.934500940325413e-07, + "loss": 0.3733501434326172, + "step": 1826 + }, + { + "epoch": 1.7900992525425807, + "grad_norm": 0.3622867465019226, + "learning_rate": 2.9077818180237693e-07, + "loss": 0.34761834144592285, + "step": 1827 + }, + { + "epoch": 1.7910795245680675, + "grad_norm": 0.3469088077545166, + "learning_rate": 2.881181250325299e-07, + "loss": 0.29729926586151123, + "step": 1828 + }, + { + "epoch": 1.7920597965935547, + "grad_norm": 0.35646477341651917, + "learning_rate": 2.8546993041969173e-07, + "loss": 0.30250295996665955, + "step": 1829 + }, + { + "epoch": 1.7930400686190417, + "grad_norm": 0.35574835538864136, + "learning_rate": 2.8283360463068786e-07, + "loss": 0.31331729888916016, + "step": 1830 + }, + { + "epoch": 1.7940203406445288, + "grad_norm": 0.33787837624549866, + "learning_rate": 2.802091543024671e-07, + "loss": 0.34680333733558655, + "step": 1831 + }, + { + "epoch": 1.795000612670016, + "grad_norm": 0.42167672514915466, + "learning_rate": 2.7759658604207705e-07, + "loss": 0.4635133743286133, + "step": 1832 + }, + { + "epoch": 1.795980884695503, + "grad_norm": 0.43681633472442627, + "learning_rate": 2.7499590642665773e-07, + "loss": 0.3901320993900299, + "step": 1833 + }, + { + "epoch": 1.79696115672099, + "grad_norm": 0.37042954564094543, + "learning_rate": 2.724071220034158e-07, + "loss": 0.36036092042922974, + "step": 1834 + }, + { + "epoch": 1.7979414287464772, + "grad_norm": 0.4252375662326813, + "learning_rate": 2.6983023928961406e-07, + "loss": 0.28743281960487366, + "step": 1835 + }, + { + "epoch": 1.7989217007719642, + "grad_norm": 0.3786608576774597, + "learning_rate": 2.6726526477254986e-07, + "loss": 0.38278865814208984, + "step": 1836 + }, + { + "epoch": 1.7999019727974512, + "grad_norm": 0.37242409586906433, + "learning_rate": 2.647122049095463e-07, + "loss": 0.345023512840271, + "step": 1837 + }, + { + "epoch": 1.8008822448229385, + "grad_norm": 0.3937869966030121, + "learning_rate": 2.621710661279253e-07, + "loss": 0.3539823889732361, + "step": 1838 + }, + { + "epoch": 1.8018625168484255, + "grad_norm": 0.37955591082572937, + "learning_rate": 2.596418548250029e-07, + "loss": 0.3501545190811157, + "step": 1839 + }, + { + "epoch": 1.8028427888739125, + "grad_norm": 0.3294471204280853, + "learning_rate": 2.571245773680647e-07, + "loss": 0.2854633033275604, + "step": 1840 + }, + { + "epoch": 1.8038230608993997, + "grad_norm": 0.38820013403892517, + "learning_rate": 2.546192400943537e-07, + "loss": 0.40697968006134033, + "step": 1841 + }, + { + "epoch": 1.8048033329248867, + "grad_norm": 0.4043434262275696, + "learning_rate": 2.52125849311054e-07, + "loss": 0.4158911108970642, + "step": 1842 + }, + { + "epoch": 1.8057836049503737, + "grad_norm": 0.46304336190223694, + "learning_rate": 2.4964441129527337e-07, + "loss": 0.5164629220962524, + "step": 1843 + }, + { + "epoch": 1.806763876975861, + "grad_norm": 0.2915133535861969, + "learning_rate": 2.4717493229402976e-07, + "loss": 0.272190123796463, + "step": 1844 + }, + { + "epoch": 1.8077441490013477, + "grad_norm": 0.37358495593070984, + "learning_rate": 2.447174185242324e-07, + "loss": 0.344493567943573, + "step": 1845 + }, + { + "epoch": 1.808724421026835, + "grad_norm": 0.37864813208580017, + "learning_rate": 2.4227187617267057e-07, + "loss": 0.375437468290329, + "step": 1846 + }, + { + "epoch": 1.8097046930523222, + "grad_norm": 0.31568974256515503, + "learning_rate": 2.398383113959929e-07, + "loss": 0.3203946650028229, + "step": 1847 + }, + { + "epoch": 1.810684965077809, + "grad_norm": 0.33784857392311096, + "learning_rate": 2.3741673032069757e-07, + "loss": 0.3141031265258789, + "step": 1848 + }, + { + "epoch": 1.8116652371032962, + "grad_norm": 0.3876346945762634, + "learning_rate": 2.3500713904311023e-07, + "loss": 0.346286803483963, + "step": 1849 + }, + { + "epoch": 1.8126455091287832, + "grad_norm": 0.3582172393798828, + "learning_rate": 2.3260954362937626e-07, + "loss": 0.3255946636199951, + "step": 1850 + }, + { + "epoch": 1.8136257811542702, + "grad_norm": 1.4689475297927856, + "learning_rate": 2.3022395011543687e-07, + "loss": 0.43359845876693726, + "step": 1851 + }, + { + "epoch": 1.8146060531797574, + "grad_norm": 0.39466992020606995, + "learning_rate": 2.2785036450702237e-07, + "loss": 0.35341185331344604, + "step": 1852 + }, + { + "epoch": 1.8155863252052444, + "grad_norm": 0.38006591796875, + "learning_rate": 2.2548879277963065e-07, + "loss": 0.36566489934921265, + "step": 1853 + }, + { + "epoch": 1.8165665972307314, + "grad_norm": 0.41100892424583435, + "learning_rate": 2.2313924087851657e-07, + "loss": 0.3486058712005615, + "step": 1854 + }, + { + "epoch": 1.8175468692562187, + "grad_norm": 0.45427221059799194, + "learning_rate": 2.2080171471867362e-07, + "loss": 0.36333462595939636, + "step": 1855 + }, + { + "epoch": 1.8185271412817057, + "grad_norm": 0.4660656452178955, + "learning_rate": 2.1847622018482283e-07, + "loss": 0.4000430405139923, + "step": 1856 + }, + { + "epoch": 1.8195074133071927, + "grad_norm": 0.3475498557090759, + "learning_rate": 2.161627631313923e-07, + "loss": 0.35434359312057495, + "step": 1857 + }, + { + "epoch": 1.82048768533268, + "grad_norm": 0.3405502140522003, + "learning_rate": 2.1386134938250645e-07, + "loss": 0.3633918762207031, + "step": 1858 + }, + { + "epoch": 1.821467957358167, + "grad_norm": 0.30488121509552, + "learning_rate": 2.1157198473197417e-07, + "loss": 0.2960091829299927, + "step": 1859 + }, + { + "epoch": 1.822448229383654, + "grad_norm": 0.30573368072509766, + "learning_rate": 2.0929467494326616e-07, + "loss": 0.2696135342121124, + "step": 1860 + }, + { + "epoch": 1.8234285014091411, + "grad_norm": 0.3861675262451172, + "learning_rate": 2.0702942574950812e-07, + "loss": 0.3925628066062927, + "step": 1861 + }, + { + "epoch": 1.8244087734346282, + "grad_norm": 0.39956194162368774, + "learning_rate": 2.047762428534622e-07, + "loss": 0.4451996684074402, + "step": 1862 + }, + { + "epoch": 1.8253890454601152, + "grad_norm": 0.4141245484352112, + "learning_rate": 2.0253513192751374e-07, + "loss": 0.48541459441185, + "step": 1863 + }, + { + "epoch": 1.8263693174856024, + "grad_norm": 0.3621307611465454, + "learning_rate": 2.0030609861365568e-07, + "loss": 0.3605848550796509, + "step": 1864 + }, + { + "epoch": 1.8273495895110892, + "grad_norm": 0.3806905746459961, + "learning_rate": 1.9808914852347817e-07, + "loss": 0.3863823115825653, + "step": 1865 + }, + { + "epoch": 1.8283298615365764, + "grad_norm": 0.40031641721725464, + "learning_rate": 1.9588428723814945e-07, + "loss": 0.38765889406204224, + "step": 1866 + }, + { + "epoch": 1.8293101335620636, + "grad_norm": 0.4227183759212494, + "learning_rate": 1.9369152030840553e-07, + "loss": 0.34702572226524353, + "step": 1867 + }, + { + "epoch": 1.8302904055875504, + "grad_norm": 0.42528119683265686, + "learning_rate": 1.9151085325453512e-07, + "loss": 0.3112640380859375, + "step": 1868 + }, + { + "epoch": 1.8312706776130376, + "grad_norm": 0.3496108949184418, + "learning_rate": 1.8934229156636453e-07, + "loss": 0.30526429414749146, + "step": 1869 + }, + { + "epoch": 1.8322509496385246, + "grad_norm": 0.4570808708667755, + "learning_rate": 1.8718584070324452e-07, + "loss": 0.43626856803894043, + "step": 1870 + }, + { + "epoch": 1.8332312216640116, + "grad_norm": 0.6757300496101379, + "learning_rate": 1.8504150609403858e-07, + "loss": 0.3831341862678528, + "step": 1871 + }, + { + "epoch": 1.8342114936894989, + "grad_norm": 0.4172285497188568, + "learning_rate": 1.8290929313710514e-07, + "loss": 0.3849405348300934, + "step": 1872 + }, + { + "epoch": 1.8351917657149859, + "grad_norm": 0.3603311777114868, + "learning_rate": 1.807892072002898e-07, + "loss": 0.3212820589542389, + "step": 1873 + }, + { + "epoch": 1.8361720377404729, + "grad_norm": 0.3588409125804901, + "learning_rate": 1.7868125362090482e-07, + "loss": 0.32618218660354614, + "step": 1874 + }, + { + "epoch": 1.8371523097659601, + "grad_norm": 0.3422310948371887, + "learning_rate": 1.765854377057219e-07, + "loss": 0.28395843505859375, + "step": 1875 + }, + { + "epoch": 1.8381325817914471, + "grad_norm": 0.3215237557888031, + "learning_rate": 1.7450176473095436e-07, + "loss": 0.3091646432876587, + "step": 1876 + }, + { + "epoch": 1.8391128538169341, + "grad_norm": 0.4790702164173126, + "learning_rate": 1.724302399422456e-07, + "loss": 0.47884172201156616, + "step": 1877 + }, + { + "epoch": 1.8400931258424214, + "grad_norm": 0.3550809621810913, + "learning_rate": 1.7037086855465902e-07, + "loss": 0.3525258004665375, + "step": 1878 + }, + { + "epoch": 1.8410733978679084, + "grad_norm": 0.3271271288394928, + "learning_rate": 1.6832365575265742e-07, + "loss": 0.3195466101169586, + "step": 1879 + }, + { + "epoch": 1.8420536698933954, + "grad_norm": 0.4773300886154175, + "learning_rate": 1.6628860669009816e-07, + "loss": 0.36375588178634644, + "step": 1880 + }, + { + "epoch": 1.8430339419188826, + "grad_norm": 0.3409133553504944, + "learning_rate": 1.6426572649021477e-07, + "loss": 0.369698166847229, + "step": 1881 + }, + { + "epoch": 1.8440142139443696, + "grad_norm": 0.47618257999420166, + "learning_rate": 1.6225502024560468e-07, + "loss": 0.44566285610198975, + "step": 1882 + }, + { + "epoch": 1.8449944859698566, + "grad_norm": 0.37478893995285034, + "learning_rate": 1.6025649301821877e-07, + "loss": 0.3461802005767822, + "step": 1883 + }, + { + "epoch": 1.8459747579953438, + "grad_norm": 0.4216592311859131, + "learning_rate": 1.582701498393474e-07, + "loss": 0.3890058398246765, + "step": 1884 + }, + { + "epoch": 1.8469550300208308, + "grad_norm": 0.3270227909088135, + "learning_rate": 1.562959957096072e-07, + "loss": 0.31345048546791077, + "step": 1885 + }, + { + "epoch": 1.8479353020463178, + "grad_norm": 0.40899965167045593, + "learning_rate": 1.543340355989287e-07, + "loss": 0.3930346965789795, + "step": 1886 + }, + { + "epoch": 1.848915574071805, + "grad_norm": 0.4171411991119385, + "learning_rate": 1.5238427444654368e-07, + "loss": 0.42073363065719604, + "step": 1887 + }, + { + "epoch": 1.8498958460972919, + "grad_norm": 0.39931097626686096, + "learning_rate": 1.5044671716097414e-07, + "loss": 0.3975702226161957, + "step": 1888 + }, + { + "epoch": 1.850876118122779, + "grad_norm": 0.3799835741519928, + "learning_rate": 1.4852136862001766e-07, + "loss": 0.3898460566997528, + "step": 1889 + }, + { + "epoch": 1.8518563901482663, + "grad_norm": 0.36510559916496277, + "learning_rate": 1.466082336707375e-07, + "loss": 0.4031357169151306, + "step": 1890 + }, + { + "epoch": 1.852836662173753, + "grad_norm": 0.33252233266830444, + "learning_rate": 1.4470731712944885e-07, + "loss": 0.38171181082725525, + "step": 1891 + }, + { + "epoch": 1.8538169341992403, + "grad_norm": 0.33753806352615356, + "learning_rate": 1.428186237817053e-07, + "loss": 0.28889352083206177, + "step": 1892 + }, + { + "epoch": 1.8547972062247273, + "grad_norm": 0.3914101719856262, + "learning_rate": 1.4094215838229176e-07, + "loss": 0.3529321551322937, + "step": 1893 + }, + { + "epoch": 1.8557774782502143, + "grad_norm": 0.3802926540374756, + "learning_rate": 1.3907792565520716e-07, + "loss": 0.358803391456604, + "step": 1894 + }, + { + "epoch": 1.8567577502757016, + "grad_norm": 0.3857244849205017, + "learning_rate": 1.372259302936546e-07, + "loss": 0.33400875329971313, + "step": 1895 + }, + { + "epoch": 1.8577380223011886, + "grad_norm": 0.37638941407203674, + "learning_rate": 1.3538617696003066e-07, + "loss": 0.3147771656513214, + "step": 1896 + }, + { + "epoch": 1.8587182943266756, + "grad_norm": 0.4054686725139618, + "learning_rate": 1.3355867028591209e-07, + "loss": 0.3243183493614197, + "step": 1897 + }, + { + "epoch": 1.8596985663521628, + "grad_norm": 0.3516264855861664, + "learning_rate": 1.3174341487204478e-07, + "loss": 0.3295627534389496, + "step": 1898 + }, + { + "epoch": 1.8606788383776498, + "grad_norm": 0.36703845858573914, + "learning_rate": 1.2994041528833267e-07, + "loss": 0.3564237058162689, + "step": 1899 + }, + { + "epoch": 1.8616591104031368, + "grad_norm": 0.37199637293815613, + "learning_rate": 1.2814967607382433e-07, + "loss": 0.3491576015949249, + "step": 1900 + }, + { + "epoch": 1.862639382428624, + "grad_norm": 0.3504694700241089, + "learning_rate": 1.263712017367036e-07, + "loss": 0.32253769040107727, + "step": 1901 + }, + { + "epoch": 1.863619654454111, + "grad_norm": 0.38024047017097473, + "learning_rate": 1.246049967542773e-07, + "loss": 0.38772034645080566, + "step": 1902 + }, + { + "epoch": 1.864599926479598, + "grad_norm": 0.3988204002380371, + "learning_rate": 1.2285106557296479e-07, + "loss": 0.35567206144332886, + "step": 1903 + }, + { + "epoch": 1.8655801985050853, + "grad_norm": 0.3725142180919647, + "learning_rate": 1.211094126082857e-07, + "loss": 0.3349960744380951, + "step": 1904 + }, + { + "epoch": 1.8665604705305723, + "grad_norm": 0.3695859909057617, + "learning_rate": 1.193800422448499e-07, + "loss": 0.3461756706237793, + "step": 1905 + }, + { + "epoch": 1.8675407425560593, + "grad_norm": 0.3805686831474304, + "learning_rate": 1.1766295883634427e-07, + "loss": 0.37669533491134644, + "step": 1906 + }, + { + "epoch": 1.8685210145815465, + "grad_norm": 0.4000875651836395, + "learning_rate": 1.1595816670552429e-07, + "loss": 0.4143308401107788, + "step": 1907 + }, + { + "epoch": 1.8695012866070333, + "grad_norm": 0.36493295431137085, + "learning_rate": 1.1426567014420297e-07, + "loss": 0.3761562705039978, + "step": 1908 + }, + { + "epoch": 1.8704815586325205, + "grad_norm": 0.3971060514450073, + "learning_rate": 1.12585473413237e-07, + "loss": 0.3754567801952362, + "step": 1909 + }, + { + "epoch": 1.8714618306580078, + "grad_norm": 0.3939042389392853, + "learning_rate": 1.1091758074252112e-07, + "loss": 0.379552960395813, + "step": 1910 + }, + { + "epoch": 1.8724421026834945, + "grad_norm": 0.39711931347846985, + "learning_rate": 1.0926199633097156e-07, + "loss": 0.3647288680076599, + "step": 1911 + }, + { + "epoch": 1.8734223747089818, + "grad_norm": 0.33604320883750916, + "learning_rate": 1.076187243465221e-07, + "loss": 0.28570079803466797, + "step": 1912 + }, + { + "epoch": 1.8744026467344688, + "grad_norm": 0.3451414704322815, + "learning_rate": 1.0598776892610685e-07, + "loss": 0.31603747606277466, + "step": 1913 + }, + { + "epoch": 1.8753829187599558, + "grad_norm": 0.4146670699119568, + "learning_rate": 1.0436913417565364e-07, + "loss": 0.37732064723968506, + "step": 1914 + }, + { + "epoch": 1.876363190785443, + "grad_norm": 0.3134925663471222, + "learning_rate": 1.0276282417007399e-07, + "loss": 0.24549098312854767, + "step": 1915 + }, + { + "epoch": 1.87734346281093, + "grad_norm": 0.346036434173584, + "learning_rate": 1.0116884295325202e-07, + "loss": 0.337798535823822, + "step": 1916 + }, + { + "epoch": 1.878323734836417, + "grad_norm": 0.34530937671661377, + "learning_rate": 9.958719453803278e-08, + "loss": 0.29336920380592346, + "step": 1917 + }, + { + "epoch": 1.8793040068619042, + "grad_norm": 0.41307905316352844, + "learning_rate": 9.801788290621505e-08, + "loss": 0.43094706535339355, + "step": 1918 + }, + { + "epoch": 1.8802842788873912, + "grad_norm": 0.41243818402290344, + "learning_rate": 9.646091200853802e-08, + "loss": 0.42707014083862305, + "step": 1919 + }, + { + "epoch": 1.8812645509128783, + "grad_norm": 0.33866050839424133, + "learning_rate": 9.491628576467516e-08, + "loss": 0.2805923819541931, + "step": 1920 + }, + { + "epoch": 1.8822448229383655, + "grad_norm": 0.4044766128063202, + "learning_rate": 9.338400806321979e-08, + "loss": 0.45327508449554443, + "step": 1921 + }, + { + "epoch": 1.8832250949638525, + "grad_norm": 0.3599652051925659, + "learning_rate": 9.186408276168012e-08, + "loss": 0.34058117866516113, + "step": 1922 + }, + { + "epoch": 1.8842053669893395, + "grad_norm": 0.4139306843280792, + "learning_rate": 9.035651368646647e-08, + "loss": 0.3773718476295471, + "step": 1923 + }, + { + "epoch": 1.8851856390148267, + "grad_norm": 0.43189549446105957, + "learning_rate": 8.886130463288178e-08, + "loss": 0.40143847465515137, + "step": 1924 + }, + { + "epoch": 1.8861659110403137, + "grad_norm": 0.3982018232345581, + "learning_rate": 8.737845936511335e-08, + "loss": 0.4041132628917694, + "step": 1925 + }, + { + "epoch": 1.8871461830658007, + "grad_norm": 0.48200032114982605, + "learning_rate": 8.590798161622227e-08, + "loss": 0.4371276795864105, + "step": 1926 + }, + { + "epoch": 1.888126455091288, + "grad_norm": 0.37154990434646606, + "learning_rate": 8.444987508813451e-08, + "loss": 0.41200995445251465, + "step": 1927 + }, + { + "epoch": 1.8891067271167747, + "grad_norm": 0.39796820282936096, + "learning_rate": 8.300414345163043e-08, + "loss": 0.3685952425003052, + "step": 1928 + }, + { + "epoch": 1.890086999142262, + "grad_norm": 0.38173747062683105, + "learning_rate": 8.157079034633974e-08, + "loss": 0.38268765807151794, + "step": 1929 + }, + { + "epoch": 1.8910672711677492, + "grad_norm": 0.3422482907772064, + "learning_rate": 8.014981938072542e-08, + "loss": 0.3192417621612549, + "step": 1930 + }, + { + "epoch": 1.892047543193236, + "grad_norm": 0.3937561810016632, + "learning_rate": 7.874123413208145e-08, + "loss": 0.3470272421836853, + "step": 1931 + }, + { + "epoch": 1.8930278152187232, + "grad_norm": 0.4789409339427948, + "learning_rate": 7.734503814651906e-08, + "loss": 0.45786482095718384, + "step": 1932 + }, + { + "epoch": 1.8940080872442102, + "grad_norm": 0.36675921082496643, + "learning_rate": 7.59612349389599e-08, + "loss": 0.3520694971084595, + "step": 1933 + }, + { + "epoch": 1.8949883592696972, + "grad_norm": 0.3945583403110504, + "learning_rate": 7.458982799312676e-08, + "loss": 0.3263518214225769, + "step": 1934 + }, + { + "epoch": 1.8959686312951844, + "grad_norm": 0.3871973156929016, + "learning_rate": 7.32308207615351e-08, + "loss": 0.4045059084892273, + "step": 1935 + }, + { + "epoch": 1.8969489033206715, + "grad_norm": 0.4343588352203369, + "learning_rate": 7.188421666548373e-08, + "loss": 0.3911435604095459, + "step": 1936 + }, + { + "epoch": 1.8979291753461585, + "grad_norm": 0.31257694959640503, + "learning_rate": 7.055001909504755e-08, + "loss": 0.2708739936351776, + "step": 1937 + }, + { + "epoch": 1.8989094473716457, + "grad_norm": 0.3348176181316376, + "learning_rate": 6.922823140906754e-08, + "loss": 0.3335874676704407, + "step": 1938 + }, + { + "epoch": 1.8998897193971327, + "grad_norm": 0.3991497755050659, + "learning_rate": 6.791885693514134e-08, + "loss": 0.36584001779556274, + "step": 1939 + }, + { + "epoch": 1.9008699914226197, + "grad_norm": 0.33452481031417847, + "learning_rate": 6.662189896961824e-08, + "loss": 0.3227638602256775, + "step": 1940 + }, + { + "epoch": 1.901850263448107, + "grad_norm": 0.3502449691295624, + "learning_rate": 6.533736077758868e-08, + "loss": 0.33033889532089233, + "step": 1941 + }, + { + "epoch": 1.902830535473594, + "grad_norm": 0.36946505308151245, + "learning_rate": 6.406524559287641e-08, + "loss": 0.35373038053512573, + "step": 1942 + }, + { + "epoch": 1.903810807499081, + "grad_norm": 0.4744099974632263, + "learning_rate": 6.280555661802857e-08, + "loss": 0.4236101806163788, + "step": 1943 + }, + { + "epoch": 1.9047910795245682, + "grad_norm": 0.47161856293678284, + "learning_rate": 6.15582970243117e-08, + "loss": 0.4352308511734009, + "step": 1944 + }, + { + "epoch": 1.9057713515500552, + "grad_norm": 0.3910539448261261, + "learning_rate": 6.032346995169968e-08, + "loss": 0.4637935161590576, + "step": 1945 + }, + { + "epoch": 1.9067516235755422, + "grad_norm": 0.3528776466846466, + "learning_rate": 5.9101078508869144e-08, + "loss": 0.3925752639770508, + "step": 1946 + }, + { + "epoch": 1.9077318956010294, + "grad_norm": 0.3184770941734314, + "learning_rate": 5.7891125773187896e-08, + "loss": 0.31664010882377625, + "step": 1947 + }, + { + "epoch": 1.9087121676265164, + "grad_norm": 0.3500765860080719, + "learning_rate": 5.669361479071156e-08, + "loss": 0.32596200704574585, + "step": 1948 + }, + { + "epoch": 1.9096924396520034, + "grad_norm": 0.42387109994888306, + "learning_rate": 5.550854857617194e-08, + "loss": 0.39374762773513794, + "step": 1949 + }, + { + "epoch": 1.9106727116774906, + "grad_norm": 0.37398621439933777, + "learning_rate": 5.4335930112972e-08, + "loss": 0.39457735419273376, + "step": 1950 + }, + { + "epoch": 1.9116529837029774, + "grad_norm": 0.364576131105423, + "learning_rate": 5.3175762353177563e-08, + "loss": 0.3248162567615509, + "step": 1951 + }, + { + "epoch": 1.9126332557284647, + "grad_norm": 0.4627110958099365, + "learning_rate": 5.202804821750951e-08, + "loss": 0.29501545429229736, + "step": 1952 + }, + { + "epoch": 1.9136135277539519, + "grad_norm": 0.40157920122146606, + "learning_rate": 5.089279059533658e-08, + "loss": 0.3362949788570404, + "step": 1953 + }, + { + "epoch": 1.9145937997794387, + "grad_norm": 0.47221657633781433, + "learning_rate": 4.9769992344668174e-08, + "loss": 0.3179032802581787, + "step": 1954 + }, + { + "epoch": 1.915574071804926, + "grad_norm": 0.3419935703277588, + "learning_rate": 4.865965629214819e-08, + "loss": 0.34885525703430176, + "step": 1955 + }, + { + "epoch": 1.916554343830413, + "grad_norm": 0.3962067663669586, + "learning_rate": 4.756178523304622e-08, + "loss": 0.36980053782463074, + "step": 1956 + }, + { + "epoch": 1.9175346158559, + "grad_norm": 0.408200204372406, + "learning_rate": 4.6476381931251366e-08, + "loss": 0.3871716558933258, + "step": 1957 + }, + { + "epoch": 1.9185148878813871, + "grad_norm": 0.4055185317993164, + "learning_rate": 4.5403449119265085e-08, + "loss": 0.41482850909233093, + "step": 1958 + }, + { + "epoch": 1.9194951599068741, + "grad_norm": 0.4378550052642822, + "learning_rate": 4.434298949819449e-08, + "loss": 0.42568671703338623, + "step": 1959 + }, + { + "epoch": 1.9204754319323611, + "grad_norm": 0.38033613562583923, + "learning_rate": 4.329500573774459e-08, + "loss": 0.36229217052459717, + "step": 1960 + }, + { + "epoch": 1.9214557039578484, + "grad_norm": 0.39758366346359253, + "learning_rate": 4.225950047621441e-08, + "loss": 0.2938281297683716, + "step": 1961 + }, + { + "epoch": 1.9224359759833354, + "grad_norm": 0.3471410274505615, + "learning_rate": 4.123647632048644e-08, + "loss": 0.3377930521965027, + "step": 1962 + }, + { + "epoch": 1.9234162480088224, + "grad_norm": 0.36092984676361084, + "learning_rate": 4.02259358460233e-08, + "loss": 0.3372440040111542, + "step": 1963 + }, + { + "epoch": 1.9243965200343096, + "grad_norm": 0.5835425853729248, + "learning_rate": 3.922788159685942e-08, + "loss": 0.4844486713409424, + "step": 1964 + }, + { + "epoch": 1.9253767920597966, + "grad_norm": 0.3920924663543701, + "learning_rate": 3.8242316085594923e-08, + "loss": 0.38885951042175293, + "step": 1965 + }, + { + "epoch": 1.9263570640852836, + "grad_norm": 0.3648390471935272, + "learning_rate": 3.726924179339009e-08, + "loss": 0.3539251387119293, + "step": 1966 + }, + { + "epoch": 1.9273373361107708, + "grad_norm": 0.34048476815223694, + "learning_rate": 3.630866116995757e-08, + "loss": 0.32399624586105347, + "step": 1967 + }, + { + "epoch": 1.9283176081362579, + "grad_norm": 0.3979117274284363, + "learning_rate": 3.536057663355852e-08, + "loss": 0.40811002254486084, + "step": 1968 + }, + { + "epoch": 1.9292978801617449, + "grad_norm": 0.3613481819629669, + "learning_rate": 3.44249905709948e-08, + "loss": 0.3490302264690399, + "step": 1969 + }, + { + "epoch": 1.930278152187232, + "grad_norm": 0.40579718351364136, + "learning_rate": 3.350190533760289e-08, + "loss": 0.41862714290618896, + "step": 1970 + }, + { + "epoch": 1.9312584242127189, + "grad_norm": 0.3858543634414673, + "learning_rate": 3.25913232572489e-08, + "loss": 0.3664434552192688, + "step": 1971 + }, + { + "epoch": 1.932238696238206, + "grad_norm": 0.35242167115211487, + "learning_rate": 3.169324662232187e-08, + "loss": 0.36104094982147217, + "step": 1972 + }, + { + "epoch": 1.9332189682636933, + "grad_norm": 0.3108178377151489, + "learning_rate": 3.080767769372939e-08, + "loss": 0.3001745343208313, + "step": 1973 + }, + { + "epoch": 1.93419924028918, + "grad_norm": 0.3411625623703003, + "learning_rate": 2.9934618700889205e-08, + "loss": 0.28352412581443787, + "step": 1974 + }, + { + "epoch": 1.9351795123146673, + "grad_norm": 0.38755854964256287, + "learning_rate": 2.907407184172706e-08, + "loss": 0.379106342792511, + "step": 1975 + }, + { + "epoch": 1.9361597843401543, + "grad_norm": 0.8897514939308167, + "learning_rate": 2.8226039282667762e-08, + "loss": 0.37476542592048645, + "step": 1976 + }, + { + "epoch": 1.9371400563656413, + "grad_norm": 0.4192017614841461, + "learning_rate": 2.7390523158633552e-08, + "loss": 0.37311655282974243, + "step": 1977 + }, + { + "epoch": 1.9381203283911286, + "grad_norm": 0.3789272904396057, + "learning_rate": 2.6567525573034102e-08, + "loss": 0.36981621384620667, + "step": 1978 + }, + { + "epoch": 1.9391006004166156, + "grad_norm": 0.3810921907424927, + "learning_rate": 2.57570485977654e-08, + "loss": 0.3803723454475403, + "step": 1979 + }, + { + "epoch": 1.9400808724421026, + "grad_norm": 0.33650895953178406, + "learning_rate": 2.495909427320198e-08, + "loss": 0.3219085931777954, + "step": 1980 + }, + { + "epoch": 1.9410611444675898, + "grad_norm": 0.41487401723861694, + "learning_rate": 2.4173664608193592e-08, + "loss": 0.46797502040863037, + "step": 1981 + }, + { + "epoch": 1.9420414164930768, + "grad_norm": 0.4950869381427765, + "learning_rate": 2.3400761580057442e-08, + "loss": 0.3930601179599762, + "step": 1982 + }, + { + "epoch": 1.9430216885185638, + "grad_norm": 0.3461202383041382, + "learning_rate": 2.264038713457706e-08, + "loss": 0.33587607741355896, + "step": 1983 + }, + { + "epoch": 1.944001960544051, + "grad_norm": 0.3365919291973114, + "learning_rate": 2.189254318599343e-08, + "loss": 0.304770827293396, + "step": 1984 + }, + { + "epoch": 1.944982232569538, + "grad_norm": 0.3526042699813843, + "learning_rate": 2.1157231617002783e-08, + "loss": 0.31271421909332275, + "step": 1985 + }, + { + "epoch": 1.945962504595025, + "grad_norm": 0.4145696759223938, + "learning_rate": 2.0434454278752126e-08, + "loss": 0.38078755140304565, + "step": 1986 + }, + { + "epoch": 1.9469427766205123, + "grad_norm": 0.2975658178329468, + "learning_rate": 1.9724212990830938e-08, + "loss": 0.2699826955795288, + "step": 1987 + }, + { + "epoch": 1.9479230486459993, + "grad_norm": 0.39280709624290466, + "learning_rate": 1.9026509541272276e-08, + "loss": 0.36080029606819153, + "step": 1988 + }, + { + "epoch": 1.9489033206714863, + "grad_norm": 0.3775005340576172, + "learning_rate": 1.834134568654333e-08, + "loss": 0.39978429675102234, + "step": 1989 + }, + { + "epoch": 1.9498835926969735, + "grad_norm": 0.4009869396686554, + "learning_rate": 1.7668723151542667e-08, + "loss": 0.34865182638168335, + "step": 1990 + }, + { + "epoch": 1.9508638647224603, + "grad_norm": 0.5271181464195251, + "learning_rate": 1.7008643629596866e-08, + "loss": 0.5259040594100952, + "step": 1991 + }, + { + "epoch": 1.9518441367479475, + "grad_norm": 0.3561205565929413, + "learning_rate": 1.6361108782456113e-08, + "loss": 0.3491579592227936, + "step": 1992 + }, + { + "epoch": 1.9528244087734348, + "grad_norm": 0.402241051197052, + "learning_rate": 1.5726120240288632e-08, + "loss": 0.3347298502922058, + "step": 1993 + }, + { + "epoch": 1.9538046807989216, + "grad_norm": 0.3009505271911621, + "learning_rate": 1.510367960167569e-08, + "loss": 0.25810128450393677, + "step": 1994 + }, + { + "epoch": 1.9547849528244088, + "grad_norm": 0.442399263381958, + "learning_rate": 1.449378843361271e-08, + "loss": 0.3368980288505554, + "step": 1995 + }, + { + "epoch": 1.955765224849896, + "grad_norm": 0.3836606740951538, + "learning_rate": 1.3896448271499275e-08, + "loss": 0.38483136892318726, + "step": 1996 + }, + { + "epoch": 1.9567454968753828, + "grad_norm": 0.3858362138271332, + "learning_rate": 1.3311660619138578e-08, + "loss": 0.37200963497161865, + "step": 1997 + }, + { + "epoch": 1.95772576890087, + "grad_norm": 0.34064731001853943, + "learning_rate": 1.2739426948732426e-08, + "loss": 0.28852540254592896, + "step": 1998 + }, + { + "epoch": 1.958706040926357, + "grad_norm": 0.3998526930809021, + "learning_rate": 1.2179748700879013e-08, + "loss": 0.3903662860393524, + "step": 1999 + }, + { + "epoch": 1.959686312951844, + "grad_norm": 0.34466320276260376, + "learning_rate": 1.1632627284566822e-08, + "loss": 0.3239680528640747, + "step": 2000 + }, + { + "epoch": 1.9606665849773313, + "grad_norm": 0.43105149269104004, + "learning_rate": 1.109806407717462e-08, + "loss": 0.3704794943332672, + "step": 2001 + }, + { + "epoch": 1.9616468570028183, + "grad_norm": 0.4494905173778534, + "learning_rate": 1.0576060424463686e-08, + "loss": 0.4241977035999298, + "step": 2002 + }, + { + "epoch": 1.9626271290283053, + "grad_norm": 0.34961533546447754, + "learning_rate": 1.006661764057837e-08, + "loss": 0.32587072253227234, + "step": 2003 + }, + { + "epoch": 1.9636074010537925, + "grad_norm": 0.41239458322525024, + "learning_rate": 9.56973700803887e-09, + "loss": 0.3496946096420288, + "step": 2004 + }, + { + "epoch": 1.9645876730792795, + "grad_norm": 0.46796709299087524, + "learning_rate": 9.085419777743465e-09, + "loss": 0.4969180226325989, + "step": 2005 + }, + { + "epoch": 1.9655679451047665, + "grad_norm": 0.3892061710357666, + "learning_rate": 8.613667168958506e-09, + "loss": 0.3607162833213806, + "step": 2006 + }, + { + "epoch": 1.9665482171302537, + "grad_norm": 0.4369899332523346, + "learning_rate": 8.15448036932176e-09, + "loss": 0.41520261764526367, + "step": 2007 + }, + { + "epoch": 1.9675284891557407, + "grad_norm": 0.37779292464256287, + "learning_rate": 7.707860534834632e-09, + "loss": 0.3675624430179596, + "step": 2008 + }, + { + "epoch": 1.9685087611812278, + "grad_norm": 0.32598474621772766, + "learning_rate": 7.273808789862724e-09, + "loss": 0.29060912132263184, + "step": 2009 + }, + { + "epoch": 1.969489033206715, + "grad_norm": 0.4254051446914673, + "learning_rate": 6.852326227130835e-09, + "loss": 0.4241909384727478, + "step": 2010 + }, + { + "epoch": 1.970469305232202, + "grad_norm": 0.3962550759315491, + "learning_rate": 6.4434139077201865e-09, + "loss": 0.38578057289123535, + "step": 2011 + }, + { + "epoch": 1.971449577257689, + "grad_norm": 0.37741607427597046, + "learning_rate": 6.0470728610673155e-09, + "loss": 0.3685004711151123, + "step": 2012 + }, + { + "epoch": 1.9724298492831762, + "grad_norm": 0.4556925594806671, + "learning_rate": 5.6633040849601865e-09, + "loss": 0.44286853075027466, + "step": 2013 + }, + { + "epoch": 1.973410121308663, + "grad_norm": 0.36857128143310547, + "learning_rate": 5.292108545536523e-09, + "loss": 0.3409087061882019, + "step": 2014 + }, + { + "epoch": 1.9743903933341502, + "grad_norm": 0.42536187171936035, + "learning_rate": 4.933487177280483e-09, + "loss": 0.4396209716796875, + "step": 2015 + }, + { + "epoch": 1.9753706653596375, + "grad_norm": 0.3863128423690796, + "learning_rate": 4.587440883021543e-09, + "loss": 0.3482202887535095, + "step": 2016 + }, + { + "epoch": 1.9763509373851242, + "grad_norm": 0.33822304010391235, + "learning_rate": 4.253970533929508e-09, + "loss": 0.3142094016075134, + "step": 2017 + }, + { + "epoch": 1.9773312094106115, + "grad_norm": 0.43342381715774536, + "learning_rate": 3.9330769695167245e-09, + "loss": 0.31652769446372986, + "step": 2018 + }, + { + "epoch": 1.9783114814360985, + "grad_norm": 0.43887975811958313, + "learning_rate": 3.6247609976319818e-09, + "loss": 0.4079042077064514, + "step": 2019 + }, + { + "epoch": 1.9792917534615855, + "grad_norm": 0.3833581507205963, + "learning_rate": 3.3290233944605067e-09, + "loss": 0.3510245680809021, + "step": 2020 + }, + { + "epoch": 1.9802720254870727, + "grad_norm": 0.3753303587436676, + "learning_rate": 3.0458649045211897e-09, + "loss": 0.334653377532959, + "step": 2021 + }, + { + "epoch": 1.9812522975125597, + "grad_norm": 0.4557226896286011, + "learning_rate": 2.7752862406654755e-09, + "loss": 0.36727383732795715, + "step": 2022 + }, + { + "epoch": 1.9822325695380467, + "grad_norm": 0.34632253646850586, + "learning_rate": 2.5172880840745873e-09, + "loss": 0.32775968313217163, + "step": 2023 + }, + { + "epoch": 1.983212841563534, + "grad_norm": 0.3431369662284851, + "learning_rate": 2.2718710842584145e-09, + "loss": 0.3316638767719269, + "step": 2024 + }, + { + "epoch": 1.984193113589021, + "grad_norm": 0.377046674489975, + "learning_rate": 2.0390358590538507e-09, + "loss": 0.34596455097198486, + "step": 2025 + }, + { + "epoch": 1.985173385614508, + "grad_norm": 2.796781301498413, + "learning_rate": 1.8187829946236802e-09, + "loss": 0.4017394185066223, + "step": 2026 + }, + { + "epoch": 1.9861536576399952, + "grad_norm": 0.3372899889945984, + "learning_rate": 1.61111304545436e-09, + "loss": 0.2676371932029724, + "step": 2027 + }, + { + "epoch": 1.9871339296654822, + "grad_norm": 0.34505340456962585, + "learning_rate": 1.4160265343549084e-09, + "loss": 0.35526472330093384, + "step": 2028 + }, + { + "epoch": 1.9881142016909692, + "grad_norm": 0.3976952135562897, + "learning_rate": 1.2335239524541298e-09, + "loss": 0.3644595146179199, + "step": 2029 + }, + { + "epoch": 1.9890944737164564, + "grad_norm": 0.33283817768096924, + "learning_rate": 1.0636057592033899e-09, + "loss": 0.3527385890483856, + "step": 2030 + }, + { + "epoch": 1.9900747457419434, + "grad_norm": 0.33022892475128174, + "learning_rate": 9.062723823710651e-10, + "loss": 0.29703113436698914, + "step": 2031 + }, + { + "epoch": 1.9910550177674304, + "grad_norm": 0.4557890295982361, + "learning_rate": 7.615242180436521e-10, + "loss": 0.4306028187274933, + "step": 2032 + }, + { + "epoch": 1.9920352897929177, + "grad_norm": 0.4146806001663208, + "learning_rate": 6.293616306246586e-10, + "loss": 0.4078826904296875, + "step": 2033 + }, + { + "epoch": 1.9930155618184044, + "grad_norm": 0.3461019992828369, + "learning_rate": 5.09784952833492e-10, + "loss": 0.33398452401161194, + "step": 2034 + }, + { + "epoch": 1.9939958338438917, + "grad_norm": 0.414501428604126, + "learning_rate": 4.027944857032395e-10, + "loss": 0.4134873151779175, + "step": 2035 + }, + { + "epoch": 1.994976105869379, + "grad_norm": 0.3774224519729614, + "learning_rate": 3.0839049858344406e-10, + "loss": 0.3409615159034729, + "step": 2036 + }, + { + "epoch": 1.9959563778948657, + "grad_norm": 0.4097723960876465, + "learning_rate": 2.265732291356626e-10, + "loss": 0.38473933935165405, + "step": 2037 + }, + { + "epoch": 1.996936649920353, + "grad_norm": 0.36617469787597656, + "learning_rate": 1.5734288333457692e-10, + "loss": 0.42465245723724365, + "step": 2038 + }, + { + "epoch": 1.99791692194584, + "grad_norm": 0.3955402374267578, + "learning_rate": 1.0069963546743833e-10, + "loss": 0.3782247304916382, + "step": 2039 + }, + { + "epoch": 1.998897193971327, + "grad_norm": 0.3959399461746216, + "learning_rate": 5.6643628134067655e-11, + "loss": 0.3760492205619812, + "step": 2040 + }, + { + "epoch": 1.9998774659968142, + "grad_norm": 0.3445758521556854, + "learning_rate": 2.5174972244634834e-11, + "loss": 0.3230331838130951, + "step": 2041 + }, + { + "epoch": 2.0, + "grad_norm": 1.267087459564209, + "learning_rate": 6.2937470224344596e-12, + "loss": 0.6405820846557617, + "step": 2042 + } + ], + "logging_steps": 1, + "max_steps": 2042, + "num_input_tokens_seen": 0, + "num_train_epochs": 2, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 2.0191018389843476e+18, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}