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": 580.9388427734375,
+      "learning_rate": 0.0,
+      "loss": 11.372468948364258,
+      "step": 1
+    },
+    {
+      "epoch": 0.0019605440509741453,
+      "grad_norm": 106.38761901855469,
+      "learning_rate": 1.6129032258064518e-07,
+      "loss": 11.29159164428711,
+      "step": 2
+    },
+    {
+      "epoch": 0.002940816076461218,
+      "grad_norm": 96.89288330078125,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 11.195578575134277,
+      "step": 3
+    },
+    {
+      "epoch": 0.003921088101948291,
+      "grad_norm": 102.1543960571289,
+      "learning_rate": 4.838709677419355e-07,
+      "loss": 11.074962615966797,
+      "step": 4
+    },
+    {
+      "epoch": 0.004901360127435363,
+      "grad_norm": 85.24600982666016,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 11.49647045135498,
+      "step": 5
+    },
+    {
+      "epoch": 0.005881632152922436,
+      "grad_norm": 677.4779052734375,
+      "learning_rate": 8.064516129032258e-07,
+      "loss": 10.837055206298828,
+      "step": 6
+    },
+    {
+      "epoch": 0.006861904178409509,
+      "grad_norm": 79.36517333984375,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 10.529215812683105,
+      "step": 7
+    },
+    {
+      "epoch": 0.007842176203896581,
+      "grad_norm": 88.23780822753906,
+      "learning_rate": 1.1290322580645162e-06,
+      "loss": 10.430451393127441,
+      "step": 8
+    },
+    {
+      "epoch": 0.008822448229383655,
+      "grad_norm": 97.00506591796875,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 9.306344032287598,
+      "step": 9
+    },
+    {
+      "epoch": 0.009802720254870726,
+      "grad_norm": 53.76038360595703,
+      "learning_rate": 1.4516129032258066e-06,
+      "loss": 9.023826599121094,
+      "step": 10
+    },
+    {
+      "epoch": 0.0107829922803578,
+      "grad_norm": 100.02516174316406,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 8.353536605834961,
+      "step": 11
+    },
+    {
+      "epoch": 0.011763264305844873,
+      "grad_norm": 64.13335418701172,
+      "learning_rate": 1.774193548387097e-06,
+      "loss": 7.085076808929443,
+      "step": 12
+    },
+    {
+      "epoch": 0.012743536331331944,
+      "grad_norm": 23.48418617248535,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 6.453302383422852,
+      "step": 13
+    },
+    {
+      "epoch": 0.013723808356819018,
+      "grad_norm": 28.481307983398438,
+      "learning_rate": 2.096774193548387e-06,
+      "loss": 6.847012042999268,
+      "step": 14
+    },
+    {
+      "epoch": 0.01470408038230609,
+      "grad_norm": 21.471084594726562,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 6.16412878036499,
+      "step": 15
+    },
+    {
+      "epoch": 0.015684352407793162,
+      "grad_norm": 16.86027717590332,
+      "learning_rate": 2.4193548387096776e-06,
+      "loss": 5.394906520843506,
+      "step": 16
+    },
+    {
+      "epoch": 0.016664624433280234,
+      "grad_norm": 17.459009170532227,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 5.069245338439941,
+      "step": 17
+    },
+    {
+      "epoch": 0.01764489645876731,
+      "grad_norm": 10.897882461547852,
+      "learning_rate": 2.7419354838709676e-06,
+      "loss": 4.876791954040527,
+      "step": 18
+    },
+    {
+      "epoch": 0.01862516848425438,
+      "grad_norm": 14.206769943237305,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 4.901210784912109,
+      "step": 19
+    },
+    {
+      "epoch": 0.019605440509741452,
+      "grad_norm": 7.508848190307617,
+      "learning_rate": 3.0645161290322584e-06,
+      "loss": 4.36484432220459,
+      "step": 20
+    },
+    {
+      "epoch": 0.020585712535228527,
+      "grad_norm": 81.67982482910156,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 4.08056640625,
+      "step": 21
+    },
+    {
+      "epoch": 0.0215659845607156,
+      "grad_norm": 8.69333553314209,
+      "learning_rate": 3.3870967741935484e-06,
+      "loss": 4.004530906677246,
+      "step": 22
+    },
+    {
+      "epoch": 0.02254625658620267,
+      "grad_norm": 6.333710193634033,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 3.9167585372924805,
+      "step": 23
+    },
+    {
+      "epoch": 0.023526528611689745,
+      "grad_norm": 5.399563789367676,
+      "learning_rate": 3.7096774193548392e-06,
+      "loss": 3.7403154373168945,
+      "step": 24
+    },
+    {
+      "epoch": 0.024506800637176817,
+      "grad_norm": 25.601049423217773,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 3.4189391136169434,
+      "step": 25
+    },
+    {
+      "epoch": 0.02548707266266389,
+      "grad_norm": 23.70180892944336,
+      "learning_rate": 4.032258064516129e-06,
+      "loss": 3.4575324058532715,
+      "step": 26
+    },
+    {
+      "epoch": 0.026467344688150964,
+      "grad_norm": 7.3810553550720215,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 3.2141077518463135,
+      "step": 27
+    },
+    {
+      "epoch": 0.027447616713638035,
+      "grad_norm": 3.960932970046997,
+      "learning_rate": 4.35483870967742e-06,
+      "loss": 2.892970085144043,
+      "step": 28
+    },
+    {
+      "epoch": 0.028427888739125107,
+      "grad_norm": 9.63996696472168,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 2.9488391876220703,
+      "step": 29
+    },
+    {
+      "epoch": 0.02940816076461218,
+      "grad_norm": 6.565059661865234,
+      "learning_rate": 4.67741935483871e-06,
+      "loss": 2.8827285766601562,
+      "step": 30
+    },
+    {
+      "epoch": 0.030388432790099253,
+      "grad_norm": 7.520769119262695,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 2.7817604541778564,
+      "step": 31
+    },
+    {
+      "epoch": 0.031368704815586325,
+      "grad_norm": 3.8251500129699707,
+      "learning_rate": 5e-06,
+      "loss": 2.6396312713623047,
+      "step": 32
+    },
+    {
+      "epoch": 0.0323489768410734,
+      "grad_norm": 4.61839485168457,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 2.982926368713379,
+      "step": 33
+    },
+    {
+      "epoch": 0.03332924886656047,
+      "grad_norm": 4.124711513519287,
+      "learning_rate": 5.322580645161291e-06,
+      "loss": 2.7103724479675293,
+      "step": 34
+    },
+    {
+      "epoch": 0.03430952089204754,
+      "grad_norm": 9.275640487670898,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 2.6467177867889404,
+      "step": 35
+    },
+    {
+      "epoch": 0.03528979291753462,
+      "grad_norm": 9.66199779510498,
+      "learning_rate": 5.645161290322582e-06,
+      "loss": 2.360792875289917,
+      "step": 36
+    },
+    {
+      "epoch": 0.036270064943021686,
+      "grad_norm": 11.543595314025879,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 2.2305736541748047,
+      "step": 37
+    },
+    {
+      "epoch": 0.03725033696850876,
+      "grad_norm": 4.072683334350586,
+      "learning_rate": 5.967741935483872e-06,
+      "loss": 2.6064844131469727,
+      "step": 38
+    },
+    {
+      "epoch": 0.038230608993995836,
+      "grad_norm": 5.190624713897705,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 2.3223533630371094,
+      "step": 39
+    },
+    {
+      "epoch": 0.039210881019482904,
+      "grad_norm": 6.913625717163086,
+      "learning_rate": 6.290322580645162e-06,
+      "loss": 1.9609746932983398,
+      "step": 40
+    },
+    {
+      "epoch": 0.04019115304496998,
+      "grad_norm": 5.53462028503418,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 2.0762758255004883,
+      "step": 41
+    },
+    {
+      "epoch": 0.041171425070457054,
+      "grad_norm": 5.241805553436279,
+      "learning_rate": 6.612903225806452e-06,
+      "loss": 1.8631705045700073,
+      "step": 42
+    },
+    {
+      "epoch": 0.04215169709594412,
+      "grad_norm": 4.304568767547607,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 2.16390061378479,
+      "step": 43
+    },
+    {
+      "epoch": 0.0431319691214312,
+      "grad_norm": 3.250955820083618,
+      "learning_rate": 6.935483870967743e-06,
+      "loss": 1.966679573059082,
+      "step": 44
+    },
+    {
+      "epoch": 0.04411224114691827,
+      "grad_norm": 4.159644603729248,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 1.6435564756393433,
+      "step": 45
+    },
+    {
+      "epoch": 0.04509251317240534,
+      "grad_norm": 4.289608001708984,
+      "learning_rate": 7.258064516129033e-06,
+      "loss": 1.889953851699829,
+      "step": 46
+    },
+    {
+      "epoch": 0.046072785197892416,
+      "grad_norm": 11.505047798156738,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 1.7259182929992676,
+      "step": 47
+    },
+    {
+      "epoch": 0.04705305722337949,
+      "grad_norm": 3.5180811882019043,
+      "learning_rate": 7.580645161290323e-06,
+      "loss": 1.7574291229248047,
+      "step": 48
+    },
+    {
+      "epoch": 0.04803332924886656,
+      "grad_norm": 4.833779811859131,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 2.000178813934326,
+      "step": 49
+    },
+    {
+      "epoch": 0.049013601274353634,
+      "grad_norm": 3.7938976287841797,
+      "learning_rate": 7.903225806451613e-06,
+      "loss": 1.8317043781280518,
+      "step": 50
+    },
+    {
+      "epoch": 0.04999387329984071,
+      "grad_norm": 23.372081756591797,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 2.134334087371826,
+      "step": 51
+    },
+    {
+      "epoch": 0.05097414532532778,
+      "grad_norm": 19.262693405151367,
+      "learning_rate": 8.225806451612904e-06,
+      "loss": 1.531740427017212,
+      "step": 52
+    },
+    {
+      "epoch": 0.05195441735081485,
+      "grad_norm": 10.475744247436523,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 1.889465093612671,
+      "step": 53
+    },
+    {
+      "epoch": 0.05293468937630193,
+      "grad_norm": 2.9210002422332764,
+      "learning_rate": 8.548387096774194e-06,
+      "loss": 1.595625400543213,
+      "step": 54
+    },
+    {
+      "epoch": 0.053914961401788995,
+      "grad_norm": 4.744402885437012,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 1.6288739442825317,
+      "step": 55
+    },
+    {
+      "epoch": 0.05489523342727607,
+      "grad_norm": 3.7411880493164062,
+      "learning_rate": 8.870967741935484e-06,
+      "loss": 1.4130423069000244,
+      "step": 56
+    },
+    {
+      "epoch": 0.05587550545276314,
+      "grad_norm": 7.154916763305664,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 1.6026225090026855,
+      "step": 57
+    },
+    {
+      "epoch": 0.05685577747825021,
+      "grad_norm": 682.0612182617188,
+      "learning_rate": 9.193548387096775e-06,
+      "loss": 1.70638906955719,
+      "step": 58
+    },
+    {
+      "epoch": 0.05783604950373729,
+      "grad_norm": 2.5724411010742188,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 1.4803117513656616,
+      "step": 59
+    },
+    {
+      "epoch": 0.05881632152922436,
+      "grad_norm": 21.43183135986328,
+      "learning_rate": 9.516129032258065e-06,
+      "loss": 1.2856636047363281,
+      "step": 60
+    },
+    {
+      "epoch": 0.05979659355471143,
+      "grad_norm": 10.6372652053833,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 1.6496026515960693,
+      "step": 61
+    },
+    {
+      "epoch": 0.06077686558019851,
+      "grad_norm": 4.41999626159668,
+      "learning_rate": 9.838709677419356e-06,
+      "loss": 1.6411583423614502,
+      "step": 62
+    },
+    {
+      "epoch": 0.061757137605685575,
+      "grad_norm": 2.860690116882324,
+      "learning_rate": 1e-05,
+      "loss": 1.5934927463531494,
+      "step": 63
+    },
+    {
+      "epoch": 0.06273740963117265,
+      "grad_norm": 2.8396387100219727,
+      "learning_rate": 9.999993706252978e-06,
+      "loss": 1.481213092803955,
+      "step": 64
+    },
+    {
+      "epoch": 0.06371768165665972,
+      "grad_norm": 21.375768661499023,
+      "learning_rate": 9.999974825027756e-06,
+      "loss": 1.484856128692627,
+      "step": 65
+    },
+    {
+      "epoch": 0.0646979536821468,
+      "grad_norm": 2.6061575412750244,
+      "learning_rate": 9.999943356371867e-06,
+      "loss": 1.3070042133331299,
+      "step": 66
+    },
+    {
+      "epoch": 0.06567822570763387,
+      "grad_norm": 3.8316309452056885,
+      "learning_rate": 9.999899300364534e-06,
+      "loss": 1.0604199171066284,
+      "step": 67
+    },
+    {
+      "epoch": 0.06665849773312094,
+      "grad_norm": 2.795164108276367,
+      "learning_rate": 9.999842657116667e-06,
+      "loss": 1.2925666570663452,
+      "step": 68
+    },
+    {
+      "epoch": 0.06763876975860801,
+      "grad_norm": 34.0756721496582,
+      "learning_rate": 9.999773426770864e-06,
+      "loss": 1.2416415214538574,
+      "step": 69
+    },
+    {
+      "epoch": 0.06861904178409509,
+      "grad_norm": 2.152639865875244,
+      "learning_rate": 9.999691609501418e-06,
+      "loss": 1.405970573425293,
+      "step": 70
+    },
+    {
+      "epoch": 0.06959931380958216,
+      "grad_norm": 2.4891347885131836,
+      "learning_rate": 9.999597205514298e-06,
+      "loss": 1.2651238441467285,
+      "step": 71
+    },
+    {
+      "epoch": 0.07057958583506924,
+      "grad_norm": 3.078615427017212,
+      "learning_rate": 9.999490215047167e-06,
+      "loss": 1.2236170768737793,
+      "step": 72
+    },
+    {
+      "epoch": 0.07155985786055631,
+      "grad_norm": 2.3321290016174316,
+      "learning_rate": 9.999370638369377e-06,
+      "loss": 1.4294530153274536,
+      "step": 73
+    },
+    {
+      "epoch": 0.07254012988604337,
+      "grad_norm": 2.775500774383545,
+      "learning_rate": 9.999238475781957e-06,
+      "loss": 1.4197056293487549,
+      "step": 74
+    },
+    {
+      "epoch": 0.07352040191153045,
+      "grad_norm": 2.4884092807769775,
+      "learning_rate": 9.99909372761763e-06,
+      "loss": 1.2769899368286133,
+      "step": 75
+    },
+    {
+      "epoch": 0.07450067393701752,
+      "grad_norm": 2.3005406856536865,
+      "learning_rate": 9.998936394240798e-06,
+      "loss": 1.5367214679718018,
+      "step": 76
+    },
+    {
+      "epoch": 0.0754809459625046,
+      "grad_norm": 10.765892028808594,
+      "learning_rate": 9.998766476047546e-06,
+      "loss": 1.1992907524108887,
+      "step": 77
+    },
+    {
+      "epoch": 0.07646121798799167,
+      "grad_norm": 5.862050533294678,
+      "learning_rate": 9.998583973465647e-06,
+      "loss": 1.2983886003494263,
+      "step": 78
+    },
+    {
+      "epoch": 0.07744149001347873,
+      "grad_norm": 7.480165481567383,
+      "learning_rate": 9.998388886954546e-06,
+      "loss": 1.3342112302780151,
+      "step": 79
+    },
+    {
+      "epoch": 0.07842176203896581,
+      "grad_norm": 3.8878355026245117,
+      "learning_rate": 9.998181217005377e-06,
+      "loss": 1.314904808998108,
+      "step": 80
+    },
+    {
+      "epoch": 0.07940203406445288,
+      "grad_norm": 2.048588752746582,
+      "learning_rate": 9.997960964140946e-06,
+      "loss": 1.3722119331359863,
+      "step": 81
+    },
+    {
+      "epoch": 0.08038230608993996,
+      "grad_norm": 2.098726272583008,
+      "learning_rate": 9.997728128915742e-06,
+      "loss": 1.2300019264221191,
+      "step": 82
+    },
+    {
+      "epoch": 0.08136257811542703,
+      "grad_norm": 103.55396270751953,
+      "learning_rate": 9.997482711915926e-06,
+      "loss": 1.267553448677063,
+      "step": 83
+    },
+    {
+      "epoch": 0.08234285014091411,
+      "grad_norm": 2.7096829414367676,
+      "learning_rate": 9.997224713759335e-06,
+      "loss": 1.5392009019851685,
+      "step": 84
+    },
+    {
+      "epoch": 0.08332312216640117,
+      "grad_norm": 5.360546588897705,
+      "learning_rate": 9.99695413509548e-06,
+      "loss": 1.378632664680481,
+      "step": 85
+    },
+    {
+      "epoch": 0.08430339419188825,
+      "grad_norm": 2.272047758102417,
+      "learning_rate": 9.99667097660554e-06,
+      "loss": 1.2816886901855469,
+      "step": 86
+    },
+    {
+      "epoch": 0.08528366621737532,
+      "grad_norm": 1.9458200931549072,
+      "learning_rate": 9.99637523900237e-06,
+      "loss": 1.3429923057556152,
+      "step": 87
+    },
+    {
+      "epoch": 0.0862639382428624,
+      "grad_norm": 5.74704122543335,
+      "learning_rate": 9.996066923030484e-06,
+      "loss": 1.1690120697021484,
+      "step": 88
+    },
+    {
+      "epoch": 0.08724421026834947,
+      "grad_norm": 4.9611496925354,
+      "learning_rate": 9.995746029466071e-06,
+      "loss": 1.1214773654937744,
+      "step": 89
+    },
+    {
+      "epoch": 0.08822448229383655,
+      "grad_norm": 3.390514373779297,
+      "learning_rate": 9.995412559116979e-06,
+      "loss": 1.0879783630371094,
+      "step": 90
+    },
+    {
+      "epoch": 0.0892047543193236,
+      "grad_norm": 33.46209716796875,
+      "learning_rate": 9.99506651282272e-06,
+      "loss": 1.2780193090438843,
+      "step": 91
+    },
+    {
+      "epoch": 0.09018502634481068,
+      "grad_norm": 3.6176626682281494,
+      "learning_rate": 9.994707891454466e-06,
+      "loss": 1.0887904167175293,
+      "step": 92
+    },
+    {
+      "epoch": 0.09116529837029776,
+      "grad_norm": 4.247608184814453,
+      "learning_rate": 9.994336695915041e-06,
+      "loss": 1.3033946752548218,
+      "step": 93
+    },
+    {
+      "epoch": 0.09214557039578483,
+      "grad_norm": 2.710637092590332,
+      "learning_rate": 9.993952927138933e-06,
+      "loss": 1.300328254699707,
+      "step": 94
+    },
+    {
+      "epoch": 0.0931258424212719,
+      "grad_norm": 4.671359539031982,
+      "learning_rate": 9.993556586092281e-06,
+      "loss": 1.0613725185394287,
+      "step": 95
+    },
+    {
+      "epoch": 0.09410611444675898,
+      "grad_norm": 1.7260868549346924,
+      "learning_rate": 9.993147673772869e-06,
+      "loss": 1.321921467781067,
+      "step": 96
+    },
+    {
+      "epoch": 0.09508638647224604,
+      "grad_norm": 134.52059936523438,
+      "learning_rate": 9.992726191210139e-06,
+      "loss": 1.0365302562713623,
+      "step": 97
+    },
+    {
+      "epoch": 0.09606665849773312,
+      "grad_norm": 1.7052913904190063,
+      "learning_rate": 9.992292139465166e-06,
+      "loss": 1.0690393447875977,
+      "step": 98
+    },
+    {
+      "epoch": 0.09704693052322019,
+      "grad_norm": 3.6903247833251953,
+      "learning_rate": 9.991845519630679e-06,
+      "loss": 1.2925593852996826,
+      "step": 99
+    },
+    {
+      "epoch": 0.09802720254870727,
+      "grad_norm": 1.9280320405960083,
+      "learning_rate": 9.991386332831042e-06,
+      "loss": 1.0765389204025269,
+      "step": 100
+    },
+    {
+      "epoch": 0.09900747457419434,
+      "grad_norm": 26.84065818786621,
+      "learning_rate": 9.990914580222258e-06,
+      "loss": 1.0379040241241455,
+      "step": 101
+    },
+    {
+      "epoch": 0.09998774659968142,
+      "grad_norm": 3.2243404388427734,
+      "learning_rate": 9.990430262991962e-06,
+      "loss": 1.3107194900512695,
+      "step": 102
+    },
+    {
+      "epoch": 0.10096801862516848,
+      "grad_norm": 1.367473840713501,
+      "learning_rate": 9.989933382359423e-06,
+      "loss": 0.964836597442627,
+      "step": 103
+    },
+    {
+      "epoch": 0.10194829065065555,
+      "grad_norm": 3.4907357692718506,
+      "learning_rate": 9.989423939575538e-06,
+      "loss": 1.317588448524475,
+      "step": 104
+    },
+    {
+      "epoch": 0.10292856267614263,
+      "grad_norm": 7.3067240715026855,
+      "learning_rate": 9.988901935922826e-06,
+      "loss": 0.9410532712936401,
+      "step": 105
+    },
+    {
+      "epoch": 0.1039088347016297,
+      "grad_norm": 1.9850404262542725,
+      "learning_rate": 9.988367372715434e-06,
+      "loss": 1.196638584136963,
+      "step": 106
+    },
+    {
+      "epoch": 0.10488910672711678,
+      "grad_norm": 2.154287099838257,
+      "learning_rate": 9.987820251299121e-06,
+      "loss": 0.9918879866600037,
+      "step": 107
+    },
+    {
+      "epoch": 0.10586937875260385,
+      "grad_norm": 3.228703737258911,
+      "learning_rate": 9.987260573051268e-06,
+      "loss": 1.137268304824829,
+      "step": 108
+    },
+    {
+      "epoch": 0.10684965077809092,
+      "grad_norm": 1.538352131843567,
+      "learning_rate": 9.986688339380863e-06,
+      "loss": 0.9699339866638184,
+      "step": 109
+    },
+    {
+      "epoch": 0.10782992280357799,
+      "grad_norm": 2.0486645698547363,
+      "learning_rate": 9.9861035517285e-06,
+      "loss": 1.0073795318603516,
+      "step": 110
+    },
+    {
+      "epoch": 0.10881019482906507,
+      "grad_norm": 65.49430084228516,
+      "learning_rate": 9.985506211566388e-06,
+      "loss": 1.0427565574645996,
+      "step": 111
+    },
+    {
+      "epoch": 0.10979046685455214,
+      "grad_norm": 2.0276129245758057,
+      "learning_rate": 9.984896320398326e-06,
+      "loss": 1.1128275394439697,
+      "step": 112
+    },
+    {
+      "epoch": 0.11077073888003922,
+      "grad_norm": 1.559835433959961,
+      "learning_rate": 9.984273879759713e-06,
+      "loss": 0.9216982126235962,
+      "step": 113
+    },
+    {
+      "epoch": 0.11175101090552628,
+      "grad_norm": 1.7905768156051636,
+      "learning_rate": 9.983638891217544e-06,
+      "loss": 1.294603705406189,
+      "step": 114
+    },
+    {
+      "epoch": 0.11273128293101335,
+      "grad_norm": 4.4755754470825195,
+      "learning_rate": 9.982991356370404e-06,
+      "loss": 1.1224995851516724,
+      "step": 115
+    },
+    {
+      "epoch": 0.11371155495650043,
+      "grad_norm": 5.663575649261475,
+      "learning_rate": 9.982331276848457e-06,
+      "loss": 1.3055403232574463,
+      "step": 116
+    },
+    {
+      "epoch": 0.1146918269819875,
+      "grad_norm": 3.212271213531494,
+      "learning_rate": 9.981658654313458e-06,
+      "loss": 0.837086021900177,
+      "step": 117
+    },
+    {
+      "epoch": 0.11567209900747458,
+      "grad_norm": 3.20328950881958,
+      "learning_rate": 9.980973490458728e-06,
+      "loss": 1.0615761280059814,
+      "step": 118
+    },
+    {
+      "epoch": 0.11665237103296165,
+      "grad_norm": 42.11716842651367,
+      "learning_rate": 9.98027578700917e-06,
+      "loss": 1.0969202518463135,
+      "step": 119
+    },
+    {
+      "epoch": 0.11763264305844871,
+      "grad_norm": 1.5393548011779785,
+      "learning_rate": 9.979565545721249e-06,
+      "loss": 0.971848726272583,
+      "step": 120
+    },
+    {
+      "epoch": 0.11861291508393579,
+      "grad_norm": 7.33202600479126,
+      "learning_rate": 9.978842768382999e-06,
+      "loss": 1.0128525495529175,
+      "step": 121
+    },
+    {
+      "epoch": 0.11959318710942286,
+      "grad_norm": 2.730177164077759,
+      "learning_rate": 9.978107456814007e-06,
+      "loss": 1.1008663177490234,
+      "step": 122
+    },
+    {
+      "epoch": 0.12057345913490994,
+      "grad_norm": 13.244009017944336,
+      "learning_rate": 9.977359612865424e-06,
+      "loss": 1.078629732131958,
+      "step": 123
+    },
+    {
+      "epoch": 0.12155373116039701,
+      "grad_norm": 1.9960155487060547,
+      "learning_rate": 9.976599238419943e-06,
+      "loss": 1.0663634538650513,
+      "step": 124
+    },
+    {
+      "epoch": 0.12253400318588409,
+      "grad_norm": 12.15828800201416,
+      "learning_rate": 9.975826335391808e-06,
+      "loss": 1.4069898128509521,
+      "step": 125
+    },
+    {
+      "epoch": 0.12351427521137115,
+      "grad_norm": 3.776693344116211,
+      "learning_rate": 9.975040905726799e-06,
+      "loss": 1.0607662200927734,
+      "step": 126
+    },
+    {
+      "epoch": 0.12449454723685822,
+      "grad_norm": 2.029813289642334,
+      "learning_rate": 9.974242951402236e-06,
+      "loss": 0.997125506401062,
+      "step": 127
+    },
+    {
+      "epoch": 0.1254748192623453,
+      "grad_norm": 3.0135178565979004,
+      "learning_rate": 9.973432474426968e-06,
+      "loss": 0.9560483694076538,
+      "step": 128
+    },
+    {
+      "epoch": 0.12645509128783236,
+      "grad_norm": 1.6492265462875366,
+      "learning_rate": 9.972609476841368e-06,
+      "loss": 1.0704450607299805,
+      "step": 129
+    },
+    {
+      "epoch": 0.12743536331331945,
+      "grad_norm": 3.0668017864227295,
+      "learning_rate": 9.971773960717333e-06,
+      "loss": 0.8618835210800171,
+      "step": 130
+    },
+    {
+      "epoch": 0.1284156353388065,
+      "grad_norm": 1.9962427616119385,
+      "learning_rate": 9.970925928158275e-06,
+      "loss": 0.9284038543701172,
+      "step": 131
+    },
+    {
+      "epoch": 0.1293959073642936,
+      "grad_norm": 4.200085639953613,
+      "learning_rate": 9.970065381299112e-06,
+      "loss": 0.9037516713142395,
+      "step": 132
+    },
+    {
+      "epoch": 0.13037617938978066,
+      "grad_norm": 4.228448390960693,
+      "learning_rate": 9.969192322306271e-06,
+      "loss": 1.0130993127822876,
+      "step": 133
+    },
+    {
+      "epoch": 0.13135645141526775,
+      "grad_norm": 1.6451573371887207,
+      "learning_rate": 9.968306753377679e-06,
+      "loss": 0.9967314004898071,
+      "step": 134
+    },
+    {
+      "epoch": 0.1323367234407548,
+      "grad_norm": 3.4750893115997314,
+      "learning_rate": 9.96740867674275e-06,
+      "loss": 1.2647136449813843,
+      "step": 135
+    },
+    {
+      "epoch": 0.13331699546624187,
+      "grad_norm": 1.7133313417434692,
+      "learning_rate": 9.966498094662398e-06,
+      "loss": 0.9861700534820557,
+      "step": 136
+    },
+    {
+      "epoch": 0.13429726749172896,
+      "grad_norm": 1.942231297492981,
+      "learning_rate": 9.965575009429006e-06,
+      "loss": 0.8320759534835815,
+      "step": 137
+    },
+    {
+      "epoch": 0.13527753951721602,
+      "grad_norm": 1.790663242340088,
+      "learning_rate": 9.964639423366442e-06,
+      "loss": 0.9279268980026245,
+      "step": 138
+    },
+    {
+      "epoch": 0.1362578115427031,
+      "grad_norm": 2.697603464126587,
+      "learning_rate": 9.963691338830045e-06,
+      "loss": 0.9657130837440491,
+      "step": 139
+    },
+    {
+      "epoch": 0.13723808356819017,
+      "grad_norm": 1.277811050415039,
+      "learning_rate": 9.962730758206612e-06,
+      "loss": 0.9654596447944641,
+      "step": 140
+    },
+    {
+      "epoch": 0.13821835559367723,
+      "grad_norm": 1.4491413831710815,
+      "learning_rate": 9.961757683914406e-06,
+      "loss": 1.1169567108154297,
+      "step": 141
+    },
+    {
+      "epoch": 0.13919862761916432,
+      "grad_norm": 1.871880054473877,
+      "learning_rate": 9.960772118403142e-06,
+      "loss": 0.9070476293563843,
+      "step": 142
+    },
+    {
+      "epoch": 0.14017889964465138,
+      "grad_norm": 2.418889045715332,
+      "learning_rate": 9.959774064153977e-06,
+      "loss": 0.7983903884887695,
+      "step": 143
+    },
+    {
+      "epoch": 0.14115917167013847,
+      "grad_norm": 2.5387730598449707,
+      "learning_rate": 9.958763523679515e-06,
+      "loss": 1.0602573156356812,
+      "step": 144
+    },
+    {
+      "epoch": 0.14213944369562553,
+      "grad_norm": 1.261715292930603,
+      "learning_rate": 9.957740499523787e-06,
+      "loss": 0.9131547212600708,
+      "step": 145
+    },
+    {
+      "epoch": 0.14311971572111262,
+      "grad_norm": 2.176605224609375,
+      "learning_rate": 9.956704994262256e-06,
+      "loss": 1.061754584312439,
+      "step": 146
+    },
+    {
+      "epoch": 0.14409998774659968,
+      "grad_norm": 1.4627416133880615,
+      "learning_rate": 9.955657010501807e-06,
+      "loss": 0.9212430119514465,
+      "step": 147
+    },
+    {
+      "epoch": 0.14508025977208674,
+      "grad_norm": 1.4561854600906372,
+      "learning_rate": 9.954596550880735e-06,
+      "loss": 0.8806812167167664,
+      "step": 148
+    },
+    {
+      "epoch": 0.14606053179757383,
+      "grad_norm": 1.4765551090240479,
+      "learning_rate": 9.95352361806875e-06,
+      "loss": 0.7166234850883484,
+      "step": 149
+    },
+    {
+      "epoch": 0.1470408038230609,
+      "grad_norm": 1.6301178932189941,
+      "learning_rate": 9.952438214766956e-06,
+      "loss": 0.8946540355682373,
+      "step": 150
+    },
+    {
+      "epoch": 0.14802107584854798,
+      "grad_norm": 1.6439906358718872,
+      "learning_rate": 9.951340343707852e-06,
+      "loss": 1.1272326707839966,
+      "step": 151
+    },
+    {
+      "epoch": 0.14900134787403504,
+      "grad_norm": 1.5681216716766357,
+      "learning_rate": 9.950230007655333e-06,
+      "loss": 0.7440442442893982,
+      "step": 152
+    },
+    {
+      "epoch": 0.1499816198995221,
+      "grad_norm": 2.383523464202881,
+      "learning_rate": 9.949107209404664e-06,
+      "loss": 1.0407934188842773,
+      "step": 153
+    },
+    {
+      "epoch": 0.1509618919250092,
+      "grad_norm": 1.7701935768127441,
+      "learning_rate": 9.94797195178249e-06,
+      "loss": 1.0672982931137085,
+      "step": 154
+    },
+    {
+      "epoch": 0.15194216395049626,
+      "grad_norm": 2.023500919342041,
+      "learning_rate": 9.946824237646823e-06,
+      "loss": 1.0460244417190552,
+      "step": 155
+    },
+    {
+      "epoch": 0.15292243597598335,
+      "grad_norm": 2.0290346145629883,
+      "learning_rate": 9.945664069887028e-06,
+      "loss": 1.1332428455352783,
+      "step": 156
+    },
+    {
+      "epoch": 0.1539027080014704,
+      "grad_norm": 1.4425419569015503,
+      "learning_rate": 9.944491451423829e-06,
+      "loss": 0.8501012325286865,
+      "step": 157
+    },
+    {
+      "epoch": 0.15488298002695747,
+      "grad_norm": 2.9868381023406982,
+      "learning_rate": 9.94330638520929e-06,
+      "loss": 0.9394078254699707,
+      "step": 158
+    },
+    {
+      "epoch": 0.15586325205244456,
+      "grad_norm": 2.930450439453125,
+      "learning_rate": 9.942108874226812e-06,
+      "loss": 0.9087206125259399,
+      "step": 159
+    },
+    {
+      "epoch": 0.15684352407793162,
+      "grad_norm": 1.3846571445465088,
+      "learning_rate": 9.940898921491132e-06,
+      "loss": 0.837488055229187,
+      "step": 160
+    },
+    {
+      "epoch": 0.1578237961034187,
+      "grad_norm": 1.1196712255477905,
+      "learning_rate": 9.9396765300483e-06,
+      "loss": 0.8358951807022095,
+      "step": 161
+    },
+    {
+      "epoch": 0.15880406812890577,
+      "grad_norm": 3.958306074142456,
+      "learning_rate": 9.938441702975689e-06,
+      "loss": 1.0169492959976196,
+      "step": 162
+    },
+    {
+      "epoch": 0.15978434015439286,
+      "grad_norm": 6.9478068351745605,
+      "learning_rate": 9.937194443381972e-06,
+      "loss": 1.1400552988052368,
+      "step": 163
+    },
+    {
+      "epoch": 0.16076461217987992,
+      "grad_norm": 9.230393409729004,
+      "learning_rate": 9.935934754407125e-06,
+      "loss": 0.9421498775482178,
+      "step": 164
+    },
+    {
+      "epoch": 0.16174488420536698,
+      "grad_norm": 1.5736265182495117,
+      "learning_rate": 9.934662639222412e-06,
+      "loss": 0.9221128821372986,
+      "step": 165
+    },
+    {
+      "epoch": 0.16272515623085407,
+      "grad_norm": 2.2992024421691895,
+      "learning_rate": 9.933378101030382e-06,
+      "loss": 0.8492938280105591,
+      "step": 166
+    },
+    {
+      "epoch": 0.16370542825634113,
+      "grad_norm": 3.105520248413086,
+      "learning_rate": 9.93208114306486e-06,
+      "loss": 0.8923220038414001,
+      "step": 167
+    },
+    {
+      "epoch": 0.16468570028182822,
+      "grad_norm": 3.1217753887176514,
+      "learning_rate": 9.930771768590934e-06,
+      "loss": 0.9330375790596008,
+      "step": 168
+    },
+    {
+      "epoch": 0.16566597230731528,
+      "grad_norm": 13.082860946655273,
+      "learning_rate": 9.929449980904952e-06,
+      "loss": 0.7983123064041138,
+      "step": 169
+    },
+    {
+      "epoch": 0.16664624433280234,
+      "grad_norm": 2.2359111309051514,
+      "learning_rate": 9.928115783334518e-06,
+      "loss": 1.0458908081054688,
+      "step": 170
+    },
+    {
+      "epoch": 0.16762651635828943,
+      "grad_norm": 1.5942684412002563,
+      "learning_rate": 9.926769179238467e-06,
+      "loss": 0.8456210494041443,
+      "step": 171
+    },
+    {
+      "epoch": 0.1686067883837765,
+      "grad_norm": 1.8771940469741821,
+      "learning_rate": 9.925410172006874e-06,
+      "loss": 0.9488444328308105,
+      "step": 172
+    },
+    {
+      "epoch": 0.16958706040926358,
+      "grad_norm": 1.629173755645752,
+      "learning_rate": 9.924038765061042e-06,
+      "loss": 1.027367115020752,
+      "step": 173
+    },
+    {
+      "epoch": 0.17056733243475064,
+      "grad_norm": 1.705886721611023,
+      "learning_rate": 9.922654961853483e-06,
+      "loss": 1.0480202436447144,
+      "step": 174
+    },
+    {
+      "epoch": 0.17154760446023773,
+      "grad_norm": 3.083059549331665,
+      "learning_rate": 9.921258765867919e-06,
+      "loss": 1.170985221862793,
+      "step": 175
+    },
+    {
+      "epoch": 0.1725278764857248,
+      "grad_norm": 1.213767647743225,
+      "learning_rate": 9.919850180619274e-06,
+      "loss": 0.9057183265686035,
+      "step": 176
+    },
+    {
+      "epoch": 0.17350814851121185,
+      "grad_norm": 2.005957841873169,
+      "learning_rate": 9.918429209653662e-06,
+      "loss": 0.8352057933807373,
+      "step": 177
+    },
+    {
+      "epoch": 0.17448842053669894,
+      "grad_norm": 1.9753361940383911,
+      "learning_rate": 9.916995856548371e-06,
+      "loss": 0.9967347383499146,
+      "step": 178
+    },
+    {
+      "epoch": 0.175468692562186,
+      "grad_norm": 3.9508841037750244,
+      "learning_rate": 9.915550124911866e-06,
+      "loss": 0.9397368431091309,
+      "step": 179
+    },
+    {
+      "epoch": 0.1764489645876731,
+      "grad_norm": 2.486034393310547,
+      "learning_rate": 9.914092018383778e-06,
+      "loss": 0.8116385340690613,
+      "step": 180
+    },
+    {
+      "epoch": 0.17742923661316015,
+      "grad_norm": 2.1286423206329346,
+      "learning_rate": 9.912621540634889e-06,
+      "loss": 0.8690676689147949,
+      "step": 181
+    },
+    {
+      "epoch": 0.1784095086386472,
+      "grad_norm": 1.70301353931427,
+      "learning_rate": 9.91113869536712e-06,
+      "loss": 0.7545929551124573,
+      "step": 182
+    },
+    {
+      "epoch": 0.1793897806641343,
+      "grad_norm": 2.2550504207611084,
+      "learning_rate": 9.909643486313533e-06,
+      "loss": 0.8710118532180786,
+      "step": 183
+    },
+    {
+      "epoch": 0.18037005268962136,
+      "grad_norm": 7.460702896118164,
+      "learning_rate": 9.908135917238321e-06,
+      "loss": 1.1188092231750488,
+      "step": 184
+    },
+    {
+      "epoch": 0.18135032471510845,
+      "grad_norm": 4.5814642906188965,
+      "learning_rate": 9.906615991936781e-06,
+      "loss": 0.8766211867332458,
+      "step": 185
+    },
+    {
+      "epoch": 0.1823305967405955,
+      "grad_norm": 1.8015472888946533,
+      "learning_rate": 9.905083714235326e-06,
+      "loss": 0.9747026562690735,
+      "step": 186
+    },
+    {
+      "epoch": 0.18331086876608257,
+      "grad_norm": 2.0149457454681396,
+      "learning_rate": 9.903539087991462e-06,
+      "loss": 1.0040225982666016,
+      "step": 187
+    },
+    {
+      "epoch": 0.18429114079156966,
+      "grad_norm": 1.5122036933898926,
+      "learning_rate": 9.901982117093786e-06,
+      "loss": 0.9926754832267761,
+      "step": 188
+    },
+    {
+      "epoch": 0.18527141281705672,
+      "grad_norm": 1.5093003511428833,
+      "learning_rate": 9.900412805461968e-06,
+      "loss": 0.6946313381195068,
+      "step": 189
+    },
+    {
+      "epoch": 0.1862516848425438,
+      "grad_norm": 0.9908679723739624,
+      "learning_rate": 9.898831157046748e-06,
+      "loss": 0.7255351543426514,
+      "step": 190
+    },
+    {
+      "epoch": 0.18723195686803087,
+      "grad_norm": 1.393106460571289,
+      "learning_rate": 9.897237175829927e-06,
+      "loss": 0.8476099967956543,
+      "step": 191
+    },
+    {
+      "epoch": 0.18821222889351796,
+      "grad_norm": 3.165613889694214,
+      "learning_rate": 9.895630865824347e-06,
+      "loss": 0.7950215935707092,
+      "step": 192
+    },
+    {
+      "epoch": 0.18919250091900502,
+      "grad_norm": 1.5488990545272827,
+      "learning_rate": 9.894012231073895e-06,
+      "loss": 0.904655933380127,
+      "step": 193
+    },
+    {
+      "epoch": 0.19017277294449209,
+      "grad_norm": 3.1914010047912598,
+      "learning_rate": 9.89238127565348e-06,
+      "loss": 0.7940858602523804,
+      "step": 194
+    },
+    {
+      "epoch": 0.19115304496997917,
+      "grad_norm": 2.0876028537750244,
+      "learning_rate": 9.890738003669029e-06,
+      "loss": 0.9836628437042236,
+      "step": 195
+    },
+    {
+      "epoch": 0.19213331699546624,
+      "grad_norm": 2.138056516647339,
+      "learning_rate": 9.88908241925748e-06,
+      "loss": 0.8696339130401611,
+      "step": 196
+    },
+    {
+      "epoch": 0.19311358902095332,
+      "grad_norm": 4.552305698394775,
+      "learning_rate": 9.887414526586764e-06,
+      "loss": 0.9055603742599487,
+      "step": 197
+    },
+    {
+      "epoch": 0.19409386104644039,
+      "grad_norm": 1.216966152191162,
+      "learning_rate": 9.885734329855798e-06,
+      "loss": 0.7843443155288696,
+      "step": 198
+    },
+    {
+      "epoch": 0.19507413307192745,
+      "grad_norm": 1.4261524677276611,
+      "learning_rate": 9.884041833294477e-06,
+      "loss": 0.891804575920105,
+      "step": 199
+    },
+    {
+      "epoch": 0.19605440509741454,
+      "grad_norm": 1.6586214303970337,
+      "learning_rate": 9.882337041163656e-06,
+      "loss": 0.894652247428894,
+      "step": 200
+    },
+    {
+      "epoch": 0.1970346771229016,
+      "grad_norm": 3.3041770458221436,
+      "learning_rate": 9.880619957755151e-06,
+      "loss": 0.8902967572212219,
+      "step": 201
+    },
+    {
+      "epoch": 0.19801494914838869,
+      "grad_norm": 1.4221687316894531,
+      "learning_rate": 9.878890587391715e-06,
+      "loss": 0.8312460780143738,
+      "step": 202
+    },
+    {
+      "epoch": 0.19899522117387575,
+      "grad_norm": 2.876361846923828,
+      "learning_rate": 9.877148934427037e-06,
+      "loss": 0.7611515522003174,
+      "step": 203
+    },
+    {
+      "epoch": 0.19997549319936284,
+      "grad_norm": 1.066075325012207,
+      "learning_rate": 9.875395003245724e-06,
+      "loss": 0.7034518718719482,
+      "step": 204
+    },
+    {
+      "epoch": 0.2009557652248499,
+      "grad_norm": 3.6058127880096436,
+      "learning_rate": 9.873628798263297e-06,
+      "loss": 1.0095345973968506,
+      "step": 205
+    },
+    {
+      "epoch": 0.20193603725033696,
+      "grad_norm": 1.9826786518096924,
+      "learning_rate": 9.871850323926178e-06,
+      "loss": 0.8382951021194458,
+      "step": 206
+    },
+    {
+      "epoch": 0.20291630927582405,
+      "grad_norm": 1.381087303161621,
+      "learning_rate": 9.870059584711668e-06,
+      "loss": 0.8321932554244995,
+      "step": 207
+    },
+    {
+      "epoch": 0.2038965813013111,
+      "grad_norm": 6.653373718261719,
+      "learning_rate": 9.868256585127956e-06,
+      "loss": 0.8264703750610352,
+      "step": 208
+    },
+    {
+      "epoch": 0.2048768533267982,
+      "grad_norm": 2.0672712326049805,
+      "learning_rate": 9.86644132971409e-06,
+      "loss": 0.8410891890525818,
+      "step": 209
+    },
+    {
+      "epoch": 0.20585712535228526,
+      "grad_norm": 2.0962629318237305,
+      "learning_rate": 9.864613823039971e-06,
+      "loss": 0.9301019310951233,
+      "step": 210
+    },
+    {
+      "epoch": 0.20683739737777232,
+      "grad_norm": 10.260763168334961,
+      "learning_rate": 9.862774069706346e-06,
+      "loss": 1.0459771156311035,
+      "step": 211
+    },
+    {
+      "epoch": 0.2078176694032594,
+      "grad_norm": 3.456390619277954,
+      "learning_rate": 9.860922074344794e-06,
+      "loss": 0.8872228860855103,
+      "step": 212
+    },
+    {
+      "epoch": 0.20879794142874647,
+      "grad_norm": 1.5585193634033203,
+      "learning_rate": 9.859057841617709e-06,
+      "loss": 0.7565791606903076,
+      "step": 213
+    },
+    {
+      "epoch": 0.20977821345423356,
+      "grad_norm": 2.6125149726867676,
+      "learning_rate": 9.857181376218295e-06,
+      "loss": 0.7839133739471436,
+      "step": 214
+    },
+    {
+      "epoch": 0.21075848547972062,
+      "grad_norm": 1.9059659242630005,
+      "learning_rate": 9.855292682870552e-06,
+      "loss": 0.8504835963249207,
+      "step": 215
+    },
+    {
+      "epoch": 0.2117387575052077,
+      "grad_norm": 1.5269825458526611,
+      "learning_rate": 9.853391766329264e-06,
+      "loss": 0.8221988677978516,
+      "step": 216
+    },
+    {
+      "epoch": 0.21271902953069477,
+      "grad_norm": 2.2962594032287598,
+      "learning_rate": 9.851478631379982e-06,
+      "loss": 0.7724093198776245,
+      "step": 217
+    },
+    {
+      "epoch": 0.21369930155618183,
+      "grad_norm": 2.4179270267486572,
+      "learning_rate": 9.849553282839025e-06,
+      "loss": 0.78898024559021,
+      "step": 218
+    },
+    {
+      "epoch": 0.21467957358166892,
+      "grad_norm": 2.7591922283172607,
+      "learning_rate": 9.847615725553457e-06,
+      "loss": 0.8101059794425964,
+      "step": 219
+    },
+    {
+      "epoch": 0.21565984560715598,
+      "grad_norm": 2.565300941467285,
+      "learning_rate": 9.845665964401072e-06,
+      "loss": 0.7674195766448975,
+      "step": 220
+    },
+    {
+      "epoch": 0.21664011763264307,
+      "grad_norm": 1.6755082607269287,
+      "learning_rate": 9.843704004290393e-06,
+      "loss": 0.8255548477172852,
+      "step": 221
+    },
+    {
+      "epoch": 0.21762038965813013,
+      "grad_norm": 1.3054996728897095,
+      "learning_rate": 9.841729850160652e-06,
+      "loss": 0.8056149482727051,
+      "step": 222
+    },
+    {
+      "epoch": 0.2186006616836172,
+      "grad_norm": 1.8252493143081665,
+      "learning_rate": 9.839743506981783e-06,
+      "loss": 0.8478525280952454,
+      "step": 223
+    },
+    {
+      "epoch": 0.21958093370910428,
+      "grad_norm": 1.1294909715652466,
+      "learning_rate": 9.837744979754397e-06,
+      "loss": 0.7057391405105591,
+      "step": 224
+    },
+    {
+      "epoch": 0.22056120573459134,
+      "grad_norm": 2.3227827548980713,
+      "learning_rate": 9.835734273509787e-06,
+      "loss": 0.8048064708709717,
+      "step": 225
+    },
+    {
+      "epoch": 0.22154147776007843,
+      "grad_norm": 2.4523065090179443,
+      "learning_rate": 9.833711393309902e-06,
+      "loss": 0.9739433526992798,
+      "step": 226
+    },
+    {
+      "epoch": 0.2225217497855655,
+      "grad_norm": 6.219913959503174,
+      "learning_rate": 9.831676344247343e-06,
+      "loss": 0.9670267105102539,
+      "step": 227
+    },
+    {
+      "epoch": 0.22350202181105255,
+      "grad_norm": 1.4539546966552734,
+      "learning_rate": 9.829629131445342e-06,
+      "loss": 0.7615249156951904,
+      "step": 228
+    },
+    {
+      "epoch": 0.22448229383653964,
+      "grad_norm": 1.6476659774780273,
+      "learning_rate": 9.827569760057755e-06,
+      "loss": 0.6248297095298767,
+      "step": 229
+    },
+    {
+      "epoch": 0.2254625658620267,
+      "grad_norm": 4.1464338302612305,
+      "learning_rate": 9.825498235269046e-06,
+      "loss": 0.859736442565918,
+      "step": 230
+    },
+    {
+      "epoch": 0.2264428378875138,
+      "grad_norm": 1.5779759883880615,
+      "learning_rate": 9.82341456229428e-06,
+      "loss": 0.7575444579124451,
+      "step": 231
+    },
+    {
+      "epoch": 0.22742310991300085,
+      "grad_norm": 2.3473963737487793,
+      "learning_rate": 9.821318746379097e-06,
+      "loss": 0.8634650707244873,
+      "step": 232
+    },
+    {
+      "epoch": 0.22840338193848794,
+      "grad_norm": 1.8142012357711792,
+      "learning_rate": 9.819210792799711e-06,
+      "loss": 0.8329586386680603,
+      "step": 233
+    },
+    {
+      "epoch": 0.229383653963975,
+      "grad_norm": 1.4023786783218384,
+      "learning_rate": 9.817090706862895e-06,
+      "loss": 0.7443281412124634,
+      "step": 234
+    },
+    {
+      "epoch": 0.23036392598946207,
+      "grad_norm": 3.477633237838745,
+      "learning_rate": 9.814958493905962e-06,
+      "loss": 0.9234720468521118,
+      "step": 235
+    },
+    {
+      "epoch": 0.23134419801494915,
+      "grad_norm": 1.6123863458633423,
+      "learning_rate": 9.812814159296757e-06,
+      "loss": 0.8642038106918335,
+      "step": 236
+    },
+    {
+      "epoch": 0.23232447004043622,
+      "grad_norm": 1.8461908102035522,
+      "learning_rate": 9.810657708433637e-06,
+      "loss": 0.9516077041625977,
+      "step": 237
+    },
+    {
+      "epoch": 0.2333047420659233,
+      "grad_norm": 2.335613965988159,
+      "learning_rate": 9.808489146745466e-06,
+      "loss": 0.7823044061660767,
+      "step": 238
+    },
+    {
+      "epoch": 0.23428501409141037,
+      "grad_norm": 0.7424708008766174,
+      "learning_rate": 9.806308479691595e-06,
+      "loss": 0.7390064001083374,
+      "step": 239
+    },
+    {
+      "epoch": 0.23526528611689743,
+      "grad_norm": 3.7687036991119385,
+      "learning_rate": 9.804115712761851e-06,
+      "loss": 1.083920955657959,
+      "step": 240
+    },
+    {
+      "epoch": 0.23624555814238452,
+      "grad_norm": 1.3474844694137573,
+      "learning_rate": 9.801910851476524e-06,
+      "loss": 0.8942563533782959,
+      "step": 241
+    },
+    {
+      "epoch": 0.23722583016787158,
+      "grad_norm": 5.237135410308838,
+      "learning_rate": 9.799693901386346e-06,
+      "loss": 1.0037472248077393,
+      "step": 242
+    },
+    {
+      "epoch": 0.23820610219335867,
+      "grad_norm": 1.1229761838912964,
+      "learning_rate": 9.797464868072489e-06,
+      "loss": 0.7325989007949829,
+      "step": 243
+    },
+    {
+      "epoch": 0.23918637421884573,
+      "grad_norm": 1.7052339315414429,
+      "learning_rate": 9.79522375714654e-06,
+      "loss": 0.7529733180999756,
+      "step": 244
+    },
+    {
+      "epoch": 0.24016664624433282,
+      "grad_norm": 1.4736284017562866,
+      "learning_rate": 9.792970574250493e-06,
+      "loss": 0.8858606815338135,
+      "step": 245
+    },
+    {
+      "epoch": 0.24114691826981988,
+      "grad_norm": 1.206372618675232,
+      "learning_rate": 9.790705325056735e-06,
+      "loss": 0.798877477645874,
+      "step": 246
+    },
+    {
+      "epoch": 0.24212719029530694,
+      "grad_norm": 1.0359853506088257,
+      "learning_rate": 9.788428015268027e-06,
+      "loss": 0.716171145439148,
+      "step": 247
+    },
+    {
+      "epoch": 0.24310746232079403,
+      "grad_norm": 1.3520716428756714,
+      "learning_rate": 9.786138650617494e-06,
+      "loss": 0.8814042806625366,
+      "step": 248
+    },
+    {
+      "epoch": 0.2440877343462811,
+      "grad_norm": 0.9526112675666809,
+      "learning_rate": 9.78383723686861e-06,
+      "loss": 0.6469528079032898,
+      "step": 249
+    },
+    {
+      "epoch": 0.24506800637176818,
+      "grad_norm": 2.102663516998291,
+      "learning_rate": 9.781523779815178e-06,
+      "loss": 0.7236974239349365,
+      "step": 250
+    },
+    {
+      "epoch": 0.24604827839725524,
+      "grad_norm": 1.304598093032837,
+      "learning_rate": 9.779198285281326e-06,
+      "loss": 0.7935341596603394,
+      "step": 251
+    },
+    {
+      "epoch": 0.2470285504227423,
+      "grad_norm": 18.215852737426758,
+      "learning_rate": 9.776860759121485e-06,
+      "loss": 0.8978430032730103,
+      "step": 252
+    },
+    {
+      "epoch": 0.2480088224482294,
+      "grad_norm": 1.975088357925415,
+      "learning_rate": 9.774511207220369e-06,
+      "loss": 0.8099026679992676,
+      "step": 253
+    },
+    {
+      "epoch": 0.24898909447371645,
+      "grad_norm": 0.9218975305557251,
+      "learning_rate": 9.772149635492979e-06,
+      "loss": 0.757601261138916,
+      "step": 254
+    },
+    {
+      "epoch": 0.24996936649920354,
+      "grad_norm": 1.0467439889907837,
+      "learning_rate": 9.769776049884564e-06,
+      "loss": 0.8007693290710449,
+      "step": 255
+    },
+    {
+      "epoch": 0.2509496385246906,
+      "grad_norm": 1.6121203899383545,
+      "learning_rate": 9.767390456370625e-06,
+      "loss": 1.0843206644058228,
+      "step": 256
+    },
+    {
+      "epoch": 0.25192991055017766,
+      "grad_norm": 1.8285597562789917,
+      "learning_rate": 9.76499286095689e-06,
+      "loss": 0.8214015960693359,
+      "step": 257
+    },
+    {
+      "epoch": 0.2529101825756647,
+      "grad_norm": 1.4345918893814087,
+      "learning_rate": 9.762583269679304e-06,
+      "loss": 0.7590874433517456,
+      "step": 258
+    },
+    {
+      "epoch": 0.25389045460115184,
+      "grad_norm": 2.7142724990844727,
+      "learning_rate": 9.760161688604008e-06,
+      "loss": 0.9466472864151001,
+      "step": 259
+    },
+    {
+      "epoch": 0.2548707266266389,
+      "grad_norm": 4.027217864990234,
+      "learning_rate": 9.75772812382733e-06,
+      "loss": 0.7247453331947327,
+      "step": 260
+    },
+    {
+      "epoch": 0.25585099865212596,
+      "grad_norm": 1.1912027597427368,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.800244927406311,
+      "step": 261
+    },
+    {
+      "epoch": 0.256831270677613,
+      "grad_norm": 2.4537627696990967,
+      "learning_rate": 9.75282506770597e-06,
+      "loss": 0.725568950176239,
+      "step": 262
+    },
+    {
+      "epoch": 0.2578115427031001,
+      "grad_norm": 2.2537238597869873,
+      "learning_rate": 9.750355588704728e-06,
+      "loss": 0.7959393858909607,
+      "step": 263
+    },
+    {
+      "epoch": 0.2587918147285872,
+      "grad_norm": 2.578726291656494,
+      "learning_rate": 9.747874150688948e-06,
+      "loss": 0.8287099599838257,
+      "step": 264
+    },
+    {
+      "epoch": 0.25977208675407426,
+      "grad_norm": 1.2124162912368774,
+      "learning_rate": 9.745380759905648e-06,
+      "loss": 0.8278694152832031,
+      "step": 265
+    },
+    {
+      "epoch": 0.2607523587795613,
+      "grad_norm": 0.9931219816207886,
+      "learning_rate": 9.742875422631937e-06,
+      "loss": 0.7235227227210999,
+      "step": 266
+    },
+    {
+      "epoch": 0.2617326308050484,
+      "grad_norm": 1.3074661493301392,
+      "learning_rate": 9.740358145174999e-06,
+      "loss": 0.8493878841400146,
+      "step": 267
+    },
+    {
+      "epoch": 0.2627129028305355,
+      "grad_norm": 0.936988890171051,
+      "learning_rate": 9.737828933872076e-06,
+      "loss": 0.7563545107841492,
+      "step": 268
+    },
+    {
+      "epoch": 0.26369317485602256,
+      "grad_norm": 4.2573065757751465,
+      "learning_rate": 9.735287795090455e-06,
+      "loss": 0.8253098130226135,
+      "step": 269
+    },
+    {
+      "epoch": 0.2646734468815096,
+      "grad_norm": 1.2360275983810425,
+      "learning_rate": 9.73273473522745e-06,
+      "loss": 0.7253192663192749,
+      "step": 270
+    },
+    {
+      "epoch": 0.2656537189069967,
+      "grad_norm": 42.30955123901367,
+      "learning_rate": 9.730169760710385e-06,
+      "loss": 0.8240077495574951,
+      "step": 271
+    },
+    {
+      "epoch": 0.26663399093248374,
+      "grad_norm": 1.0315252542495728,
+      "learning_rate": 9.727592877996585e-06,
+      "loss": 0.7828395962715149,
+      "step": 272
+    },
+    {
+      "epoch": 0.26761426295797086,
+      "grad_norm": 0.9821422100067139,
+      "learning_rate": 9.725004093573343e-06,
+      "loss": 0.8393453359603882,
+      "step": 273
+    },
+    {
+      "epoch": 0.2685945349834579,
+      "grad_norm": 1.0683647394180298,
+      "learning_rate": 9.722403413957924e-06,
+      "loss": 0.7098627686500549,
+      "step": 274
+    },
+    {
+      "epoch": 0.269574807008945,
+      "grad_norm": 4.975259780883789,
+      "learning_rate": 9.719790845697534e-06,
+      "loss": 0.8404012322425842,
+      "step": 275
+    },
+    {
+      "epoch": 0.27055507903443204,
+      "grad_norm": 1.2321093082427979,
+      "learning_rate": 9.717166395369312e-06,
+      "loss": 0.7505289316177368,
+      "step": 276
+    },
+    {
+      "epoch": 0.2715353510599191,
+      "grad_norm": 3.933621406555176,
+      "learning_rate": 9.71453006958031e-06,
+      "loss": 0.7256272435188293,
+      "step": 277
+    },
+    {
+      "epoch": 0.2725156230854062,
+      "grad_norm": 1.2206323146820068,
+      "learning_rate": 9.711881874967471e-06,
+      "loss": 0.5751657485961914,
+      "step": 278
+    },
+    {
+      "epoch": 0.2734958951108933,
+      "grad_norm": 6.056179046630859,
+      "learning_rate": 9.709221818197626e-06,
+      "loss": 0.8268659114837646,
+      "step": 279
+    },
+    {
+      "epoch": 0.27447616713638034,
+      "grad_norm": 21.996158599853516,
+      "learning_rate": 9.70654990596746e-06,
+      "loss": 0.9641490578651428,
+      "step": 280
+    },
+    {
+      "epoch": 0.2754564391618674,
+      "grad_norm": 1.5312501192092896,
+      "learning_rate": 9.703866145003512e-06,
+      "loss": 0.7989124655723572,
+      "step": 281
+    },
+    {
+      "epoch": 0.27643671118735447,
+      "grad_norm": 1.9520635604858398,
+      "learning_rate": 9.701170542062149e-06,
+      "loss": 0.7586979866027832,
+      "step": 282
+    },
+    {
+      "epoch": 0.2774169832128416,
+      "grad_norm": 0.9791378378868103,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.8906596899032593,
+      "step": 283
+    },
+    {
+      "epoch": 0.27839725523832864,
+      "grad_norm": 2.6486432552337646,
+      "learning_rate": 9.69574383742167e-06,
+      "loss": 0.6601815819740295,
+      "step": 284
+    },
+    {
+      "epoch": 0.2793775272638157,
+      "grad_norm": 3.991563081741333,
+      "learning_rate": 9.69301274938428e-06,
+      "loss": 0.886470377445221,
+      "step": 285
+    },
+    {
+      "epoch": 0.28035779928930277,
+      "grad_norm": 1.9946955442428589,
+      "learning_rate": 9.690269846692882e-06,
+      "loss": 0.8359981775283813,
+      "step": 286
+    },
+    {
+      "epoch": 0.28133807131478983,
+      "grad_norm": 1.330568552017212,
+      "learning_rate": 9.687515136252732e-06,
+      "loss": 0.8143620491027832,
+      "step": 287
+    },
+    {
+      "epoch": 0.28231834334027694,
+      "grad_norm": 1.4931341409683228,
+      "learning_rate": 9.68474862499881e-06,
+      "loss": 0.7600244879722595,
+      "step": 288
+    },
+    {
+      "epoch": 0.283298615365764,
+      "grad_norm": 0.9717527031898499,
+      "learning_rate": 9.681970319895804e-06,
+      "loss": 0.8248692750930786,
+      "step": 289
+    },
+    {
+      "epoch": 0.28427888739125107,
+      "grad_norm": 1.4346030950546265,
+      "learning_rate": 9.679180227938095e-06,
+      "loss": 0.7979452610015869,
+      "step": 290
+    },
+    {
+      "epoch": 0.28525915941673813,
+      "grad_norm": 2.0320849418640137,
+      "learning_rate": 9.676378356149733e-06,
+      "loss": 0.8522430062294006,
+      "step": 291
+    },
+    {
+      "epoch": 0.28623943144222525,
+      "grad_norm": 0.9154717326164246,
+      "learning_rate": 9.67356471158443e-06,
+      "loss": 0.6991317272186279,
+      "step": 292
+    },
+    {
+      "epoch": 0.2872197034677123,
+      "grad_norm": 1.685329794883728,
+      "learning_rate": 9.670739301325534e-06,
+      "loss": 0.7803480625152588,
+      "step": 293
+    },
+    {
+      "epoch": 0.28819997549319937,
+      "grad_norm": 1.2144747972488403,
+      "learning_rate": 9.667902132486009e-06,
+      "loss": 0.9544454216957092,
+      "step": 294
+    },
+    {
+      "epoch": 0.28918024751868643,
+      "grad_norm": 0.9177585244178772,
+      "learning_rate": 9.665053212208426e-06,
+      "loss": 0.7026416659355164,
+      "step": 295
+    },
+    {
+      "epoch": 0.2901605195441735,
+      "grad_norm": 1.2642945051193237,
+      "learning_rate": 9.662192547664938e-06,
+      "loss": 0.7789238691329956,
+      "step": 296
+    },
+    {
+      "epoch": 0.2911407915696606,
+      "grad_norm": 1.527679443359375,
+      "learning_rate": 9.659320146057263e-06,
+      "loss": 0.7460813522338867,
+      "step": 297
+    },
+    {
+      "epoch": 0.29212106359514767,
+      "grad_norm": 1.5328047275543213,
+      "learning_rate": 9.65643601461667e-06,
+      "loss": 0.6149399280548096,
+      "step": 298
+    },
+    {
+      "epoch": 0.29310133562063473,
+      "grad_norm": 2.002063512802124,
+      "learning_rate": 9.653540160603956e-06,
+      "loss": 0.7808958292007446,
+      "step": 299
+    },
+    {
+      "epoch": 0.2940816076461218,
+      "grad_norm": 1.3209649324417114,
+      "learning_rate": 9.650632591309431e-06,
+      "loss": 0.8286932706832886,
+      "step": 300
+    },
+    {
+      "epoch": 0.29506187967160885,
+      "grad_norm": 1.1301060914993286,
+      "learning_rate": 9.647713314052896e-06,
+      "loss": 0.7079054713249207,
+      "step": 301
+    },
+    {
+      "epoch": 0.29604215169709597,
+      "grad_norm": 1.2198007106781006,
+      "learning_rate": 9.64478233618363e-06,
+      "loss": 0.7337124347686768,
+      "step": 302
+    },
+    {
+      "epoch": 0.29702242372258303,
+      "grad_norm": 1.8013590574264526,
+      "learning_rate": 9.641839665080363e-06,
+      "loss": 0.7348734140396118,
+      "step": 303
+    },
+    {
+      "epoch": 0.2980026957480701,
+      "grad_norm": 6.982769966125488,
+      "learning_rate": 9.63888530815127e-06,
+      "loss": 0.621437668800354,
+      "step": 304
+    },
+    {
+      "epoch": 0.29898296777355715,
+      "grad_norm": 0.8037109971046448,
+      "learning_rate": 9.635919272833938e-06,
+      "loss": 0.7044026851654053,
+      "step": 305
+    },
+    {
+      "epoch": 0.2999632397990442,
+      "grad_norm": 2.649951696395874,
+      "learning_rate": 9.632941566595357e-06,
+      "loss": 0.8717630505561829,
+      "step": 306
+    },
+    {
+      "epoch": 0.30094351182453133,
+      "grad_norm": 0.7429581880569458,
+      "learning_rate": 9.629952196931902e-06,
+      "loss": 0.6706403493881226,
+      "step": 307
+    },
+    {
+      "epoch": 0.3019237838500184,
+      "grad_norm": 2.517219066619873,
+      "learning_rate": 9.626951171369306e-06,
+      "loss": 0.7474154829978943,
+      "step": 308
+    },
+    {
+      "epoch": 0.30290405587550545,
+      "grad_norm": 3.993215560913086,
+      "learning_rate": 9.623938497462647e-06,
+      "loss": 1.0605274438858032,
+      "step": 309
+    },
+    {
+      "epoch": 0.3038843279009925,
+      "grad_norm": 0.7498021125793457,
+      "learning_rate": 9.620914182796327e-06,
+      "loss": 0.6114619970321655,
+      "step": 310
+    },
+    {
+      "epoch": 0.3048645999264796,
+      "grad_norm": 1.1178343296051025,
+      "learning_rate": 9.617878234984056e-06,
+      "loss": 0.7127967476844788,
+      "step": 311
+    },
+    {
+      "epoch": 0.3058448719519667,
+      "grad_norm": 1.370303988456726,
+      "learning_rate": 9.61483066166883e-06,
+      "loss": 0.6828451156616211,
+      "step": 312
+    },
+    {
+      "epoch": 0.30682514397745375,
+      "grad_norm": 1.6824547052383423,
+      "learning_rate": 9.611771470522908e-06,
+      "loss": 0.7666938304901123,
+      "step": 313
+    },
+    {
+      "epoch": 0.3078054160029408,
+      "grad_norm": 1.178693175315857,
+      "learning_rate": 9.608700669247803e-06,
+      "loss": 0.7262972593307495,
+      "step": 314
+    },
+    {
+      "epoch": 0.3087856880284279,
+      "grad_norm": 0.8273369669914246,
+      "learning_rate": 9.60561826557425e-06,
+      "loss": 0.663546085357666,
+      "step": 315
+    },
+    {
+      "epoch": 0.30976596005391493,
+      "grad_norm": 1.2372169494628906,
+      "learning_rate": 9.602524267262202e-06,
+      "loss": 0.7528340220451355,
+      "step": 316
+    },
+    {
+      "epoch": 0.31074623207940205,
+      "grad_norm": 1.1937190294265747,
+      "learning_rate": 9.599418682100793e-06,
+      "loss": 0.6922539472579956,
+      "step": 317
+    },
+    {
+      "epoch": 0.3117265041048891,
+      "grad_norm": 1.4707770347595215,
+      "learning_rate": 9.596301517908329e-06,
+      "loss": 0.6598691940307617,
+      "step": 318
+    },
+    {
+      "epoch": 0.3127067761303762,
+      "grad_norm": 2.283562183380127,
+      "learning_rate": 9.59317278253227e-06,
+      "loss": 0.8772923946380615,
+      "step": 319
+    },
+    {
+      "epoch": 0.31368704815586324,
+      "grad_norm": 3.324998378753662,
+      "learning_rate": 9.5900324838492e-06,
+      "loss": 0.7775747776031494,
+      "step": 320
+    },
+    {
+      "epoch": 0.31466732018135035,
+      "grad_norm": 0.9688576459884644,
+      "learning_rate": 9.586880629764817e-06,
+      "loss": 0.6990121603012085,
+      "step": 321
+    },
+    {
+      "epoch": 0.3156475922068374,
+      "grad_norm": 1.3204386234283447,
+      "learning_rate": 9.583717228213914e-06,
+      "loss": 0.7537082433700562,
+      "step": 322
+    },
+    {
+      "epoch": 0.3166278642323245,
+      "grad_norm": 1.9989577531814575,
+      "learning_rate": 9.580542287160348e-06,
+      "loss": 0.792009711265564,
+      "step": 323
+    },
+    {
+      "epoch": 0.31760813625781154,
+      "grad_norm": 14.248479843139648,
+      "learning_rate": 9.577355814597031e-06,
+      "loss": 0.8783087730407715,
+      "step": 324
+    },
+    {
+      "epoch": 0.3185884082832986,
+      "grad_norm": 2.7614834308624268,
+      "learning_rate": 9.574157818545902e-06,
+      "loss": 0.6729799509048462,
+      "step": 325
+    },
+    {
+      "epoch": 0.3195686803087857,
+      "grad_norm": 1.4884456396102905,
+      "learning_rate": 9.570948307057912e-06,
+      "loss": 0.857717752456665,
+      "step": 326
+    },
+    {
+      "epoch": 0.3205489523342728,
+      "grad_norm": 2.3030965328216553,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 0.6739245057106018,
+      "step": 327
+    },
+    {
+      "epoch": 0.32152922435975984,
+      "grad_norm": 1.0925297737121582,
+      "learning_rate": 9.564494770120089e-06,
+      "loss": 0.7250769734382629,
+      "step": 328
+    },
+    {
+      "epoch": 0.3225094963852469,
+      "grad_norm": 1.2971023321151733,
+      "learning_rate": 9.561250760917026e-06,
+      "loss": 0.782243013381958,
+      "step": 329
+    },
+    {
+      "epoch": 0.32348976841073396,
+      "grad_norm": 1.5687350034713745,
+      "learning_rate": 9.557995268770609e-06,
+      "loss": 0.8246535658836365,
+      "step": 330
+    },
+    {
+      "epoch": 0.3244700404362211,
+      "grad_norm": 0.8873333930969238,
+      "learning_rate": 9.554728301876525e-06,
+      "loss": 0.6622074842453003,
+      "step": 331
+    },
+    {
+      "epoch": 0.32545031246170814,
+      "grad_norm": 1.0537322759628296,
+      "learning_rate": 9.551449868459371e-06,
+      "loss": 0.7784616351127625,
+      "step": 332
+    },
+    {
+      "epoch": 0.3264305844871952,
+      "grad_norm": 1.1330640316009521,
+      "learning_rate": 9.548159976772593e-06,
+      "loss": 0.7197192907333374,
+      "step": 333
+    },
+    {
+      "epoch": 0.32741085651268226,
+      "grad_norm": 2.125290632247925,
+      "learning_rate": 9.544858635098491e-06,
+      "loss": 0.7126935124397278,
+      "step": 334
+    },
+    {
+      "epoch": 0.3283911285381693,
+      "grad_norm": 8.362874984741211,
+      "learning_rate": 9.541545851748186e-06,
+      "loss": 0.8333726525306702,
+      "step": 335
+    },
+    {
+      "epoch": 0.32937140056365644,
+      "grad_norm": 1.6455880403518677,
+      "learning_rate": 9.538221635061611e-06,
+      "loss": 0.6324223279953003,
+      "step": 336
+    },
+    {
+      "epoch": 0.3303516725891435,
+      "grad_norm": 0.9797875881195068,
+      "learning_rate": 9.534885993407474e-06,
+      "loss": 0.7844573259353638,
+      "step": 337
+    },
+    {
+      "epoch": 0.33133194461463056,
+      "grad_norm": 1.4511393308639526,
+      "learning_rate": 9.531538935183252e-06,
+      "loss": 0.7679675817489624,
+      "step": 338
+    },
+    {
+      "epoch": 0.3323122166401176,
+      "grad_norm": 3.444230794906616,
+      "learning_rate": 9.528180468815155e-06,
+      "loss": 0.7836833000183105,
+      "step": 339
+    },
+    {
+      "epoch": 0.3332924886656047,
+      "grad_norm": 0.6860651969909668,
+      "learning_rate": 9.524810602758122e-06,
+      "loss": 0.6795933246612549,
+      "step": 340
+    },
+    {
+      "epoch": 0.3342727606910918,
+      "grad_norm": 1.147053599357605,
+      "learning_rate": 9.521429345495787e-06,
+      "loss": 0.7233742475509644,
+      "step": 341
+    },
+    {
+      "epoch": 0.33525303271657886,
+      "grad_norm": 0.8374312520027161,
+      "learning_rate": 9.518036705540459e-06,
+      "loss": 0.9334558248519897,
+      "step": 342
+    },
+    {
+      "epoch": 0.3362333047420659,
+      "grad_norm": 1.3453978300094604,
+      "learning_rate": 9.514632691433108e-06,
+      "loss": 0.7483043670654297,
+      "step": 343
+    },
+    {
+      "epoch": 0.337213576767553,
+      "grad_norm": 3.7183032035827637,
+      "learning_rate": 9.51121731174333e-06,
+      "loss": 0.7183725833892822,
+      "step": 344
+    },
+    {
+      "epoch": 0.33819384879304004,
+      "grad_norm": 1.340848445892334,
+      "learning_rate": 9.507790575069347e-06,
+      "loss": 0.8598052859306335,
+      "step": 345
+    },
+    {
+      "epoch": 0.33917412081852716,
+      "grad_norm": 1.9590427875518799,
+      "learning_rate": 9.504352490037957e-06,
+      "loss": 0.6911303400993347,
+      "step": 346
+    },
+    {
+      "epoch": 0.3401543928440142,
+      "grad_norm": 2.716817855834961,
+      "learning_rate": 9.50090306530454e-06,
+      "loss": 0.8543134331703186,
+      "step": 347
+    },
+    {
+      "epoch": 0.3411346648695013,
+      "grad_norm": 1.2692991495132446,
+      "learning_rate": 9.497442309553017e-06,
+      "loss": 0.6810178756713867,
+      "step": 348
+    },
+    {
+      "epoch": 0.34211493689498834,
+      "grad_norm": 1.1566184759140015,
+      "learning_rate": 9.493970231495836e-06,
+      "loss": 0.8954340815544128,
+      "step": 349
+    },
+    {
+      "epoch": 0.34309520892047546,
+      "grad_norm": 0.6450574398040771,
+      "learning_rate": 9.490486839873949e-06,
+      "loss": 0.8001822233200073,
+      "step": 350
+    },
+    {
+      "epoch": 0.3440754809459625,
+      "grad_norm": 1.5686252117156982,
+      "learning_rate": 9.486992143456792e-06,
+      "loss": 0.6919950246810913,
+      "step": 351
+    },
+    {
+      "epoch": 0.3450557529714496,
+      "grad_norm": 1.5829793214797974,
+      "learning_rate": 9.483486151042258e-06,
+      "loss": 0.6850660443305969,
+      "step": 352
+    },
+    {
+      "epoch": 0.34603602499693664,
+      "grad_norm": 0.7141197323799133,
+      "learning_rate": 9.47996887145668e-06,
+      "loss": 0.5947696566581726,
+      "step": 353
+    },
+    {
+      "epoch": 0.3470162970224237,
+      "grad_norm": 1.7379820346832275,
+      "learning_rate": 9.476440313554803e-06,
+      "loss": 0.7049946784973145,
+      "step": 354
+    },
+    {
+      "epoch": 0.3479965690479108,
+      "grad_norm": 1.9138275384902954,
+      "learning_rate": 9.47290048621977e-06,
+      "loss": 0.6962699890136719,
+      "step": 355
+    },
+    {
+      "epoch": 0.3489768410733979,
+      "grad_norm": 1.8334511518478394,
+      "learning_rate": 9.469349398363088e-06,
+      "loss": 0.7085521221160889,
+      "step": 356
+    },
+    {
+      "epoch": 0.34995711309888494,
+      "grad_norm": 3.0309836864471436,
+      "learning_rate": 9.46578705892462e-06,
+      "loss": 0.717523992061615,
+      "step": 357
+    },
+    {
+      "epoch": 0.350937385124372,
+      "grad_norm": 3.3971755504608154,
+      "learning_rate": 9.46221347687255e-06,
+      "loss": 0.8374325037002563,
+      "step": 358
+    },
+    {
+      "epoch": 0.35191765714985906,
+      "grad_norm": 1.0974339246749878,
+      "learning_rate": 9.458628661203368e-06,
+      "loss": 0.6612426042556763,
+      "step": 359
+    },
+    {
+      "epoch": 0.3528979291753462,
+      "grad_norm": 0.7320294976234436,
+      "learning_rate": 9.45503262094184e-06,
+      "loss": 0.7445319890975952,
+      "step": 360
+    },
+    {
+      "epoch": 0.35387820120083324,
+      "grad_norm": 1.5363261699676514,
+      "learning_rate": 9.451425365140997e-06,
+      "loss": 0.6951749324798584,
+      "step": 361
+    },
+    {
+      "epoch": 0.3548584732263203,
+      "grad_norm": 0.8943214416503906,
+      "learning_rate": 9.447806902882098e-06,
+      "loss": 0.6090914011001587,
+      "step": 362
+    },
+    {
+      "epoch": 0.35583874525180736,
+      "grad_norm": 1.3308911323547363,
+      "learning_rate": 9.444177243274619e-06,
+      "loss": 0.6834264993667603,
+      "step": 363
+    },
+    {
+      "epoch": 0.3568190172772944,
+      "grad_norm": 1.1766867637634277,
+      "learning_rate": 9.440536395456223e-06,
+      "loss": 0.7620858550071716,
+      "step": 364
+    },
+    {
+      "epoch": 0.35779928930278154,
+      "grad_norm": 1.5522148609161377,
+      "learning_rate": 9.43688436859274e-06,
+      "loss": 0.7585674524307251,
+      "step": 365
+    },
+    {
+      "epoch": 0.3587795613282686,
+      "grad_norm": 1.0016453266143799,
+      "learning_rate": 9.433221171878144e-06,
+      "loss": 0.6107903122901917,
+      "step": 366
+    },
+    {
+      "epoch": 0.35975983335375566,
+      "grad_norm": 0.7920359969139099,
+      "learning_rate": 9.429546814534528e-06,
+      "loss": 0.7666431665420532,
+      "step": 367
+    },
+    {
+      "epoch": 0.3607401053792427,
+      "grad_norm": 1.033385157585144,
+      "learning_rate": 9.425861305812083e-06,
+      "loss": 0.9204304218292236,
+      "step": 368
+    },
+    {
+      "epoch": 0.3617203774047298,
+      "grad_norm": 0.80854332447052,
+      "learning_rate": 9.422164654989073e-06,
+      "loss": 0.7669047117233276,
+      "step": 369
+    },
+    {
+      "epoch": 0.3627006494302169,
+      "grad_norm": 0.937046229839325,
+      "learning_rate": 9.41845687137181e-06,
+      "loss": 0.8097209930419922,
+      "step": 370
+    },
+    {
+      "epoch": 0.36368092145570396,
+      "grad_norm": 1.8544846773147583,
+      "learning_rate": 9.414737964294636e-06,
+      "loss": 0.6807209253311157,
+      "step": 371
+    },
+    {
+      "epoch": 0.364661193481191,
+      "grad_norm": 1.2624070644378662,
+      "learning_rate": 9.411007943119895e-06,
+      "loss": 1.0650994777679443,
+      "step": 372
+    },
+    {
+      "epoch": 0.3656414655066781,
+      "grad_norm": 0.7455968856811523,
+      "learning_rate": 9.40726681723791e-06,
+      "loss": 0.8050444722175598,
+      "step": 373
+    },
+    {
+      "epoch": 0.36662173753216515,
+      "grad_norm": 1.0443443059921265,
+      "learning_rate": 9.403514596066964e-06,
+      "loss": 0.6994483470916748,
+      "step": 374
+    },
+    {
+      "epoch": 0.36760200955765227,
+      "grad_norm": 1.5019422769546509,
+      "learning_rate": 9.399751289053267e-06,
+      "loss": 0.7326818704605103,
+      "step": 375
+    },
+    {
+      "epoch": 0.3685822815831393,
+      "grad_norm": 2.751922607421875,
+      "learning_rate": 9.39597690567094e-06,
+      "loss": 0.5795458555221558,
+      "step": 376
+    },
+    {
+      "epoch": 0.3695625536086264,
+      "grad_norm": 0.8304553627967834,
+      "learning_rate": 9.392191455421989e-06,
+      "loss": 0.6351127624511719,
+      "step": 377
+    },
+    {
+      "epoch": 0.37054282563411345,
+      "grad_norm": 0.8712226152420044,
+      "learning_rate": 9.388394947836278e-06,
+      "loss": 0.6603379845619202,
+      "step": 378
+    },
+    {
+      "epoch": 0.37152309765960057,
+      "grad_norm": 1.5914474725723267,
+      "learning_rate": 9.384587392471516e-06,
+      "loss": 0.629761815071106,
+      "step": 379
+    },
+    {
+      "epoch": 0.3725033696850876,
+      "grad_norm": 1.3343446254730225,
+      "learning_rate": 9.380768798913213e-06,
+      "loss": 0.7801318168640137,
+      "step": 380
+    },
+    {
+      "epoch": 0.3734836417105747,
+      "grad_norm": 1.5806853771209717,
+      "learning_rate": 9.376939176774678e-06,
+      "loss": 0.7173468470573425,
+      "step": 381
+    },
+    {
+      "epoch": 0.37446391373606175,
+      "grad_norm": 0.9605811238288879,
+      "learning_rate": 9.37309853569698e-06,
+      "loss": 0.6645132303237915,
+      "step": 382
+    },
+    {
+      "epoch": 0.3754441857615488,
+      "grad_norm": 1.8220621347427368,
+      "learning_rate": 9.369246885348926e-06,
+      "loss": 0.7864338159561157,
+      "step": 383
+    },
+    {
+      "epoch": 0.3764244577870359,
+      "grad_norm": 0.7504281401634216,
+      "learning_rate": 9.365384235427042e-06,
+      "loss": 0.7322210073471069,
+      "step": 384
+    },
+    {
+      "epoch": 0.377404729812523,
+      "grad_norm": 1.0645418167114258,
+      "learning_rate": 9.361510595655545e-06,
+      "loss": 0.8192920684814453,
+      "step": 385
+    },
+    {
+      "epoch": 0.37838500183801005,
+      "grad_norm": 2.9109084606170654,
+      "learning_rate": 9.357625975786318e-06,
+      "loss": 0.7149112820625305,
+      "step": 386
+    },
+    {
+      "epoch": 0.3793652738634971,
+      "grad_norm": 4.40610933303833,
+      "learning_rate": 9.353730385598887e-06,
+      "loss": 0.7068130970001221,
+      "step": 387
+    },
+    {
+      "epoch": 0.38034554588898417,
+      "grad_norm": 1.2079412937164307,
+      "learning_rate": 9.349823834900396e-06,
+      "loss": 0.6572710871696472,
+      "step": 388
+    },
+    {
+      "epoch": 0.3813258179144713,
+      "grad_norm": 0.9596279859542847,
+      "learning_rate": 9.345906333525582e-06,
+      "loss": 0.6638326644897461,
+      "step": 389
+    },
+    {
+      "epoch": 0.38230608993995835,
+      "grad_norm": 1.1863784790039062,
+      "learning_rate": 9.341977891336748e-06,
+      "loss": 0.6906521320343018,
+      "step": 390
+    },
+    {
+      "epoch": 0.3832863619654454,
+      "grad_norm": 1.057320237159729,
+      "learning_rate": 9.338038518223746e-06,
+      "loss": 0.6660881638526917,
+      "step": 391
+    },
+    {
+      "epoch": 0.38426663399093247,
+      "grad_norm": 0.9026414155960083,
+      "learning_rate": 9.334088224103942e-06,
+      "loss": 0.6333972215652466,
+      "step": 392
+    },
+    {
+      "epoch": 0.38524690601641953,
+      "grad_norm": 4.749232292175293,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.8200359344482422,
+      "step": 393
+    },
+    {
+      "epoch": 0.38622717804190665,
+      "grad_norm": 1.2350492477416992,
+      "learning_rate": 9.326154912650834e-06,
+      "loss": 0.6198328137397766,
+      "step": 394
+    },
+    {
+      "epoch": 0.3872074500673937,
+      "grad_norm": 1.130714774131775,
+      "learning_rate": 9.322171915289635e-06,
+      "loss": 0.7639663219451904,
+      "step": 395
+    },
+    {
+      "epoch": 0.38818772209288077,
+      "grad_norm": 3.739182472229004,
+      "learning_rate": 9.318178036865786e-06,
+      "loss": 0.6643421649932861,
+      "step": 396
+    },
+    {
+      "epoch": 0.38916799411836783,
+      "grad_norm": 1.3137704133987427,
+      "learning_rate": 9.314173287433874e-06,
+      "loss": 0.8814802169799805,
+      "step": 397
+    },
+    {
+      "epoch": 0.3901482661438549,
+      "grad_norm": 1.0812374353408813,
+      "learning_rate": 9.310157677075847e-06,
+      "loss": 0.67060387134552,
+      "step": 398
+    },
+    {
+      "epoch": 0.391128538169342,
+      "grad_norm": 1.2764666080474854,
+      "learning_rate": 9.306131215901004e-06,
+      "loss": 0.7572896480560303,
+      "step": 399
+    },
+    {
+      "epoch": 0.39210881019482907,
+      "grad_norm": 1.349245309829712,
+      "learning_rate": 9.302093914045953e-06,
+      "loss": 0.5145754814147949,
+      "step": 400
+    },
+    {
+      "epoch": 0.39308908222031613,
+      "grad_norm": 0.7296897172927856,
+      "learning_rate": 9.298045781674595e-06,
+      "loss": 0.6878324747085571,
+      "step": 401
+    },
+    {
+      "epoch": 0.3940693542458032,
+      "grad_norm": 1.7980632781982422,
+      "learning_rate": 9.293986828978106e-06,
+      "loss": 0.7511922121047974,
+      "step": 402
+    },
+    {
+      "epoch": 0.3950496262712903,
+      "grad_norm": 0.7240265011787415,
+      "learning_rate": 9.289917066174887e-06,
+      "loss": 0.7222871780395508,
+      "step": 403
+    },
+    {
+      "epoch": 0.39602989829677737,
+      "grad_norm": 1.208769679069519,
+      "learning_rate": 9.285836503510562e-06,
+      "loss": 0.6468597054481506,
+      "step": 404
+    },
+    {
+      "epoch": 0.39701017032226443,
+      "grad_norm": 0.801677942276001,
+      "learning_rate": 9.281745151257946e-06,
+      "loss": 0.6962947845458984,
+      "step": 405
+    },
+    {
+      "epoch": 0.3979904423477515,
+      "grad_norm": 1.0546627044677734,
+      "learning_rate": 9.277643019717011e-06,
+      "loss": 0.8856452107429504,
+      "step": 406
+    },
+    {
+      "epoch": 0.39897071437323856,
+      "grad_norm": 1.4623268842697144,
+      "learning_rate": 9.273530119214868e-06,
+      "loss": 0.6555129289627075,
+      "step": 407
+    },
+    {
+      "epoch": 0.39995098639872567,
+      "grad_norm": 3.620544672012329,
+      "learning_rate": 9.269406460105742e-06,
+      "loss": 0.6803292632102966,
+      "step": 408
+    },
+    {
+      "epoch": 0.40093125842421273,
+      "grad_norm": 1.0600348711013794,
+      "learning_rate": 9.265272052770936e-06,
+      "loss": 0.6204216480255127,
+      "step": 409
+    },
+    {
+      "epoch": 0.4019115304496998,
+      "grad_norm": 0.794247567653656,
+      "learning_rate": 9.261126907618818e-06,
+      "loss": 0.6076223254203796,
+      "step": 410
+    },
+    {
+      "epoch": 0.40289180247518686,
+      "grad_norm": 0.931538462638855,
+      "learning_rate": 9.256971035084786e-06,
+      "loss": 0.7662779092788696,
+      "step": 411
+    },
+    {
+      "epoch": 0.4038720745006739,
+      "grad_norm": 1.3982493877410889,
+      "learning_rate": 9.252804445631243e-06,
+      "loss": 0.6401995420455933,
+      "step": 412
+    },
+    {
+      "epoch": 0.40485234652616103,
+      "grad_norm": 1.445826530456543,
+      "learning_rate": 9.248627149747573e-06,
+      "loss": 0.8899269700050354,
+      "step": 413
+    },
+    {
+      "epoch": 0.4058326185516481,
+      "grad_norm": 1.0429189205169678,
+      "learning_rate": 9.244439157950114e-06,
+      "loss": 0.8805820941925049,
+      "step": 414
+    },
+    {
+      "epoch": 0.40681289057713516,
+      "grad_norm": 1.0017132759094238,
+      "learning_rate": 9.24024048078213e-06,
+      "loss": 0.6045535802841187,
+      "step": 415
+    },
+    {
+      "epoch": 0.4077931626026222,
+      "grad_norm": 0.6645440459251404,
+      "learning_rate": 9.236031128813786e-06,
+      "loss": 0.6933586597442627,
+      "step": 416
+    },
+    {
+      "epoch": 0.4087734346281093,
+      "grad_norm": 93.61671447753906,
+      "learning_rate": 9.231811112642121e-06,
+      "loss": 0.7168632745742798,
+      "step": 417
+    },
+    {
+      "epoch": 0.4097537066535964,
+      "grad_norm": 0.8288508057594299,
+      "learning_rate": 9.227580442891022e-06,
+      "loss": 0.6648313403129578,
+      "step": 418
+    },
+    {
+      "epoch": 0.41073397867908346,
+      "grad_norm": 1.1719741821289062,
+      "learning_rate": 9.223339130211194e-06,
+      "loss": 0.7090386152267456,
+      "step": 419
+    },
+    {
+      "epoch": 0.4117142507045705,
+      "grad_norm": 1.9822778701782227,
+      "learning_rate": 9.219087185280132e-06,
+      "loss": 0.8285101056098938,
+      "step": 420
+    },
+    {
+      "epoch": 0.4126945227300576,
+      "grad_norm": 1.3712502717971802,
+      "learning_rate": 9.214824618802108e-06,
+      "loss": 0.7979877591133118,
+      "step": 421
+    },
+    {
+      "epoch": 0.41367479475554464,
+      "grad_norm": 1.2938177585601807,
+      "learning_rate": 9.210551441508126e-06,
+      "loss": 0.5769352316856384,
+      "step": 422
+    },
+    {
+      "epoch": 0.41465506678103176,
+      "grad_norm": 1.4776637554168701,
+      "learning_rate": 9.206267664155906e-06,
+      "loss": 0.5997432470321655,
+      "step": 423
+    },
+    {
+      "epoch": 0.4156353388065188,
+      "grad_norm": 1.2265496253967285,
+      "learning_rate": 9.201973297529852e-06,
+      "loss": 0.6579384803771973,
+      "step": 424
+    },
+    {
+      "epoch": 0.4166156108320059,
+      "grad_norm": 0.7943096160888672,
+      "learning_rate": 9.197668352441025e-06,
+      "loss": 0.6409897804260254,
+      "step": 425
+    },
+    {
+      "epoch": 0.41759588285749294,
+      "grad_norm": 1.6348960399627686,
+      "learning_rate": 9.193352839727122e-06,
+      "loss": 0.6308920383453369,
+      "step": 426
+    },
+    {
+      "epoch": 0.41857615488298,
+      "grad_norm": 0.8862689137458801,
+      "learning_rate": 9.189026770252437e-06,
+      "loss": 0.6797462105751038,
+      "step": 427
+    },
+    {
+      "epoch": 0.4195564269084671,
+      "grad_norm": 1.3907949924468994,
+      "learning_rate": 9.18469015490785e-06,
+      "loss": 0.7980557680130005,
+      "step": 428
+    },
+    {
+      "epoch": 0.4205366989339542,
+      "grad_norm": 0.8909097909927368,
+      "learning_rate": 9.18034300461078e-06,
+      "loss": 0.753312349319458,
+      "step": 429
+    },
+    {
+      "epoch": 0.42151697095944124,
+      "grad_norm": 0.7411423921585083,
+      "learning_rate": 9.175985330305176e-06,
+      "loss": 0.7175651788711548,
+      "step": 430
+    },
+    {
+      "epoch": 0.4224972429849283,
+      "grad_norm": 0.6301830410957336,
+      "learning_rate": 9.171617142961477e-06,
+      "loss": 0.6610926389694214,
+      "step": 431
+    },
+    {
+      "epoch": 0.4234775150104154,
+      "grad_norm": 1.1687572002410889,
+      "learning_rate": 9.167238453576589e-06,
+      "loss": 0.7067232131958008,
+      "step": 432
+    },
+    {
+      "epoch": 0.4244577870359025,
+      "grad_norm": 1.6193435192108154,
+      "learning_rate": 9.162849273173857e-06,
+      "loss": 0.6969735026359558,
+      "step": 433
+    },
+    {
+      "epoch": 0.42543805906138954,
+      "grad_norm": 1.0947290658950806,
+      "learning_rate": 9.158449612803039e-06,
+      "loss": 0.8598254323005676,
+      "step": 434
+    },
+    {
+      "epoch": 0.4264183310868766,
+      "grad_norm": 1.9992629289627075,
+      "learning_rate": 9.154039483540273e-06,
+      "loss": 0.7109705805778503,
+      "step": 435
+    },
+    {
+      "epoch": 0.42739860311236366,
+      "grad_norm": 2.1095035076141357,
+      "learning_rate": 9.149618896488056e-06,
+      "loss": 0.6843716502189636,
+      "step": 436
+    },
+    {
+      "epoch": 0.4283788751378508,
+      "grad_norm": 2.1968045234680176,
+      "learning_rate": 9.145187862775208e-06,
+      "loss": 0.5287095308303833,
+      "step": 437
+    },
+    {
+      "epoch": 0.42935914716333784,
+      "grad_norm": 5.562580585479736,
+      "learning_rate": 9.140746393556853e-06,
+      "loss": 0.7133256196975708,
+      "step": 438
+    },
+    {
+      "epoch": 0.4303394191888249,
+      "grad_norm": 1.312368392944336,
+      "learning_rate": 9.136294500014387e-06,
+      "loss": 0.8046531677246094,
+      "step": 439
+    },
+    {
+      "epoch": 0.43131969121431196,
+      "grad_norm": 0.6666929125785828,
+      "learning_rate": 9.13183219335544e-06,
+      "loss": 0.6948230266571045,
+      "step": 440
+    },
+    {
+      "epoch": 0.432299963239799,
+      "grad_norm": 0.6986605525016785,
+      "learning_rate": 9.12735948481387e-06,
+      "loss": 0.7104353308677673,
+      "step": 441
+    },
+    {
+      "epoch": 0.43328023526528614,
+      "grad_norm": 0.9380354881286621,
+      "learning_rate": 9.122876385649713e-06,
+      "loss": 0.6305464506149292,
+      "step": 442
+    },
+    {
+      "epoch": 0.4342605072907732,
+      "grad_norm": 0.8888136148452759,
+      "learning_rate": 9.118382907149164e-06,
+      "loss": 0.7499136924743652,
+      "step": 443
+    },
+    {
+      "epoch": 0.43524077931626026,
+      "grad_norm": 29.4472713470459,
+      "learning_rate": 9.113879060624554e-06,
+      "loss": 0.670158326625824,
+      "step": 444
+    },
+    {
+      "epoch": 0.4362210513417473,
+      "grad_norm": 1.226319432258606,
+      "learning_rate": 9.109364857414306e-06,
+      "loss": 0.6893974542617798,
+      "step": 445
+    },
+    {
+      "epoch": 0.4372013233672344,
+      "grad_norm": 1.0378698110580444,
+      "learning_rate": 9.104840308882925e-06,
+      "loss": 0.7525972127914429,
+      "step": 446
+    },
+    {
+      "epoch": 0.4381815953927215,
+      "grad_norm": 1.8934987783432007,
+      "learning_rate": 9.100305426420957e-06,
+      "loss": 0.7389995455741882,
+      "step": 447
+    },
+    {
+      "epoch": 0.43916186741820856,
+      "grad_norm": 1.879409909248352,
+      "learning_rate": 9.09576022144496e-06,
+      "loss": 0.6494981646537781,
+      "step": 448
+    },
+    {
+      "epoch": 0.4401421394436956,
+      "grad_norm": 0.8278953433036804,
+      "learning_rate": 9.091204705397485e-06,
+      "loss": 0.633147120475769,
+      "step": 449
+    },
+    {
+      "epoch": 0.4411224114691827,
+      "grad_norm": 1.0337613821029663,
+      "learning_rate": 9.086638889747034e-06,
+      "loss": 0.7052793502807617,
+      "step": 450
+    },
+    {
+      "epoch": 0.44210268349466975,
+      "grad_norm": 0.772998034954071,
+      "learning_rate": 9.08206278598805e-06,
+      "loss": 0.4826599955558777,
+      "step": 451
+    },
+    {
+      "epoch": 0.44308295552015686,
+      "grad_norm": 1.0598688125610352,
+      "learning_rate": 9.077476405640862e-06,
+      "loss": 0.7920869588851929,
+      "step": 452
+    },
+    {
+      "epoch": 0.4440632275456439,
+      "grad_norm": 1.3796448707580566,
+      "learning_rate": 9.07287976025168e-06,
+      "loss": 0.7280728816986084,
+      "step": 453
+    },
+    {
+      "epoch": 0.445043499571131,
+      "grad_norm": 0.8409003615379333,
+      "learning_rate": 9.06827286139255e-06,
+      "loss": 0.6662689447402954,
+      "step": 454
+    },
+    {
+      "epoch": 0.44602377159661805,
+      "grad_norm": 1.071440577507019,
+      "learning_rate": 9.06365572066134e-06,
+      "loss": 0.7975555658340454,
+      "step": 455
+    },
+    {
+      "epoch": 0.4470040436221051,
+      "grad_norm": 1.3031781911849976,
+      "learning_rate": 9.059028349681693e-06,
+      "loss": 0.6684004068374634,
+      "step": 456
+    },
+    {
+      "epoch": 0.4479843156475922,
+      "grad_norm": 0.8597187995910645,
+      "learning_rate": 9.05439076010301e-06,
+      "loss": 0.5784738659858704,
+      "step": 457
+    },
+    {
+      "epoch": 0.4489645876730793,
+      "grad_norm": 1.0262107849121094,
+      "learning_rate": 9.04974296360042e-06,
+      "loss": 0.7092287540435791,
+      "step": 458
+    },
+    {
+      "epoch": 0.44994485969856635,
+      "grad_norm": 5.217550277709961,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.697614312171936,
+      "step": 459
+    },
+    {
+      "epoch": 0.4509251317240534,
+      "grad_norm": 1.5341706275939941,
+      "learning_rate": 9.040416796652459e-06,
+      "loss": 0.7592578530311584,
+      "step": 460
+    },
+    {
+      "epoch": 0.4519054037495405,
+      "grad_norm": 1.3850138187408447,
+      "learning_rate": 9.035738449685707e-06,
+      "loss": 0.6280115842819214,
+      "step": 461
+    },
+    {
+      "epoch": 0.4528856757750276,
+      "grad_norm": 5.405229091644287,
+      "learning_rate": 9.031049942752215e-06,
+      "loss": 0.8402045965194702,
+      "step": 462
+    },
+    {
+      "epoch": 0.45386594780051465,
+      "grad_norm": 1.6132272481918335,
+      "learning_rate": 9.026351287655294e-06,
+      "loss": 0.6735565066337585,
+      "step": 463
+    },
+    {
+      "epoch": 0.4548462198260017,
+      "grad_norm": 2.0643248558044434,
+      "learning_rate": 9.021642496223802e-06,
+      "loss": 0.6931006908416748,
+      "step": 464
+    },
+    {
+      "epoch": 0.45582649185148877,
+      "grad_norm": 1.0031558275222778,
+      "learning_rate": 9.016923580312114e-06,
+      "loss": 0.646364688873291,
+      "step": 465
+    },
+    {
+      "epoch": 0.4568067638769759,
+      "grad_norm": 0.6422945261001587,
+      "learning_rate": 9.012194551800098e-06,
+      "loss": 0.631780207157135,
+      "step": 466
+    },
+    {
+      "epoch": 0.45778703590246295,
+      "grad_norm": 1.1849571466445923,
+      "learning_rate": 9.007455422593077e-06,
+      "loss": 0.7650151252746582,
+      "step": 467
+    },
+    {
+      "epoch": 0.45876730792795,
+      "grad_norm": 2.384178400039673,
+      "learning_rate": 9.002706204621802e-06,
+      "loss": 0.6328111886978149,
+      "step": 468
+    },
+    {
+      "epoch": 0.45974757995343707,
+      "grad_norm": 1.2209985256195068,
+      "learning_rate": 8.997946909842426e-06,
+      "loss": 0.6249936819076538,
+      "step": 469
+    },
+    {
+      "epoch": 0.46072785197892413,
+      "grad_norm": 1.7335615158081055,
+      "learning_rate": 8.993177550236464e-06,
+      "loss": 0.6562061905860901,
+      "step": 470
+    },
+    {
+      "epoch": 0.46170812400441125,
+      "grad_norm": 0.7083120942115784,
+      "learning_rate": 8.988398137810778e-06,
+      "loss": 0.6238952875137329,
+      "step": 471
+    },
+    {
+      "epoch": 0.4626883960298983,
+      "grad_norm": 0.7924184799194336,
+      "learning_rate": 8.983608684597529e-06,
+      "loss": 0.741826057434082,
+      "step": 472
+    },
+    {
+      "epoch": 0.46366866805538537,
+      "grad_norm": 2.031369686126709,
+      "learning_rate": 8.978809202654161e-06,
+      "loss": 0.7883443832397461,
+      "step": 473
+    },
+    {
+      "epoch": 0.46464894008087243,
+      "grad_norm": 1.0610408782958984,
+      "learning_rate": 8.973999704063365e-06,
+      "loss": 0.7168911099433899,
+      "step": 474
+    },
+    {
+      "epoch": 0.4656292121063595,
+      "grad_norm": 0.630119264125824,
+      "learning_rate": 8.969180200933048e-06,
+      "loss": 0.8153673410415649,
+      "step": 475
+    },
+    {
+      "epoch": 0.4666094841318466,
+      "grad_norm": 0.7618991732597351,
+      "learning_rate": 8.9643507053963e-06,
+      "loss": 0.6666956543922424,
+      "step": 476
+    },
+    {
+      "epoch": 0.46758975615733367,
+      "grad_norm": 1.4962886571884155,
+      "learning_rate": 8.959511229611377e-06,
+      "loss": 0.8878688216209412,
+      "step": 477
+    },
+    {
+      "epoch": 0.46857002818282073,
+      "grad_norm": 0.809596061706543,
+      "learning_rate": 8.954661785761648e-06,
+      "loss": 0.7555930018424988,
+      "step": 478
+    },
+    {
+      "epoch": 0.4695503002083078,
+      "grad_norm": 0.6643649339675903,
+      "learning_rate": 8.949802386055582e-06,
+      "loss": 0.8297539949417114,
+      "step": 479
+    },
+    {
+      "epoch": 0.47053057223379485,
+      "grad_norm": 0.6694355010986328,
+      "learning_rate": 8.944933042726715e-06,
+      "loss": 0.7027547359466553,
+      "step": 480
+    },
+    {
+      "epoch": 0.47151084425928197,
+      "grad_norm": 1.3136928081512451,
+      "learning_rate": 8.94005376803361e-06,
+      "loss": 0.7977560758590698,
+      "step": 481
+    },
+    {
+      "epoch": 0.47249111628476903,
+      "grad_norm": 1.2374402284622192,
+      "learning_rate": 8.935164574259838e-06,
+      "loss": 0.7727890610694885,
+      "step": 482
+    },
+    {
+      "epoch": 0.4734713883102561,
+      "grad_norm": 0.6985059976577759,
+      "learning_rate": 8.930265473713939e-06,
+      "loss": 0.7002763748168945,
+      "step": 483
+    },
+    {
+      "epoch": 0.47445166033574315,
+      "grad_norm": 0.7153370976448059,
+      "learning_rate": 8.925356478729389e-06,
+      "loss": 0.6531966924667358,
+      "step": 484
+    },
+    {
+      "epoch": 0.4754319323612302,
+      "grad_norm": 3.023678779602051,
+      "learning_rate": 8.92043760166458e-06,
+      "loss": 0.7042747735977173,
+      "step": 485
+    },
+    {
+      "epoch": 0.47641220438671733,
+      "grad_norm": 0.653689980506897,
+      "learning_rate": 8.915508854902778e-06,
+      "loss": 0.5918471217155457,
+      "step": 486
+    },
+    {
+      "epoch": 0.4773924764122044,
+      "grad_norm": 1.532910943031311,
+      "learning_rate": 8.910570250852098e-06,
+      "loss": 0.5075575113296509,
+      "step": 487
+    },
+    {
+      "epoch": 0.47837274843769145,
+      "grad_norm": 0.7907715439796448,
+      "learning_rate": 8.905621801945467e-06,
+      "loss": 0.686042070388794,
+      "step": 488
+    },
+    {
+      "epoch": 0.4793530204631785,
+      "grad_norm": 1.0896567106246948,
+      "learning_rate": 8.900663520640605e-06,
+      "loss": 0.7102211713790894,
+      "step": 489
+    },
+    {
+      "epoch": 0.48033329248866563,
+      "grad_norm": 0.8049248456954956,
+      "learning_rate": 8.895695419419973e-06,
+      "loss": 0.9622594118118286,
+      "step": 490
+    },
+    {
+      "epoch": 0.4813135645141527,
+      "grad_norm": 1.256439208984375,
+      "learning_rate": 8.890717510790763e-06,
+      "loss": 0.6511240005493164,
+      "step": 491
+    },
+    {
+      "epoch": 0.48229383653963975,
+      "grad_norm": 1.679789662361145,
+      "learning_rate": 8.885729807284855e-06,
+      "loss": 0.7414318323135376,
+      "step": 492
+    },
+    {
+      "epoch": 0.4832741085651268,
+      "grad_norm": 0.930332601070404,
+      "learning_rate": 8.880732321458785e-06,
+      "loss": 0.6837952136993408,
+      "step": 493
+    },
+    {
+      "epoch": 0.4842543805906139,
+      "grad_norm": 1.4138216972351074,
+      "learning_rate": 8.875725065893717e-06,
+      "loss": 0.6818543672561646,
+      "step": 494
+    },
+    {
+      "epoch": 0.485234652616101,
+      "grad_norm": 0.6490036249160767,
+      "learning_rate": 8.870708053195414e-06,
+      "loss": 0.7358658313751221,
+      "step": 495
+    },
+    {
+      "epoch": 0.48621492464158805,
+      "grad_norm": 58.39289855957031,
+      "learning_rate": 8.865681295994195e-06,
+      "loss": 0.7429838180541992,
+      "step": 496
+    },
+    {
+      "epoch": 0.4871951966670751,
+      "grad_norm": 1.1145416498184204,
+      "learning_rate": 8.860644806944917e-06,
+      "loss": 0.5583229660987854,
+      "step": 497
+    },
+    {
+      "epoch": 0.4881754686925622,
+      "grad_norm": 0.9933784008026123,
+      "learning_rate": 8.85559859872694e-06,
+      "loss": 0.7169632911682129,
+      "step": 498
+    },
+    {
+      "epoch": 0.48915574071804924,
+      "grad_norm": 1.079107642173767,
+      "learning_rate": 8.850542684044078e-06,
+      "loss": 0.6038336753845215,
+      "step": 499
+    },
+    {
+      "epoch": 0.49013601274353635,
+      "grad_norm": 0.8761396408081055,
+      "learning_rate": 8.845477075624598e-06,
+      "loss": 0.6701620221138,
+      "step": 500
+    },
+    {
+      "epoch": 0.4911162847690234,
+      "grad_norm": 0.6637868285179138,
+      "learning_rate": 8.84040178622116e-06,
+      "loss": 0.7567762136459351,
+      "step": 501
+    },
+    {
+      "epoch": 0.4920965567945105,
+      "grad_norm": 2.0560238361358643,
+      "learning_rate": 8.835316828610801e-06,
+      "loss": 0.7823659777641296,
+      "step": 502
+    },
+    {
+      "epoch": 0.49307682881999754,
+      "grad_norm": 0.772253692150116,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.7584106922149658,
+      "step": 503
+    },
+    {
+      "epoch": 0.4940571008454846,
+      "grad_norm": 0.8275479078292847,
+      "learning_rate": 8.825117959999117e-06,
+      "loss": 0.5793484449386597,
+      "step": 504
+    },
+    {
+      "epoch": 0.4950373728709717,
+      "grad_norm": 0.9391909241676331,
+      "learning_rate": 8.820004074673433e-06,
+      "loss": 0.7406161427497864,
+      "step": 505
+    },
+    {
+      "epoch": 0.4960176448964588,
+      "grad_norm": 1.1446795463562012,
+      "learning_rate": 8.814880572492043e-06,
+      "loss": 0.6543604135513306,
+      "step": 506
+    },
+    {
+      "epoch": 0.49699791692194584,
+      "grad_norm": 0.9661545157432556,
+      "learning_rate": 8.809747466353356e-06,
+      "loss": 0.8787537217140198,
+      "step": 507
+    },
+    {
+      "epoch": 0.4979781889474329,
+      "grad_norm": 0.9429580569267273,
+      "learning_rate": 8.804604769179958e-06,
+      "loss": 0.5396296381950378,
+      "step": 508
+    },
+    {
+      "epoch": 0.49895846097291996,
+      "grad_norm": 3.152589797973633,
+      "learning_rate": 8.799452493918586e-06,
+      "loss": 0.8350993394851685,
+      "step": 509
+    },
+    {
+      "epoch": 0.4999387329984071,
+      "grad_norm": 0.9701963663101196,
+      "learning_rate": 8.794290653540086e-06,
+      "loss": 0.8153249025344849,
+      "step": 510
+    },
+    {
+      "epoch": 0.5009190050238941,
+      "grad_norm": 0.8648568987846375,
+      "learning_rate": 8.789119261039385e-06,
+      "loss": 0.5465856790542603,
+      "step": 511
+    },
+    {
+      "epoch": 0.5018992770493812,
+      "grad_norm": 3.4678969383239746,
+      "learning_rate": 8.783938329435458e-06,
+      "loss": 0.596847653388977,
+      "step": 512
+    },
+    {
+      "epoch": 0.5028795490748683,
+      "grad_norm": 1.1874382495880127,
+      "learning_rate": 8.778747871771293e-06,
+      "loss": 0.6162095069885254,
+      "step": 513
+    },
+    {
+      "epoch": 0.5038598211003553,
+      "grad_norm": 0.6420893669128418,
+      "learning_rate": 8.773547901113862e-06,
+      "loss": 0.6212583780288696,
+      "step": 514
+    },
+    {
+      "epoch": 0.5048400931258424,
+      "grad_norm": 0.6174635887145996,
+      "learning_rate": 8.768338430554083e-06,
+      "loss": 0.6156861782073975,
+      "step": 515
+    },
+    {
+      "epoch": 0.5058203651513294,
+      "grad_norm": 0.8119685649871826,
+      "learning_rate": 8.763119473206795e-06,
+      "loss": 0.7010501623153687,
+      "step": 516
+    },
+    {
+      "epoch": 0.5068006371768166,
+      "grad_norm": 0.8714487552642822,
+      "learning_rate": 8.757891042210713e-06,
+      "loss": 0.8170113563537598,
+      "step": 517
+    },
+    {
+      "epoch": 0.5077809092023037,
+      "grad_norm": 1.6170095205307007,
+      "learning_rate": 8.752653150728412e-06,
+      "loss": 0.6287787556648254,
+      "step": 518
+    },
+    {
+      "epoch": 0.5087611812277907,
+      "grad_norm": 1.0796136856079102,
+      "learning_rate": 8.747405811946272e-06,
+      "loss": 0.5999889373779297,
+      "step": 519
+    },
+    {
+      "epoch": 0.5097414532532778,
+      "grad_norm": 0.7786392569541931,
+      "learning_rate": 8.742149039074463e-06,
+      "loss": 0.5764816999435425,
+      "step": 520
+    },
+    {
+      "epoch": 0.5107217252787648,
+      "grad_norm": 0.7494414448738098,
+      "learning_rate": 8.736882845346906e-06,
+      "loss": 0.778597891330719,
+      "step": 521
+    },
+    {
+      "epoch": 0.5117019973042519,
+      "grad_norm": 1.2446948289871216,
+      "learning_rate": 8.731607244021236e-06,
+      "loss": 0.693345844745636,
+      "step": 522
+    },
+    {
+      "epoch": 0.512682269329739,
+      "grad_norm": 1.7622649669647217,
+      "learning_rate": 8.726322248378775e-06,
+      "loss": 0.7455325126647949,
+      "step": 523
+    },
+    {
+      "epoch": 0.513662541355226,
+      "grad_norm": 0.804429292678833,
+      "learning_rate": 8.72102787172449e-06,
+      "loss": 0.6732499599456787,
+      "step": 524
+    },
+    {
+      "epoch": 0.5146428133807132,
+      "grad_norm": 1.209266185760498,
+      "learning_rate": 8.715724127386971e-06,
+      "loss": 0.7108303308486938,
+      "step": 525
+    },
+    {
+      "epoch": 0.5156230854062002,
+      "grad_norm": 0.9574494957923889,
+      "learning_rate": 8.710411028718388e-06,
+      "loss": 0.5963371396064758,
+      "step": 526
+    },
+    {
+      "epoch": 0.5166033574316873,
+      "grad_norm": 1.329498291015625,
+      "learning_rate": 8.705088589094458e-06,
+      "loss": 0.5417718887329102,
+      "step": 527
+    },
+    {
+      "epoch": 0.5175836294571744,
+      "grad_norm": 1.1467244625091553,
+      "learning_rate": 8.69975682191442e-06,
+      "loss": 0.7108482718467712,
+      "step": 528
+    },
+    {
+      "epoch": 0.5185639014826614,
+      "grad_norm": 1.128373622894287,
+      "learning_rate": 8.69441574060099e-06,
+      "loss": 0.6100748181343079,
+      "step": 529
+    },
+    {
+      "epoch": 0.5195441735081485,
+      "grad_norm": 0.5896838307380676,
+      "learning_rate": 8.68906535860033e-06,
+      "loss": 0.6295212507247925,
+      "step": 530
+    },
+    {
+      "epoch": 0.5205244455336356,
+      "grad_norm": 0.6560076475143433,
+      "learning_rate": 8.683705689382025e-06,
+      "loss": 0.7203050851821899,
+      "step": 531
+    },
+    {
+      "epoch": 0.5215047175591226,
+      "grad_norm": 0.9406687021255493,
+      "learning_rate": 8.678336746439036e-06,
+      "loss": 0.7831158638000488,
+      "step": 532
+    },
+    {
+      "epoch": 0.5224849895846098,
+      "grad_norm": 1.2101908922195435,
+      "learning_rate": 8.672958543287666e-06,
+      "loss": 0.6862603425979614,
+      "step": 533
+    },
+    {
+      "epoch": 0.5234652616100968,
+      "grad_norm": 1.2099987268447876,
+      "learning_rate": 8.667571093467542e-06,
+      "loss": 0.6591293215751648,
+      "step": 534
+    },
+    {
+      "epoch": 0.5244455336355839,
+      "grad_norm": 0.9246074557304382,
+      "learning_rate": 8.662174410541556e-06,
+      "loss": 0.7038129568099976,
+      "step": 535
+    },
+    {
+      "epoch": 0.525425805661071,
+      "grad_norm": 0.7076787352561951,
+      "learning_rate": 8.656768508095853e-06,
+      "loss": 0.6776260137557983,
+      "step": 536
+    },
+    {
+      "epoch": 0.526406077686558,
+      "grad_norm": 0.8369212746620178,
+      "learning_rate": 8.651353399739787e-06,
+      "loss": 0.7117469310760498,
+      "step": 537
+    },
+    {
+      "epoch": 0.5273863497120451,
+      "grad_norm": 1.4560307264328003,
+      "learning_rate": 8.645929099105886e-06,
+      "loss": 0.6573420763015747,
+      "step": 538
+    },
+    {
+      "epoch": 0.5283666217375321,
+      "grad_norm": 1.4285420179367065,
+      "learning_rate": 8.640495619849821e-06,
+      "loss": 0.7345580458641052,
+      "step": 539
+    },
+    {
+      "epoch": 0.5293468937630192,
+      "grad_norm": 0.915946900844574,
+      "learning_rate": 8.63505297565037e-06,
+      "loss": 0.6241335272789001,
+      "step": 540
+    },
+    {
+      "epoch": 0.5303271657885064,
+      "grad_norm": 0.930578887462616,
+      "learning_rate": 8.629601180209382e-06,
+      "loss": 0.711024284362793,
+      "step": 541
+    },
+    {
+      "epoch": 0.5313074378139934,
+      "grad_norm": 0.6595850586891174,
+      "learning_rate": 8.624140247251745e-06,
+      "loss": 0.630510687828064,
+      "step": 542
+    },
+    {
+      "epoch": 0.5322877098394805,
+      "grad_norm": 0.9870410561561584,
+      "learning_rate": 8.61867019052535e-06,
+      "loss": 0.5863507986068726,
+      "step": 543
+    },
+    {
+      "epoch": 0.5332679818649675,
+      "grad_norm": 0.8339908719062805,
+      "learning_rate": 8.613191023801065e-06,
+      "loss": 0.6086613535881042,
+      "step": 544
+    },
+    {
+      "epoch": 0.5342482538904546,
+      "grad_norm": 1.0823789834976196,
+      "learning_rate": 8.607702760872679e-06,
+      "loss": 0.6681934595108032,
+      "step": 545
+    },
+    {
+      "epoch": 0.5352285259159417,
+      "grad_norm": 0.7219529747962952,
+      "learning_rate": 8.602205415556889e-06,
+      "loss": 0.6179887056350708,
+      "step": 546
+    },
+    {
+      "epoch": 0.5362087979414287,
+      "grad_norm": 1.165724754333496,
+      "learning_rate": 8.596699001693257e-06,
+      "loss": 0.5813823938369751,
+      "step": 547
+    },
+    {
+      "epoch": 0.5371890699669158,
+      "grad_norm": 0.7914849519729614,
+      "learning_rate": 8.591183533144172e-06,
+      "loss": 0.6549311876296997,
+      "step": 548
+    },
+    {
+      "epoch": 0.5381693419924028,
+      "grad_norm": 1.8187445402145386,
+      "learning_rate": 8.585659023794818e-06,
+      "loss": 0.5614180564880371,
+      "step": 549
+    },
+    {
+      "epoch": 0.53914961401789,
+      "grad_norm": 0.8814314007759094,
+      "learning_rate": 8.580125487553143e-06,
+      "loss": 0.6609047055244446,
+      "step": 550
+    },
+    {
+      "epoch": 0.5401298860433771,
+      "grad_norm": 0.6639443039894104,
+      "learning_rate": 8.574582938349818e-06,
+      "loss": 0.6361253261566162,
+      "step": 551
+    },
+    {
+      "epoch": 0.5411101580688641,
+      "grad_norm": 0.628507673740387,
+      "learning_rate": 8.569031390138202e-06,
+      "loss": 0.7554982900619507,
+      "step": 552
+    },
+    {
+      "epoch": 0.5420904300943512,
+      "grad_norm": 0.869846522808075,
+      "learning_rate": 8.563470856894316e-06,
+      "loss": 0.6098361015319824,
+      "step": 553
+    },
+    {
+      "epoch": 0.5430707021198382,
+      "grad_norm": 0.9109020233154297,
+      "learning_rate": 8.557901352616788e-06,
+      "loss": 0.7274138331413269,
+      "step": 554
+    },
+    {
+      "epoch": 0.5440509741453253,
+      "grad_norm": 1.5429188013076782,
+      "learning_rate": 8.552322891326846e-06,
+      "loss": 0.6473425030708313,
+      "step": 555
+    },
+    {
+      "epoch": 0.5450312461708124,
+      "grad_norm": 0.7426435947418213,
+      "learning_rate": 8.546735487068253e-06,
+      "loss": 0.6691425442695618,
+      "step": 556
+    },
+    {
+      "epoch": 0.5460115181962995,
+      "grad_norm": 1.1588733196258545,
+      "learning_rate": 8.541139153907296e-06,
+      "loss": 0.6738032102584839,
+      "step": 557
+    },
+    {
+      "epoch": 0.5469917902217866,
+      "grad_norm": 0.7680075168609619,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.564139723777771,
+      "step": 558
+    },
+    {
+      "epoch": 0.5479720622472736,
+      "grad_norm": 0.6689730286598206,
+      "learning_rate": 8.529919757255783e-06,
+      "loss": 0.6075266599655151,
+      "step": 559
+    },
+    {
+      "epoch": 0.5489523342727607,
+      "grad_norm": 5.509249210357666,
+      "learning_rate": 8.524296722010042e-06,
+      "loss": 0.6835139393806458,
+      "step": 560
+    },
+    {
+      "epoch": 0.5499326062982478,
+      "grad_norm": 0.8517597317695618,
+      "learning_rate": 8.518664814351502e-06,
+      "loss": 0.6909633874893188,
+      "step": 561
+    },
+    {
+      "epoch": 0.5509128783237348,
+      "grad_norm": 0.738439679145813,
+      "learning_rate": 8.513024048458484e-06,
+      "loss": 0.6094111204147339,
+      "step": 562
+    },
+    {
+      "epoch": 0.5518931503492219,
+      "grad_norm": 0.7751814126968384,
+      "learning_rate": 8.507374438531606e-06,
+      "loss": 0.7162400484085083,
+      "step": 563
+    },
+    {
+      "epoch": 0.5528734223747089,
+      "grad_norm": 1.2148354053497314,
+      "learning_rate": 8.501715998793758e-06,
+      "loss": 0.5730451941490173,
+      "step": 564
+    },
+    {
+      "epoch": 0.553853694400196,
+      "grad_norm": 0.6777694821357727,
+      "learning_rate": 8.496048743490053e-06,
+      "loss": 0.6548066139221191,
+      "step": 565
+    },
+    {
+      "epoch": 0.5548339664256832,
+      "grad_norm": 0.6004366278648376,
+      "learning_rate": 8.490372686887802e-06,
+      "loss": 0.6705085635185242,
+      "step": 566
+    },
+    {
+      "epoch": 0.5558142384511702,
+      "grad_norm": 0.9465246796607971,
+      "learning_rate": 8.48468784327647e-06,
+      "loss": 0.762035608291626,
+      "step": 567
+    },
+    {
+      "epoch": 0.5567945104766573,
+      "grad_norm": 0.7760327458381653,
+      "learning_rate": 8.478994226967638e-06,
+      "loss": 0.6888554096221924,
+      "step": 568
+    },
+    {
+      "epoch": 0.5577747825021443,
+      "grad_norm": 1.0672935247421265,
+      "learning_rate": 8.473291852294986e-06,
+      "loss": 0.6906014084815979,
+      "step": 569
+    },
+    {
+      "epoch": 0.5587550545276314,
+      "grad_norm": 1.2181823253631592,
+      "learning_rate": 8.467580733614233e-06,
+      "loss": 0.7269997000694275,
+      "step": 570
+    },
+    {
+      "epoch": 0.5597353265531185,
+      "grad_norm": 0.8513377904891968,
+      "learning_rate": 8.461860885303116e-06,
+      "loss": 0.6398078203201294,
+      "step": 571
+    },
+    {
+      "epoch": 0.5607155985786055,
+      "grad_norm": 5.012400150299072,
+      "learning_rate": 8.45613232176134e-06,
+      "loss": 0.6001594066619873,
+      "step": 572
+    },
+    {
+      "epoch": 0.5616958706040927,
+      "grad_norm": 1.0870821475982666,
+      "learning_rate": 8.450395057410561e-06,
+      "loss": 0.6807757616043091,
+      "step": 573
+    },
+    {
+      "epoch": 0.5626761426295797,
+      "grad_norm": 1.2243040800094604,
+      "learning_rate": 8.444649106694335e-06,
+      "loss": 0.7422083616256714,
+      "step": 574
+    },
+    {
+      "epoch": 0.5636564146550668,
+      "grad_norm": 0.7392075061798096,
+      "learning_rate": 8.438894484078086e-06,
+      "loss": 0.5492063760757446,
+      "step": 575
+    },
+    {
+      "epoch": 0.5646366866805539,
+      "grad_norm": 0.892780601978302,
+      "learning_rate": 8.433131204049067e-06,
+      "loss": 0.7455860376358032,
+      "step": 576
+    },
+    {
+      "epoch": 0.5656169587060409,
+      "grad_norm": 0.6653965711593628,
+      "learning_rate": 8.427359281116335e-06,
+      "loss": 0.5985487103462219,
+      "step": 577
+    },
+    {
+      "epoch": 0.566597230731528,
+      "grad_norm": 0.9835824966430664,
+      "learning_rate": 8.421578729810693e-06,
+      "loss": 0.7031208276748657,
+      "step": 578
+    },
+    {
+      "epoch": 0.567577502757015,
+      "grad_norm": 0.9250913262367249,
+      "learning_rate": 8.415789564684673e-06,
+      "loss": 0.8197964429855347,
+      "step": 579
+    },
+    {
+      "epoch": 0.5685577747825021,
+      "grad_norm": 0.7315778732299805,
+      "learning_rate": 8.409991800312493e-06,
+      "loss": 0.6521624326705933,
+      "step": 580
+    },
+    {
+      "epoch": 0.5695380468079893,
+      "grad_norm": 0.9971609115600586,
+      "learning_rate": 8.404185451290017e-06,
+      "loss": 0.744938850402832,
+      "step": 581
+    },
+    {
+      "epoch": 0.5705183188334763,
+      "grad_norm": 18.52988052368164,
+      "learning_rate": 8.398370532234723e-06,
+      "loss": 0.8215644955635071,
+      "step": 582
+    },
+    {
+      "epoch": 0.5714985908589634,
+      "grad_norm": 0.7019656896591187,
+      "learning_rate": 8.392547057785662e-06,
+      "loss": 0.4670473039150238,
+      "step": 583
+    },
+    {
+      "epoch": 0.5724788628844505,
+      "grad_norm": 1.5505694150924683,
+      "learning_rate": 8.386715042603424e-06,
+      "loss": 0.6690788865089417,
+      "step": 584
+    },
+    {
+      "epoch": 0.5734591349099375,
+      "grad_norm": 1.6945151090621948,
+      "learning_rate": 8.380874501370098e-06,
+      "loss": 0.6558585166931152,
+      "step": 585
+    },
+    {
+      "epoch": 0.5744394069354246,
+      "grad_norm": 1.2926048040390015,
+      "learning_rate": 8.375025448789242e-06,
+      "loss": 0.5685740113258362,
+      "step": 586
+    },
+    {
+      "epoch": 0.5754196789609116,
+      "grad_norm": 1.8694062232971191,
+      "learning_rate": 8.36916789958584e-06,
+      "loss": 0.6317107677459717,
+      "step": 587
+    },
+    {
+      "epoch": 0.5763999509863987,
+      "grad_norm": 3.3498735427856445,
+      "learning_rate": 8.363301868506264e-06,
+      "loss": 0.7119582891464233,
+      "step": 588
+    },
+    {
+      "epoch": 0.5773802230118859,
+      "grad_norm": 1.1095170974731445,
+      "learning_rate": 8.357427370318239e-06,
+      "loss": 0.6322152018547058,
+      "step": 589
+    },
+    {
+      "epoch": 0.5783604950373729,
+      "grad_norm": 3.1217989921569824,
+      "learning_rate": 8.351544419810808e-06,
+      "loss": 0.5786447525024414,
+      "step": 590
+    },
+    {
+      "epoch": 0.57934076706286,
+      "grad_norm": 2.517881155014038,
+      "learning_rate": 8.345653031794292e-06,
+      "loss": 0.641181468963623,
+      "step": 591
+    },
+    {
+      "epoch": 0.580321039088347,
+      "grad_norm": 0.6871888637542725,
+      "learning_rate": 8.339753221100254e-06,
+      "loss": 0.6676073670387268,
+      "step": 592
+    },
+    {
+      "epoch": 0.5813013111138341,
+      "grad_norm": 0.8354384899139404,
+      "learning_rate": 8.33384500258146e-06,
+      "loss": 0.8618124127388,
+      "step": 593
+    },
+    {
+      "epoch": 0.5822815831393212,
+      "grad_norm": 1.581214427947998,
+      "learning_rate": 8.327928391111841e-06,
+      "loss": 0.6307835578918457,
+      "step": 594
+    },
+    {
+      "epoch": 0.5832618551648082,
+      "grad_norm": 1.9953625202178955,
+      "learning_rate": 8.322003401586463e-06,
+      "loss": 0.739708423614502,
+      "step": 595
+    },
+    {
+      "epoch": 0.5842421271902953,
+      "grad_norm": 1.117095947265625,
+      "learning_rate": 8.316070048921477e-06,
+      "loss": 0.5766733884811401,
+      "step": 596
+    },
+    {
+      "epoch": 0.5852223992157823,
+      "grad_norm": 2.9105641841888428,
+      "learning_rate": 8.310128348054093e-06,
+      "loss": 0.5194284915924072,
+      "step": 597
+    },
+    {
+      "epoch": 0.5862026712412695,
+      "grad_norm": 2.1335606575012207,
+      "learning_rate": 8.304178313942536e-06,
+      "loss": 0.6000288128852844,
+      "step": 598
+    },
+    {
+      "epoch": 0.5871829432667566,
+      "grad_norm": 1.72444748878479,
+      "learning_rate": 8.298219961566008e-06,
+      "loss": 0.6127111911773682,
+      "step": 599
+    },
+    {
+      "epoch": 0.5881632152922436,
+      "grad_norm": 1.1741032600402832,
+      "learning_rate": 8.292253305924656e-06,
+      "loss": 0.6842024326324463,
+      "step": 600
+    },
+    {
+      "epoch": 0.5891434873177307,
+      "grad_norm": 0.9328085780143738,
+      "learning_rate": 8.286278362039527e-06,
+      "loss": 0.7023148536682129,
+      "step": 601
+    },
+    {
+      "epoch": 0.5901237593432177,
+      "grad_norm": 0.6796233654022217,
+      "learning_rate": 8.280295144952537e-06,
+      "loss": 0.7303265333175659,
+      "step": 602
+    },
+    {
+      "epoch": 0.5911040313687048,
+      "grad_norm": 1.214897632598877,
+      "learning_rate": 8.274303669726427e-06,
+      "loss": 0.570577085018158,
+      "step": 603
+    },
+    {
+      "epoch": 0.5920843033941919,
+      "grad_norm": 0.8431791663169861,
+      "learning_rate": 8.268303951444727e-06,
+      "loss": 0.5560767650604248,
+      "step": 604
+    },
+    {
+      "epoch": 0.5930645754196789,
+      "grad_norm": 0.7330615520477295,
+      "learning_rate": 8.262296005211722e-06,
+      "loss": 0.5567612051963806,
+      "step": 605
+    },
+    {
+      "epoch": 0.5940448474451661,
+      "grad_norm": 1.3167418241500854,
+      "learning_rate": 8.25627984615241e-06,
+      "loss": 0.7696295380592346,
+      "step": 606
+    },
+    {
+      "epoch": 0.5950251194706531,
+      "grad_norm": 0.9472094178199768,
+      "learning_rate": 8.250255489412464e-06,
+      "loss": 0.6157063245773315,
+      "step": 607
+    },
+    {
+      "epoch": 0.5960053914961402,
+      "grad_norm": 0.7250748872756958,
+      "learning_rate": 8.244222950158194e-06,
+      "loss": 0.7223548889160156,
+      "step": 608
+    },
+    {
+      "epoch": 0.5969856635216273,
+      "grad_norm": 0.7567031979560852,
+      "learning_rate": 8.238182243576512e-06,
+      "loss": 0.5250760912895203,
+      "step": 609
+    },
+    {
+      "epoch": 0.5979659355471143,
+      "grad_norm": 7.1886396408081055,
+      "learning_rate": 8.232133384874887e-06,
+      "loss": 0.6256387829780579,
+      "step": 610
+    },
+    {
+      "epoch": 0.5989462075726014,
+      "grad_norm": 0.8025698661804199,
+      "learning_rate": 8.226076389281316e-06,
+      "loss": 0.6919379234313965,
+      "step": 611
+    },
+    {
+      "epoch": 0.5999264795980884,
+      "grad_norm": 0.6828694343566895,
+      "learning_rate": 8.220011272044276e-06,
+      "loss": 0.6849675178527832,
+      "step": 612
+    },
+    {
+      "epoch": 0.6009067516235755,
+      "grad_norm": 0.7576507925987244,
+      "learning_rate": 8.213938048432697e-06,
+      "loss": 0.6458473801612854,
+      "step": 613
+    },
+    {
+      "epoch": 0.6018870236490627,
+      "grad_norm": 0.6589511036872864,
+      "learning_rate": 8.207856733735908e-06,
+      "loss": 0.6043916940689087,
+      "step": 614
+    },
+    {
+      "epoch": 0.6028672956745497,
+      "grad_norm": 1.3640341758728027,
+      "learning_rate": 8.201767343263612e-06,
+      "loss": 0.6255053281784058,
+      "step": 615
+    },
+    {
+      "epoch": 0.6038475677000368,
+      "grad_norm": 0.67889004945755,
+      "learning_rate": 8.195669892345842e-06,
+      "loss": 0.5921801328659058,
+      "step": 616
+    },
+    {
+      "epoch": 0.6048278397255238,
+      "grad_norm": 0.707345187664032,
+      "learning_rate": 8.189564396332927e-06,
+      "loss": 0.6156002283096313,
+      "step": 617
+    },
+    {
+      "epoch": 0.6058081117510109,
+      "grad_norm": 0.8241771459579468,
+      "learning_rate": 8.183450870595443e-06,
+      "loss": 0.5504246354103088,
+      "step": 618
+    },
+    {
+      "epoch": 0.606788383776498,
+      "grad_norm": 0.7092338800430298,
+      "learning_rate": 8.177329330524182e-06,
+      "loss": 0.5916330814361572,
+      "step": 619
+    },
+    {
+      "epoch": 0.607768655801985,
+      "grad_norm": 0.6089708209037781,
+      "learning_rate": 8.171199791530117e-06,
+      "loss": 0.6382315754890442,
+      "step": 620
+    },
+    {
+      "epoch": 0.6087489278274721,
+      "grad_norm": 0.9124857783317566,
+      "learning_rate": 8.165062269044353e-06,
+      "loss": 0.5988181829452515,
+      "step": 621
+    },
+    {
+      "epoch": 0.6097291998529591,
+      "grad_norm": 1.3915491104125977,
+      "learning_rate": 8.158916778518097e-06,
+      "loss": 0.6557927131652832,
+      "step": 622
+    },
+    {
+      "epoch": 0.6107094718784463,
+      "grad_norm": 0.9545561075210571,
+      "learning_rate": 8.152763335422612e-06,
+      "loss": 0.6603102087974548,
+      "step": 623
+    },
+    {
+      "epoch": 0.6116897439039334,
+      "grad_norm": 0.6929757595062256,
+      "learning_rate": 8.146601955249187e-06,
+      "loss": 0.6477068066596985,
+      "step": 624
+    },
+    {
+      "epoch": 0.6126700159294204,
+      "grad_norm": 0.7381318211555481,
+      "learning_rate": 8.140432653509089e-06,
+      "loss": 0.635793924331665,
+      "step": 625
+    },
+    {
+      "epoch": 0.6136502879549075,
+      "grad_norm": 1.5131570100784302,
+      "learning_rate": 8.134255445733524e-06,
+      "loss": 0.7369328141212463,
+      "step": 626
+    },
+    {
+      "epoch": 0.6146305599803945,
+      "grad_norm": 0.7864885330200195,
+      "learning_rate": 8.128070347473609e-06,
+      "loss": 0.6652648448944092,
+      "step": 627
+    },
+    {
+      "epoch": 0.6156108320058816,
+      "grad_norm": 0.5698316693305969,
+      "learning_rate": 8.121877374300318e-06,
+      "loss": 0.583881139755249,
+      "step": 628
+    },
+    {
+      "epoch": 0.6165911040313687,
+      "grad_norm": 0.7332568764686584,
+      "learning_rate": 8.115676541804456e-06,
+      "loss": 0.5860457420349121,
+      "step": 629
+    },
+    {
+      "epoch": 0.6175713760568557,
+      "grad_norm": 0.7971566915512085,
+      "learning_rate": 8.109467865596612e-06,
+      "loss": 0.5995970368385315,
+      "step": 630
+    },
+    {
+      "epoch": 0.6185516480823429,
+      "grad_norm": 1.5517250299453735,
+      "learning_rate": 8.10325136130712e-06,
+      "loss": 0.6478885412216187,
+      "step": 631
+    },
+    {
+      "epoch": 0.6195319201078299,
+      "grad_norm": 0.930991530418396,
+      "learning_rate": 8.097027044586021e-06,
+      "loss": 0.8530893325805664,
+      "step": 632
+    },
+    {
+      "epoch": 0.620512192133317,
+      "grad_norm": 0.862178385257721,
+      "learning_rate": 8.090794931103026e-06,
+      "loss": 0.6163445711135864,
+      "step": 633
+    },
+    {
+      "epoch": 0.6214924641588041,
+      "grad_norm": 0.7164607048034668,
+      "learning_rate": 8.084555036547475e-06,
+      "loss": 0.6941037178039551,
+      "step": 634
+    },
+    {
+      "epoch": 0.6224727361842911,
+      "grad_norm": 1.1507890224456787,
+      "learning_rate": 8.078307376628292e-06,
+      "loss": 0.5343766212463379,
+      "step": 635
+    },
+    {
+      "epoch": 0.6234530082097782,
+      "grad_norm": 0.7418875694274902,
+      "learning_rate": 8.072051967073955e-06,
+      "loss": 0.6268494129180908,
+      "step": 636
+    },
+    {
+      "epoch": 0.6244332802352652,
+      "grad_norm": 0.985659122467041,
+      "learning_rate": 8.065788823632451e-06,
+      "loss": 0.4631681442260742,
+      "step": 637
+    },
+    {
+      "epoch": 0.6254135522607523,
+      "grad_norm": 0.8198592662811279,
+      "learning_rate": 8.059517962071234e-06,
+      "loss": 0.5277544856071472,
+      "step": 638
+    },
+    {
+      "epoch": 0.6263938242862395,
+      "grad_norm": 0.8775269389152527,
+      "learning_rate": 8.053239398177191e-06,
+      "loss": 0.6032053232192993,
+      "step": 639
+    },
+    {
+      "epoch": 0.6273740963117265,
+      "grad_norm": 0.7492501139640808,
+      "learning_rate": 8.046953147756601e-06,
+      "loss": 0.6798757314682007,
+      "step": 640
+    },
+    {
+      "epoch": 0.6283543683372136,
+      "grad_norm": 0.9301172494888306,
+      "learning_rate": 8.04065922663509e-06,
+      "loss": 0.5427758693695068,
+      "step": 641
+    },
+    {
+      "epoch": 0.6293346403627007,
+      "grad_norm": 1.4323341846466064,
+      "learning_rate": 8.034357650657597e-06,
+      "loss": 0.8664883375167847,
+      "step": 642
+    },
+    {
+      "epoch": 0.6303149123881877,
+      "grad_norm": 0.7746670246124268,
+      "learning_rate": 8.028048435688333e-06,
+      "loss": 0.5990991592407227,
+      "step": 643
+    },
+    {
+      "epoch": 0.6312951844136748,
+      "grad_norm": 0.6360200047492981,
+      "learning_rate": 8.02173159761074e-06,
+      "loss": 0.6479265689849854,
+      "step": 644
+    },
+    {
+      "epoch": 0.6322754564391618,
+      "grad_norm": 0.9644507765769958,
+      "learning_rate": 8.015407152327448e-06,
+      "loss": 0.7435730695724487,
+      "step": 645
+    },
+    {
+      "epoch": 0.633255728464649,
+      "grad_norm": 1.2824594974517822,
+      "learning_rate": 8.009075115760243e-06,
+      "loss": 0.7728074789047241,
+      "step": 646
+    },
+    {
+      "epoch": 0.6342360004901361,
+      "grad_norm": 1.1604210138320923,
+      "learning_rate": 8.002735503850016e-06,
+      "loss": 0.596459686756134,
+      "step": 647
+    },
+    {
+      "epoch": 0.6352162725156231,
+      "grad_norm": 0.8743922114372253,
+      "learning_rate": 7.996388332556735e-06,
+      "loss": 0.7101293802261353,
+      "step": 648
+    },
+    {
+      "epoch": 0.6361965445411102,
+      "grad_norm": 0.9315487742424011,
+      "learning_rate": 7.990033617859396e-06,
+      "loss": 0.5783198475837708,
+      "step": 649
+    },
+    {
+      "epoch": 0.6371768165665972,
+      "grad_norm": 0.6622239351272583,
+      "learning_rate": 7.983671375755986e-06,
+      "loss": 0.6646347641944885,
+      "step": 650
+    },
+    {
+      "epoch": 0.6381570885920843,
+      "grad_norm": 0.5275871753692627,
+      "learning_rate": 7.97730162226344e-06,
+      "loss": 0.7172477841377258,
+      "step": 651
+    },
+    {
+      "epoch": 0.6391373606175714,
+      "grad_norm": 0.7323616743087769,
+      "learning_rate": 7.970924373417608e-06,
+      "loss": 0.6967377662658691,
+      "step": 652
+    },
+    {
+      "epoch": 0.6401176326430584,
+      "grad_norm": 1.7099934816360474,
+      "learning_rate": 7.964539645273204e-06,
+      "loss": 0.5868270397186279,
+      "step": 653
+    },
+    {
+      "epoch": 0.6410979046685455,
+      "grad_norm": 0.665709376335144,
+      "learning_rate": 7.958147453903773e-06,
+      "loss": 0.5133366584777832,
+      "step": 654
+    },
+    {
+      "epoch": 0.6420781766940326,
+      "grad_norm": 0.6716167330741882,
+      "learning_rate": 7.951747815401651e-06,
+      "loss": 0.6083835363388062,
+      "step": 655
+    },
+    {
+      "epoch": 0.6430584487195197,
+      "grad_norm": 0.6983871459960938,
+      "learning_rate": 7.945340745877919e-06,
+      "loss": 0.6565608382225037,
+      "step": 656
+    },
+    {
+      "epoch": 0.6440387207450068,
+      "grad_norm": 2.223074436187744,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.7368947267532349,
+      "step": 657
+    },
+    {
+      "epoch": 0.6450189927704938,
+      "grad_norm": 0.8114476203918457,
+      "learning_rate": 7.932504378303452e-06,
+      "loss": 0.7846980094909668,
+      "step": 658
+    },
+    {
+      "epoch": 0.6459992647959809,
+      "grad_norm": 0.6098480820655823,
+      "learning_rate": 7.92607511256826e-06,
+      "loss": 0.5550116300582886,
+      "step": 659
+    },
+    {
+      "epoch": 0.6469795368214679,
+      "grad_norm": 0.8538875579833984,
+      "learning_rate": 7.919638480442453e-06,
+      "loss": 0.6363775730133057,
+      "step": 660
+    },
+    {
+      "epoch": 0.647959808846955,
+      "grad_norm": 0.6441235542297363,
+      "learning_rate": 7.913194498130252e-06,
+      "loss": 0.5846317410469055,
+      "step": 661
+    },
+    {
+      "epoch": 0.6489400808724421,
+      "grad_norm": 0.6625627279281616,
+      "learning_rate": 7.90674318185437e-06,
+      "loss": 0.6980340480804443,
+      "step": 662
+    },
+    {
+      "epoch": 0.6499203528979292,
+      "grad_norm": 0.8105193972587585,
+      "learning_rate": 7.900284547855992e-06,
+      "loss": 0.6991692781448364,
+      "step": 663
+    },
+    {
+      "epoch": 0.6509006249234163,
+      "grad_norm": 0.74409419298172,
+      "learning_rate": 7.893818612394717e-06,
+      "loss": 0.6157128810882568,
+      "step": 664
+    },
+    {
+      "epoch": 0.6518808969489033,
+      "grad_norm": 0.9216258525848389,
+      "learning_rate": 7.887345391748533e-06,
+      "loss": 0.6117196679115295,
+      "step": 665
+    },
+    {
+      "epoch": 0.6528611689743904,
+      "grad_norm": 0.7538702487945557,
+      "learning_rate": 7.880864902213765e-06,
+      "loss": 0.7590198516845703,
+      "step": 666
+    },
+    {
+      "epoch": 0.6538414409998775,
+      "grad_norm": 2.2449469566345215,
+      "learning_rate": 7.874377160105037e-06,
+      "loss": 0.6009644865989685,
+      "step": 667
+    },
+    {
+      "epoch": 0.6548217130253645,
+      "grad_norm": 1.5578954219818115,
+      "learning_rate": 7.86788218175523e-06,
+      "loss": 0.7116764783859253,
+      "step": 668
+    },
+    {
+      "epoch": 0.6558019850508516,
+      "grad_norm": 0.8876961469650269,
+      "learning_rate": 7.861379983515449e-06,
+      "loss": 0.7099578380584717,
+      "step": 669
+    },
+    {
+      "epoch": 0.6567822570763386,
+      "grad_norm": 0.7350718379020691,
+      "learning_rate": 7.854870581754967e-06,
+      "loss": 0.5960590839385986,
+      "step": 670
+    },
+    {
+      "epoch": 0.6577625291018258,
+      "grad_norm": 1.4135644435882568,
+      "learning_rate": 7.848353992861195e-06,
+      "loss": 0.6619967818260193,
+      "step": 671
+    },
+    {
+      "epoch": 0.6587428011273129,
+      "grad_norm": 0.6925951242446899,
+      "learning_rate": 7.841830233239638e-06,
+      "loss": 0.5737698674201965,
+      "step": 672
+    },
+    {
+      "epoch": 0.6597230731527999,
+      "grad_norm": 1.133914589881897,
+      "learning_rate": 7.835299319313854e-06,
+      "loss": 0.5855647921562195,
+      "step": 673
+    },
+    {
+      "epoch": 0.660703345178287,
+      "grad_norm": 0.5980855822563171,
+      "learning_rate": 7.82876126752541e-06,
+      "loss": 0.5699212551116943,
+      "step": 674
+    },
+    {
+      "epoch": 0.661683617203774,
+      "grad_norm": 0.5689190030097961,
+      "learning_rate": 7.822216094333847e-06,
+      "loss": 0.5055632591247559,
+      "step": 675
+    },
+    {
+      "epoch": 0.6626638892292611,
+      "grad_norm": 0.6792899966239929,
+      "learning_rate": 7.815663816216626e-06,
+      "loss": 0.6879778504371643,
+      "step": 676
+    },
+    {
+      "epoch": 0.6636441612547482,
+      "grad_norm": 0.803128182888031,
+      "learning_rate": 7.8091044496691e-06,
+      "loss": 0.6372441053390503,
+      "step": 677
+    },
+    {
+      "epoch": 0.6646244332802352,
+      "grad_norm": 0.5344311594963074,
+      "learning_rate": 7.80253801120447e-06,
+      "loss": 0.6114630699157715,
+      "step": 678
+    },
+    {
+      "epoch": 0.6656047053057224,
+      "grad_norm": 0.6020611524581909,
+      "learning_rate": 7.795964517353734e-06,
+      "loss": 0.5540463328361511,
+      "step": 679
+    },
+    {
+      "epoch": 0.6665849773312094,
+      "grad_norm": 0.6257390379905701,
+      "learning_rate": 7.789383984665658e-06,
+      "loss": 0.5714541673660278,
+      "step": 680
+    },
+    {
+      "epoch": 0.6675652493566965,
+      "grad_norm": 0.8412136435508728,
+      "learning_rate": 7.782796429706721e-06,
+      "loss": 0.5719906091690063,
+      "step": 681
+    },
+    {
+      "epoch": 0.6685455213821836,
+      "grad_norm": 0.9053896069526672,
+      "learning_rate": 7.776201869061086e-06,
+      "loss": 0.6683034896850586,
+      "step": 682
+    },
+    {
+      "epoch": 0.6695257934076706,
+      "grad_norm": 0.4834095537662506,
+      "learning_rate": 7.769600319330553e-06,
+      "loss": 0.5566085577011108,
+      "step": 683
+    },
+    {
+      "epoch": 0.6705060654331577,
+      "grad_norm": 0.7962729930877686,
+      "learning_rate": 7.762991797134513e-06,
+      "loss": 0.6366957426071167,
+      "step": 684
+    },
+    {
+      "epoch": 0.6714863374586447,
+      "grad_norm": 0.5554247498512268,
+      "learning_rate": 7.756376319109917e-06,
+      "loss": 0.5735347270965576,
+      "step": 685
+    },
+    {
+      "epoch": 0.6724666094841318,
+      "grad_norm": 0.6228365302085876,
+      "learning_rate": 7.74975390191122e-06,
+      "loss": 0.6469916105270386,
+      "step": 686
+    },
+    {
+      "epoch": 0.673446881509619,
+      "grad_norm": 0.907863199710846,
+      "learning_rate": 7.743124562210351e-06,
+      "loss": 0.7663567662239075,
+      "step": 687
+    },
+    {
+      "epoch": 0.674427153535106,
+      "grad_norm": 0.9878220558166504,
+      "learning_rate": 7.736488316696663e-06,
+      "loss": 0.4941222369670868,
+      "step": 688
+    },
+    {
+      "epoch": 0.6754074255605931,
+      "grad_norm": 1.0419323444366455,
+      "learning_rate": 7.729845182076896e-06,
+      "loss": 0.7166256904602051,
+      "step": 689
+    },
+    {
+      "epoch": 0.6763876975860801,
+      "grad_norm": 0.9776319861412048,
+      "learning_rate": 7.723195175075136e-06,
+      "loss": 0.6621251106262207,
+      "step": 690
+    },
+    {
+      "epoch": 0.6773679696115672,
+      "grad_norm": 1.2711231708526611,
+      "learning_rate": 7.716538312432767e-06,
+      "loss": 0.6442791223526001,
+      "step": 691
+    },
+    {
+      "epoch": 0.6783482416370543,
+      "grad_norm": 1.2766711711883545,
+      "learning_rate": 7.70987461090843e-06,
+      "loss": 0.6728129982948303,
+      "step": 692
+    },
+    {
+      "epoch": 0.6793285136625413,
+      "grad_norm": 0.7948386073112488,
+      "learning_rate": 7.703204087277989e-06,
+      "loss": 0.6724798679351807,
+      "step": 693
+    },
+    {
+      "epoch": 0.6803087856880284,
+      "grad_norm": 1.0042457580566406,
+      "learning_rate": 7.696526758334477e-06,
+      "loss": 0.6475642919540405,
+      "step": 694
+    },
+    {
+      "epoch": 0.6812890577135156,
+      "grad_norm": 0.6241769790649414,
+      "learning_rate": 7.689842640888063e-06,
+      "loss": 0.497187077999115,
+      "step": 695
+    },
+    {
+      "epoch": 0.6822693297390026,
+      "grad_norm": 0.5805015563964844,
+      "learning_rate": 7.683151751766005e-06,
+      "loss": 0.5927690267562866,
+      "step": 696
+    },
+    {
+      "epoch": 0.6832496017644897,
+      "grad_norm": 1.0481382608413696,
+      "learning_rate": 7.676454107812608e-06,
+      "loss": 0.8189921975135803,
+      "step": 697
+    },
+    {
+      "epoch": 0.6842298737899767,
+      "grad_norm": 0.95531165599823,
+      "learning_rate": 7.669749725889182e-06,
+      "loss": 0.6346918344497681,
+      "step": 698
+    },
+    {
+      "epoch": 0.6852101458154638,
+      "grad_norm": 0.787591814994812,
+      "learning_rate": 7.663038622873999e-06,
+      "loss": 0.6715888381004333,
+      "step": 699
+    },
+    {
+      "epoch": 0.6861904178409509,
+      "grad_norm": 0.8717363476753235,
+      "learning_rate": 7.656320815662257e-06,
+      "loss": 0.7057673931121826,
+      "step": 700
+    },
+    {
+      "epoch": 0.6871706898664379,
+      "grad_norm": 0.6337246894836426,
+      "learning_rate": 7.649596321166024e-06,
+      "loss": 0.6414722800254822,
+      "step": 701
+    },
+    {
+      "epoch": 0.688150961891925,
+      "grad_norm": 0.7646953463554382,
+      "learning_rate": 7.64286515631421e-06,
+      "loss": 0.555882453918457,
+      "step": 702
+    },
+    {
+      "epoch": 0.689131233917412,
+      "grad_norm": 1.1629198789596558,
+      "learning_rate": 7.636127338052513e-06,
+      "loss": 0.6782732605934143,
+      "step": 703
+    },
+    {
+      "epoch": 0.6901115059428992,
+      "grad_norm": 0.7114580273628235,
+      "learning_rate": 7.629382883343381e-06,
+      "loss": 0.6305976510047913,
+      "step": 704
+    },
+    {
+      "epoch": 0.6910917779683863,
+      "grad_norm": 0.5025056600570679,
+      "learning_rate": 7.622631809165972e-06,
+      "loss": 0.451313853263855,
+      "step": 705
+    },
+    {
+      "epoch": 0.6920720499938733,
+      "grad_norm": 0.8512868285179138,
+      "learning_rate": 7.615874132516108e-06,
+      "loss": 0.44753578305244446,
+      "step": 706
+    },
+    {
+      "epoch": 0.6930523220193604,
+      "grad_norm": 1.3802733421325684,
+      "learning_rate": 7.60910987040623e-06,
+      "loss": 0.5664129853248596,
+      "step": 707
+    },
+    {
+      "epoch": 0.6940325940448474,
+      "grad_norm": 0.8319457769393921,
+      "learning_rate": 7.602339039865362e-06,
+      "loss": 0.5700501799583435,
+      "step": 708
+    },
+    {
+      "epoch": 0.6950128660703345,
+      "grad_norm": 1.5555126667022705,
+      "learning_rate": 7.595561657939061e-06,
+      "loss": 0.6250100135803223,
+      "step": 709
+    },
+    {
+      "epoch": 0.6959931380958216,
+      "grad_norm": 0.9013434648513794,
+      "learning_rate": 7.588777741689376e-06,
+      "loss": 0.689163863658905,
+      "step": 710
+    },
+    {
+      "epoch": 0.6969734101213086,
+      "grad_norm": 0.6376290321350098,
+      "learning_rate": 7.5819873081948105e-06,
+      "loss": 0.6546298265457153,
+      "step": 711
+    },
+    {
+      "epoch": 0.6979536821467958,
+      "grad_norm": 2.4558827877044678,
+      "learning_rate": 7.575190374550272e-06,
+      "loss": 0.555228054523468,
+      "step": 712
+    },
+    {
+      "epoch": 0.6989339541722828,
+      "grad_norm": 0.7567102313041687,
+      "learning_rate": 7.568386957867033e-06,
+      "loss": 0.6386421918869019,
+      "step": 713
+    },
+    {
+      "epoch": 0.6999142261977699,
+      "grad_norm": 0.7928363084793091,
+      "learning_rate": 7.561577075272686e-06,
+      "loss": 0.677485466003418,
+      "step": 714
+    },
+    {
+      "epoch": 0.700894498223257,
+      "grad_norm": 0.8352997899055481,
+      "learning_rate": 7.554760743911104e-06,
+      "loss": 0.5606400966644287,
+      "step": 715
+    },
+    {
+      "epoch": 0.701874770248744,
+      "grad_norm": 1.2804988622665405,
+      "learning_rate": 7.5479379809423905e-06,
+      "loss": 0.6543707847595215,
+      "step": 716
+    },
+    {
+      "epoch": 0.7028550422742311,
+      "grad_norm": 0.9014558792114258,
+      "learning_rate": 7.541108803542846e-06,
+      "loss": 0.6200656890869141,
+      "step": 717
+    },
+    {
+      "epoch": 0.7038353142997181,
+      "grad_norm": 0.9673658013343811,
+      "learning_rate": 7.534273228904916e-06,
+      "loss": 0.5967596769332886,
+      "step": 718
+    },
+    {
+      "epoch": 0.7048155863252052,
+      "grad_norm": 0.47720956802368164,
+      "learning_rate": 7.527431274237149e-06,
+      "loss": 0.48230111598968506,
+      "step": 719
+    },
+    {
+      "epoch": 0.7057958583506924,
+      "grad_norm": 0.5929463505744934,
+      "learning_rate": 7.52058295676416e-06,
+      "loss": 0.6315179467201233,
+      "step": 720
+    },
+    {
+      "epoch": 0.7067761303761794,
+      "grad_norm": 0.9534806609153748,
+      "learning_rate": 7.5137282937265796e-06,
+      "loss": 0.6589516401290894,
+      "step": 721
+    },
+    {
+      "epoch": 0.7077564024016665,
+      "grad_norm": 3.043043375015259,
+      "learning_rate": 7.506867302381016e-06,
+      "loss": 0.5525612235069275,
+      "step": 722
+    },
+    {
+      "epoch": 0.7087366744271535,
+      "grad_norm": 0.9205632209777832,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.5240005254745483,
+      "step": 723
+    },
+    {
+      "epoch": 0.7097169464526406,
+      "grad_norm": 1.0340361595153809,
+      "learning_rate": 7.493126403871964e-06,
+      "loss": 0.5224882364273071,
+      "step": 724
+    },
+    {
+      "epoch": 0.7106972184781277,
+      "grad_norm": 1.0017213821411133,
+      "learning_rate": 7.486246531301178e-06,
+      "loss": 0.717761754989624,
+      "step": 725
+    },
+    {
+      "epoch": 0.7116774905036147,
+      "grad_norm": 0.9500141739845276,
+      "learning_rate": 7.479360399607707e-06,
+      "loss": 0.5689233541488647,
+      "step": 726
+    },
+    {
+      "epoch": 0.7126577625291018,
+      "grad_norm": 3.168675184249878,
+      "learning_rate": 7.472468026127385e-06,
+      "loss": 0.6222257018089294,
+      "step": 727
+    },
+    {
+      "epoch": 0.7136380345545889,
+      "grad_norm": 0.6307007074356079,
+      "learning_rate": 7.465569428211752e-06,
+      "loss": 0.6851769089698792,
+      "step": 728
+    },
+    {
+      "epoch": 0.714618306580076,
+      "grad_norm": 0.9678178429603577,
+      "learning_rate": 7.45866462322802e-06,
+      "loss": 0.5859109163284302,
+      "step": 729
+    },
+    {
+      "epoch": 0.7155985786055631,
+      "grad_norm": 6.1496381759643555,
+      "learning_rate": 7.451753628559027e-06,
+      "loss": 0.5398833751678467,
+      "step": 730
+    },
+    {
+      "epoch": 0.7165788506310501,
+      "grad_norm": 1.2634036540985107,
+      "learning_rate": 7.444836461603195e-06,
+      "loss": 0.6970493793487549,
+      "step": 731
+    },
+    {
+      "epoch": 0.7175591226565372,
+      "grad_norm": 0.8196023106575012,
+      "learning_rate": 7.437913139774482e-06,
+      "loss": 0.5129099488258362,
+      "step": 732
+    },
+    {
+      "epoch": 0.7185393946820242,
+      "grad_norm": 0.6801417469978333,
+      "learning_rate": 7.430983680502344e-06,
+      "loss": 0.5269551277160645,
+      "step": 733
+    },
+    {
+      "epoch": 0.7195196667075113,
+      "grad_norm": 0.6960332989692688,
+      "learning_rate": 7.424048101231687e-06,
+      "loss": 0.5466389656066895,
+      "step": 734
+    },
+    {
+      "epoch": 0.7204999387329984,
+      "grad_norm": 3.183863878250122,
+      "learning_rate": 7.4171064194228196e-06,
+      "loss": 0.603979229927063,
+      "step": 735
+    },
+    {
+      "epoch": 0.7214802107584855,
+      "grad_norm": 0.6343901753425598,
+      "learning_rate": 7.410158652551421e-06,
+      "loss": 0.6457838416099548,
+      "step": 736
+    },
+    {
+      "epoch": 0.7224604827839726,
+      "grad_norm": 0.7226002812385559,
+      "learning_rate": 7.403204818108487e-06,
+      "loss": 0.7031099796295166,
+      "step": 737
+    },
+    {
+      "epoch": 0.7234407548094596,
+      "grad_norm": 1.1366685628890991,
+      "learning_rate": 7.396244933600285e-06,
+      "loss": 0.6742706298828125,
+      "step": 738
+    },
+    {
+      "epoch": 0.7244210268349467,
+      "grad_norm": 2.4916179180145264,
+      "learning_rate": 7.3892790165483164e-06,
+      "loss": 0.6171988248825073,
+      "step": 739
+    },
+    {
+      "epoch": 0.7254012988604338,
+      "grad_norm": 0.55182945728302,
+      "learning_rate": 7.382307084489269e-06,
+      "loss": 0.6048471331596375,
+      "step": 740
+    },
+    {
+      "epoch": 0.7263815708859208,
+      "grad_norm": 0.4613666534423828,
+      "learning_rate": 7.3753291549749764e-06,
+      "loss": 0.6115667819976807,
+      "step": 741
+    },
+    {
+      "epoch": 0.7273618429114079,
+      "grad_norm": 0.7300680875778198,
+      "learning_rate": 7.368345245572362e-06,
+      "loss": 0.49636924266815186,
+      "step": 742
+    },
+    {
+      "epoch": 0.7283421149368949,
+      "grad_norm": 0.4753544330596924,
+      "learning_rate": 7.361355373863415e-06,
+      "loss": 0.5302106738090515,
+      "step": 743
+    },
+    {
+      "epoch": 0.729322386962382,
+      "grad_norm": 0.5730991363525391,
+      "learning_rate": 7.354359557445125e-06,
+      "loss": 0.6674627065658569,
+      "step": 744
+    },
+    {
+      "epoch": 0.7303026589878692,
+      "grad_norm": 0.6844426989555359,
+      "learning_rate": 7.347357813929455e-06,
+      "loss": 0.7921134233474731,
+      "step": 745
+    },
+    {
+      "epoch": 0.7312829310133562,
+      "grad_norm": 0.6181908249855042,
+      "learning_rate": 7.3403501609432835e-06,
+      "loss": 0.6121218800544739,
+      "step": 746
+    },
+    {
+      "epoch": 0.7322632030388433,
+      "grad_norm": 0.6758370995521545,
+      "learning_rate": 7.333336616128369e-06,
+      "loss": 0.647600531578064,
+      "step": 747
+    },
+    {
+      "epoch": 0.7332434750643303,
+      "grad_norm": 0.7577820420265198,
+      "learning_rate": 7.326317197141304e-06,
+      "loss": 0.7386564612388611,
+      "step": 748
+    },
+    {
+      "epoch": 0.7342237470898174,
+      "grad_norm": 0.5706998705863953,
+      "learning_rate": 7.319291921653464e-06,
+      "loss": 0.6458116173744202,
+      "step": 749
+    },
+    {
+      "epoch": 0.7352040191153045,
+      "grad_norm": 5.349646091461182,
+      "learning_rate": 7.312260807350975e-06,
+      "loss": 0.6188918352127075,
+      "step": 750
+    },
+    {
+      "epoch": 0.7361842911407915,
+      "grad_norm": 1.882954478263855,
+      "learning_rate": 7.305223871934657e-06,
+      "loss": 0.6038079857826233,
+      "step": 751
+    },
+    {
+      "epoch": 0.7371645631662787,
+      "grad_norm": 0.6516304612159729,
+      "learning_rate": 7.298181133119987e-06,
+      "loss": 0.5557756423950195,
+      "step": 752
+    },
+    {
+      "epoch": 0.7381448351917658,
+      "grad_norm": 0.6706615686416626,
+      "learning_rate": 7.291132608637053e-06,
+      "loss": 0.6859948039054871,
+      "step": 753
+    },
+    {
+      "epoch": 0.7391251072172528,
+      "grad_norm": 0.6638997793197632,
+      "learning_rate": 7.284078316230504e-06,
+      "loss": 0.5381072759628296,
+      "step": 754
+    },
+    {
+      "epoch": 0.7401053792427399,
+      "grad_norm": 0.5846869945526123,
+      "learning_rate": 7.2770182736595164e-06,
+      "loss": 0.5414190888404846,
+      "step": 755
+    },
+    {
+      "epoch": 0.7410856512682269,
+      "grad_norm": 0.4703490734100342,
+      "learning_rate": 7.269952498697734e-06,
+      "loss": 0.5949432849884033,
+      "step": 756
+    },
+    {
+      "epoch": 0.742065923293714,
+      "grad_norm": 0.6069402694702148,
+      "learning_rate": 7.262881009133242e-06,
+      "loss": 0.7321566939353943,
+      "step": 757
+    },
+    {
+      "epoch": 0.7430461953192011,
+      "grad_norm": 0.704056441783905,
+      "learning_rate": 7.255803822768504e-06,
+      "loss": 0.5396850109100342,
+      "step": 758
+    },
+    {
+      "epoch": 0.7440264673446881,
+      "grad_norm": 0.7957848310470581,
+      "learning_rate": 7.24872095742033e-06,
+      "loss": 0.6128697395324707,
+      "step": 759
+    },
+    {
+      "epoch": 0.7450067393701753,
+      "grad_norm": 0.7028953433036804,
+      "learning_rate": 7.241632430919823e-06,
+      "loss": 0.6348029375076294,
+      "step": 760
+    },
+    {
+      "epoch": 0.7459870113956623,
+      "grad_norm": 0.46351173520088196,
+      "learning_rate": 7.234538261112342e-06,
+      "loss": 0.4864131808280945,
+      "step": 761
+    },
+    {
+      "epoch": 0.7469672834211494,
+      "grad_norm": 0.9782418608665466,
+      "learning_rate": 7.227438465857448e-06,
+      "loss": 0.7021412253379822,
+      "step": 762
+    },
+    {
+      "epoch": 0.7479475554466365,
+      "grad_norm": 0.6365218162536621,
+      "learning_rate": 7.2203330630288714e-06,
+      "loss": 0.4935818016529083,
+      "step": 763
+    },
+    {
+      "epoch": 0.7489278274721235,
+      "grad_norm": 0.847671627998352,
+      "learning_rate": 7.213222070514453e-06,
+      "loss": 0.8174105882644653,
+      "step": 764
+    },
+    {
+      "epoch": 0.7499080994976106,
+      "grad_norm": 0.606162428855896,
+      "learning_rate": 7.206105506216107e-06,
+      "loss": 0.5682967305183411,
+      "step": 765
+    },
+    {
+      "epoch": 0.7508883715230976,
+      "grad_norm": 0.7045398950576782,
+      "learning_rate": 7.198983388049779e-06,
+      "loss": 0.6105291843414307,
+      "step": 766
+    },
+    {
+      "epoch": 0.7518686435485847,
+      "grad_norm": 0.8413256406784058,
+      "learning_rate": 7.191855733945388e-06,
+      "loss": 0.5956174731254578,
+      "step": 767
+    },
+    {
+      "epoch": 0.7528489155740719,
+      "grad_norm": 20.123249053955078,
+      "learning_rate": 7.1847225618467975e-06,
+      "loss": 0.47062593698501587,
+      "step": 768
+    },
+    {
+      "epoch": 0.7538291875995589,
+      "grad_norm": 0.7928285002708435,
+      "learning_rate": 7.177583889711763e-06,
+      "loss": 0.6014308333396912,
+      "step": 769
+    },
+    {
+      "epoch": 0.754809459625046,
+      "grad_norm": 4.023245811462402,
+      "learning_rate": 7.170439735511876e-06,
+      "loss": 0.6861827373504639,
+      "step": 770
+    },
+    {
+      "epoch": 0.755789731650533,
+      "grad_norm": 1.9669843912124634,
+      "learning_rate": 7.163290117232542e-06,
+      "loss": 0.6137866377830505,
+      "step": 771
+    },
+    {
+      "epoch": 0.7567700036760201,
+      "grad_norm": 0.6258137226104736,
+      "learning_rate": 7.1561350528729145e-06,
+      "loss": 0.6574512124061584,
+      "step": 772
+    },
+    {
+      "epoch": 0.7577502757015072,
+      "grad_norm": 0.7114458680152893,
+      "learning_rate": 7.148974560445859e-06,
+      "loss": 0.48910653591156006,
+      "step": 773
+    },
+    {
+      "epoch": 0.7587305477269942,
+      "grad_norm": 1.8172527551651,
+      "learning_rate": 7.1418086579779075e-06,
+      "loss": 0.5452988743782043,
+      "step": 774
+    },
+    {
+      "epoch": 0.7597108197524813,
+      "grad_norm": 0.7762359976768494,
+      "learning_rate": 7.1346373635092095e-06,
+      "loss": 0.6227539777755737,
+      "step": 775
+    },
+    {
+      "epoch": 0.7606910917779683,
+      "grad_norm": 0.745168149471283,
+      "learning_rate": 7.127460695093493e-06,
+      "loss": 0.6921566128730774,
+      "step": 776
+    },
+    {
+      "epoch": 0.7616713638034555,
+      "grad_norm": 0.8893136382102966,
+      "learning_rate": 7.12027867079801e-06,
+      "loss": 0.6578104496002197,
+      "step": 777
+    },
+    {
+      "epoch": 0.7626516358289426,
+      "grad_norm": 0.5235515832901001,
+      "learning_rate": 7.113091308703498e-06,
+      "loss": 0.6317405104637146,
+      "step": 778
+    },
+    {
+      "epoch": 0.7636319078544296,
+      "grad_norm": 2.2910876274108887,
+      "learning_rate": 7.105898626904134e-06,
+      "loss": 0.6121262311935425,
+      "step": 779
+    },
+    {
+      "epoch": 0.7646121798799167,
+      "grad_norm": 1.0871317386627197,
+      "learning_rate": 7.098700643507485e-06,
+      "loss": 0.6684179306030273,
+      "step": 780
+    },
+    {
+      "epoch": 0.7655924519054037,
+      "grad_norm": 0.5349941253662109,
+      "learning_rate": 7.0914973766344645e-06,
+      "loss": 0.4995436668395996,
+      "step": 781
+    },
+    {
+      "epoch": 0.7665727239308908,
+      "grad_norm": 0.9247681498527527,
+      "learning_rate": 7.0842888444192905e-06,
+      "loss": 0.7629621028900146,
+      "step": 782
+    },
+    {
+      "epoch": 0.7675529959563779,
+      "grad_norm": 0.8464233875274658,
+      "learning_rate": 7.0770750650094335e-06,
+      "loss": 0.681257426738739,
+      "step": 783
+    },
+    {
+      "epoch": 0.7685332679818649,
+      "grad_norm": 0.8439611792564392,
+      "learning_rate": 7.069856056565573e-06,
+      "loss": 0.5528218746185303,
+      "step": 784
+    },
+    {
+      "epoch": 0.7695135400073521,
+      "grad_norm": 1.050178050994873,
+      "learning_rate": 7.062631837261556e-06,
+      "loss": 0.5813249349594116,
+      "step": 785
+    },
+    {
+      "epoch": 0.7704938120328391,
+      "grad_norm": 0.7861621975898743,
+      "learning_rate": 7.055402425284346e-06,
+      "loss": 0.5101428031921387,
+      "step": 786
+    },
+    {
+      "epoch": 0.7714740840583262,
+      "grad_norm": 1.0121400356292725,
+      "learning_rate": 7.048167838833977e-06,
+      "loss": 0.6809993982315063,
+      "step": 787
+    },
+    {
+      "epoch": 0.7724543560838133,
+      "grad_norm": 0.9147869348526001,
+      "learning_rate": 7.040928096123516e-06,
+      "loss": 0.6002492904663086,
+      "step": 788
+    },
+    {
+      "epoch": 0.7734346281093003,
+      "grad_norm": 1.0097702741622925,
+      "learning_rate": 7.033683215379002e-06,
+      "loss": 0.615148663520813,
+      "step": 789
+    },
+    {
+      "epoch": 0.7744149001347874,
+      "grad_norm": 0.5456814765930176,
+      "learning_rate": 7.026433214839416e-06,
+      "loss": 0.5866764783859253,
+      "step": 790
+    },
+    {
+      "epoch": 0.7753951721602744,
+      "grad_norm": 0.5823632478713989,
+      "learning_rate": 7.019178112756625e-06,
+      "loss": 0.5176118016242981,
+      "step": 791
+    },
+    {
+      "epoch": 0.7763754441857615,
+      "grad_norm": 0.9243921637535095,
+      "learning_rate": 7.011917927395341e-06,
+      "loss": 0.5832178592681885,
+      "step": 792
+    },
+    {
+      "epoch": 0.7773557162112487,
+      "grad_norm": 0.6878734827041626,
+      "learning_rate": 7.004652677033069e-06,
+      "loss": 0.5880787968635559,
+      "step": 793
+    },
+    {
+      "epoch": 0.7783359882367357,
+      "grad_norm": 0.6831730008125305,
+      "learning_rate": 6.997382379960072e-06,
+      "loss": 0.6205973625183105,
+      "step": 794
+    },
+    {
+      "epoch": 0.7793162602622228,
+      "grad_norm": 1.2210882902145386,
+      "learning_rate": 6.990107054479313e-06,
+      "loss": 0.5620274543762207,
+      "step": 795
+    },
+    {
+      "epoch": 0.7802965322877098,
+      "grad_norm": 0.9477983713150024,
+      "learning_rate": 6.9828267189064125e-06,
+      "loss": 0.5950782895088196,
+      "step": 796
+    },
+    {
+      "epoch": 0.7812768043131969,
+      "grad_norm": 1.2986098527908325,
+      "learning_rate": 6.9755413915696105e-06,
+      "loss": 0.6701383590698242,
+      "step": 797
+    },
+    {
+      "epoch": 0.782257076338684,
+      "grad_norm": 0.7791178226470947,
+      "learning_rate": 6.968251090809708e-06,
+      "loss": 0.4881432056427002,
+      "step": 798
+    },
+    {
+      "epoch": 0.783237348364171,
+      "grad_norm": 1.7391661405563354,
+      "learning_rate": 6.960955834980028e-06,
+      "loss": 0.554238498210907,
+      "step": 799
+    },
+    {
+      "epoch": 0.7842176203896581,
+      "grad_norm": 0.5904809832572937,
+      "learning_rate": 6.953655642446368e-06,
+      "loss": 0.6785469651222229,
+      "step": 800
+    },
+    {
+      "epoch": 0.7851978924151451,
+      "grad_norm": 0.5747798085212708,
+      "learning_rate": 6.946350531586959e-06,
+      "loss": 0.5274198651313782,
+      "step": 801
+    },
+    {
+      "epoch": 0.7861781644406323,
+      "grad_norm": 0.6243445873260498,
+      "learning_rate": 6.9390405207924014e-06,
+      "loss": 0.6013908386230469,
+      "step": 802
+    },
+    {
+      "epoch": 0.7871584364661194,
+      "grad_norm": 0.901065468788147,
+      "learning_rate": 6.931725628465643e-06,
+      "loss": 0.5571209788322449,
+      "step": 803
+    },
+    {
+      "epoch": 0.7881387084916064,
+      "grad_norm": 0.8468716144561768,
+      "learning_rate": 6.9244058730219176e-06,
+      "loss": 0.679326057434082,
+      "step": 804
+    },
+    {
+      "epoch": 0.7891189805170935,
+      "grad_norm": 0.5513537526130676,
+      "learning_rate": 6.917081272888697e-06,
+      "loss": 0.4129813313484192,
+      "step": 805
+    },
+    {
+      "epoch": 0.7900992525425806,
+      "grad_norm": 1.3474310636520386,
+      "learning_rate": 6.909751846505657e-06,
+      "loss": 0.524448812007904,
+      "step": 806
+    },
+    {
+      "epoch": 0.7910795245680676,
+      "grad_norm": 0.6401530504226685,
+      "learning_rate": 6.902417612324615e-06,
+      "loss": 0.6002922058105469,
+      "step": 807
+    },
+    {
+      "epoch": 0.7920597965935547,
+      "grad_norm": 1.391119122505188,
+      "learning_rate": 6.895078588809503e-06,
+      "loss": 0.5608537197113037,
+      "step": 808
+    },
+    {
+      "epoch": 0.7930400686190417,
+      "grad_norm": 0.6057695150375366,
+      "learning_rate": 6.887734794436301e-06,
+      "loss": 0.4683951139450073,
+      "step": 809
+    },
+    {
+      "epoch": 0.7940203406445289,
+      "grad_norm": 0.6945317387580872,
+      "learning_rate": 6.880386247692999e-06,
+      "loss": 0.46992582082748413,
+      "step": 810
+    },
+    {
+      "epoch": 0.795000612670016,
+      "grad_norm": 0.7487401366233826,
+      "learning_rate": 6.873032967079562e-06,
+      "loss": 0.5497885942459106,
+      "step": 811
+    },
+    {
+      "epoch": 0.795980884695503,
+      "grad_norm": 0.6409137845039368,
+      "learning_rate": 6.865674971107859e-06,
+      "loss": 0.649165153503418,
+      "step": 812
+    },
+    {
+      "epoch": 0.7969611567209901,
+      "grad_norm": 1.9870905876159668,
+      "learning_rate": 6.858312278301638e-06,
+      "loss": 0.6488982439041138,
+      "step": 813
+    },
+    {
+      "epoch": 0.7979414287464771,
+      "grad_norm": 0.5581271052360535,
+      "learning_rate": 6.850944907196472e-06,
+      "loss": 0.5292209386825562,
+      "step": 814
+    },
+    {
+      "epoch": 0.7989217007719642,
+      "grad_norm": 0.6206526756286621,
+      "learning_rate": 6.8435728763397045e-06,
+      "loss": 0.5159962773323059,
+      "step": 815
+    },
+    {
+      "epoch": 0.7999019727974513,
+      "grad_norm": 0.9332653880119324,
+      "learning_rate": 6.836196204290417e-06,
+      "loss": 0.6101655960083008,
+      "step": 816
+    },
+    {
+      "epoch": 0.8008822448229383,
+      "grad_norm": 0.696185290813446,
+      "learning_rate": 6.828814909619374e-06,
+      "loss": 0.6255779266357422,
+      "step": 817
+    },
+    {
+      "epoch": 0.8018625168484255,
+      "grad_norm": 0.599087119102478,
+      "learning_rate": 6.821429010908972e-06,
+      "loss": 0.5408052206039429,
+      "step": 818
+    },
+    {
+      "epoch": 0.8028427888739125,
+      "grad_norm": 0.9013589024543762,
+      "learning_rate": 6.814038526753205e-06,
+      "loss": 0.5954138040542603,
+      "step": 819
+    },
+    {
+      "epoch": 0.8038230608993996,
+      "grad_norm": 0.8954412341117859,
+      "learning_rate": 6.806643475757608e-06,
+      "loss": 0.5609475374221802,
+      "step": 820
+    },
+    {
+      "epoch": 0.8048033329248867,
+      "grad_norm": 0.9954658150672913,
+      "learning_rate": 6.799243876539213e-06,
+      "loss": 0.5928156971931458,
+      "step": 821
+    },
+    {
+      "epoch": 0.8057836049503737,
+      "grad_norm": 0.8288171887397766,
+      "learning_rate": 6.7918397477265e-06,
+      "loss": 0.6964554786682129,
+      "step": 822
+    },
+    {
+      "epoch": 0.8067638769758608,
+      "grad_norm": 1.528401494026184,
+      "learning_rate": 6.78443110795936e-06,
+      "loss": 0.607907772064209,
+      "step": 823
+    },
+    {
+      "epoch": 0.8077441490013478,
+      "grad_norm": 0.47465401887893677,
+      "learning_rate": 6.77701797588903e-06,
+      "loss": 0.542430579662323,
+      "step": 824
+    },
+    {
+      "epoch": 0.808724421026835,
+      "grad_norm": 0.6082414984703064,
+      "learning_rate": 6.76960037017806e-06,
+      "loss": 0.5040891766548157,
+      "step": 825
+    },
+    {
+      "epoch": 0.8097046930523221,
+      "grad_norm": 0.9670936465263367,
+      "learning_rate": 6.762178309500267e-06,
+      "loss": 0.7020210027694702,
+      "step": 826
+    },
+    {
+      "epoch": 0.8106849650778091,
+      "grad_norm": 0.5746657848358154,
+      "learning_rate": 6.75475181254068e-06,
+      "loss": 0.6425360441207886,
+      "step": 827
+    },
+    {
+      "epoch": 0.8116652371032962,
+      "grad_norm": 0.472930908203125,
+      "learning_rate": 6.747320897995493e-06,
+      "loss": 0.6312180757522583,
+      "step": 828
+    },
+    {
+      "epoch": 0.8126455091287832,
+      "grad_norm": 0.5002622008323669,
+      "learning_rate": 6.739885584572026e-06,
+      "loss": 0.5433318614959717,
+      "step": 829
+    },
+    {
+      "epoch": 0.8136257811542703,
+      "grad_norm": 2.922562837600708,
+      "learning_rate": 6.732445890988675e-06,
+      "loss": 0.7362682819366455,
+      "step": 830
+    },
+    {
+      "epoch": 0.8146060531797574,
+      "grad_norm": 0.6053969264030457,
+      "learning_rate": 6.725001835974854e-06,
+      "loss": 0.5773118734359741,
+      "step": 831
+    },
+    {
+      "epoch": 0.8155863252052444,
+      "grad_norm": 1.244458556175232,
+      "learning_rate": 6.717553438270966e-06,
+      "loss": 0.5248075723648071,
+      "step": 832
+    },
+    {
+      "epoch": 0.8165665972307316,
+      "grad_norm": 4.066065311431885,
+      "learning_rate": 6.710100716628345e-06,
+      "loss": 0.6107986569404602,
+      "step": 833
+    },
+    {
+      "epoch": 0.8175468692562186,
+      "grad_norm": 13.074731826782227,
+      "learning_rate": 6.702643689809205e-06,
+      "loss": 0.4682711958885193,
+      "step": 834
+    },
+    {
+      "epoch": 0.8185271412817057,
+      "grad_norm": 0.9204456806182861,
+      "learning_rate": 6.695182376586603e-06,
+      "loss": 0.5581796765327454,
+      "step": 835
+    },
+    {
+      "epoch": 0.8195074133071928,
+      "grad_norm": 2.913454294204712,
+      "learning_rate": 6.68771679574439e-06,
+      "loss": 0.6707707643508911,
+      "step": 836
+    },
+    {
+      "epoch": 0.8204876853326798,
+      "grad_norm": 1.1219816207885742,
+      "learning_rate": 6.680246966077151e-06,
+      "loss": 0.5806658267974854,
+      "step": 837
+    },
+    {
+      "epoch": 0.8214679573581669,
+      "grad_norm": 0.6837668418884277,
+      "learning_rate": 6.672772906390177e-06,
+      "loss": 0.5435550808906555,
+      "step": 838
+    },
+    {
+      "epoch": 0.8224482293836539,
+      "grad_norm": 1.678803563117981,
+      "learning_rate": 6.665294635499404e-06,
+      "loss": 0.48199865221977234,
+      "step": 839
+    },
+    {
+      "epoch": 0.823428501409141,
+      "grad_norm": 0.831813633441925,
+      "learning_rate": 6.65781217223137e-06,
+      "loss": 0.6230430006980896,
+      "step": 840
+    },
+    {
+      "epoch": 0.8244087734346282,
+      "grad_norm": 0.5748403072357178,
+      "learning_rate": 6.650325535423166e-06,
+      "loss": 0.7263023853302002,
+      "step": 841
+    },
+    {
+      "epoch": 0.8253890454601152,
+      "grad_norm": 1.0182629823684692,
+      "learning_rate": 6.642834743922393e-06,
+      "loss": 0.7895938754081726,
+      "step": 842
+    },
+    {
+      "epoch": 0.8263693174856023,
+      "grad_norm": 1.411647081375122,
+      "learning_rate": 6.635339816587109e-06,
+      "loss": 0.5766178965568542,
+      "step": 843
+    },
+    {
+      "epoch": 0.8273495895110893,
+      "grad_norm": 0.7775676250457764,
+      "learning_rate": 6.627840772285784e-06,
+      "loss": 0.5816385746002197,
+      "step": 844
+    },
+    {
+      "epoch": 0.8283298615365764,
+      "grad_norm": 0.9636262059211731,
+      "learning_rate": 6.6203376298972535e-06,
+      "loss": 0.6016203165054321,
+      "step": 845
+    },
+    {
+      "epoch": 0.8293101335620635,
+      "grad_norm": 3.9955432415008545,
+      "learning_rate": 6.612830408310671e-06,
+      "loss": 0.7095068097114563,
+      "step": 846
+    },
+    {
+      "epoch": 0.8302904055875505,
+      "grad_norm": 0.6053678393363953,
+      "learning_rate": 6.605319126425455e-06,
+      "loss": 0.6205288171768188,
+      "step": 847
+    },
+    {
+      "epoch": 0.8312706776130376,
+      "grad_norm": 0.6421292424201965,
+      "learning_rate": 6.5978038031512505e-06,
+      "loss": 0.47240081429481506,
+      "step": 848
+    },
+    {
+      "epoch": 0.8322509496385246,
+      "grad_norm": 2.9203879833221436,
+      "learning_rate": 6.590284457407876e-06,
+      "loss": 0.7266073226928711,
+      "step": 849
+    },
+    {
+      "epoch": 0.8332312216640118,
+      "grad_norm": 1.0975518226623535,
+      "learning_rate": 6.582761108125272e-06,
+      "loss": 0.5229194760322571,
+      "step": 850
+    },
+    {
+      "epoch": 0.8342114936894989,
+      "grad_norm": 0.49383655190467834,
+      "learning_rate": 6.5752337742434644e-06,
+      "loss": 0.4620560109615326,
+      "step": 851
+    },
+    {
+      "epoch": 0.8351917657149859,
+      "grad_norm": 0.6703282594680786,
+      "learning_rate": 6.567702474712507e-06,
+      "loss": 0.6520070433616638,
+      "step": 852
+    },
+    {
+      "epoch": 0.836172037740473,
+      "grad_norm": 0.5750573873519897,
+      "learning_rate": 6.560167228492436e-06,
+      "loss": 0.5389132499694824,
+      "step": 853
+    },
+    {
+      "epoch": 0.83715230976596,
+      "grad_norm": 0.5545781850814819,
+      "learning_rate": 6.552628054553225e-06,
+      "loss": 0.5768039226531982,
+      "step": 854
+    },
+    {
+      "epoch": 0.8381325817914471,
+      "grad_norm": 1.8055278062820435,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.6694058179855347,
+      "step": 855
+    },
+    {
+      "epoch": 0.8391128538169342,
+      "grad_norm": 0.6390528082847595,
+      "learning_rate": 6.537537999446673e-06,
+      "loss": 0.584903359413147,
+      "step": 856
+    },
+    {
+      "epoch": 0.8400931258424212,
+      "grad_norm": 0.9968768358230591,
+      "learning_rate": 6.529987156268527e-06,
+      "loss": 0.5555717945098877,
+      "step": 857
+    },
+    {
+      "epoch": 0.8410733978679084,
+      "grad_norm": 0.5672370195388794,
+      "learning_rate": 6.522432461349536e-06,
+      "loss": 0.5726813077926636,
+      "step": 858
+    },
+    {
+      "epoch": 0.8420536698933954,
+      "grad_norm": 0.662235677242279,
+      "learning_rate": 6.514873933708637e-06,
+      "loss": 0.7376776933670044,
+      "step": 859
+    },
+    {
+      "epoch": 0.8430339419188825,
+      "grad_norm": 0.752988874912262,
+      "learning_rate": 6.507311592374415e-06,
+      "loss": 0.5993643999099731,
+      "step": 860
+    },
+    {
+      "epoch": 0.8440142139443696,
+      "grad_norm": 0.8915222883224487,
+      "learning_rate": 6.499745456385054e-06,
+      "loss": 0.6790447235107422,
+      "step": 861
+    },
+    {
+      "epoch": 0.8449944859698566,
+      "grad_norm": 0.8085651993751526,
+      "learning_rate": 6.4921755447882926e-06,
+      "loss": 0.5463986396789551,
+      "step": 862
+    },
+    {
+      "epoch": 0.8459747579953437,
+      "grad_norm": 1.4450210332870483,
+      "learning_rate": 6.484601876641375e-06,
+      "loss": 0.6264548897743225,
+      "step": 863
+    },
+    {
+      "epoch": 0.8469550300208308,
+      "grad_norm": 0.671293318271637,
+      "learning_rate": 6.477024471011001e-06,
+      "loss": 0.5638301372528076,
+      "step": 864
+    },
+    {
+      "epoch": 0.8479353020463178,
+      "grad_norm": 0.5809031128883362,
+      "learning_rate": 6.469443346973281e-06,
+      "loss": 0.5892854928970337,
+      "step": 865
+    },
+    {
+      "epoch": 0.848915574071805,
+      "grad_norm": 0.7964317202568054,
+      "learning_rate": 6.461858523613684e-06,
+      "loss": 0.5828635096549988,
+      "step": 866
+    },
+    {
+      "epoch": 0.849895846097292,
+      "grad_norm": 1.2660839557647705,
+      "learning_rate": 6.454270020026996e-06,
+      "loss": 0.6912301778793335,
+      "step": 867
+    },
+    {
+      "epoch": 0.8508761181227791,
+      "grad_norm": 0.6901545524597168,
+      "learning_rate": 6.446677855317265e-06,
+      "loss": 0.526343584060669,
+      "step": 868
+    },
+    {
+      "epoch": 0.8518563901482662,
+      "grad_norm": 0.6633983254432678,
+      "learning_rate": 6.439082048597755e-06,
+      "loss": 0.5672087669372559,
+      "step": 869
+    },
+    {
+      "epoch": 0.8528366621737532,
+      "grad_norm": 0.5235927104949951,
+      "learning_rate": 6.431482618990902e-06,
+      "loss": 0.5292974710464478,
+      "step": 870
+    },
+    {
+      "epoch": 0.8538169341992403,
+      "grad_norm": 0.75315260887146,
+      "learning_rate": 6.423879585628262e-06,
+      "loss": 0.7237055897712708,
+      "step": 871
+    },
+    {
+      "epoch": 0.8547972062247273,
+      "grad_norm": 1.7529493570327759,
+      "learning_rate": 6.41627296765046e-06,
+      "loss": 0.7014877200126648,
+      "step": 872
+    },
+    {
+      "epoch": 0.8557774782502144,
+      "grad_norm": 0.7256580591201782,
+      "learning_rate": 6.408662784207149e-06,
+      "loss": 0.5885106325149536,
+      "step": 873
+    },
+    {
+      "epoch": 0.8567577502757016,
+      "grad_norm": 1.1877896785736084,
+      "learning_rate": 6.401049054456957e-06,
+      "loss": 0.7587547302246094,
+      "step": 874
+    },
+    {
+      "epoch": 0.8577380223011886,
+      "grad_norm": 0.765882670879364,
+      "learning_rate": 6.39343179756744e-06,
+      "loss": 0.507827877998352,
+      "step": 875
+    },
+    {
+      "epoch": 0.8587182943266757,
+      "grad_norm": 0.6027549505233765,
+      "learning_rate": 6.385811032715031e-06,
+      "loss": 0.5204265117645264,
+      "step": 876
+    },
+    {
+      "epoch": 0.8596985663521627,
+      "grad_norm": 1.0691670179367065,
+      "learning_rate": 6.378186779084996e-06,
+      "loss": 0.6082635521888733,
+      "step": 877
+    },
+    {
+      "epoch": 0.8606788383776498,
+      "grad_norm": 0.667834460735321,
+      "learning_rate": 6.370559055871389e-06,
+      "loss": 0.6502209901809692,
+      "step": 878
+    },
+    {
+      "epoch": 0.8616591104031369,
+      "grad_norm": 15.34140682220459,
+      "learning_rate": 6.362927882276991e-06,
+      "loss": 0.7983745336532593,
+      "step": 879
+    },
+    {
+      "epoch": 0.8626393824286239,
+      "grad_norm": 1.1628925800323486,
+      "learning_rate": 6.355293277513271e-06,
+      "loss": 0.5471218228340149,
+      "step": 880
+    },
+    {
+      "epoch": 0.863619654454111,
+      "grad_norm": 0.5984328985214233,
+      "learning_rate": 6.34765526080034e-06,
+      "loss": 0.49805784225463867,
+      "step": 881
+    },
+    {
+      "epoch": 0.864599926479598,
+      "grad_norm": 0.8155985474586487,
+      "learning_rate": 6.3400138513668955e-06,
+      "loss": 0.5052790641784668,
+      "step": 882
+    },
+    {
+      "epoch": 0.8655801985050852,
+      "grad_norm": 2.3906466960906982,
+      "learning_rate": 6.332369068450175e-06,
+      "loss": 0.617497444152832,
+      "step": 883
+    },
+    {
+      "epoch": 0.8665604705305723,
+      "grad_norm": 0.40641558170318604,
+      "learning_rate": 6.324720931295913e-06,
+      "loss": 0.42097628116607666,
+      "step": 884
+    },
+    {
+      "epoch": 0.8675407425560593,
+      "grad_norm": 0.7153171300888062,
+      "learning_rate": 6.317069459158284e-06,
+      "loss": 0.6523081064224243,
+      "step": 885
+    },
+    {
+      "epoch": 0.8685210145815464,
+      "grad_norm": 1.6543089151382446,
+      "learning_rate": 6.3094146712998595e-06,
+      "loss": 0.6268733739852905,
+      "step": 886
+    },
+    {
+      "epoch": 0.8695012866070334,
+      "grad_norm": 0.6385596394538879,
+      "learning_rate": 6.301756586991561e-06,
+      "loss": 0.5713005065917969,
+      "step": 887
+    },
+    {
+      "epoch": 0.8704815586325205,
+      "grad_norm": 0.6074492931365967,
+      "learning_rate": 6.294095225512604e-06,
+      "loss": 0.5398827791213989,
+      "step": 888
+    },
+    {
+      "epoch": 0.8714618306580076,
+      "grad_norm": 0.8076753616333008,
+      "learning_rate": 6.286430606150458e-06,
+      "loss": 0.6593376398086548,
+      "step": 889
+    },
+    {
+      "epoch": 0.8724421026834946,
+      "grad_norm": 0.5496326088905334,
+      "learning_rate": 6.278762748200795e-06,
+      "loss": 0.5054129362106323,
+      "step": 890
+    },
+    {
+      "epoch": 0.8734223747089818,
+      "grad_norm": 0.7481637001037598,
+      "learning_rate": 6.271091670967437e-06,
+      "loss": 0.6245748996734619,
+      "step": 891
+    },
+    {
+      "epoch": 0.8744026467344688,
+      "grad_norm": 0.6720854043960571,
+      "learning_rate": 6.2634173937623075e-06,
+      "loss": 0.5316181182861328,
+      "step": 892
+    },
+    {
+      "epoch": 0.8753829187599559,
+      "grad_norm": 2.7728631496429443,
+      "learning_rate": 6.255739935905396e-06,
+      "loss": 0.5387269258499146,
+      "step": 893
+    },
+    {
+      "epoch": 0.876363190785443,
+      "grad_norm": 0.5457331538200378,
+      "learning_rate": 6.248059316724693e-06,
+      "loss": 0.48916563391685486,
+      "step": 894
+    },
+    {
+      "epoch": 0.87734346281093,
+      "grad_norm": 0.5553867220878601,
+      "learning_rate": 6.240375555556145e-06,
+      "loss": 0.6687355637550354,
+      "step": 895
+    },
+    {
+      "epoch": 0.8783237348364171,
+      "grad_norm": 0.46050673723220825,
+      "learning_rate": 6.232688671743613e-06,
+      "loss": 0.5231783390045166,
+      "step": 896
+    },
+    {
+      "epoch": 0.8793040068619041,
+      "grad_norm": 1.1934031248092651,
+      "learning_rate": 6.22499868463882e-06,
+      "loss": 0.5726327896118164,
+      "step": 897
+    },
+    {
+      "epoch": 0.8802842788873912,
+      "grad_norm": 9.92513370513916,
+      "learning_rate": 6.217305613601296e-06,
+      "loss": 0.7107685804367065,
+      "step": 898
+    },
+    {
+      "epoch": 0.8812645509128784,
+      "grad_norm": 0.4932812750339508,
+      "learning_rate": 6.209609477998339e-06,
+      "loss": 0.6587767601013184,
+      "step": 899
+    },
+    {
+      "epoch": 0.8822448229383654,
+      "grad_norm": 1.0200669765472412,
+      "learning_rate": 6.201910297204963e-06,
+      "loss": 0.522339403629303,
+      "step": 900
+    },
+    {
+      "epoch": 0.8832250949638525,
+      "grad_norm": 2.011092185974121,
+      "learning_rate": 6.194208090603845e-06,
+      "loss": 0.5590274930000305,
+      "step": 901
+    },
+    {
+      "epoch": 0.8842053669893395,
+      "grad_norm": 0.6383436322212219,
+      "learning_rate": 6.186502877585281e-06,
+      "loss": 0.5745048522949219,
+      "step": 902
+    },
+    {
+      "epoch": 0.8851856390148266,
+      "grad_norm": 1.2666497230529785,
+      "learning_rate": 6.178794677547138e-06,
+      "loss": 0.5858358144760132,
+      "step": 903
+    },
+    {
+      "epoch": 0.8861659110403137,
+      "grad_norm": 0.8475280404090881,
+      "learning_rate": 6.171083509894795e-06,
+      "loss": 0.5306216478347778,
+      "step": 904
+    },
+    {
+      "epoch": 0.8871461830658007,
+      "grad_norm": 0.7674664258956909,
+      "learning_rate": 6.163369394041112e-06,
+      "loss": 0.5390942096710205,
+      "step": 905
+    },
+    {
+      "epoch": 0.8881264550912878,
+      "grad_norm": 0.7527076601982117,
+      "learning_rate": 6.155652349406366e-06,
+      "loss": 0.532594621181488,
+      "step": 906
+    },
+    {
+      "epoch": 0.8891067271167749,
+      "grad_norm": 0.6020624041557312,
+      "learning_rate": 6.1479323954182055e-06,
+      "loss": 0.5256073474884033,
+      "step": 907
+    },
+    {
+      "epoch": 0.890086999142262,
+      "grad_norm": 0.7182669043540955,
+      "learning_rate": 6.140209551511609e-06,
+      "loss": 0.6302467584609985,
+      "step": 908
+    },
+    {
+      "epoch": 0.8910672711677491,
+      "grad_norm": 0.6637997031211853,
+      "learning_rate": 6.132483837128823e-06,
+      "loss": 0.5175154805183411,
+      "step": 909
+    },
+    {
+      "epoch": 0.8920475431932361,
+      "grad_norm": 0.9244447946548462,
+      "learning_rate": 6.124755271719326e-06,
+      "loss": 0.855939507484436,
+      "step": 910
+    },
+    {
+      "epoch": 0.8930278152187232,
+      "grad_norm": 0.6451686024665833,
+      "learning_rate": 6.1170238747397715e-06,
+      "loss": 0.6661945581436157,
+      "step": 911
+    },
+    {
+      "epoch": 0.8940080872442102,
+      "grad_norm": 0.6926910281181335,
+      "learning_rate": 6.109289665653944e-06,
+      "loss": 0.545681357383728,
+      "step": 912
+    },
+    {
+      "epoch": 0.8949883592696973,
+      "grad_norm": 0.5391589403152466,
+      "learning_rate": 6.101552663932704e-06,
+      "loss": 0.569096565246582,
+      "step": 913
+    },
+    {
+      "epoch": 0.8959686312951844,
+      "grad_norm": 0.8027747869491577,
+      "learning_rate": 6.093812889053945e-06,
+      "loss": 0.6496536135673523,
+      "step": 914
+    },
+    {
+      "epoch": 0.8969489033206715,
+      "grad_norm": 0.9151626229286194,
+      "learning_rate": 6.08607036050254e-06,
+      "loss": 0.5546607375144958,
+      "step": 915
+    },
+    {
+      "epoch": 0.8979291753461586,
+      "grad_norm": 0.5226544737815857,
+      "learning_rate": 6.078325097770296e-06,
+      "loss": 0.5756404399871826,
+      "step": 916
+    },
+    {
+      "epoch": 0.8989094473716457,
+      "grad_norm": 0.666583240032196,
+      "learning_rate": 6.070577120355903e-06,
+      "loss": 0.7767114639282227,
+      "step": 917
+    },
+    {
+      "epoch": 0.8998897193971327,
+      "grad_norm": 0.8170400857925415,
+      "learning_rate": 6.062826447764883e-06,
+      "loss": 0.5552890300750732,
+      "step": 918
+    },
+    {
+      "epoch": 0.9008699914226198,
+      "grad_norm": 0.6850262880325317,
+      "learning_rate": 6.055073099509549e-06,
+      "loss": 0.6183463335037231,
+      "step": 919
+    },
+    {
+      "epoch": 0.9018502634481068,
+      "grad_norm": 0.6853513717651367,
+      "learning_rate": 6.047317095108943e-06,
+      "loss": 0.6733920574188232,
+      "step": 920
+    },
+    {
+      "epoch": 0.9028305354735939,
+      "grad_norm": 0.5335755348205566,
+      "learning_rate": 6.039558454088796e-06,
+      "loss": 0.45112305879592896,
+      "step": 921
+    },
+    {
+      "epoch": 0.903810807499081,
+      "grad_norm": 1.1216295957565308,
+      "learning_rate": 6.031797195981481e-06,
+      "loss": 0.659801721572876,
+      "step": 922
+    },
+    {
+      "epoch": 0.904791079524568,
+      "grad_norm": 2.0858795642852783,
+      "learning_rate": 6.024033340325954e-06,
+      "loss": 0.5316166877746582,
+      "step": 923
+    },
+    {
+      "epoch": 0.9057713515500552,
+      "grad_norm": 0.48735520243644714,
+      "learning_rate": 6.016266906667711e-06,
+      "loss": 0.5399875044822693,
+      "step": 924
+    },
+    {
+      "epoch": 0.9067516235755422,
+      "grad_norm": 0.5014345049858093,
+      "learning_rate": 6.0084979145587444e-06,
+      "loss": 0.5551764965057373,
+      "step": 925
+    },
+    {
+      "epoch": 0.9077318956010293,
+      "grad_norm": 3.2125260829925537,
+      "learning_rate": 6.000726383557476e-06,
+      "loss": 0.5554008483886719,
+      "step": 926
+    },
+    {
+      "epoch": 0.9087121676265164,
+      "grad_norm": 1.5546711683273315,
+      "learning_rate": 5.9929523332287275e-06,
+      "loss": 0.554888129234314,
+      "step": 927
+    },
+    {
+      "epoch": 0.9096924396520034,
+      "grad_norm": 9.836438179016113,
+      "learning_rate": 5.985175783143666e-06,
+      "loss": 0.5780972242355347,
+      "step": 928
+    },
+    {
+      "epoch": 0.9106727116774905,
+      "grad_norm": 1.0174751281738281,
+      "learning_rate": 5.977396752879742e-06,
+      "loss": 0.6562635898590088,
+      "step": 929
+    },
+    {
+      "epoch": 0.9116529837029775,
+      "grad_norm": 0.5529032349586487,
+      "learning_rate": 5.9696152620206575e-06,
+      "loss": 0.48356837034225464,
+      "step": 930
+    },
+    {
+      "epoch": 0.9126332557284647,
+      "grad_norm": 0.7021100521087646,
+      "learning_rate": 5.961831330156306e-06,
+      "loss": 0.712357759475708,
+      "step": 931
+    },
+    {
+      "epoch": 0.9136135277539518,
+      "grad_norm": 0.6379647850990295,
+      "learning_rate": 5.954044976882725e-06,
+      "loss": 0.6436705589294434,
+      "step": 932
+    },
+    {
+      "epoch": 0.9145937997794388,
+      "grad_norm": 0.4843108057975769,
+      "learning_rate": 5.946256221802052e-06,
+      "loss": 0.5778604745864868,
+      "step": 933
+    },
+    {
+      "epoch": 0.9155740718049259,
+      "grad_norm": 0.6174368858337402,
+      "learning_rate": 5.938465084522468e-06,
+      "loss": 0.6862471103668213,
+      "step": 934
+    },
+    {
+      "epoch": 0.9165543438304129,
+      "grad_norm": 0.48853597044944763,
+      "learning_rate": 5.930671584658151e-06,
+      "loss": 0.5211154222488403,
+      "step": 935
+    },
+    {
+      "epoch": 0.9175346158559,
+      "grad_norm": 0.6832510828971863,
+      "learning_rate": 5.922875741829227e-06,
+      "loss": 0.627961277961731,
+      "step": 936
+    },
+    {
+      "epoch": 0.9185148878813871,
+      "grad_norm": 0.664499044418335,
+      "learning_rate": 5.915077575661723e-06,
+      "loss": 0.728675365447998,
+      "step": 937
+    },
+    {
+      "epoch": 0.9194951599068741,
+      "grad_norm": 0.5297572016716003,
+      "learning_rate": 5.907277105787513e-06,
+      "loss": 0.4794204533100128,
+      "step": 938
+    },
+    {
+      "epoch": 0.9204754319323613,
+      "grad_norm": 0.5751549005508423,
+      "learning_rate": 5.89947435184427e-06,
+      "loss": 0.43880611658096313,
+      "step": 939
+    },
+    {
+      "epoch": 0.9214557039578483,
+      "grad_norm": 0.5993437170982361,
+      "learning_rate": 5.891669333475416e-06,
+      "loss": 0.7502422332763672,
+      "step": 940
+    },
+    {
+      "epoch": 0.9224359759833354,
+      "grad_norm": 0.802736222743988,
+      "learning_rate": 5.883862070330079e-06,
+      "loss": 0.6673043966293335,
+      "step": 941
+    },
+    {
+      "epoch": 0.9234162480088225,
+      "grad_norm": 0.7638604044914246,
+      "learning_rate": 5.876052582063031e-06,
+      "loss": 0.5863341093063354,
+      "step": 942
+    },
+    {
+      "epoch": 0.9243965200343095,
+      "grad_norm": 0.4516844153404236,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.4866067171096802,
+      "step": 943
+    },
+    {
+      "epoch": 0.9253767920597966,
+      "grad_norm": 0.566494345664978,
+      "learning_rate": 5.860427008810872e-06,
+      "loss": 0.5502532720565796,
+      "step": 944
+    },
+    {
+      "epoch": 0.9263570640852836,
+      "grad_norm": 0.5122462511062622,
+      "learning_rate": 5.85261096316312e-06,
+      "loss": 0.5924001932144165,
+      "step": 945
+    },
+    {
+      "epoch": 0.9273373361107707,
+      "grad_norm": 0.7986511588096619,
+      "learning_rate": 5.844792771068284e-06,
+      "loss": 0.6251932382583618,
+      "step": 946
+    },
+    {
+      "epoch": 0.9283176081362579,
+      "grad_norm": 0.4851377308368683,
+      "learning_rate": 5.8369724522086545e-06,
+      "loss": 0.5673865079879761,
+      "step": 947
+    },
+    {
+      "epoch": 0.9292978801617449,
+      "grad_norm": 0.9102888107299805,
+      "learning_rate": 5.829150026271871e-06,
+      "loss": 0.7148391008377075,
+      "step": 948
+    },
+    {
+      "epoch": 0.930278152187232,
+      "grad_norm": 0.5190656185150146,
+      "learning_rate": 5.821325512950886e-06,
+      "loss": 0.5720458030700684,
+      "step": 949
+    },
+    {
+      "epoch": 0.931258424212719,
+      "grad_norm": 1.644549012184143,
+      "learning_rate": 5.8134989319439004e-06,
+      "loss": 0.7064533233642578,
+      "step": 950
+    },
+    {
+      "epoch": 0.9322386962382061,
+      "grad_norm": 0.936130166053772,
+      "learning_rate": 5.805670302954322e-06,
+      "loss": 0.5619934797286987,
+      "step": 951
+    },
+    {
+      "epoch": 0.9332189682636932,
+      "grad_norm": 0.6011050343513489,
+      "learning_rate": 5.797839645690715e-06,
+      "loss": 0.5513787865638733,
+      "step": 952
+    },
+    {
+      "epoch": 0.9341992402891802,
+      "grad_norm": 0.8352797031402588,
+      "learning_rate": 5.79000697986675e-06,
+      "loss": 0.5585759282112122,
+      "step": 953
+    },
+    {
+      "epoch": 0.9351795123146673,
+      "grad_norm": 1.2676984071731567,
+      "learning_rate": 5.782172325201155e-06,
+      "loss": 0.5337406992912292,
+      "step": 954
+    },
+    {
+      "epoch": 0.9361597843401543,
+      "grad_norm": 1.0544131994247437,
+      "learning_rate": 5.774335701417662e-06,
+      "loss": 0.5163780450820923,
+      "step": 955
+    },
+    {
+      "epoch": 0.9371400563656415,
+      "grad_norm": 0.5444765090942383,
+      "learning_rate": 5.766497128244963e-06,
+      "loss": 0.6283679604530334,
+      "step": 956
+    },
+    {
+      "epoch": 0.9381203283911286,
+      "grad_norm": 0.726844310760498,
+      "learning_rate": 5.758656625416659e-06,
+      "loss": 0.5578550100326538,
+      "step": 957
+    },
+    {
+      "epoch": 0.9391006004166156,
+      "grad_norm": 2.33665132522583,
+      "learning_rate": 5.750814212671202e-06,
+      "loss": 0.6489031314849854,
+      "step": 958
+    },
+    {
+      "epoch": 0.9400808724421027,
+      "grad_norm": 0.8335084915161133,
+      "learning_rate": 5.7429699097518585e-06,
+      "loss": 0.592069149017334,
+      "step": 959
+    },
+    {
+      "epoch": 0.9410611444675897,
+      "grad_norm": 0.5806710124015808,
+      "learning_rate": 5.735123736406655e-06,
+      "loss": 0.5590806007385254,
+      "step": 960
+    },
+    {
+      "epoch": 0.9420414164930768,
+      "grad_norm": 0.6549690365791321,
+      "learning_rate": 5.727275712388318e-06,
+      "loss": 0.540640115737915,
+      "step": 961
+    },
+    {
+      "epoch": 0.9430216885185639,
+      "grad_norm": 0.9472928643226624,
+      "learning_rate": 5.719425857454242e-06,
+      "loss": 0.7817751169204712,
+      "step": 962
+    },
+    {
+      "epoch": 0.9440019605440509,
+      "grad_norm": 1.01449453830719,
+      "learning_rate": 5.711574191366427e-06,
+      "loss": 0.493563175201416,
+      "step": 963
+    },
+    {
+      "epoch": 0.9449822325695381,
+      "grad_norm": 0.7543380260467529,
+      "learning_rate": 5.7037207338914326e-06,
+      "loss": 0.57488614320755,
+      "step": 964
+    },
+    {
+      "epoch": 0.9459625045950251,
+      "grad_norm": 0.5324261784553528,
+      "learning_rate": 5.695865504800328e-06,
+      "loss": 0.6612128019332886,
+      "step": 965
+    },
+    {
+      "epoch": 0.9469427766205122,
+      "grad_norm": 0.6752833127975464,
+      "learning_rate": 5.688008523868646e-06,
+      "loss": 0.5913141965866089,
+      "step": 966
+    },
+    {
+      "epoch": 0.9479230486459993,
+      "grad_norm": 1.4944877624511719,
+      "learning_rate": 5.680149810876322e-06,
+      "loss": 0.676797091960907,
+      "step": 967
+    },
+    {
+      "epoch": 0.9489033206714863,
+      "grad_norm": 0.733349621295929,
+      "learning_rate": 5.6722893856076596e-06,
+      "loss": 0.5253714323043823,
+      "step": 968
+    },
+    {
+      "epoch": 0.9498835926969734,
+      "grad_norm": 0.6789211630821228,
+      "learning_rate": 5.664427267851271e-06,
+      "loss": 0.5896862745285034,
+      "step": 969
+    },
+    {
+      "epoch": 0.9508638647224604,
+      "grad_norm": 0.497837632894516,
+      "learning_rate": 5.656563477400026e-06,
+      "loss": 0.6182793378829956,
+      "step": 970
+    },
+    {
+      "epoch": 0.9518441367479475,
+      "grad_norm": 0.7501582503318787,
+      "learning_rate": 5.648698034051009e-06,
+      "loss": 0.5923135876655579,
+      "step": 971
+    },
+    {
+      "epoch": 0.9528244087734347,
+      "grad_norm": 0.789541482925415,
+      "learning_rate": 5.640830957605466e-06,
+      "loss": 0.5664257407188416,
+      "step": 972
+    },
+    {
+      "epoch": 0.9538046807989217,
+      "grad_norm": 0.7743480801582336,
+      "learning_rate": 5.632962267868747e-06,
+      "loss": 0.5279930830001831,
+      "step": 973
+    },
+    {
+      "epoch": 0.9547849528244088,
+      "grad_norm": 0.6490453481674194,
+      "learning_rate": 5.625091984650273e-06,
+      "loss": 0.6161771416664124,
+      "step": 974
+    },
+    {
+      "epoch": 0.9557652248498959,
+      "grad_norm": 0.666648805141449,
+      "learning_rate": 5.617220127763474e-06,
+      "loss": 0.5054368376731873,
+      "step": 975
+    },
+    {
+      "epoch": 0.9567454968753829,
+      "grad_norm": 0.8474366664886475,
+      "learning_rate": 5.609346717025738e-06,
+      "loss": 0.5411797165870667,
+      "step": 976
+    },
+    {
+      "epoch": 0.95772576890087,
+      "grad_norm": 0.8270455002784729,
+      "learning_rate": 5.601471772258368e-06,
+      "loss": 0.5817526578903198,
+      "step": 977
+    },
+    {
+      "epoch": 0.958706040926357,
+      "grad_norm": 0.5137153267860413,
+      "learning_rate": 5.593595313286526e-06,
+      "loss": 0.46354103088378906,
+      "step": 978
+    },
+    {
+      "epoch": 0.9596863129518441,
+      "grad_norm": 0.6770734190940857,
+      "learning_rate": 5.585717359939192e-06,
+      "loss": 0.5301814079284668,
+      "step": 979
+    },
+    {
+      "epoch": 0.9606665849773313,
+      "grad_norm": 0.5311392545700073,
+      "learning_rate": 5.5778379320491e-06,
+      "loss": 0.46160927414894104,
+      "step": 980
+    },
+    {
+      "epoch": 0.9616468570028183,
+      "grad_norm": 0.6499536633491516,
+      "learning_rate": 5.569957049452703e-06,
+      "loss": 0.7306463122367859,
+      "step": 981
+    },
+    {
+      "epoch": 0.9626271290283054,
+      "grad_norm": 0.500795304775238,
+      "learning_rate": 5.562074731990115e-06,
+      "loss": 0.5398798584938049,
+      "step": 982
+    },
+    {
+      "epoch": 0.9636074010537924,
+      "grad_norm": 0.5290957093238831,
+      "learning_rate": 5.5541909995050554e-06,
+      "loss": 0.5360432863235474,
+      "step": 983
+    },
+    {
+      "epoch": 0.9645876730792795,
+      "grad_norm": 0.8447754383087158,
+      "learning_rate": 5.546305871844816e-06,
+      "loss": 0.4349919557571411,
+      "step": 984
+    },
+    {
+      "epoch": 0.9655679451047666,
+      "grad_norm": 0.5076740384101868,
+      "learning_rate": 5.538419368860196e-06,
+      "loss": 0.48006004095077515,
+      "step": 985
+    },
+    {
+      "epoch": 0.9665482171302536,
+      "grad_norm": 0.5041320323944092,
+      "learning_rate": 5.530531510405454e-06,
+      "loss": 0.48883479833602905,
+      "step": 986
+    },
+    {
+      "epoch": 0.9675284891557407,
+      "grad_norm": 0.5577536225318909,
+      "learning_rate": 5.522642316338268e-06,
+      "loss": 0.5165166854858398,
+      "step": 987
+    },
+    {
+      "epoch": 0.9685087611812278,
+      "grad_norm": 14.343788146972656,
+      "learning_rate": 5.514751806519673e-06,
+      "loss": 0.6178335547447205,
+      "step": 988
+    },
+    {
+      "epoch": 0.9694890332067149,
+      "grad_norm": 2.3618218898773193,
+      "learning_rate": 5.506860000814017e-06,
+      "loss": 0.4987364709377289,
+      "step": 989
+    },
+    {
+      "epoch": 0.970469305232202,
+      "grad_norm": 0.6319774985313416,
+      "learning_rate": 5.498966919088914e-06,
+      "loss": 0.5759574770927429,
+      "step": 990
+    },
+    {
+      "epoch": 0.971449577257689,
+      "grad_norm": 1.0980327129364014,
+      "learning_rate": 5.491072581215186e-06,
+      "loss": 0.5760068893432617,
+      "step": 991
+    },
+    {
+      "epoch": 0.9724298492831761,
+      "grad_norm": 0.45583245158195496,
+      "learning_rate": 5.48317700706682e-06,
+      "loss": 0.5371822118759155,
+      "step": 992
+    },
+    {
+      "epoch": 0.9734101213086631,
+      "grad_norm": 0.5124374628067017,
+      "learning_rate": 5.475280216520913e-06,
+      "loss": 0.4932123124599457,
+      "step": 993
+    },
+    {
+      "epoch": 0.9743903933341502,
+      "grad_norm": 0.6532335877418518,
+      "learning_rate": 5.467382229457629e-06,
+      "loss": 0.5507386922836304,
+      "step": 994
+    },
+    {
+      "epoch": 0.9753706653596373,
+      "grad_norm": 0.7889341711997986,
+      "learning_rate": 5.459483065760138e-06,
+      "loss": 0.5716214179992676,
+      "step": 995
+    },
+    {
+      "epoch": 0.9763509373851244,
+      "grad_norm": 0.47389283776283264,
+      "learning_rate": 5.451582745314576e-06,
+      "loss": 0.5770868062973022,
+      "step": 996
+    },
+    {
+      "epoch": 0.9773312094106115,
+      "grad_norm": 1.6698098182678223,
+      "learning_rate": 5.443681288009991e-06,
+      "loss": 0.5604249238967896,
+      "step": 997
+    },
+    {
+      "epoch": 0.9783114814360985,
+      "grad_norm": 0.6366322636604309,
+      "learning_rate": 5.435778713738292e-06,
+      "loss": 0.5466155409812927,
+      "step": 998
+    },
+    {
+      "epoch": 0.9792917534615856,
+      "grad_norm": 0.7743147611618042,
+      "learning_rate": 5.4278750423942e-06,
+      "loss": 0.6183242797851562,
+      "step": 999
+    },
+    {
+      "epoch": 0.9802720254870727,
+      "grad_norm": 0.7067610025405884,
+      "learning_rate": 5.419970293875197e-06,
+      "loss": 0.4826086163520813,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9812522975125597,
+      "grad_norm": 0.6085331439971924,
+      "learning_rate": 5.412064488081482e-06,
+      "loss": 0.5776460766792297,
+      "step": 1001
+    },
+    {
+      "epoch": 0.9822325695380468,
+      "grad_norm": 1.9325439929962158,
+      "learning_rate": 5.404157644915907e-06,
+      "loss": 0.5454663038253784,
+      "step": 1002
+    },
+    {
+      "epoch": 0.9832128415635338,
+      "grad_norm": 0.6339237093925476,
+      "learning_rate": 5.396249784283943e-06,
+      "loss": 0.5601510405540466,
+      "step": 1003
+    },
+    {
+      "epoch": 0.984193113589021,
+      "grad_norm": 0.5168980956077576,
+      "learning_rate": 5.388340926093619e-06,
+      "loss": 0.5439151525497437,
+      "step": 1004
+    },
+    {
+      "epoch": 0.9851733856145081,
+      "grad_norm": 0.6750264167785645,
+      "learning_rate": 5.380431090255475e-06,
+      "loss": 0.550171434879303,
+      "step": 1005
+    },
+    {
+      "epoch": 0.9861536576399951,
+      "grad_norm": 0.8704385161399841,
+      "learning_rate": 5.372520296682516e-06,
+      "loss": 0.7121069431304932,
+      "step": 1006
+    },
+    {
+      "epoch": 0.9871339296654822,
+      "grad_norm": 0.6224149465560913,
+      "learning_rate": 5.364608565290154e-06,
+      "loss": 0.5001834630966187,
+      "step": 1007
+    },
+    {
+      "epoch": 0.9881142016909692,
+      "grad_norm": 0.7595095038414001,
+      "learning_rate": 5.356695915996162e-06,
+      "loss": 0.55515456199646,
+      "step": 1008
+    },
+    {
+      "epoch": 0.9890944737164563,
+      "grad_norm": 5.989940643310547,
+      "learning_rate": 5.348782368720627e-06,
+      "loss": 0.5854626893997192,
+      "step": 1009
+    },
+    {
+      "epoch": 0.9900747457419434,
+      "grad_norm": 0.4549104571342468,
+      "learning_rate": 5.340867943385895e-06,
+      "loss": 0.5413213968276978,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9910550177674304,
+      "grad_norm": 0.7498558759689331,
+      "learning_rate": 5.33295265991652e-06,
+      "loss": 0.6045414805412292,
+      "step": 1011
+    },
+    {
+      "epoch": 0.9920352897929176,
+      "grad_norm": 0.6960429549217224,
+      "learning_rate": 5.325036538239221e-06,
+      "loss": 0.4991673231124878,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9930155618184046,
+      "grad_norm": 2.0619726181030273,
+      "learning_rate": 5.317119598282823e-06,
+      "loss": 0.4863162636756897,
+      "step": 1013
+    },
+    {
+      "epoch": 0.9939958338438917,
+      "grad_norm": 0.8567239046096802,
+      "learning_rate": 5.3092018599782155e-06,
+      "loss": 0.6333738565444946,
+      "step": 1014
+    },
+    {
+      "epoch": 0.9949761058693788,
+      "grad_norm": 0.8139933943748474,
+      "learning_rate": 5.301283343258293e-06,
+      "loss": 0.6447440385818481,
+      "step": 1015
+    },
+    {
+      "epoch": 0.9959563778948658,
+      "grad_norm": 0.8827837109565735,
+      "learning_rate": 5.293364068057912e-06,
+      "loss": 0.6062864065170288,
+      "step": 1016
+    },
+    {
+      "epoch": 0.9969366499203529,
+      "grad_norm": 0.5673131346702576,
+      "learning_rate": 5.285444054313841e-06,
+      "loss": 0.3972030580043793,
+      "step": 1017
+    },
+    {
+      "epoch": 0.9979169219458399,
+      "grad_norm": 0.806973934173584,
+      "learning_rate": 5.277523321964701e-06,
+      "loss": 0.6702872514724731,
+      "step": 1018
+    },
+    {
+      "epoch": 0.998897193971327,
+      "grad_norm": 0.7192707061767578,
+      "learning_rate": 5.26960189095093e-06,
+      "loss": 0.456107497215271,
+      "step": 1019
+    },
+    {
+      "epoch": 0.9998774659968142,
+      "grad_norm": 0.6452010273933411,
+      "learning_rate": 5.2616797812147205e-06,
+      "loss": 0.47958672046661377,
+      "step": 1020
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6038899421691895,
+      "learning_rate": 5.253757012699972e-06,
+      "loss": 0.7019495964050293,
+      "step": 1021
+    },
+    {
+      "epoch": 1.000980272025487,
+      "grad_norm": 1.4249541759490967,
+      "learning_rate": 5.245833605352246e-06,
+      "loss": 0.5797180533409119,
+      "step": 1022
+    },
+    {
+      "epoch": 1.0019605440509742,
+      "grad_norm": 0.8366342782974243,
+      "learning_rate": 5.237909579118713e-06,
+      "loss": 0.5079513192176819,
+      "step": 1023
+    },
+    {
+      "epoch": 1.0029408160764612,
+      "grad_norm": 0.67219078540802,
+      "learning_rate": 5.2299849539480965e-06,
+      "loss": 0.5283257365226746,
+      "step": 1024
+    },
+    {
+      "epoch": 1.0039210881019482,
+      "grad_norm": 0.7875847816467285,
+      "learning_rate": 5.2220597497906315e-06,
+      "loss": 0.4881901144981384,
+      "step": 1025
+    },
+    {
+      "epoch": 1.0049013601274355,
+      "grad_norm": 0.48576492071151733,
+      "learning_rate": 5.214133986598014e-06,
+      "loss": 0.45480456948280334,
+      "step": 1026
+    },
+    {
+      "epoch": 1.0058816321529225,
+      "grad_norm": 0.6740761399269104,
+      "learning_rate": 5.206207684323337e-06,
+      "loss": 0.3906821012496948,
+      "step": 1027
+    },
+    {
+      "epoch": 1.0068619041784095,
+      "grad_norm": 0.704173743724823,
+      "learning_rate": 5.198280862921062e-06,
+      "loss": 0.51959228515625,
+      "step": 1028
+    },
+    {
+      "epoch": 1.0078421762038965,
+      "grad_norm": 0.6730834245681763,
+      "learning_rate": 5.190353542346951e-06,
+      "loss": 0.4134294390678406,
+      "step": 1029
+    },
+    {
+      "epoch": 1.0088224482293837,
+      "grad_norm": 0.7609149217605591,
+      "learning_rate": 5.182425742558024e-06,
+      "loss": 0.531853199005127,
+      "step": 1030
+    },
+    {
+      "epoch": 1.0098027202548707,
+      "grad_norm": 1.0709447860717773,
+      "learning_rate": 5.174497483512506e-06,
+      "loss": 0.5824171304702759,
+      "step": 1031
+    },
+    {
+      "epoch": 1.0107829922803577,
+      "grad_norm": 0.5358075499534607,
+      "learning_rate": 5.166568785169781e-06,
+      "loss": 0.5012086629867554,
+      "step": 1032
+    },
+    {
+      "epoch": 1.011763264305845,
+      "grad_norm": 0.6626541614532471,
+      "learning_rate": 5.15863966749034e-06,
+      "loss": 0.45648837089538574,
+      "step": 1033
+    },
+    {
+      "epoch": 1.012743536331332,
+      "grad_norm": 0.5781713128089905,
+      "learning_rate": 5.1507101504357225e-06,
+      "loss": 0.5260370373725891,
+      "step": 1034
+    },
+    {
+      "epoch": 1.013723808356819,
+      "grad_norm": 0.5141299366950989,
+      "learning_rate": 5.142780253968481e-06,
+      "loss": 0.38620617985725403,
+      "step": 1035
+    },
+    {
+      "epoch": 1.0147040803823062,
+      "grad_norm": 0.8716086149215698,
+      "learning_rate": 5.134849998052123e-06,
+      "loss": 0.5863468647003174,
+      "step": 1036
+    },
+    {
+      "epoch": 1.0156843524077932,
+      "grad_norm": 0.7488311529159546,
+      "learning_rate": 5.126919402651053e-06,
+      "loss": 0.49834364652633667,
+      "step": 1037
+    },
+    {
+      "epoch": 1.0166646244332802,
+      "grad_norm": 1.071315050125122,
+      "learning_rate": 5.118988487730537e-06,
+      "loss": 0.5789937973022461,
+      "step": 1038
+    },
+    {
+      "epoch": 1.0176448964587672,
+      "grad_norm": 0.65739506483078,
+      "learning_rate": 5.111057273256648e-06,
+      "loss": 0.5703906416893005,
+      "step": 1039
+    },
+    {
+      "epoch": 1.0186251684842544,
+      "grad_norm": 1.0825657844543457,
+      "learning_rate": 5.103125779196204e-06,
+      "loss": 0.4941566586494446,
+      "step": 1040
+    },
+    {
+      "epoch": 1.0196054405097414,
+      "grad_norm": 0.537731409072876,
+      "learning_rate": 5.095194025516733e-06,
+      "loss": 0.47552376985549927,
+      "step": 1041
+    },
+    {
+      "epoch": 1.0205857125352285,
+      "grad_norm": 0.6944141387939453,
+      "learning_rate": 5.087262032186418e-06,
+      "loss": 0.6037413477897644,
+      "step": 1042
+    },
+    {
+      "epoch": 1.0215659845607157,
+      "grad_norm": 1.6507412195205688,
+      "learning_rate": 5.07932981917404e-06,
+      "loss": 0.5054140090942383,
+      "step": 1043
+    },
+    {
+      "epoch": 1.0225462565862027,
+      "grad_norm": 0.44498687982559204,
+      "learning_rate": 5.071397406448937e-06,
+      "loss": 0.4743750989437103,
+      "step": 1044
+    },
+    {
+      "epoch": 1.0235265286116897,
+      "grad_norm": 0.6090741753578186,
+      "learning_rate": 5.063464813980948e-06,
+      "loss": 0.48587140440940857,
+      "step": 1045
+    },
+    {
+      "epoch": 1.024506800637177,
+      "grad_norm": 0.840766429901123,
+      "learning_rate": 5.055532061740367e-06,
+      "loss": 0.39757728576660156,
+      "step": 1046
+    },
+    {
+      "epoch": 1.025487072662664,
+      "grad_norm": 0.6779752373695374,
+      "learning_rate": 5.0475991696978844e-06,
+      "loss": 0.5216217041015625,
+      "step": 1047
+    },
+    {
+      "epoch": 1.026467344688151,
+      "grad_norm": 0.5362389087677002,
+      "learning_rate": 5.039666157824549e-06,
+      "loss": 0.49866318702697754,
+      "step": 1048
+    },
+    {
+      "epoch": 1.027447616713638,
+      "grad_norm": 1.6441469192504883,
+      "learning_rate": 5.03173304609171e-06,
+      "loss": 0.47683894634246826,
+      "step": 1049
+    },
+    {
+      "epoch": 1.0284278887391252,
+      "grad_norm": 0.6118367910385132,
+      "learning_rate": 5.023799854470963e-06,
+      "loss": 0.5168311595916748,
+      "step": 1050
+    },
+    {
+      "epoch": 1.0294081607646122,
+      "grad_norm": 0.41360053420066833,
+      "learning_rate": 5.015866602934112e-06,
+      "loss": 0.3936558663845062,
+      "step": 1051
+    },
+    {
+      "epoch": 1.0303884327900992,
+      "grad_norm": 0.6367619037628174,
+      "learning_rate": 5.007933311453108e-06,
+      "loss": 0.6046751737594604,
+      "step": 1052
+    },
+    {
+      "epoch": 1.0313687048155864,
+      "grad_norm": 0.6235337853431702,
+      "learning_rate": 5e-06,
+      "loss": 0.4950873553752899,
+      "step": 1053
+    },
+    {
+      "epoch": 1.0323489768410734,
+      "grad_norm": 0.5964008569717407,
+      "learning_rate": 4.992066688546895e-06,
+      "loss": 0.4607120156288147,
+      "step": 1054
+    },
+    {
+      "epoch": 1.0333292488665604,
+      "grad_norm": 1.0708956718444824,
+      "learning_rate": 4.984133397065889e-06,
+      "loss": 0.48211315274238586,
+      "step": 1055
+    },
+    {
+      "epoch": 1.0343095208920476,
+      "grad_norm": 0.5160314440727234,
+      "learning_rate": 4.976200145529039e-06,
+      "loss": 0.4449300765991211,
+      "step": 1056
+    },
+    {
+      "epoch": 1.0352897929175346,
+      "grad_norm": 0.8167364001274109,
+      "learning_rate": 4.9682669539082914e-06,
+      "loss": 0.5578494071960449,
+      "step": 1057
+    },
+    {
+      "epoch": 1.0362700649430217,
+      "grad_norm": 0.4616822302341461,
+      "learning_rate": 4.960333842175453e-06,
+      "loss": 0.4051743149757385,
+      "step": 1058
+    },
+    {
+      "epoch": 1.0372503369685087,
+      "grad_norm": 2.050485134124756,
+      "learning_rate": 4.952400830302117e-06,
+      "loss": 0.5753476023674011,
+      "step": 1059
+    },
+    {
+      "epoch": 1.0382306089939959,
+      "grad_norm": 1.6075116395950317,
+      "learning_rate": 4.944467938259636e-06,
+      "loss": 0.37830492854118347,
+      "step": 1060
+    },
+    {
+      "epoch": 1.039210881019483,
+      "grad_norm": 0.7246462106704712,
+      "learning_rate": 4.936535186019053e-06,
+      "loss": 0.4199037551879883,
+      "step": 1061
+    },
+    {
+      "epoch": 1.04019115304497,
+      "grad_norm": 1.4789761304855347,
+      "learning_rate": 4.928602593551065e-06,
+      "loss": 0.571158766746521,
+      "step": 1062
+    },
+    {
+      "epoch": 1.0411714250704571,
+      "grad_norm": 0.5212058424949646,
+      "learning_rate": 4.9206701808259605e-06,
+      "loss": 0.48598021268844604,
+      "step": 1063
+    },
+    {
+      "epoch": 1.0421516970959441,
+      "grad_norm": 0.8929732441902161,
+      "learning_rate": 4.9127379678135825e-06,
+      "loss": 0.5332722067832947,
+      "step": 1064
+    },
+    {
+      "epoch": 1.0431319691214311,
+      "grad_norm": 0.6291391849517822,
+      "learning_rate": 4.904805974483267e-06,
+      "loss": 0.5221359133720398,
+      "step": 1065
+    },
+    {
+      "epoch": 1.0441122411469184,
+      "grad_norm": 0.7792730331420898,
+      "learning_rate": 4.8968742208037976e-06,
+      "loss": 0.398730993270874,
+      "step": 1066
+    },
+    {
+      "epoch": 1.0450925131724054,
+      "grad_norm": 0.7644619345664978,
+      "learning_rate": 4.888942726743353e-06,
+      "loss": 0.3855770230293274,
+      "step": 1067
+    },
+    {
+      "epoch": 1.0460727851978924,
+      "grad_norm": 1.4235645532608032,
+      "learning_rate": 4.881011512269464e-06,
+      "loss": 0.7365342378616333,
+      "step": 1068
+    },
+    {
+      "epoch": 1.0470530572233794,
+      "grad_norm": 0.5539320111274719,
+      "learning_rate": 4.873080597348948e-06,
+      "loss": 0.4274979531764984,
+      "step": 1069
+    },
+    {
+      "epoch": 1.0480333292488666,
+      "grad_norm": 1.9417518377304077,
+      "learning_rate": 4.86515000194788e-06,
+      "loss": 0.5061432123184204,
+      "step": 1070
+    },
+    {
+      "epoch": 1.0490136012743536,
+      "grad_norm": 0.9011236429214478,
+      "learning_rate": 4.85721974603152e-06,
+      "loss": 0.6233201026916504,
+      "step": 1071
+    },
+    {
+      "epoch": 1.0499938732998406,
+      "grad_norm": 2.24048113822937,
+      "learning_rate": 4.849289849564278e-06,
+      "loss": 0.45891186594963074,
+      "step": 1072
+    },
+    {
+      "epoch": 1.0509741453253278,
+      "grad_norm": 2.29797101020813,
+      "learning_rate": 4.841360332509663e-06,
+      "loss": 0.4944881796836853,
+      "step": 1073
+    },
+    {
+      "epoch": 1.0519544173508149,
+      "grad_norm": 0.6497139930725098,
+      "learning_rate": 4.83343121483022e-06,
+      "loss": 0.541831374168396,
+      "step": 1074
+    },
+    {
+      "epoch": 1.0529346893763019,
+      "grad_norm": 0.568021297454834,
+      "learning_rate": 4.825502516487497e-06,
+      "loss": 0.5159993171691895,
+      "step": 1075
+    },
+    {
+      "epoch": 1.053914961401789,
+      "grad_norm": 0.5896922945976257,
+      "learning_rate": 4.817574257441978e-06,
+      "loss": 0.4934307336807251,
+      "step": 1076
+    },
+    {
+      "epoch": 1.054895233427276,
+      "grad_norm": 0.5019741654396057,
+      "learning_rate": 4.809646457653051e-06,
+      "loss": 0.46182769536972046,
+      "step": 1077
+    },
+    {
+      "epoch": 1.055875505452763,
+      "grad_norm": 0.682341992855072,
+      "learning_rate": 4.801719137078939e-06,
+      "loss": 0.3881131410598755,
+      "step": 1078
+    },
+    {
+      "epoch": 1.0568557774782503,
+      "grad_norm": 0.6213642954826355,
+      "learning_rate": 4.793792315676665e-06,
+      "loss": 0.5301284790039062,
+      "step": 1079
+    },
+    {
+      "epoch": 1.0578360495037373,
+      "grad_norm": 0.5075803399085999,
+      "learning_rate": 4.785866013401989e-06,
+      "loss": 0.4713447093963623,
+      "step": 1080
+    },
+    {
+      "epoch": 1.0588163215292243,
+      "grad_norm": 0.7492678761482239,
+      "learning_rate": 4.777940250209369e-06,
+      "loss": 0.46998438239097595,
+      "step": 1081
+    },
+    {
+      "epoch": 1.0597965935547113,
+      "grad_norm": 0.5904183387756348,
+      "learning_rate": 4.770015046051904e-06,
+      "loss": 0.5474088788032532,
+      "step": 1082
+    },
+    {
+      "epoch": 1.0607768655801986,
+      "grad_norm": 0.8297621607780457,
+      "learning_rate": 4.762090420881289e-06,
+      "loss": 0.4612366259098053,
+      "step": 1083
+    },
+    {
+      "epoch": 1.0617571376056856,
+      "grad_norm": 1.3703441619873047,
+      "learning_rate": 4.754166394647755e-06,
+      "loss": 0.4448715150356293,
+      "step": 1084
+    },
+    {
+      "epoch": 1.0627374096311726,
+      "grad_norm": 0.5170440673828125,
+      "learning_rate": 4.74624298730003e-06,
+      "loss": 0.4783724546432495,
+      "step": 1085
+    },
+    {
+      "epoch": 1.0637176816566598,
+      "grad_norm": 1.0383613109588623,
+      "learning_rate": 4.738320218785281e-06,
+      "loss": 0.5514822006225586,
+      "step": 1086
+    },
+    {
+      "epoch": 1.0646979536821468,
+      "grad_norm": 0.5789095163345337,
+      "learning_rate": 4.7303981090490715e-06,
+      "loss": 0.5083911418914795,
+      "step": 1087
+    },
+    {
+      "epoch": 1.0656782257076338,
+      "grad_norm": 0.5016994476318359,
+      "learning_rate": 4.7224766780353005e-06,
+      "loss": 0.4580415189266205,
+      "step": 1088
+    },
+    {
+      "epoch": 1.0666584977331208,
+      "grad_norm": 0.8807493448257446,
+      "learning_rate": 4.71455594568616e-06,
+      "loss": 0.5211396217346191,
+      "step": 1089
+    },
+    {
+      "epoch": 1.067638769758608,
+      "grad_norm": 0.46059858798980713,
+      "learning_rate": 4.706635931942089e-06,
+      "loss": 0.4311555027961731,
+      "step": 1090
+    },
+    {
+      "epoch": 1.068619041784095,
+      "grad_norm": 2.1164276599884033,
+      "learning_rate": 4.6987166567417085e-06,
+      "loss": 0.5832147598266602,
+      "step": 1091
+    },
+    {
+      "epoch": 1.069599313809582,
+      "grad_norm": 0.756169855594635,
+      "learning_rate": 4.690798140021787e-06,
+      "loss": 0.3836706876754761,
+      "step": 1092
+    },
+    {
+      "epoch": 1.0705795858350693,
+      "grad_norm": 0.7354283928871155,
+      "learning_rate": 4.682880401717178e-06,
+      "loss": 0.48054513335227966,
+      "step": 1093
+    },
+    {
+      "epoch": 1.0715598578605563,
+      "grad_norm": 0.45950162410736084,
+      "learning_rate": 4.674963461760782e-06,
+      "loss": 0.42868906259536743,
+      "step": 1094
+    },
+    {
+      "epoch": 1.0725401298860433,
+      "grad_norm": 0.5602439641952515,
+      "learning_rate": 4.667047340083481e-06,
+      "loss": 0.42878490686416626,
+      "step": 1095
+    },
+    {
+      "epoch": 1.0735204019115305,
+      "grad_norm": 0.7466726303100586,
+      "learning_rate": 4.659132056614108e-06,
+      "loss": 0.546418309211731,
+      "step": 1096
+    },
+    {
+      "epoch": 1.0745006739370175,
+      "grad_norm": 0.5399389266967773,
+      "learning_rate": 4.651217631279374e-06,
+      "loss": 0.5127849578857422,
+      "step": 1097
+    },
+    {
+      "epoch": 1.0754809459625045,
+      "grad_norm": 0.483035147190094,
+      "learning_rate": 4.643304084003839e-06,
+      "loss": 0.33804774284362793,
+      "step": 1098
+    },
+    {
+      "epoch": 1.0764612179879918,
+      "grad_norm": 0.853963315486908,
+      "learning_rate": 4.635391434709847e-06,
+      "loss": 0.45693209767341614,
+      "step": 1099
+    },
+    {
+      "epoch": 1.0774414900134788,
+      "grad_norm": 0.5202552676200867,
+      "learning_rate": 4.627479703317485e-06,
+      "loss": 0.4849308133125305,
+      "step": 1100
+    },
+    {
+      "epoch": 1.0784217620389658,
+      "grad_norm": 1.0684430599212646,
+      "learning_rate": 4.619568909744524e-06,
+      "loss": 0.46725156903266907,
+      "step": 1101
+    },
+    {
+      "epoch": 1.0794020340644528,
+      "grad_norm": 0.7909156084060669,
+      "learning_rate": 4.611659073906382e-06,
+      "loss": 0.49704810976982117,
+      "step": 1102
+    },
+    {
+      "epoch": 1.08038230608994,
+      "grad_norm": 0.920854926109314,
+      "learning_rate": 4.603750215716057e-06,
+      "loss": 0.41881412267684937,
+      "step": 1103
+    },
+    {
+      "epoch": 1.081362578115427,
+      "grad_norm": 0.7200092077255249,
+      "learning_rate": 4.595842355084094e-06,
+      "loss": 0.4722629189491272,
+      "step": 1104
+    },
+    {
+      "epoch": 1.082342850140914,
+      "grad_norm": 0.5527382493019104,
+      "learning_rate": 4.587935511918521e-06,
+      "loss": 0.5283772349357605,
+      "step": 1105
+    },
+    {
+      "epoch": 1.0833231221664013,
+      "grad_norm": 0.8212664127349854,
+      "learning_rate": 4.580029706124804e-06,
+      "loss": 0.4981229901313782,
+      "step": 1106
+    },
+    {
+      "epoch": 1.0843033941918883,
+      "grad_norm": 0.5677878856658936,
+      "learning_rate": 4.572124957605803e-06,
+      "loss": 0.4896043539047241,
+      "step": 1107
+    },
+    {
+      "epoch": 1.0852836662173753,
+      "grad_norm": 0.5342493057250977,
+      "learning_rate": 4.564221286261709e-06,
+      "loss": 0.439292311668396,
+      "step": 1108
+    },
+    {
+      "epoch": 1.0862639382428625,
+      "grad_norm": 0.6431173086166382,
+      "learning_rate": 4.55631871199001e-06,
+      "loss": 0.4506379961967468,
+      "step": 1109
+    },
+    {
+      "epoch": 1.0872442102683495,
+      "grad_norm": 0.5269262194633484,
+      "learning_rate": 4.548417254685425e-06,
+      "loss": 0.6768068671226501,
+      "step": 1110
+    },
+    {
+      "epoch": 1.0882244822938365,
+      "grad_norm": 0.9696868658065796,
+      "learning_rate": 4.5405169342398634e-06,
+      "loss": 0.40201422572135925,
+      "step": 1111
+    },
+    {
+      "epoch": 1.0892047543193235,
+      "grad_norm": 0.6789999008178711,
+      "learning_rate": 4.5326177705423725e-06,
+      "loss": 0.4404922127723694,
+      "step": 1112
+    },
+    {
+      "epoch": 1.0901850263448107,
+      "grad_norm": 1.6775376796722412,
+      "learning_rate": 4.524719783479088e-06,
+      "loss": 0.493441104888916,
+      "step": 1113
+    },
+    {
+      "epoch": 1.0911652983702977,
+      "grad_norm": 0.65725177526474,
+      "learning_rate": 4.5168229929331805e-06,
+      "loss": 0.49768978357315063,
+      "step": 1114
+    },
+    {
+      "epoch": 1.0921455703957847,
+      "grad_norm": 0.5085949301719666,
+      "learning_rate": 4.5089274187848144e-06,
+      "loss": 0.47643017768859863,
+      "step": 1115
+    },
+    {
+      "epoch": 1.093125842421272,
+      "grad_norm": 0.5621051788330078,
+      "learning_rate": 4.501033080911087e-06,
+      "loss": 0.5799874663352966,
+      "step": 1116
+    },
+    {
+      "epoch": 1.094106114446759,
+      "grad_norm": 0.6580293774604797,
+      "learning_rate": 4.493139999185984e-06,
+      "loss": 0.4764883518218994,
+      "step": 1117
+    },
+    {
+      "epoch": 1.095086386472246,
+      "grad_norm": 0.7277138829231262,
+      "learning_rate": 4.485248193480328e-06,
+      "loss": 0.5280158519744873,
+      "step": 1118
+    },
+    {
+      "epoch": 1.0960666584977332,
+      "grad_norm": 0.7329255938529968,
+      "learning_rate": 4.477357683661734e-06,
+      "loss": 0.6371022462844849,
+      "step": 1119
+    },
+    {
+      "epoch": 1.0970469305232202,
+      "grad_norm": 0.9120189547538757,
+      "learning_rate": 4.469468489594546e-06,
+      "loss": 0.42339026927948,
+      "step": 1120
+    },
+    {
+      "epoch": 1.0980272025487072,
+      "grad_norm": 0.45920926332473755,
+      "learning_rate": 4.461580631139806e-06,
+      "loss": 0.4648767411708832,
+      "step": 1121
+    },
+    {
+      "epoch": 1.0990074745741945,
+      "grad_norm": 0.6987278461456299,
+      "learning_rate": 4.453694128155187e-06,
+      "loss": 0.4336951673030853,
+      "step": 1122
+    },
+    {
+      "epoch": 1.0999877465996815,
+      "grad_norm": 0.4945792555809021,
+      "learning_rate": 4.445809000494945e-06,
+      "loss": 0.4956054091453552,
+      "step": 1123
+    },
+    {
+      "epoch": 1.1009680186251685,
+      "grad_norm": 0.44705942273139954,
+      "learning_rate": 4.4379252680098875e-06,
+      "loss": 0.4850808084011078,
+      "step": 1124
+    },
+    {
+      "epoch": 1.1019482906506555,
+      "grad_norm": 2.327049493789673,
+      "learning_rate": 4.430042950547298e-06,
+      "loss": 0.48761698603630066,
+      "step": 1125
+    },
+    {
+      "epoch": 1.1029285626761427,
+      "grad_norm": 0.5656946301460266,
+      "learning_rate": 4.4221620679509016e-06,
+      "loss": 0.4539868235588074,
+      "step": 1126
+    },
+    {
+      "epoch": 1.1039088347016297,
+      "grad_norm": 0.9848067164421082,
+      "learning_rate": 4.414282640060809e-06,
+      "loss": 0.46509498357772827,
+      "step": 1127
+    },
+    {
+      "epoch": 1.1048891067271167,
+      "grad_norm": 0.4271516501903534,
+      "learning_rate": 4.4064046867134755e-06,
+      "loss": 0.36360645294189453,
+      "step": 1128
+    },
+    {
+      "epoch": 1.105869378752604,
+      "grad_norm": 0.8738290667533875,
+      "learning_rate": 4.398528227741634e-06,
+      "loss": 0.41328540444374084,
+      "step": 1129
+    },
+    {
+      "epoch": 1.106849650778091,
+      "grad_norm": 0.6625762581825256,
+      "learning_rate": 4.390653282974264e-06,
+      "loss": 0.41522017121315,
+      "step": 1130
+    },
+    {
+      "epoch": 1.107829922803578,
+      "grad_norm": 0.8081544041633606,
+      "learning_rate": 4.382779872236527e-06,
+      "loss": 0.49105754494667053,
+      "step": 1131
+    },
+    {
+      "epoch": 1.108810194829065,
+      "grad_norm": 0.8454148769378662,
+      "learning_rate": 4.374908015349728e-06,
+      "loss": 0.5055080056190491,
+      "step": 1132
+    },
+    {
+      "epoch": 1.1097904668545522,
+      "grad_norm": 0.48247644305229187,
+      "learning_rate": 4.367037732131254e-06,
+      "loss": 0.4331636130809784,
+      "step": 1133
+    },
+    {
+      "epoch": 1.1107707388800392,
+      "grad_norm": 0.8663395643234253,
+      "learning_rate": 4.359169042394537e-06,
+      "loss": 0.4387267827987671,
+      "step": 1134
+    },
+    {
+      "epoch": 1.1117510109055262,
+      "grad_norm": 0.6425599455833435,
+      "learning_rate": 4.3513019659489916e-06,
+      "loss": 0.5296391844749451,
+      "step": 1135
+    },
+    {
+      "epoch": 1.1127312829310134,
+      "grad_norm": 0.5669677257537842,
+      "learning_rate": 4.343436522599975e-06,
+      "loss": 0.5115692615509033,
+      "step": 1136
+    },
+    {
+      "epoch": 1.1137115549565004,
+      "grad_norm": 0.9621993899345398,
+      "learning_rate": 4.33557273214873e-06,
+      "loss": 0.47891098260879517,
+      "step": 1137
+    },
+    {
+      "epoch": 1.1146918269819874,
+      "grad_norm": 0.5834046602249146,
+      "learning_rate": 4.327710614392341e-06,
+      "loss": 0.44402241706848145,
+      "step": 1138
+    },
+    {
+      "epoch": 1.1156720990074747,
+      "grad_norm": 0.6605415344238281,
+      "learning_rate": 4.319850189123681e-06,
+      "loss": 0.4194111227989197,
+      "step": 1139
+    },
+    {
+      "epoch": 1.1166523710329617,
+      "grad_norm": 0.8602020144462585,
+      "learning_rate": 4.3119914761313566e-06,
+      "loss": 0.44912153482437134,
+      "step": 1140
+    },
+    {
+      "epoch": 1.1176326430584487,
+      "grad_norm": 0.5399560332298279,
+      "learning_rate": 4.304134495199675e-06,
+      "loss": 0.38555246591567993,
+      "step": 1141
+    },
+    {
+      "epoch": 1.118612915083936,
+      "grad_norm": 1.1265947818756104,
+      "learning_rate": 4.296279266108569e-06,
+      "loss": 0.4807998239994049,
+      "step": 1142
+    },
+    {
+      "epoch": 1.119593187109423,
+      "grad_norm": 0.4873620569705963,
+      "learning_rate": 4.2884258086335755e-06,
+      "loss": 0.48006612062454224,
+      "step": 1143
+    },
+    {
+      "epoch": 1.12057345913491,
+      "grad_norm": 0.5415102243423462,
+      "learning_rate": 4.28057414254576e-06,
+      "loss": 0.582271933555603,
+      "step": 1144
+    },
+    {
+      "epoch": 1.121553731160397,
+      "grad_norm": 0.5199561715126038,
+      "learning_rate": 4.272724287611684e-06,
+      "loss": 0.5169079303741455,
+      "step": 1145
+    },
+    {
+      "epoch": 1.1225340031858841,
+      "grad_norm": 0.5009452104568481,
+      "learning_rate": 4.264876263593347e-06,
+      "loss": 0.48657190799713135,
+      "step": 1146
+    },
+    {
+      "epoch": 1.1235142752113711,
+      "grad_norm": 0.5618320107460022,
+      "learning_rate": 4.257030090248142e-06,
+      "loss": 0.5304977297782898,
+      "step": 1147
+    },
+    {
+      "epoch": 1.1244945472368582,
+      "grad_norm": 0.7813818454742432,
+      "learning_rate": 4.249185787328798e-06,
+      "loss": 0.4170266091823578,
+      "step": 1148
+    },
+    {
+      "epoch": 1.1254748192623454,
+      "grad_norm": 0.9779244065284729,
+      "learning_rate": 4.241343374583343e-06,
+      "loss": 0.4659832715988159,
+      "step": 1149
+    },
+    {
+      "epoch": 1.1264550912878324,
+      "grad_norm": 0.4752320647239685,
+      "learning_rate": 4.233502871755036e-06,
+      "loss": 0.5189892649650574,
+      "step": 1150
+    },
+    {
+      "epoch": 1.1274353633133194,
+      "grad_norm": 0.5039144158363342,
+      "learning_rate": 4.225664298582339e-06,
+      "loss": 0.44341498613357544,
+      "step": 1151
+    },
+    {
+      "epoch": 1.1284156353388064,
+      "grad_norm": 0.7455763220787048,
+      "learning_rate": 4.217827674798845e-06,
+      "loss": 0.5803717970848083,
+      "step": 1152
+    },
+    {
+      "epoch": 1.1293959073642936,
+      "grad_norm": 0.6575899720191956,
+      "learning_rate": 4.209993020133251e-06,
+      "loss": 0.497933030128479,
+      "step": 1153
+    },
+    {
+      "epoch": 1.1303761793897806,
+      "grad_norm": 0.562065601348877,
+      "learning_rate": 4.202160354309286e-06,
+      "loss": 0.48490002751350403,
+      "step": 1154
+    },
+    {
+      "epoch": 1.1313564514152676,
+      "grad_norm": 0.452420711517334,
+      "learning_rate": 4.194329697045681e-06,
+      "loss": 0.4651474356651306,
+      "step": 1155
+    },
+    {
+      "epoch": 1.1323367234407549,
+      "grad_norm": 1.0015565156936646,
+      "learning_rate": 4.186501068056102e-06,
+      "loss": 0.5368155241012573,
+      "step": 1156
+    },
+    {
+      "epoch": 1.1333169954662419,
+      "grad_norm": 0.40889620780944824,
+      "learning_rate": 4.178674487049116e-06,
+      "loss": 0.43562960624694824,
+      "step": 1157
+    },
+    {
+      "epoch": 1.1342972674917289,
+      "grad_norm": 0.504135251045227,
+      "learning_rate": 4.1708499737281305e-06,
+      "loss": 0.46995699405670166,
+      "step": 1158
+    },
+    {
+      "epoch": 1.135277539517216,
+      "grad_norm": 1.0381016731262207,
+      "learning_rate": 4.163027547791347e-06,
+      "loss": 0.42162442207336426,
+      "step": 1159
+    },
+    {
+      "epoch": 1.1362578115427031,
+      "grad_norm": 0.4613876938819885,
+      "learning_rate": 4.1552072289317185e-06,
+      "loss": 0.3633173704147339,
+      "step": 1160
+    },
+    {
+      "epoch": 1.1372380835681901,
+      "grad_norm": 1.463637113571167,
+      "learning_rate": 4.147389036836881e-06,
+      "loss": 0.4073065519332886,
+      "step": 1161
+    },
+    {
+      "epoch": 1.1382183555936773,
+      "grad_norm": 0.8064992427825928,
+      "learning_rate": 4.139572991189131e-06,
+      "loss": 0.5470760464668274,
+      "step": 1162
+    },
+    {
+      "epoch": 1.1391986276191644,
+      "grad_norm": 0.6078717112541199,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.514742910861969,
+      "step": 1163
+    },
+    {
+      "epoch": 1.1401788996446514,
+      "grad_norm": 0.7199503779411316,
+      "learning_rate": 4.12394741793697e-06,
+      "loss": 0.6024434566497803,
+      "step": 1164
+    },
+    {
+      "epoch": 1.1411591716701386,
+      "grad_norm": 0.743129312992096,
+      "learning_rate": 4.116137929669921e-06,
+      "loss": 0.48926013708114624,
+      "step": 1165
+    },
+    {
+      "epoch": 1.1421394436956256,
+      "grad_norm": 0.49870219826698303,
+      "learning_rate": 4.1083306665245855e-06,
+      "loss": 0.5160895586013794,
+      "step": 1166
+    },
+    {
+      "epoch": 1.1431197157211126,
+      "grad_norm": 0.5480358600616455,
+      "learning_rate": 4.100525648155731e-06,
+      "loss": 0.4644427001476288,
+      "step": 1167
+    },
+    {
+      "epoch": 1.1440999877465996,
+      "grad_norm": 1.1541972160339355,
+      "learning_rate": 4.092722894212488e-06,
+      "loss": 0.6033247709274292,
+      "step": 1168
+    },
+    {
+      "epoch": 1.1450802597720868,
+      "grad_norm": 0.5516543388366699,
+      "learning_rate": 4.084922424338277e-06,
+      "loss": 0.5139232873916626,
+      "step": 1169
+    },
+    {
+      "epoch": 1.1460605317975738,
+      "grad_norm": 0.7862409353256226,
+      "learning_rate": 4.077124258170774e-06,
+      "loss": 0.4857257008552551,
+      "step": 1170
+    },
+    {
+      "epoch": 1.1470408038230608,
+      "grad_norm": 0.5861995220184326,
+      "learning_rate": 4.06932841534185e-06,
+      "loss": 0.4120182991027832,
+      "step": 1171
+    },
+    {
+      "epoch": 1.148021075848548,
+      "grad_norm": 0.5404655933380127,
+      "learning_rate": 4.061534915477533e-06,
+      "loss": 0.39241641759872437,
+      "step": 1172
+    },
+    {
+      "epoch": 1.149001347874035,
+      "grad_norm": 0.4459927976131439,
+      "learning_rate": 4.053743778197951e-06,
+      "loss": 0.4598627984523773,
+      "step": 1173
+    },
+    {
+      "epoch": 1.149981619899522,
+      "grad_norm": 0.6284507513046265,
+      "learning_rate": 4.045955023117276e-06,
+      "loss": 0.4963238537311554,
+      "step": 1174
+    },
+    {
+      "epoch": 1.150961891925009,
+      "grad_norm": 0.6696841716766357,
+      "learning_rate": 4.038168669843698e-06,
+      "loss": 0.5655513405799866,
+      "step": 1175
+    },
+    {
+      "epoch": 1.1519421639504963,
+      "grad_norm": 0.6391484141349792,
+      "learning_rate": 4.030384737979345e-06,
+      "loss": 0.5078308582305908,
+      "step": 1176
+    },
+    {
+      "epoch": 1.1529224359759833,
+      "grad_norm": 0.5808607339859009,
+      "learning_rate": 4.02260324712026e-06,
+      "loss": 0.4951038956642151,
+      "step": 1177
+    },
+    {
+      "epoch": 1.1539027080014703,
+      "grad_norm": 0.558118999004364,
+      "learning_rate": 4.014824216856336e-06,
+      "loss": 0.5149471759796143,
+      "step": 1178
+    },
+    {
+      "epoch": 1.1548829800269576,
+      "grad_norm": 0.4732533097267151,
+      "learning_rate": 4.007047666771274e-06,
+      "loss": 0.4792274534702301,
+      "step": 1179
+    },
+    {
+      "epoch": 1.1558632520524446,
+      "grad_norm": 1.5609807968139648,
+      "learning_rate": 3.999273616442526e-06,
+      "loss": 0.47155439853668213,
+      "step": 1180
+    },
+    {
+      "epoch": 1.1568435240779316,
+      "grad_norm": 0.5191420912742615,
+      "learning_rate": 3.991502085441259e-06,
+      "loss": 0.4385320842266083,
+      "step": 1181
+    },
+    {
+      "epoch": 1.1578237961034188,
+      "grad_norm": 0.5935909748077393,
+      "learning_rate": 3.9837330933322895e-06,
+      "loss": 0.5085321664810181,
+      "step": 1182
+    },
+    {
+      "epoch": 1.1588040681289058,
+      "grad_norm": 0.5560732483863831,
+      "learning_rate": 3.975966659674048e-06,
+      "loss": 0.4215100407600403,
+      "step": 1183
+    },
+    {
+      "epoch": 1.1597843401543928,
+      "grad_norm": 0.4994036555290222,
+      "learning_rate": 3.968202804018519e-06,
+      "loss": 0.4886816143989563,
+      "step": 1184
+    },
+    {
+      "epoch": 1.16076461217988,
+      "grad_norm": 1.0786913633346558,
+      "learning_rate": 3.960441545911205e-06,
+      "loss": 0.42260122299194336,
+      "step": 1185
+    },
+    {
+      "epoch": 1.161744884205367,
+      "grad_norm": 0.49047064781188965,
+      "learning_rate": 3.952682904891058e-06,
+      "loss": 0.4785465598106384,
+      "step": 1186
+    },
+    {
+      "epoch": 1.162725156230854,
+      "grad_norm": 0.5353580117225647,
+      "learning_rate": 3.944926900490452e-06,
+      "loss": 0.5321719646453857,
+      "step": 1187
+    },
+    {
+      "epoch": 1.163705428256341,
+      "grad_norm": 0.66249680519104,
+      "learning_rate": 3.937173552235117e-06,
+      "loss": 0.5973975658416748,
+      "step": 1188
+    },
+    {
+      "epoch": 1.1646857002818283,
+      "grad_norm": 0.6254546046257019,
+      "learning_rate": 3.929422879644099e-06,
+      "loss": 0.4747315049171448,
+      "step": 1189
+    },
+    {
+      "epoch": 1.1656659723073153,
+      "grad_norm": 0.8109750151634216,
+      "learning_rate": 3.921674902229706e-06,
+      "loss": 0.40968257188796997,
+      "step": 1190
+    },
+    {
+      "epoch": 1.1666462443328023,
+      "grad_norm": 0.4648740589618683,
+      "learning_rate": 3.913929639497462e-06,
+      "loss": 0.46284130215644836,
+      "step": 1191
+    },
+    {
+      "epoch": 1.1676265163582895,
+      "grad_norm": 0.4730297029018402,
+      "learning_rate": 3.906187110946058e-06,
+      "loss": 0.46130192279815674,
+      "step": 1192
+    },
+    {
+      "epoch": 1.1686067883837765,
+      "grad_norm": 0.5263428688049316,
+      "learning_rate": 3.898447336067297e-06,
+      "loss": 0.40865224599838257,
+      "step": 1193
+    },
+    {
+      "epoch": 1.1695870604092635,
+      "grad_norm": 0.6213173866271973,
+      "learning_rate": 3.890710334346058e-06,
+      "loss": 0.5853030681610107,
+      "step": 1194
+    },
+    {
+      "epoch": 1.1705673324347505,
+      "grad_norm": 0.5107208490371704,
+      "learning_rate": 3.882976125260229e-06,
+      "loss": 0.4625176191329956,
+      "step": 1195
+    },
+    {
+      "epoch": 1.1715476044602378,
+      "grad_norm": 0.4959002733230591,
+      "learning_rate": 3.875244728280676e-06,
+      "loss": 0.47776761651039124,
+      "step": 1196
+    },
+    {
+      "epoch": 1.1725278764857248,
+      "grad_norm": 1.2493647336959839,
+      "learning_rate": 3.867516162871177e-06,
+      "loss": 0.4697044789791107,
+      "step": 1197
+    },
+    {
+      "epoch": 1.1735081485112118,
+      "grad_norm": 0.47801104187965393,
+      "learning_rate": 3.859790448488394e-06,
+      "loss": 0.5257695913314819,
+      "step": 1198
+    },
+    {
+      "epoch": 1.174488420536699,
+      "grad_norm": 0.8754520416259766,
+      "learning_rate": 3.8520676045817945e-06,
+      "loss": 0.49145370721817017,
+      "step": 1199
+    },
+    {
+      "epoch": 1.175468692562186,
+      "grad_norm": 0.5753970146179199,
+      "learning_rate": 3.844347650593635e-06,
+      "loss": 0.5001497268676758,
+      "step": 1200
+    },
+    {
+      "epoch": 1.176448964587673,
+      "grad_norm": 1.1341899633407593,
+      "learning_rate": 3.8366306059588885e-06,
+      "loss": 0.41793498396873474,
+      "step": 1201
+    },
+    {
+      "epoch": 1.1774292366131602,
+      "grad_norm": 0.8537812829017639,
+      "learning_rate": 3.828916490105206e-06,
+      "loss": 0.5769152045249939,
+      "step": 1202
+    },
+    {
+      "epoch": 1.1784095086386472,
+      "grad_norm": 0.6759680509567261,
+      "learning_rate": 3.821205322452863e-06,
+      "loss": 0.5450162291526794,
+      "step": 1203
+    },
+    {
+      "epoch": 1.1793897806641342,
+      "grad_norm": 0.6544007658958435,
+      "learning_rate": 3.8134971224147204e-06,
+      "loss": 0.5008829832077026,
+      "step": 1204
+    },
+    {
+      "epoch": 1.1803700526896215,
+      "grad_norm": 0.7752823829650879,
+      "learning_rate": 3.8057919093961554e-06,
+      "loss": 0.549090564250946,
+      "step": 1205
+    },
+    {
+      "epoch": 1.1813503247151085,
+      "grad_norm": 0.6783880591392517,
+      "learning_rate": 3.798089702795038e-06,
+      "loss": 0.47776558995246887,
+      "step": 1206
+    },
+    {
+      "epoch": 1.1823305967405955,
+      "grad_norm": 0.7233131527900696,
+      "learning_rate": 3.790390522001662e-06,
+      "loss": 0.5543979406356812,
+      "step": 1207
+    },
+    {
+      "epoch": 1.1833108687660825,
+      "grad_norm": 0.5482250452041626,
+      "learning_rate": 3.782694386398706e-06,
+      "loss": 0.443537175655365,
+      "step": 1208
+    },
+    {
+      "epoch": 1.1842911407915697,
+      "grad_norm": 0.7086508274078369,
+      "learning_rate": 3.775001315361183e-06,
+      "loss": 0.5620536208152771,
+      "step": 1209
+    },
+    {
+      "epoch": 1.1852714128170567,
+      "grad_norm": 0.992607057094574,
+      "learning_rate": 3.7673113282563883e-06,
+      "loss": 0.5365359783172607,
+      "step": 1210
+    },
+    {
+      "epoch": 1.1862516848425437,
+      "grad_norm": 0.7321624159812927,
+      "learning_rate": 3.7596244444438577e-06,
+      "loss": 0.4467947483062744,
+      "step": 1211
+    },
+    {
+      "epoch": 1.187231956868031,
+      "grad_norm": 0.504638135433197,
+      "learning_rate": 3.7519406832753085e-06,
+      "loss": 0.4542618989944458,
+      "step": 1212
+    },
+    {
+      "epoch": 1.188212228893518,
+      "grad_norm": 0.5391073226928711,
+      "learning_rate": 3.7442600640946045e-06,
+      "loss": 0.3927634358406067,
+      "step": 1213
+    },
+    {
+      "epoch": 1.189192500919005,
+      "grad_norm": 0.4917605221271515,
+      "learning_rate": 3.7365826062376933e-06,
+      "loss": 0.40807273983955383,
+      "step": 1214
+    },
+    {
+      "epoch": 1.190172772944492,
+      "grad_norm": 0.5705227255821228,
+      "learning_rate": 3.7289083290325668e-06,
+      "loss": 0.47283175587654114,
+      "step": 1215
+    },
+    {
+      "epoch": 1.1911530449699792,
+      "grad_norm": 0.6518253684043884,
+      "learning_rate": 3.721237251799206e-06,
+      "loss": 0.5323647260665894,
+      "step": 1216
+    },
+    {
+      "epoch": 1.1921333169954662,
+      "grad_norm": 0.6583929061889648,
+      "learning_rate": 3.7135693938495433e-06,
+      "loss": 0.5231755971908569,
+      "step": 1217
+    },
+    {
+      "epoch": 1.1931135890209532,
+      "grad_norm": 0.6381376385688782,
+      "learning_rate": 3.705904774487396e-06,
+      "loss": 0.4587666392326355,
+      "step": 1218
+    },
+    {
+      "epoch": 1.1940938610464404,
+      "grad_norm": 0.7738186120986938,
+      "learning_rate": 3.69824341300844e-06,
+      "loss": 0.5320371985435486,
+      "step": 1219
+    },
+    {
+      "epoch": 1.1950741330719274,
+      "grad_norm": 0.5309854745864868,
+      "learning_rate": 3.6905853287001405e-06,
+      "loss": 0.5248380899429321,
+      "step": 1220
+    },
+    {
+      "epoch": 1.1960544050974145,
+      "grad_norm": 0.4635276794433594,
+      "learning_rate": 3.682930540841717e-06,
+      "loss": 0.43631255626678467,
+      "step": 1221
+    },
+    {
+      "epoch": 1.1970346771229017,
+      "grad_norm": 0.7179809212684631,
+      "learning_rate": 3.675279068704087e-06,
+      "loss": 0.5215682983398438,
+      "step": 1222
+    },
+    {
+      "epoch": 1.1980149491483887,
+      "grad_norm": 0.9172608256340027,
+      "learning_rate": 3.667630931549826e-06,
+      "loss": 0.4576943516731262,
+      "step": 1223
+    },
+    {
+      "epoch": 1.1989952211738757,
+      "grad_norm": 3.4810826778411865,
+      "learning_rate": 3.6599861486331074e-06,
+      "loss": 0.5567039847373962,
+      "step": 1224
+    },
+    {
+      "epoch": 1.199975493199363,
+      "grad_norm": 0.540644645690918,
+      "learning_rate": 3.6523447391996613e-06,
+      "loss": 0.5110781192779541,
+      "step": 1225
+    },
+    {
+      "epoch": 1.20095576522485,
+      "grad_norm": 0.5207131505012512,
+      "learning_rate": 3.6447067224867306e-06,
+      "loss": 0.4437496066093445,
+      "step": 1226
+    },
+    {
+      "epoch": 1.201936037250337,
+      "grad_norm": 0.5812633037567139,
+      "learning_rate": 3.637072117723012e-06,
+      "loss": 0.5272185206413269,
+      "step": 1227
+    },
+    {
+      "epoch": 1.2029163092758242,
+      "grad_norm": 0.546802282333374,
+      "learning_rate": 3.6294409441286133e-06,
+      "loss": 0.5426909923553467,
+      "step": 1228
+    },
+    {
+      "epoch": 1.2038965813013112,
+      "grad_norm": 2.622368574142456,
+      "learning_rate": 3.6218132209150047e-06,
+      "loss": 0.37105417251586914,
+      "step": 1229
+    },
+    {
+      "epoch": 1.2048768533267982,
+      "grad_norm": 0.5903920531272888,
+      "learning_rate": 3.6141889672849728e-06,
+      "loss": 0.4802726209163666,
+      "step": 1230
+    },
+    {
+      "epoch": 1.2058571253522852,
+      "grad_norm": 0.4096813499927521,
+      "learning_rate": 3.606568202432562e-06,
+      "loss": 0.3984874188899994,
+      "step": 1231
+    },
+    {
+      "epoch": 1.2068373973777724,
+      "grad_norm": 0.9228672385215759,
+      "learning_rate": 3.598950945543045e-06,
+      "loss": 0.48613429069519043,
+      "step": 1232
+    },
+    {
+      "epoch": 1.2078176694032594,
+      "grad_norm": 0.4981374144554138,
+      "learning_rate": 3.5913372157928515e-06,
+      "loss": 0.4454062283039093,
+      "step": 1233
+    },
+    {
+      "epoch": 1.2087979414287464,
+      "grad_norm": 0.6354205012321472,
+      "learning_rate": 3.5837270323495416e-06,
+      "loss": 0.5434946417808533,
+      "step": 1234
+    },
+    {
+      "epoch": 1.2097782134542336,
+      "grad_norm": 0.6049605011940002,
+      "learning_rate": 3.5761204143717387e-06,
+      "loss": 0.4497336149215698,
+      "step": 1235
+    },
+    {
+      "epoch": 1.2107584854797206,
+      "grad_norm": 1.4385194778442383,
+      "learning_rate": 3.568517381009099e-06,
+      "loss": 0.48973768949508667,
+      "step": 1236
+    },
+    {
+      "epoch": 1.2117387575052077,
+      "grad_norm": 0.476318895816803,
+      "learning_rate": 3.560917951402245e-06,
+      "loss": 0.4358331263065338,
+      "step": 1237
+    },
+    {
+      "epoch": 1.2127190295306947,
+      "grad_norm": 1.6453044414520264,
+      "learning_rate": 3.553322144682737e-06,
+      "loss": 0.47369587421417236,
+      "step": 1238
+    },
+    {
+      "epoch": 1.2136993015561819,
+      "grad_norm": 1.8284611701965332,
+      "learning_rate": 3.5457299799730047e-06,
+      "loss": 0.4425419867038727,
+      "step": 1239
+    },
+    {
+      "epoch": 1.214679573581669,
+      "grad_norm": 0.7361161112785339,
+      "learning_rate": 3.538141476386317e-06,
+      "loss": 0.48916521668434143,
+      "step": 1240
+    },
+    {
+      "epoch": 1.215659845607156,
+      "grad_norm": 0.5542389154434204,
+      "learning_rate": 3.5305566530267217e-06,
+      "loss": 0.5181790590286255,
+      "step": 1241
+    },
+    {
+      "epoch": 1.2166401176326431,
+      "grad_norm": 0.666354238986969,
+      "learning_rate": 3.5229755289890005e-06,
+      "loss": 0.5331695675849915,
+      "step": 1242
+    },
+    {
+      "epoch": 1.2176203896581301,
+      "grad_norm": 0.6894227862358093,
+      "learning_rate": 3.5153981233586277e-06,
+      "loss": 0.5200810432434082,
+      "step": 1243
+    },
+    {
+      "epoch": 1.2186006616836171,
+      "grad_norm": 0.4610001742839813,
+      "learning_rate": 3.507824455211708e-06,
+      "loss": 0.4264940023422241,
+      "step": 1244
+    },
+    {
+      "epoch": 1.2195809337091044,
+      "grad_norm": 0.45207440853118896,
+      "learning_rate": 3.5002545436149478e-06,
+      "loss": 0.362376868724823,
+      "step": 1245
+    },
+    {
+      "epoch": 1.2205612057345914,
+      "grad_norm": 0.7733064293861389,
+      "learning_rate": 3.4926884076255864e-06,
+      "loss": 0.43666115403175354,
+      "step": 1246
+    },
+    {
+      "epoch": 1.2215414777600784,
+      "grad_norm": 0.4350748658180237,
+      "learning_rate": 3.4851260662913643e-06,
+      "loss": 0.4533810019493103,
+      "step": 1247
+    },
+    {
+      "epoch": 1.2225217497855656,
+      "grad_norm": 3.093027114868164,
+      "learning_rate": 3.477567538650466e-06,
+      "loss": 0.5103939771652222,
+      "step": 1248
+    },
+    {
+      "epoch": 1.2235020218110526,
+      "grad_norm": 0.7805991768836975,
+      "learning_rate": 3.470012843731476e-06,
+      "loss": 0.5876331925392151,
+      "step": 1249
+    },
+    {
+      "epoch": 1.2244822938365396,
+      "grad_norm": 0.8977651596069336,
+      "learning_rate": 3.4624620005533273e-06,
+      "loss": 0.37214231491088867,
+      "step": 1250
+    },
+    {
+      "epoch": 1.2254625658620266,
+      "grad_norm": 0.5769979953765869,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.4435306191444397,
+      "step": 1251
+    },
+    {
+      "epoch": 1.2264428378875138,
+      "grad_norm": 0.5545756816864014,
+      "learning_rate": 3.4473719454467747e-06,
+      "loss": 0.42642131447792053,
+      "step": 1252
+    },
+    {
+      "epoch": 1.2274231099130009,
+      "grad_norm": 0.4968145489692688,
+      "learning_rate": 3.439832771507565e-06,
+      "loss": 0.41257011890411377,
+      "step": 1253
+    },
+    {
+      "epoch": 1.2284033819384879,
+      "grad_norm": 0.53887939453125,
+      "learning_rate": 3.432297525287495e-06,
+      "loss": 0.4947076439857483,
+      "step": 1254
+    },
+    {
+      "epoch": 1.229383653963975,
+      "grad_norm": 2.3414499759674072,
+      "learning_rate": 3.4247662257565372e-06,
+      "loss": 0.49269381165504456,
+      "step": 1255
+    },
+    {
+      "epoch": 1.230363925989462,
+      "grad_norm": 0.5277802348136902,
+      "learning_rate": 3.4172388918747278e-06,
+      "loss": 0.42014068365097046,
+      "step": 1256
+    },
+    {
+      "epoch": 1.231344198014949,
+      "grad_norm": 0.5019338130950928,
+      "learning_rate": 3.4097155425921256e-06,
+      "loss": 0.44505324959754944,
+      "step": 1257
+    },
+    {
+      "epoch": 1.232324470040436,
+      "grad_norm": 0.6259382367134094,
+      "learning_rate": 3.402196196848751e-06,
+      "loss": 0.48224300146102905,
+      "step": 1258
+    },
+    {
+      "epoch": 1.2333047420659233,
+      "grad_norm": 0.6104170083999634,
+      "learning_rate": 3.394680873574546e-06,
+      "loss": 0.4728332757949829,
+      "step": 1259
+    },
+    {
+      "epoch": 1.2342850140914103,
+      "grad_norm": 0.7429087162017822,
+      "learning_rate": 3.3871695916893314e-06,
+      "loss": 0.4325198233127594,
+      "step": 1260
+    },
+    {
+      "epoch": 1.2352652861168973,
+      "grad_norm": 0.5760616660118103,
+      "learning_rate": 3.3796623701027477e-06,
+      "loss": 0.4270816743373871,
+      "step": 1261
+    },
+    {
+      "epoch": 1.2362455581423846,
+      "grad_norm": 0.4719271659851074,
+      "learning_rate": 3.372159227714218e-06,
+      "loss": 0.46249717473983765,
+      "step": 1262
+    },
+    {
+      "epoch": 1.2372258301678716,
+      "grad_norm": 0.513340175151825,
+      "learning_rate": 3.3646601834128924e-06,
+      "loss": 0.4586717486381531,
+      "step": 1263
+    },
+    {
+      "epoch": 1.2382061021933586,
+      "grad_norm": 1.1043552160263062,
+      "learning_rate": 3.357165256077608e-06,
+      "loss": 0.39324066042900085,
+      "step": 1264
+    },
+    {
+      "epoch": 1.2391863742188458,
+      "grad_norm": 0.6222000122070312,
+      "learning_rate": 3.349674464576834e-06,
+      "loss": 0.4483102262020111,
+      "step": 1265
+    },
+    {
+      "epoch": 1.2401666462443328,
+      "grad_norm": 0.45501989126205444,
+      "learning_rate": 3.3421878277686315e-06,
+      "loss": 0.4764709174633026,
+      "step": 1266
+    },
+    {
+      "epoch": 1.2411469182698198,
+      "grad_norm": 0.5954030156135559,
+      "learning_rate": 3.3347053645005965e-06,
+      "loss": 0.5492681860923767,
+      "step": 1267
+    },
+    {
+      "epoch": 1.242127190295307,
+      "grad_norm": 0.6396109461784363,
+      "learning_rate": 3.3272270936098243e-06,
+      "loss": 0.44812387228012085,
+      "step": 1268
+    },
+    {
+      "epoch": 1.243107462320794,
+      "grad_norm": 0.51876300573349,
+      "learning_rate": 3.319753033922849e-06,
+      "loss": 0.5012656450271606,
+      "step": 1269
+    },
+    {
+      "epoch": 1.244087734346281,
+      "grad_norm": 0.6712484359741211,
+      "learning_rate": 3.3122832042556118e-06,
+      "loss": 0.48526322841644287,
+      "step": 1270
+    },
+    {
+      "epoch": 1.2450680063717683,
+      "grad_norm": 0.5675247311592102,
+      "learning_rate": 3.3048176234133967e-06,
+      "loss": 0.43196529150009155,
+      "step": 1271
+    },
+    {
+      "epoch": 1.2460482783972553,
+      "grad_norm": 0.5418819785118103,
+      "learning_rate": 3.297356310190797e-06,
+      "loss": 0.45385661721229553,
+      "step": 1272
+    },
+    {
+      "epoch": 1.2470285504227423,
+      "grad_norm": 0.7696358561515808,
+      "learning_rate": 3.289899283371657e-06,
+      "loss": 0.4452989101409912,
+      "step": 1273
+    },
+    {
+      "epoch": 1.2480088224482293,
+      "grad_norm": 0.5821284055709839,
+      "learning_rate": 3.282446561729035e-06,
+      "loss": 0.4642988443374634,
+      "step": 1274
+    },
+    {
+      "epoch": 1.2489890944737165,
+      "grad_norm": 0.5464370846748352,
+      "learning_rate": 3.274998164025148e-06,
+      "loss": 0.4268155097961426,
+      "step": 1275
+    },
+    {
+      "epoch": 1.2499693664992035,
+      "grad_norm": 0.5983017086982727,
+      "learning_rate": 3.267554109011327e-06,
+      "loss": 0.497368723154068,
+      "step": 1276
+    },
+    {
+      "epoch": 1.2509496385246905,
+      "grad_norm": 0.4701131284236908,
+      "learning_rate": 3.260114415427975e-06,
+      "loss": 0.404036283493042,
+      "step": 1277
+    },
+    {
+      "epoch": 1.2519299105501775,
+      "grad_norm": 0.5047106146812439,
+      "learning_rate": 3.252679102004509e-06,
+      "loss": 0.4623900055885315,
+      "step": 1278
+    },
+    {
+      "epoch": 1.2529101825756648,
+      "grad_norm": 0.7616816759109497,
+      "learning_rate": 3.2452481874593234e-06,
+      "loss": 0.5825682282447815,
+      "step": 1279
+    },
+    {
+      "epoch": 1.2538904546011518,
+      "grad_norm": 0.6160778999328613,
+      "learning_rate": 3.2378216904997345e-06,
+      "loss": 0.4850836992263794,
+      "step": 1280
+    },
+    {
+      "epoch": 1.2548707266266388,
+      "grad_norm": 0.4300895631313324,
+      "learning_rate": 3.230399629821942e-06,
+      "loss": 0.4843904376029968,
+      "step": 1281
+    },
+    {
+      "epoch": 1.255850998652126,
+      "grad_norm": 0.5318914651870728,
+      "learning_rate": 3.2229820241109722e-06,
+      "loss": 0.49085891246795654,
+      "step": 1282
+    },
+    {
+      "epoch": 1.256831270677613,
+      "grad_norm": 0.6523246765136719,
+      "learning_rate": 3.2155688920406415e-06,
+      "loss": 0.41093993186950684,
+      "step": 1283
+    },
+    {
+      "epoch": 1.2578115427031,
+      "grad_norm": 0.7310000061988831,
+      "learning_rate": 3.2081602522734987e-06,
+      "loss": 0.5332520604133606,
+      "step": 1284
+    },
+    {
+      "epoch": 1.2587918147285873,
+      "grad_norm": 0.7484921813011169,
+      "learning_rate": 3.200756123460788e-06,
+      "loss": 0.5165744423866272,
+      "step": 1285
+    },
+    {
+      "epoch": 1.2597720867540743,
+      "grad_norm": 0.5102726221084595,
+      "learning_rate": 3.1933565242423925e-06,
+      "loss": 0.42199623584747314,
+      "step": 1286
+    },
+    {
+      "epoch": 1.2607523587795613,
+      "grad_norm": 0.5948237180709839,
+      "learning_rate": 3.1859614732467957e-06,
+      "loss": 0.5722297430038452,
+      "step": 1287
+    },
+    {
+      "epoch": 1.2617326308050485,
+      "grad_norm": 0.738662600517273,
+      "learning_rate": 3.178570989091028e-06,
+      "loss": 0.52826988697052,
+      "step": 1288
+    },
+    {
+      "epoch": 1.2627129028305355,
+      "grad_norm": 0.4065679609775543,
+      "learning_rate": 3.171185090380628e-06,
+      "loss": 0.44927310943603516,
+      "step": 1289
+    },
+    {
+      "epoch": 1.2636931748560225,
+      "grad_norm": 0.48834338784217834,
+      "learning_rate": 3.1638037957095836e-06,
+      "loss": 0.45962879061698914,
+      "step": 1290
+    },
+    {
+      "epoch": 1.2646734468815097,
+      "grad_norm": 0.4958083927631378,
+      "learning_rate": 3.156427123660297e-06,
+      "loss": 0.4813717007637024,
+      "step": 1291
+    },
+    {
+      "epoch": 1.2656537189069967,
+      "grad_norm": 0.47404876351356506,
+      "learning_rate": 3.1490550928035312e-06,
+      "loss": 0.45314693450927734,
+      "step": 1292
+    },
+    {
+      "epoch": 1.2666339909324837,
+      "grad_norm": 0.5284994840621948,
+      "learning_rate": 3.141687721698363e-06,
+      "loss": 0.4616464078426361,
+      "step": 1293
+    },
+    {
+      "epoch": 1.267614262957971,
+      "grad_norm": 0.4745299220085144,
+      "learning_rate": 3.1343250288921438e-06,
+      "loss": 0.3958478271961212,
+      "step": 1294
+    },
+    {
+      "epoch": 1.268594534983458,
+      "grad_norm": 0.7007741332054138,
+      "learning_rate": 3.12696703292044e-06,
+      "loss": 0.5411830544471741,
+      "step": 1295
+    },
+    {
+      "epoch": 1.269574807008945,
+      "grad_norm": 0.48585832118988037,
+      "learning_rate": 3.119613752307002e-06,
+      "loss": 0.48058146238327026,
+      "step": 1296
+    },
+    {
+      "epoch": 1.270555079034432,
+      "grad_norm": 1.5168288946151733,
+      "learning_rate": 3.1122652055637014e-06,
+      "loss": 0.48774874210357666,
+      "step": 1297
+    },
+    {
+      "epoch": 1.271535351059919,
+      "grad_norm": 0.5308241844177246,
+      "learning_rate": 3.104921411190499e-06,
+      "loss": 0.34147417545318604,
+      "step": 1298
+    },
+    {
+      "epoch": 1.2725156230854062,
+      "grad_norm": 0.8927557468414307,
+      "learning_rate": 3.097582387675385e-06,
+      "loss": 0.44527333974838257,
+      "step": 1299
+    },
+    {
+      "epoch": 1.2734958951108932,
+      "grad_norm": 0.49227195978164673,
+      "learning_rate": 3.0902481534943464e-06,
+      "loss": 0.45440614223480225,
+      "step": 1300
+    },
+    {
+      "epoch": 1.2744761671363802,
+      "grad_norm": 0.735521674156189,
+      "learning_rate": 3.0829187271113035e-06,
+      "loss": 0.4908064603805542,
+      "step": 1301
+    },
+    {
+      "epoch": 1.2754564391618675,
+      "grad_norm": 2.5072503089904785,
+      "learning_rate": 3.075594126978084e-06,
+      "loss": 0.5805734395980835,
+      "step": 1302
+    },
+    {
+      "epoch": 1.2764367111873545,
+      "grad_norm": 0.5549875497817993,
+      "learning_rate": 3.0682743715343565e-06,
+      "loss": 0.3929234743118286,
+      "step": 1303
+    },
+    {
+      "epoch": 1.2774169832128415,
+      "grad_norm": 0.5165079236030579,
+      "learning_rate": 3.0609594792076e-06,
+      "loss": 0.4107758402824402,
+      "step": 1304
+    },
+    {
+      "epoch": 1.2783972552383287,
+      "grad_norm": 0.4315003454685211,
+      "learning_rate": 3.053649468413043e-06,
+      "loss": 0.44753432273864746,
+      "step": 1305
+    },
+    {
+      "epoch": 1.2793775272638157,
+      "grad_norm": 0.5202887058258057,
+      "learning_rate": 3.0463443575536324e-06,
+      "loss": 0.5366567373275757,
+      "step": 1306
+    },
+    {
+      "epoch": 1.2803577992893027,
+      "grad_norm": 0.8168291449546814,
+      "learning_rate": 3.0390441650199727e-06,
+      "loss": 0.409548819065094,
+      "step": 1307
+    },
+    {
+      "epoch": 1.28133807131479,
+      "grad_norm": 0.5473346710205078,
+      "learning_rate": 3.0317489091902936e-06,
+      "loss": 0.45940572023391724,
+      "step": 1308
+    },
+    {
+      "epoch": 1.282318343340277,
+      "grad_norm": 0.469739705324173,
+      "learning_rate": 3.0244586084303908e-06,
+      "loss": 0.44988882541656494,
+      "step": 1309
+    },
+    {
+      "epoch": 1.283298615365764,
+      "grad_norm": 0.9798917770385742,
+      "learning_rate": 3.0171732810935887e-06,
+      "loss": 0.43008363246917725,
+      "step": 1310
+    },
+    {
+      "epoch": 1.2842788873912512,
+      "grad_norm": 0.5055895447731018,
+      "learning_rate": 3.0098929455206905e-06,
+      "loss": 0.40093767642974854,
+      "step": 1311
+    },
+    {
+      "epoch": 1.2852591594167382,
+      "grad_norm": 0.7303021550178528,
+      "learning_rate": 3.00261762003993e-06,
+      "loss": 0.5003647208213806,
+      "step": 1312
+    },
+    {
+      "epoch": 1.2862394314422252,
+      "grad_norm": 0.5726895928382874,
+      "learning_rate": 2.995347322966933e-06,
+      "loss": 0.549673855304718,
+      "step": 1313
+    },
+    {
+      "epoch": 1.2872197034677124,
+      "grad_norm": 0.6732115149497986,
+      "learning_rate": 2.9880820726046613e-06,
+      "loss": 0.4318796694278717,
+      "step": 1314
+    },
+    {
+      "epoch": 1.2881999754931994,
+      "grad_norm": 0.663090705871582,
+      "learning_rate": 2.980821887243377e-06,
+      "loss": 0.5061198472976685,
+      "step": 1315
+    },
+    {
+      "epoch": 1.2891802475186864,
+      "grad_norm": 0.5475381016731262,
+      "learning_rate": 2.9735667851605855e-06,
+      "loss": 0.493241548538208,
+      "step": 1316
+    },
+    {
+      "epoch": 1.2901605195441734,
+      "grad_norm": 0.4337247312068939,
+      "learning_rate": 2.966316784621e-06,
+      "loss": 0.42421814799308777,
+      "step": 1317
+    },
+    {
+      "epoch": 1.2911407915696607,
+      "grad_norm": 0.514087438583374,
+      "learning_rate": 2.959071903876486e-06,
+      "loss": 0.43121546506881714,
+      "step": 1318
+    },
+    {
+      "epoch": 1.2921210635951477,
+      "grad_norm": 0.647024393081665,
+      "learning_rate": 2.951832161166024e-06,
+      "loss": 0.3900883197784424,
+      "step": 1319
+    },
+    {
+      "epoch": 1.2931013356206347,
+      "grad_norm": 0.7032902240753174,
+      "learning_rate": 2.9445975747156548e-06,
+      "loss": 0.5442631244659424,
+      "step": 1320
+    },
+    {
+      "epoch": 1.2940816076461217,
+      "grad_norm": 0.48699951171875,
+      "learning_rate": 2.937368162738445e-06,
+      "loss": 0.39778363704681396,
+      "step": 1321
+    },
+    {
+      "epoch": 1.295061879671609,
+      "grad_norm": 0.6491889953613281,
+      "learning_rate": 2.9301439434344275e-06,
+      "loss": 0.5215015411376953,
+      "step": 1322
+    },
+    {
+      "epoch": 1.296042151697096,
+      "grad_norm": 0.6458990573883057,
+      "learning_rate": 2.9229249349905686e-06,
+      "loss": 0.3307855725288391,
+      "step": 1323
+    },
+    {
+      "epoch": 1.297022423722583,
+      "grad_norm": 0.5985997319221497,
+      "learning_rate": 2.9157111555807103e-06,
+      "loss": 0.39830899238586426,
+      "step": 1324
+    },
+    {
+      "epoch": 1.2980026957480701,
+      "grad_norm": 0.539268970489502,
+      "learning_rate": 2.9085026233655367e-06,
+      "loss": 0.40604180097579956,
+      "step": 1325
+    },
+    {
+      "epoch": 1.2989829677735572,
+      "grad_norm": 0.46744582056999207,
+      "learning_rate": 2.901299356492516e-06,
+      "loss": 0.4269559383392334,
+      "step": 1326
+    },
+    {
+      "epoch": 1.2999632397990442,
+      "grad_norm": 0.5766334533691406,
+      "learning_rate": 2.8941013730958674e-06,
+      "loss": 0.48035719990730286,
+      "step": 1327
+    },
+    {
+      "epoch": 1.3009435118245314,
+      "grad_norm": 0.8332285284996033,
+      "learning_rate": 2.886908691296504e-06,
+      "loss": 0.46889883279800415,
+      "step": 1328
+    },
+    {
+      "epoch": 1.3019237838500184,
+      "grad_norm": 0.7617515921592712,
+      "learning_rate": 2.8797213292019927e-06,
+      "loss": 0.4631807208061218,
+      "step": 1329
+    },
+    {
+      "epoch": 1.3029040558755054,
+      "grad_norm": 2.276003837585449,
+      "learning_rate": 2.8725393049065098e-06,
+      "loss": 0.47461408376693726,
+      "step": 1330
+    },
+    {
+      "epoch": 1.3038843279009926,
+      "grad_norm": 0.541693925857544,
+      "learning_rate": 2.8653626364907918e-06,
+      "loss": 0.46481406688690186,
+      "step": 1331
+    },
+    {
+      "epoch": 1.3048645999264796,
+      "grad_norm": 0.4667285680770874,
+      "learning_rate": 2.8581913420220954e-06,
+      "loss": 0.390758216381073,
+      "step": 1332
+    },
+    {
+      "epoch": 1.3058448719519666,
+      "grad_norm": 0.546916663646698,
+      "learning_rate": 2.851025439554142e-06,
+      "loss": 0.4840351641178131,
+      "step": 1333
+    },
+    {
+      "epoch": 1.3068251439774539,
+      "grad_norm": 0.5572726726531982,
+      "learning_rate": 2.8438649471270876e-06,
+      "loss": 0.3909565210342407,
+      "step": 1334
+    },
+    {
+      "epoch": 1.3078054160029409,
+      "grad_norm": 0.5014473795890808,
+      "learning_rate": 2.8367098827674575e-06,
+      "loss": 0.4729064106941223,
+      "step": 1335
+    },
+    {
+      "epoch": 1.3087856880284279,
+      "grad_norm": 0.38745760917663574,
+      "learning_rate": 2.829560264488126e-06,
+      "loss": 0.3734501600265503,
+      "step": 1336
+    },
+    {
+      "epoch": 1.3097659600539149,
+      "grad_norm": 0.829460084438324,
+      "learning_rate": 2.82241611028824e-06,
+      "loss": 0.6040037274360657,
+      "step": 1337
+    },
+    {
+      "epoch": 1.310746232079402,
+      "grad_norm": 1.5882498025894165,
+      "learning_rate": 2.8152774381532033e-06,
+      "loss": 0.5499056577682495,
+      "step": 1338
+    },
+    {
+      "epoch": 1.3117265041048891,
+      "grad_norm": 0.9027197957038879,
+      "learning_rate": 2.8081442660546126e-06,
+      "loss": 0.3664601743221283,
+      "step": 1339
+    },
+    {
+      "epoch": 1.3127067761303761,
+      "grad_norm": 0.46631717681884766,
+      "learning_rate": 2.8010166119502238e-06,
+      "loss": 0.43814074993133545,
+      "step": 1340
+    },
+    {
+      "epoch": 1.3136870481558631,
+      "grad_norm": 0.4109143316745758,
+      "learning_rate": 2.7938944937838924e-06,
+      "loss": 0.35841983556747437,
+      "step": 1341
+    },
+    {
+      "epoch": 1.3146673201813504,
+      "grad_norm": 0.541437029838562,
+      "learning_rate": 2.7867779294855502e-06,
+      "loss": 0.435494601726532,
+      "step": 1342
+    },
+    {
+      "epoch": 1.3156475922068374,
+      "grad_norm": 0.6770660877227783,
+      "learning_rate": 2.7796669369711294e-06,
+      "loss": 0.5311322808265686,
+      "step": 1343
+    },
+    {
+      "epoch": 1.3166278642323244,
+      "grad_norm": 1.3199328184127808,
+      "learning_rate": 2.7725615341425526e-06,
+      "loss": 0.4997367262840271,
+      "step": 1344
+    },
+    {
+      "epoch": 1.3176081362578116,
+      "grad_norm": 0.5961549878120422,
+      "learning_rate": 2.7654617388876612e-06,
+      "loss": 0.5458905696868896,
+      "step": 1345
+    },
+    {
+      "epoch": 1.3185884082832986,
+      "grad_norm": 0.5188878774642944,
+      "learning_rate": 2.758367569080178e-06,
+      "loss": 0.4911271929740906,
+      "step": 1346
+    },
+    {
+      "epoch": 1.3195686803087856,
+      "grad_norm": 0.6204545497894287,
+      "learning_rate": 2.751279042579672e-06,
+      "loss": 0.5132603049278259,
+      "step": 1347
+    },
+    {
+      "epoch": 1.3205489523342728,
+      "grad_norm": 1.7074207067489624,
+      "learning_rate": 2.744196177231498e-06,
+      "loss": 0.36391371488571167,
+      "step": 1348
+    },
+    {
+      "epoch": 1.3215292243597598,
+      "grad_norm": 0.5542020201683044,
+      "learning_rate": 2.7371189908667604e-06,
+      "loss": 0.47545522451400757,
+      "step": 1349
+    },
+    {
+      "epoch": 1.3225094963852468,
+      "grad_norm": 0.5207110643386841,
+      "learning_rate": 2.7300475013022666e-06,
+      "loss": 0.45872098207473755,
+      "step": 1350
+    },
+    {
+      "epoch": 1.323489768410734,
+      "grad_norm": 0.5920431017875671,
+      "learning_rate": 2.722981726340487e-06,
+      "loss": 0.45546063780784607,
+      "step": 1351
+    },
+    {
+      "epoch": 1.324470040436221,
+      "grad_norm": 0.5728334188461304,
+      "learning_rate": 2.715921683769496e-06,
+      "loss": 0.42139214277267456,
+      "step": 1352
+    },
+    {
+      "epoch": 1.325450312461708,
+      "grad_norm": 0.49201035499572754,
+      "learning_rate": 2.708867391362948e-06,
+      "loss": 0.4226726293563843,
+      "step": 1353
+    },
+    {
+      "epoch": 1.3264305844871953,
+      "grad_norm": 5.129342079162598,
+      "learning_rate": 2.7018188668800123e-06,
+      "loss": 0.4804314970970154,
+      "step": 1354
+    },
+    {
+      "epoch": 1.3274108565126823,
+      "grad_norm": 0.9133296012878418,
+      "learning_rate": 2.694776128065345e-06,
+      "loss": 0.4127538204193115,
+      "step": 1355
+    },
+    {
+      "epoch": 1.3283911285381693,
+      "grad_norm": 0.9942786693572998,
+      "learning_rate": 2.687739192649026e-06,
+      "loss": 0.4222072660923004,
+      "step": 1356
+    },
+    {
+      "epoch": 1.3293714005636565,
+      "grad_norm": 0.5572077631950378,
+      "learning_rate": 2.6807080783465376e-06,
+      "loss": 0.3931940793991089,
+      "step": 1357
+    },
+    {
+      "epoch": 1.3303516725891436,
+      "grad_norm": 0.559248685836792,
+      "learning_rate": 2.673682802858697e-06,
+      "loss": 0.4120192527770996,
+      "step": 1358
+    },
+    {
+      "epoch": 1.3313319446146306,
+      "grad_norm": 0.5777015686035156,
+      "learning_rate": 2.6666633838716317e-06,
+      "loss": 0.5096394419670105,
+      "step": 1359
+    },
+    {
+      "epoch": 1.3323122166401176,
+      "grad_norm": 3.8931586742401123,
+      "learning_rate": 2.659649839056716e-06,
+      "loss": 0.4330391585826874,
+      "step": 1360
+    },
+    {
+      "epoch": 1.3332924886656046,
+      "grad_norm": 0.6878312826156616,
+      "learning_rate": 2.6526421860705474e-06,
+      "loss": 0.46365708112716675,
+      "step": 1361
+    },
+    {
+      "epoch": 1.3342727606910918,
+      "grad_norm": 0.5937561392784119,
+      "learning_rate": 2.645640442554878e-06,
+      "loss": 0.3641710877418518,
+      "step": 1362
+    },
+    {
+      "epoch": 1.3352530327165788,
+      "grad_norm": 0.5706226229667664,
+      "learning_rate": 2.6386446261365874e-06,
+      "loss": 0.4197372794151306,
+      "step": 1363
+    },
+    {
+      "epoch": 1.3362333047420658,
+      "grad_norm": 1.8058125972747803,
+      "learning_rate": 2.63165475442764e-06,
+      "loss": 0.43173325061798096,
+      "step": 1364
+    },
+    {
+      "epoch": 1.337213576767553,
+      "grad_norm": 0.5859666466712952,
+      "learning_rate": 2.6246708450250256e-06,
+      "loss": 0.4868532419204712,
+      "step": 1365
+    },
+    {
+      "epoch": 1.33819384879304,
+      "grad_norm": 0.427780419588089,
+      "learning_rate": 2.617692915510732e-06,
+      "loss": 0.3842759430408478,
+      "step": 1366
+    },
+    {
+      "epoch": 1.339174120818527,
+      "grad_norm": 0.5964685678482056,
+      "learning_rate": 2.6107209834516857e-06,
+      "loss": 0.4351728558540344,
+      "step": 1367
+    },
+    {
+      "epoch": 1.3401543928440143,
+      "grad_norm": 1.3115756511688232,
+      "learning_rate": 2.603755066399718e-06,
+      "loss": 0.4745458662509918,
+      "step": 1368
+    },
+    {
+      "epoch": 1.3411346648695013,
+      "grad_norm": 0.5060911774635315,
+      "learning_rate": 2.5967951818915137e-06,
+      "loss": 0.42621201276779175,
+      "step": 1369
+    },
+    {
+      "epoch": 1.3421149368949883,
+      "grad_norm": 0.5071215629577637,
+      "learning_rate": 2.5898413474485797e-06,
+      "loss": 0.4589035212993622,
+      "step": 1370
+    },
+    {
+      "epoch": 1.3430952089204755,
+      "grad_norm": 0.731763482093811,
+      "learning_rate": 2.5828935805771804e-06,
+      "loss": 0.551716148853302,
+      "step": 1371
+    },
+    {
+      "epoch": 1.3440754809459625,
+      "grad_norm": 0.5851214528083801,
+      "learning_rate": 2.5759518987683154e-06,
+      "loss": 0.4210929870605469,
+      "step": 1372
+    },
+    {
+      "epoch": 1.3450557529714495,
+      "grad_norm": 0.7077540755271912,
+      "learning_rate": 2.5690163194976576e-06,
+      "loss": 0.41452324390411377,
+      "step": 1373
+    },
+    {
+      "epoch": 1.3460360249969368,
+      "grad_norm": 0.5633883476257324,
+      "learning_rate": 2.56208686022552e-06,
+      "loss": 0.47126954793930054,
+      "step": 1374
+    },
+    {
+      "epoch": 1.3470162970224238,
+      "grad_norm": 0.5696465969085693,
+      "learning_rate": 2.5551635383968063e-06,
+      "loss": 0.37556034326553345,
+      "step": 1375
+    },
+    {
+      "epoch": 1.3479965690479108,
+      "grad_norm": 0.4966161549091339,
+      "learning_rate": 2.5482463714409744e-06,
+      "loss": 0.3975297212600708,
+      "step": 1376
+    },
+    {
+      "epoch": 1.348976841073398,
+      "grad_norm": 0.6480564475059509,
+      "learning_rate": 2.5413353767719805e-06,
+      "loss": 0.4691520929336548,
+      "step": 1377
+    },
+    {
+      "epoch": 1.349957113098885,
+      "grad_norm": 0.6175550222396851,
+      "learning_rate": 2.5344305717882487e-06,
+      "loss": 0.37845751643180847,
+      "step": 1378
+    },
+    {
+      "epoch": 1.350937385124372,
+      "grad_norm": 0.4501942992210388,
+      "learning_rate": 2.527531973872617e-06,
+      "loss": 0.38910311460494995,
+      "step": 1379
+    },
+    {
+      "epoch": 1.351917657149859,
+      "grad_norm": 0.8925672173500061,
+      "learning_rate": 2.520639600392295e-06,
+      "loss": 0.4976675510406494,
+      "step": 1380
+    },
+    {
+      "epoch": 1.3528979291753462,
+      "grad_norm": 0.4999285936355591,
+      "learning_rate": 2.5137534686988265e-06,
+      "loss": 0.4947189390659332,
+      "step": 1381
+    },
+    {
+      "epoch": 1.3538782012008332,
+      "grad_norm": 0.7632034420967102,
+      "learning_rate": 2.5068735961280367e-06,
+      "loss": 0.35181957483291626,
+      "step": 1382
+    },
+    {
+      "epoch": 1.3548584732263202,
+      "grad_norm": 6.416927814483643,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.46852946281433105,
+      "step": 1383
+    },
+    {
+      "epoch": 1.3558387452518073,
+      "grad_norm": 0.5609520673751831,
+      "learning_rate": 2.493132697618986e-06,
+      "loss": 0.3739578127861023,
+      "step": 1384
+    },
+    {
+      "epoch": 1.3568190172772945,
+      "grad_norm": 4.267370223999023,
+      "learning_rate": 2.486271706273421e-06,
+      "loss": 0.3682084381580353,
+      "step": 1385
+    },
+    {
+      "epoch": 1.3577992893027815,
+      "grad_norm": 0.6486363410949707,
+      "learning_rate": 2.4794170432358415e-06,
+      "loss": 0.4416516125202179,
+      "step": 1386
+    },
+    {
+      "epoch": 1.3587795613282685,
+      "grad_norm": 0.714850664138794,
+      "learning_rate": 2.4725687257628533e-06,
+      "loss": 0.49150222539901733,
+      "step": 1387
+    },
+    {
+      "epoch": 1.3597598333537557,
+      "grad_norm": 0.48163819313049316,
+      "learning_rate": 2.465726771095086e-06,
+      "loss": 0.45457136631011963,
+      "step": 1388
+    },
+    {
+      "epoch": 1.3607401053792427,
+      "grad_norm": 0.4883973300457001,
+      "learning_rate": 2.4588911964571557e-06,
+      "loss": 0.45393669605255127,
+      "step": 1389
+    },
+    {
+      "epoch": 1.3617203774047297,
+      "grad_norm": 5.476712226867676,
+      "learning_rate": 2.4520620190576094e-06,
+      "loss": 0.4825793504714966,
+      "step": 1390
+    },
+    {
+      "epoch": 1.362700649430217,
+      "grad_norm": 0.7358740568161011,
+      "learning_rate": 2.445239256088898e-06,
+      "loss": 0.42132365703582764,
+      "step": 1391
+    },
+    {
+      "epoch": 1.363680921455704,
+      "grad_norm": 0.4577557146549225,
+      "learning_rate": 2.4384229247273154e-06,
+      "loss": 0.41393178701400757,
+      "step": 1392
+    },
+    {
+      "epoch": 1.364661193481191,
+      "grad_norm": 0.5404036641120911,
+      "learning_rate": 2.4316130421329696e-06,
+      "loss": 0.4729655683040619,
+      "step": 1393
+    },
+    {
+      "epoch": 1.3656414655066782,
+      "grad_norm": 1.5235482454299927,
+      "learning_rate": 2.424809625449729e-06,
+      "loss": 0.4065011143684387,
+      "step": 1394
+    },
+    {
+      "epoch": 1.3666217375321652,
+      "grad_norm": 0.6296142339706421,
+      "learning_rate": 2.418012691805191e-06,
+      "loss": 0.5198254585266113,
+      "step": 1395
+    },
+    {
+      "epoch": 1.3676020095576522,
+      "grad_norm": 0.4470357894897461,
+      "learning_rate": 2.411222258310626e-06,
+      "loss": 0.4393328130245209,
+      "step": 1396
+    },
+    {
+      "epoch": 1.3685822815831394,
+      "grad_norm": 0.5446082353591919,
+      "learning_rate": 2.404438342060941e-06,
+      "loss": 0.598816454410553,
+      "step": 1397
+    },
+    {
+      "epoch": 1.3695625536086264,
+      "grad_norm": 0.4778836965560913,
+      "learning_rate": 2.3976609601346395e-06,
+      "loss": 0.4302779734134674,
+      "step": 1398
+    },
+    {
+      "epoch": 1.3705428256341134,
+      "grad_norm": 0.3977547883987427,
+      "learning_rate": 2.3908901295937713e-06,
+      "loss": 0.4966021776199341,
+      "step": 1399
+    },
+    {
+      "epoch": 1.3715230976596007,
+      "grad_norm": 0.398596853017807,
+      "learning_rate": 2.3841258674838948e-06,
+      "loss": 0.4515528976917267,
+      "step": 1400
+    },
+    {
+      "epoch": 1.3725033696850877,
+      "grad_norm": 0.5539444088935852,
+      "learning_rate": 2.3773681908340284e-06,
+      "loss": 0.6254281997680664,
+      "step": 1401
+    },
+    {
+      "epoch": 1.3734836417105747,
+      "grad_norm": 3.7672197818756104,
+      "learning_rate": 2.3706171166566204e-06,
+      "loss": 0.4072408974170685,
+      "step": 1402
+    },
+    {
+      "epoch": 1.3744639137360617,
+      "grad_norm": 1.2522873878479004,
+      "learning_rate": 2.363872661947488e-06,
+      "loss": 0.49421191215515137,
+      "step": 1403
+    },
+    {
+      "epoch": 1.3754441857615487,
+      "grad_norm": 0.6121653914451599,
+      "learning_rate": 2.3571348436857906e-06,
+      "loss": 0.39878523349761963,
+      "step": 1404
+    },
+    {
+      "epoch": 1.376424457787036,
+      "grad_norm": 0.5867156386375427,
+      "learning_rate": 2.3504036788339763e-06,
+      "loss": 0.570867657661438,
+      "step": 1405
+    },
+    {
+      "epoch": 1.377404729812523,
+      "grad_norm": 0.4516869783401489,
+      "learning_rate": 2.343679184337745e-06,
+      "loss": 0.46326032280921936,
+      "step": 1406
+    },
+    {
+      "epoch": 1.37838500183801,
+      "grad_norm": 0.5083329677581787,
+      "learning_rate": 2.3369613771260006e-06,
+      "loss": 0.556128978729248,
+      "step": 1407
+    },
+    {
+      "epoch": 1.3793652738634972,
+      "grad_norm": 0.5572146773338318,
+      "learning_rate": 2.33025027411082e-06,
+      "loss": 0.4042825698852539,
+      "step": 1408
+    },
+    {
+      "epoch": 1.3803455458889842,
+      "grad_norm": 0.4711366891860962,
+      "learning_rate": 2.323545892187393e-06,
+      "loss": 0.39403754472732544,
+      "step": 1409
+    },
+    {
+      "epoch": 1.3813258179144712,
+      "grad_norm": 0.794239342212677,
+      "learning_rate": 2.3168482482339955e-06,
+      "loss": 0.549608588218689,
+      "step": 1410
+    },
+    {
+      "epoch": 1.3823060899399584,
+      "grad_norm": 0.8837180137634277,
+      "learning_rate": 2.310157359111938e-06,
+      "loss": 0.46352821588516235,
+      "step": 1411
+    },
+    {
+      "epoch": 1.3832863619654454,
+      "grad_norm": 0.7324177622795105,
+      "learning_rate": 2.3034732416655247e-06,
+      "loss": 0.5387417674064636,
+      "step": 1412
+    },
+    {
+      "epoch": 1.3842666339909324,
+      "grad_norm": 0.6411469578742981,
+      "learning_rate": 2.296795912722014e-06,
+      "loss": 0.5159369707107544,
+      "step": 1413
+    },
+    {
+      "epoch": 1.3852469060164196,
+      "grad_norm": 1.0545029640197754,
+      "learning_rate": 2.2901253890915714e-06,
+      "loss": 0.4342513680458069,
+      "step": 1414
+    },
+    {
+      "epoch": 1.3862271780419066,
+      "grad_norm": 0.559226930141449,
+      "learning_rate": 2.2834616875672362e-06,
+      "loss": 0.48139363527297974,
+      "step": 1415
+    },
+    {
+      "epoch": 1.3872074500673937,
+      "grad_norm": 0.7985560894012451,
+      "learning_rate": 2.2768048249248648e-06,
+      "loss": 0.4348641037940979,
+      "step": 1416
+    },
+    {
+      "epoch": 1.3881877220928809,
+      "grad_norm": 1.1319375038146973,
+      "learning_rate": 2.2701548179231048e-06,
+      "loss": 0.4141901731491089,
+      "step": 1417
+    },
+    {
+      "epoch": 1.3891679941183679,
+      "grad_norm": 0.614105224609375,
+      "learning_rate": 2.263511683303339e-06,
+      "loss": 0.5976753830909729,
+      "step": 1418
+    },
+    {
+      "epoch": 1.390148266143855,
+      "grad_norm": 0.5740746259689331,
+      "learning_rate": 2.2568754377896516e-06,
+      "loss": 0.38955509662628174,
+      "step": 1419
+    },
+    {
+      "epoch": 1.3911285381693421,
+      "grad_norm": 0.5201787948608398,
+      "learning_rate": 2.2502460980887807e-06,
+      "loss": 0.548778772354126,
+      "step": 1420
+    },
+    {
+      "epoch": 1.3921088101948291,
+      "grad_norm": 1.040671706199646,
+      "learning_rate": 2.2436236808900846e-06,
+      "loss": 0.5054548382759094,
+      "step": 1421
+    },
+    {
+      "epoch": 1.3930890822203161,
+      "grad_norm": 0.6859439611434937,
+      "learning_rate": 2.2370082028654867e-06,
+      "loss": 0.5040316581726074,
+      "step": 1422
+    },
+    {
+      "epoch": 1.3940693542458031,
+      "grad_norm": 0.6289520263671875,
+      "learning_rate": 2.230399680669449e-06,
+      "loss": 0.48368340730667114,
+      "step": 1423
+    },
+    {
+      "epoch": 1.3950496262712904,
+      "grad_norm": 1.0040881633758545,
+      "learning_rate": 2.223798130938916e-06,
+      "loss": 0.4963570833206177,
+      "step": 1424
+    },
+    {
+      "epoch": 1.3960298982967774,
+      "grad_norm": 0.518243670463562,
+      "learning_rate": 2.2172035702932828e-06,
+      "loss": 0.5369514226913452,
+      "step": 1425
+    },
+    {
+      "epoch": 1.3970101703222644,
+      "grad_norm": 0.5900828242301941,
+      "learning_rate": 2.210616015334344e-06,
+      "loss": 0.4837069809436798,
+      "step": 1426
+    },
+    {
+      "epoch": 1.3979904423477514,
+      "grad_norm": 2.983966827392578,
+      "learning_rate": 2.204035482646267e-06,
+      "loss": 0.4561392664909363,
+      "step": 1427
+    },
+    {
+      "epoch": 1.3989707143732386,
+      "grad_norm": 0.7214106321334839,
+      "learning_rate": 2.1974619887955294e-06,
+      "loss": 0.4364432394504547,
+      "step": 1428
+    },
+    {
+      "epoch": 1.3999509863987256,
+      "grad_norm": 0.4587176740169525,
+      "learning_rate": 2.190895550330899e-06,
+      "loss": 0.4763728380203247,
+      "step": 1429
+    },
+    {
+      "epoch": 1.4009312584242126,
+      "grad_norm": 1.6016161441802979,
+      "learning_rate": 2.1843361837833755e-06,
+      "loss": 0.4800844192504883,
+      "step": 1430
+    },
+    {
+      "epoch": 1.4019115304496998,
+      "grad_norm": 1.2828893661499023,
+      "learning_rate": 2.1777839056661555e-06,
+      "loss": 0.5026894807815552,
+      "step": 1431
+    },
+    {
+      "epoch": 1.4028918024751869,
+      "grad_norm": 0.554818868637085,
+      "learning_rate": 2.171238732474591e-06,
+      "loss": 0.5827724933624268,
+      "step": 1432
+    },
+    {
+      "epoch": 1.4038720745006739,
+      "grad_norm": 0.517440915107727,
+      "learning_rate": 2.1647006806861472e-06,
+      "loss": 0.44045934081077576,
+      "step": 1433
+    },
+    {
+      "epoch": 1.404852346526161,
+      "grad_norm": 0.5633241534233093,
+      "learning_rate": 2.1581697667603634e-06,
+      "loss": 0.4855632781982422,
+      "step": 1434
+    },
+    {
+      "epoch": 1.405832618551648,
+      "grad_norm": 0.5263170599937439,
+      "learning_rate": 2.1516460071388062e-06,
+      "loss": 0.4401085376739502,
+      "step": 1435
+    },
+    {
+      "epoch": 1.406812890577135,
+      "grad_norm": 0.7453410625457764,
+      "learning_rate": 2.145129418245035e-06,
+      "loss": 0.4680941700935364,
+      "step": 1436
+    },
+    {
+      "epoch": 1.4077931626026223,
+      "grad_norm": 5.099278450012207,
+      "learning_rate": 2.1386200164845527e-06,
+      "loss": 0.49586793780326843,
+      "step": 1437
+    },
+    {
+      "epoch": 1.4087734346281093,
+      "grad_norm": 0.9944934248924255,
+      "learning_rate": 2.132117818244771e-06,
+      "loss": 0.4978284239768982,
+      "step": 1438
+    },
+    {
+      "epoch": 1.4097537066535963,
+      "grad_norm": 0.7622154355049133,
+      "learning_rate": 2.125622839894964e-06,
+      "loss": 0.46654218435287476,
+      "step": 1439
+    },
+    {
+      "epoch": 1.4107339786790836,
+      "grad_norm": 0.567703366279602,
+      "learning_rate": 2.1191350977862362e-06,
+      "loss": 0.36492305994033813,
+      "step": 1440
+    },
+    {
+      "epoch": 1.4117142507045706,
+      "grad_norm": 0.40764304995536804,
+      "learning_rate": 2.1126546082514665e-06,
+      "loss": 0.4005526304244995,
+      "step": 1441
+    },
+    {
+      "epoch": 1.4126945227300576,
+      "grad_norm": 0.7328307628631592,
+      "learning_rate": 2.106181387605283e-06,
+      "loss": 0.5773950219154358,
+      "step": 1442
+    },
+    {
+      "epoch": 1.4136747947555446,
+      "grad_norm": 0.5250221490859985,
+      "learning_rate": 2.09971545214401e-06,
+      "loss": 0.4971471130847931,
+      "step": 1443
+    },
+    {
+      "epoch": 1.4146550667810318,
+      "grad_norm": 0.4912668466567993,
+      "learning_rate": 2.0932568181456314e-06,
+      "loss": 0.4325234293937683,
+      "step": 1444
+    },
+    {
+      "epoch": 1.4156353388065188,
+      "grad_norm": 0.5247968435287476,
+      "learning_rate": 2.086805501869749e-06,
+      "loss": 0.37813645601272583,
+      "step": 1445
+    },
+    {
+      "epoch": 1.4166156108320058,
+      "grad_norm": 0.4732372462749481,
+      "learning_rate": 2.080361519557548e-06,
+      "loss": 0.36437588930130005,
+      "step": 1446
+    },
+    {
+      "epoch": 1.4175958828574928,
+      "grad_norm": 0.7685982584953308,
+      "learning_rate": 2.073924887431744e-06,
+      "loss": 0.5005449056625366,
+      "step": 1447
+    },
+    {
+      "epoch": 1.41857615488298,
+      "grad_norm": 0.5845059752464294,
+      "learning_rate": 2.0674956216965484e-06,
+      "loss": 0.3143443167209625,
+      "step": 1448
+    },
+    {
+      "epoch": 1.419556426908467,
+      "grad_norm": 0.766771137714386,
+      "learning_rate": 2.061073738537635e-06,
+      "loss": 0.445101261138916,
+      "step": 1449
+    },
+    {
+      "epoch": 1.420536698933954,
+      "grad_norm": 0.518658459186554,
+      "learning_rate": 2.054659254122084e-06,
+      "loss": 0.49086207151412964,
+      "step": 1450
+    },
+    {
+      "epoch": 1.4215169709594413,
+      "grad_norm": 0.5539177060127258,
+      "learning_rate": 2.0482521845983522e-06,
+      "loss": 0.5017172694206238,
+      "step": 1451
+    },
+    {
+      "epoch": 1.4224972429849283,
+      "grad_norm": 0.42391201853752136,
+      "learning_rate": 2.0418525460962283e-06,
+      "loss": 0.34360209107398987,
+      "step": 1452
+    },
+    {
+      "epoch": 1.4234775150104153,
+      "grad_norm": 0.5405973792076111,
+      "learning_rate": 2.0354603547267985e-06,
+      "loss": 0.45871347188949585,
+      "step": 1453
+    },
+    {
+      "epoch": 1.4244577870359025,
+      "grad_norm": 0.857297420501709,
+      "learning_rate": 2.029075626582393e-06,
+      "loss": 0.39155319333076477,
+      "step": 1454
+    },
+    {
+      "epoch": 1.4254380590613895,
+      "grad_norm": 0.6033921837806702,
+      "learning_rate": 2.0226983777365604e-06,
+      "loss": 0.49306613206863403,
+      "step": 1455
+    },
+    {
+      "epoch": 1.4264183310868765,
+      "grad_norm": 0.6118445992469788,
+      "learning_rate": 2.0163286242440155e-06,
+      "loss": 0.5592046976089478,
+      "step": 1456
+    },
+    {
+      "epoch": 1.4273986031123638,
+      "grad_norm": 0.9509575366973877,
+      "learning_rate": 2.009966382140606e-06,
+      "loss": 0.3417438268661499,
+      "step": 1457
+    },
+    {
+      "epoch": 1.4283788751378508,
+      "grad_norm": 10.196175575256348,
+      "learning_rate": 2.0036116674432653e-06,
+      "loss": 0.5368284583091736,
+      "step": 1458
+    },
+    {
+      "epoch": 1.4293591471633378,
+      "grad_norm": 0.762797474861145,
+      "learning_rate": 1.9972644961499853e-06,
+      "loss": 0.45527252554893494,
+      "step": 1459
+    },
+    {
+      "epoch": 1.430339419188825,
+      "grad_norm": 0.9837111830711365,
+      "learning_rate": 1.990924884239758e-06,
+      "loss": 0.507952868938446,
+      "step": 1460
+    },
+    {
+      "epoch": 1.431319691214312,
+      "grad_norm": 0.48013070225715637,
+      "learning_rate": 1.9845928476725522e-06,
+      "loss": 0.42762818932533264,
+      "step": 1461
+    },
+    {
+      "epoch": 1.432299963239799,
+      "grad_norm": 0.5173972249031067,
+      "learning_rate": 1.9782684023892613e-06,
+      "loss": 0.49466589093208313,
+      "step": 1462
+    },
+    {
+      "epoch": 1.4332802352652863,
+      "grad_norm": 0.547926127910614,
+      "learning_rate": 1.971951564311668e-06,
+      "loss": 0.41896069049835205,
+      "step": 1463
+    },
+    {
+      "epoch": 1.4342605072907733,
+      "grad_norm": 0.8163740634918213,
+      "learning_rate": 1.9656423493424047e-06,
+      "loss": 0.41507893800735474,
+      "step": 1464
+    },
+    {
+      "epoch": 1.4352407793162603,
+      "grad_norm": 0.7145270109176636,
+      "learning_rate": 1.959340773364911e-06,
+      "loss": 0.5674030184745789,
+      "step": 1465
+    },
+    {
+      "epoch": 1.4362210513417473,
+      "grad_norm": 1.2919292449951172,
+      "learning_rate": 1.953046852243401e-06,
+      "loss": 0.3920708894729614,
+      "step": 1466
+    },
+    {
+      "epoch": 1.4372013233672343,
+      "grad_norm": 1.0375070571899414,
+      "learning_rate": 1.946760601822809e-06,
+      "loss": 0.4383500814437866,
+      "step": 1467
+    },
+    {
+      "epoch": 1.4381815953927215,
+      "grad_norm": 0.6753938794136047,
+      "learning_rate": 1.9404820379287677e-06,
+      "loss": 0.4244039058685303,
+      "step": 1468
+    },
+    {
+      "epoch": 1.4391618674182085,
+      "grad_norm": 0.5957536697387695,
+      "learning_rate": 1.9342111763675512e-06,
+      "loss": 0.4169045686721802,
+      "step": 1469
+    },
+    {
+      "epoch": 1.4401421394436955,
+      "grad_norm": 0.6528060436248779,
+      "learning_rate": 1.927948032926047e-06,
+      "loss": 0.44928866624832153,
+      "step": 1470
+    },
+    {
+      "epoch": 1.4411224114691827,
+      "grad_norm": 0.4714356064796448,
+      "learning_rate": 1.9216926233717087e-06,
+      "loss": 0.46567779779434204,
+      "step": 1471
+    },
+    {
+      "epoch": 1.4421026834946697,
+      "grad_norm": 0.5790466666221619,
+      "learning_rate": 1.915444963452527e-06,
+      "loss": 0.4531933665275574,
+      "step": 1472
+    },
+    {
+      "epoch": 1.4430829555201568,
+      "grad_norm": 0.5103557109832764,
+      "learning_rate": 1.9092050688969736e-06,
+      "loss": 0.4188461899757385,
+      "step": 1473
+    },
+    {
+      "epoch": 1.444063227545644,
+      "grad_norm": 0.4897090792655945,
+      "learning_rate": 1.9029729554139798e-06,
+      "loss": 0.4001561999320984,
+      "step": 1474
+    },
+    {
+      "epoch": 1.445043499571131,
+      "grad_norm": 0.5606241226196289,
+      "learning_rate": 1.8967486386928819e-06,
+      "loss": 0.5655632019042969,
+      "step": 1475
+    },
+    {
+      "epoch": 1.446023771596618,
+      "grad_norm": 0.6278600096702576,
+      "learning_rate": 1.89053213440339e-06,
+      "loss": 0.5264230370521545,
+      "step": 1476
+    },
+    {
+      "epoch": 1.4470040436221052,
+      "grad_norm": 0.6341464519500732,
+      "learning_rate": 1.8843234581955444e-06,
+      "loss": 0.41547489166259766,
+      "step": 1477
+    },
+    {
+      "epoch": 1.4479843156475922,
+      "grad_norm": 1.0373629331588745,
+      "learning_rate": 1.8781226256996838e-06,
+      "loss": 0.4739918112754822,
+      "step": 1478
+    },
+    {
+      "epoch": 1.4489645876730792,
+      "grad_norm": 0.9013898968696594,
+      "learning_rate": 1.8719296525263925e-06,
+      "loss": 0.41171786189079285,
+      "step": 1479
+    },
+    {
+      "epoch": 1.4499448596985665,
+      "grad_norm": 0.42370179295539856,
+      "learning_rate": 1.865744554266477e-06,
+      "loss": 0.4420316815376282,
+      "step": 1480
+    },
+    {
+      "epoch": 1.4509251317240535,
+      "grad_norm": 0.7803822159767151,
+      "learning_rate": 1.859567346490913e-06,
+      "loss": 0.535688042640686,
+      "step": 1481
+    },
+    {
+      "epoch": 1.4519054037495405,
+      "grad_norm": 0.502344012260437,
+      "learning_rate": 1.8533980447508138e-06,
+      "loss": 0.4048832654953003,
+      "step": 1482
+    },
+    {
+      "epoch": 1.4528856757750277,
+      "grad_norm": 0.6584186553955078,
+      "learning_rate": 1.8472366645773892e-06,
+      "loss": 0.393306702375412,
+      "step": 1483
+    },
+    {
+      "epoch": 1.4538659478005147,
+      "grad_norm": 0.5892468094825745,
+      "learning_rate": 1.8410832214819046e-06,
+      "loss": 0.39462411403656006,
+      "step": 1484
+    },
+    {
+      "epoch": 1.4548462198260017,
+      "grad_norm": 0.4266345500946045,
+      "learning_rate": 1.8349377309556487e-06,
+      "loss": 0.3833182454109192,
+      "step": 1485
+    },
+    {
+      "epoch": 1.4558264918514887,
+      "grad_norm": 0.6392761468887329,
+      "learning_rate": 1.8288002084698842e-06,
+      "loss": 0.40881553292274475,
+      "step": 1486
+    },
+    {
+      "epoch": 1.456806763876976,
+      "grad_norm": 0.5003628134727478,
+      "learning_rate": 1.8226706694758194e-06,
+      "loss": 0.4233506917953491,
+      "step": 1487
+    },
+    {
+      "epoch": 1.457787035902463,
+      "grad_norm": 1.2619991302490234,
+      "learning_rate": 1.8165491294045596e-06,
+      "loss": 0.49281424283981323,
+      "step": 1488
+    },
+    {
+      "epoch": 1.45876730792795,
+      "grad_norm": 0.5291329026222229,
+      "learning_rate": 1.810435603667075e-06,
+      "loss": 0.49675044417381287,
+      "step": 1489
+    },
+    {
+      "epoch": 1.459747579953437,
+      "grad_norm": 0.9046873450279236,
+      "learning_rate": 1.8043301076541575e-06,
+      "loss": 0.5736644268035889,
+      "step": 1490
+    },
+    {
+      "epoch": 1.4607278519789242,
+      "grad_norm": 0.5413832068443298,
+      "learning_rate": 1.798232656736389e-06,
+      "loss": 0.5760975480079651,
+      "step": 1491
+    },
+    {
+      "epoch": 1.4617081240044112,
+      "grad_norm": 0.8667757511138916,
+      "learning_rate": 1.792143266264092e-06,
+      "loss": 0.4172160029411316,
+      "step": 1492
+    },
+    {
+      "epoch": 1.4626883960298982,
+      "grad_norm": 0.5858352780342102,
+      "learning_rate": 1.7860619515673034e-06,
+      "loss": 0.4682886600494385,
+      "step": 1493
+    },
+    {
+      "epoch": 1.4636686680553854,
+      "grad_norm": 0.43407562375068665,
+      "learning_rate": 1.7799887279557238e-06,
+      "loss": 0.42815470695495605,
+      "step": 1494
+    },
+    {
+      "epoch": 1.4646489400808724,
+      "grad_norm": 0.5090461373329163,
+      "learning_rate": 1.7739236107186858e-06,
+      "loss": 0.4821891784667969,
+      "step": 1495
+    },
+    {
+      "epoch": 1.4656292121063594,
+      "grad_norm": 0.9978723526000977,
+      "learning_rate": 1.767866615125114e-06,
+      "loss": 0.36190080642700195,
+      "step": 1496
+    },
+    {
+      "epoch": 1.4666094841318467,
+      "grad_norm": 0.6038608551025391,
+      "learning_rate": 1.7618177564234907e-06,
+      "loss": 0.5596791505813599,
+      "step": 1497
+    },
+    {
+      "epoch": 1.4675897561573337,
+      "grad_norm": 0.6541353464126587,
+      "learning_rate": 1.7557770498418087e-06,
+      "loss": 0.4910792410373688,
+      "step": 1498
+    },
+    {
+      "epoch": 1.4685700281828207,
+      "grad_norm": 0.5470070242881775,
+      "learning_rate": 1.7497445105875377e-06,
+      "loss": 0.3734048306941986,
+      "step": 1499
+    },
+    {
+      "epoch": 1.469550300208308,
+      "grad_norm": 0.4323618412017822,
+      "learning_rate": 1.7437201538475916e-06,
+      "loss": 0.437979519367218,
+      "step": 1500
+    },
+    {
+      "epoch": 1.470530572233795,
+      "grad_norm": 0.6428908705711365,
+      "learning_rate": 1.7377039947882802e-06,
+      "loss": 0.437639057636261,
+      "step": 1501
+    },
+    {
+      "epoch": 1.471510844259282,
+      "grad_norm": 0.5872750878334045,
+      "learning_rate": 1.7316960485552758e-06,
+      "loss": 0.42362987995147705,
+      "step": 1502
+    },
+    {
+      "epoch": 1.4724911162847691,
+      "grad_norm": 0.7475934028625488,
+      "learning_rate": 1.7256963302735752e-06,
+      "loss": 0.45222389698028564,
+      "step": 1503
+    },
+    {
+      "epoch": 1.4734713883102561,
+      "grad_norm": 0.39684009552001953,
+      "learning_rate": 1.7197048550474643e-06,
+      "loss": 0.38549649715423584,
+      "step": 1504
+    },
+    {
+      "epoch": 1.4744516603357432,
+      "grad_norm": 0.48198840022087097,
+      "learning_rate": 1.7137216379604727e-06,
+      "loss": 0.4066298007965088,
+      "step": 1505
+    },
+    {
+      "epoch": 1.4754319323612302,
+      "grad_norm": 0.6070873141288757,
+      "learning_rate": 1.7077466940753446e-06,
+      "loss": 0.4166245460510254,
+      "step": 1506
+    },
+    {
+      "epoch": 1.4764122043867174,
+      "grad_norm": 0.3964442014694214,
+      "learning_rate": 1.7017800384339928e-06,
+      "loss": 0.4458141028881073,
+      "step": 1507
+    },
+    {
+      "epoch": 1.4773924764122044,
+      "grad_norm": 0.4519728720188141,
+      "learning_rate": 1.6958216860574661e-06,
+      "loss": 0.35702240467071533,
+      "step": 1508
+    },
+    {
+      "epoch": 1.4783727484376914,
+      "grad_norm": 0.7520145773887634,
+      "learning_rate": 1.6898716519459074e-06,
+      "loss": 0.4385332763195038,
+      "step": 1509
+    },
+    {
+      "epoch": 1.4793530204631784,
+      "grad_norm": 0.7575941681861877,
+      "learning_rate": 1.683929951078524e-06,
+      "loss": 0.5006380081176758,
+      "step": 1510
+    },
+    {
+      "epoch": 1.4803332924886656,
+      "grad_norm": 0.5183686017990112,
+      "learning_rate": 1.6779965984135376e-06,
+      "loss": 0.5007272958755493,
+      "step": 1511
+    },
+    {
+      "epoch": 1.4813135645141526,
+      "grad_norm": 0.5692034363746643,
+      "learning_rate": 1.6720716088881595e-06,
+      "loss": 0.44675499200820923,
+      "step": 1512
+    },
+    {
+      "epoch": 1.4822938365396396,
+      "grad_norm": 0.4922841787338257,
+      "learning_rate": 1.6661549974185426e-06,
+      "loss": 0.368874192237854,
+      "step": 1513
+    },
+    {
+      "epoch": 1.4832741085651269,
+      "grad_norm": 2.0443663597106934,
+      "learning_rate": 1.6602467788997483e-06,
+      "loss": 0.4594436287879944,
+      "step": 1514
+    },
+    {
+      "epoch": 1.4842543805906139,
+      "grad_norm": 0.49635154008865356,
+      "learning_rate": 1.6543469682057105e-06,
+      "loss": 0.42359083890914917,
+      "step": 1515
+    },
+    {
+      "epoch": 1.4852346526161009,
+      "grad_norm": 0.6353366374969482,
+      "learning_rate": 1.648455580189194e-06,
+      "loss": 0.5048133134841919,
+      "step": 1516
+    },
+    {
+      "epoch": 1.486214924641588,
+      "grad_norm": 0.5531353950500488,
+      "learning_rate": 1.6425726296817634e-06,
+      "loss": 0.4163086414337158,
+      "step": 1517
+    },
+    {
+      "epoch": 1.4871951966670751,
+      "grad_norm": 0.39068838953971863,
+      "learning_rate": 1.6366981314937374e-06,
+      "loss": 0.3796992897987366,
+      "step": 1518
+    },
+    {
+      "epoch": 1.4881754686925621,
+      "grad_norm": 0.6331008672714233,
+      "learning_rate": 1.6308321004141609e-06,
+      "loss": 0.4423547089099884,
+      "step": 1519
+    },
+    {
+      "epoch": 1.4891557407180493,
+      "grad_norm": 0.393364816904068,
+      "learning_rate": 1.624974551210759e-06,
+      "loss": 0.3307700455188751,
+      "step": 1520
+    },
+    {
+      "epoch": 1.4901360127435364,
+      "grad_norm": 0.5085770487785339,
+      "learning_rate": 1.6191254986299044e-06,
+      "loss": 0.4397742748260498,
+      "step": 1521
+    },
+    {
+      "epoch": 1.4911162847690234,
+      "grad_norm": 0.4870603382587433,
+      "learning_rate": 1.6132849573965781e-06,
+      "loss": 0.37178534269332886,
+      "step": 1522
+    },
+    {
+      "epoch": 1.4920965567945106,
+      "grad_norm": 0.8178430199623108,
+      "learning_rate": 1.6074529422143398e-06,
+      "loss": 0.4749937057495117,
+      "step": 1523
+    },
+    {
+      "epoch": 1.4930768288199976,
+      "grad_norm": 1.0519657135009766,
+      "learning_rate": 1.6016294677652772e-06,
+      "loss": 0.540347695350647,
+      "step": 1524
+    },
+    {
+      "epoch": 1.4940571008454846,
+      "grad_norm": 0.6694144606590271,
+      "learning_rate": 1.5958145487099829e-06,
+      "loss": 0.45180442929267883,
+      "step": 1525
+    },
+    {
+      "epoch": 1.4950373728709718,
+      "grad_norm": 0.4629520773887634,
+      "learning_rate": 1.5900081996875083e-06,
+      "loss": 0.40680521726608276,
+      "step": 1526
+    },
+    {
+      "epoch": 1.4960176448964588,
+      "grad_norm": 0.590732753276825,
+      "learning_rate": 1.5842104353153286e-06,
+      "loss": 0.5453869104385376,
+      "step": 1527
+    },
+    {
+      "epoch": 1.4969979169219458,
+      "grad_norm": 0.42295655608177185,
+      "learning_rate": 1.5784212701893088e-06,
+      "loss": 0.37142181396484375,
+      "step": 1528
+    },
+    {
+      "epoch": 1.4979781889474328,
+      "grad_norm": 0.41932907700538635,
+      "learning_rate": 1.5726407188836672e-06,
+      "loss": 0.394989937543869,
+      "step": 1529
+    },
+    {
+      "epoch": 1.4989584609729198,
+      "grad_norm": 0.5797555446624756,
+      "learning_rate": 1.566868795950932e-06,
+      "loss": 0.38030505180358887,
+      "step": 1530
+    },
+    {
+      "epoch": 1.499938732998407,
+      "grad_norm": 0.5518311858177185,
+      "learning_rate": 1.561105515921915e-06,
+      "loss": 0.5279092788696289,
+      "step": 1531
+    },
+    {
+      "epoch": 1.500919005023894,
+      "grad_norm": 8.930876731872559,
+      "learning_rate": 1.5553508933056665e-06,
+      "loss": 0.40353113412857056,
+      "step": 1532
+    },
+    {
+      "epoch": 1.501899277049381,
+      "grad_norm": 0.4371642768383026,
+      "learning_rate": 1.549604942589441e-06,
+      "loss": 0.4437949061393738,
+      "step": 1533
+    },
+    {
+      "epoch": 1.5028795490748683,
+      "grad_norm": 0.5751242637634277,
+      "learning_rate": 1.543867678238663e-06,
+      "loss": 0.5385478734970093,
+      "step": 1534
+    },
+    {
+      "epoch": 1.5038598211003553,
+      "grad_norm": 0.673567533493042,
+      "learning_rate": 1.5381391146968866e-06,
+      "loss": 0.5047478675842285,
+      "step": 1535
+    },
+    {
+      "epoch": 1.5048400931258423,
+      "grad_norm": 0.5048635005950928,
+      "learning_rate": 1.5324192663857673e-06,
+      "loss": 0.4440262019634247,
+      "step": 1536
+    },
+    {
+      "epoch": 1.5058203651513296,
+      "grad_norm": 0.5095667839050293,
+      "learning_rate": 1.5267081477050132e-06,
+      "loss": 0.402218759059906,
+      "step": 1537
+    },
+    {
+      "epoch": 1.5068006371768166,
+      "grad_norm": 0.5917513966560364,
+      "learning_rate": 1.521005773032362e-06,
+      "loss": 0.5613376498222351,
+      "step": 1538
+    },
+    {
+      "epoch": 1.5077809092023036,
+      "grad_norm": 0.834902822971344,
+      "learning_rate": 1.5153121567235334e-06,
+      "loss": 0.4810364246368408,
+      "step": 1539
+    },
+    {
+      "epoch": 1.5087611812277908,
+      "grad_norm": 0.589628279209137,
+      "learning_rate": 1.5096273131121997e-06,
+      "loss": 0.41616058349609375,
+      "step": 1540
+    },
+    {
+      "epoch": 1.5097414532532778,
+      "grad_norm": 0.42739951610565186,
+      "learning_rate": 1.5039512565099468e-06,
+      "loss": 0.4171817898750305,
+      "step": 1541
+    },
+    {
+      "epoch": 1.5107217252787648,
+      "grad_norm": 0.7249782085418701,
+      "learning_rate": 1.4982840012062428e-06,
+      "loss": 0.36407607793807983,
+      "step": 1542
+    },
+    {
+      "epoch": 1.511701997304252,
+      "grad_norm": 0.615429699420929,
+      "learning_rate": 1.4926255614683931e-06,
+      "loss": 0.4324997067451477,
+      "step": 1543
+    },
+    {
+      "epoch": 1.512682269329739,
+      "grad_norm": 0.5160788297653198,
+      "learning_rate": 1.4869759515415167e-06,
+      "loss": 0.49377626180648804,
+      "step": 1544
+    },
+    {
+      "epoch": 1.513662541355226,
+      "grad_norm": 0.4412658214569092,
+      "learning_rate": 1.4813351856484981e-06,
+      "loss": 0.42214444279670715,
+      "step": 1545
+    },
+    {
+      "epoch": 1.5146428133807133,
+      "grad_norm": 0.5633094310760498,
+      "learning_rate": 1.4757032779899588e-06,
+      "loss": 0.4925824999809265,
+      "step": 1546
+    },
+    {
+      "epoch": 1.5156230854062,
+      "grad_norm": 0.6167675256729126,
+      "learning_rate": 1.470080242744218e-06,
+      "loss": 0.3916054964065552,
+      "step": 1547
+    },
+    {
+      "epoch": 1.5166033574316873,
+      "grad_norm": 0.5591668486595154,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.43860334157943726,
+      "step": 1548
+    },
+    {
+      "epoch": 1.5175836294571745,
+      "grad_norm": 0.7692162394523621,
+      "learning_rate": 1.458860846092705e-06,
+      "loss": 0.4547274708747864,
+      "step": 1549
+    },
+    {
+      "epoch": 1.5185639014826613,
+      "grad_norm": 0.7949987053871155,
+      "learning_rate": 1.453264512931748e-06,
+      "loss": 0.41712847352027893,
+      "step": 1550
+    },
+    {
+      "epoch": 1.5195441735081485,
+      "grad_norm": 0.5771681070327759,
+      "learning_rate": 1.4476771086731567e-06,
+      "loss": 0.47331464290618896,
+      "step": 1551
+    },
+    {
+      "epoch": 1.5205244455336357,
+      "grad_norm": 0.7947396636009216,
+      "learning_rate": 1.4420986473832127e-06,
+      "loss": 0.4071407914161682,
+      "step": 1552
+    },
+    {
+      "epoch": 1.5215047175591225,
+      "grad_norm": 0.40888887643814087,
+      "learning_rate": 1.4365291431056871e-06,
+      "loss": 0.41103702783584595,
+      "step": 1553
+    },
+    {
+      "epoch": 1.5224849895846098,
+      "grad_norm": 1.4202901124954224,
+      "learning_rate": 1.4309686098617975e-06,
+      "loss": 0.5387922525405884,
+      "step": 1554
+    },
+    {
+      "epoch": 1.5234652616100968,
+      "grad_norm": 0.4778110384941101,
+      "learning_rate": 1.4254170616501828e-06,
+      "loss": 0.42781347036361694,
+      "step": 1555
+    },
+    {
+      "epoch": 1.5244455336355838,
+      "grad_norm": 0.4557672441005707,
+      "learning_rate": 1.4198745124468565e-06,
+      "loss": 0.36109858751296997,
+      "step": 1556
+    },
+    {
+      "epoch": 1.525425805661071,
+      "grad_norm": 0.4702933132648468,
+      "learning_rate": 1.4143409762051829e-06,
+      "loss": 0.37791088223457336,
+      "step": 1557
+    },
+    {
+      "epoch": 1.526406077686558,
+      "grad_norm": 2.0401318073272705,
+      "learning_rate": 1.4088164668558302e-06,
+      "loss": 0.4928564429283142,
+      "step": 1558
+    },
+    {
+      "epoch": 1.527386349712045,
+      "grad_norm": 0.9257914423942566,
+      "learning_rate": 1.4033009983067454e-06,
+      "loss": 0.4056940972805023,
+      "step": 1559
+    },
+    {
+      "epoch": 1.5283666217375322,
+      "grad_norm": 0.5786461234092712,
+      "learning_rate": 1.3977945844431118e-06,
+      "loss": 0.3608434796333313,
+      "step": 1560
+    },
+    {
+      "epoch": 1.5293468937630192,
+      "grad_norm": 1.2767441272735596,
+      "learning_rate": 1.3922972391273226e-06,
+      "loss": 0.501929521560669,
+      "step": 1561
+    },
+    {
+      "epoch": 1.5303271657885062,
+      "grad_norm": 0.517837405204773,
+      "learning_rate": 1.3868089761989356e-06,
+      "loss": 0.3641575872898102,
+      "step": 1562
+    },
+    {
+      "epoch": 1.5313074378139935,
+      "grad_norm": 0.5195152163505554,
+      "learning_rate": 1.3813298094746491e-06,
+      "loss": 0.4160193204879761,
+      "step": 1563
+    },
+    {
+      "epoch": 1.5322877098394805,
+      "grad_norm": 0.7131319046020508,
+      "learning_rate": 1.375859752748257e-06,
+      "loss": 0.5987261533737183,
+      "step": 1564
+    },
+    {
+      "epoch": 1.5332679818649675,
+      "grad_norm": 0.49722662568092346,
+      "learning_rate": 1.3703988197906209e-06,
+      "loss": 0.5047547221183777,
+      "step": 1565
+    },
+    {
+      "epoch": 1.5342482538904547,
+      "grad_norm": 0.4824052155017853,
+      "learning_rate": 1.3649470243496327e-06,
+      "loss": 0.47263020277023315,
+      "step": 1566
+    },
+    {
+      "epoch": 1.5352285259159417,
+      "grad_norm": 0.5364777445793152,
+      "learning_rate": 1.3595043801501794e-06,
+      "loss": 0.39895063638687134,
+      "step": 1567
+    },
+    {
+      "epoch": 1.5362087979414287,
+      "grad_norm": 0.4986257255077362,
+      "learning_rate": 1.354070900894115e-06,
+      "loss": 0.39921334385871887,
+      "step": 1568
+    },
+    {
+      "epoch": 1.537189069966916,
+      "grad_norm": 0.7853311896324158,
+      "learning_rate": 1.3486466002602133e-06,
+      "loss": 0.4113119840621948,
+      "step": 1569
+    },
+    {
+      "epoch": 1.5381693419924027,
+      "grad_norm": 0.4957791566848755,
+      "learning_rate": 1.3432314919041478e-06,
+      "loss": 0.45643454790115356,
+      "step": 1570
+    },
+    {
+      "epoch": 1.53914961401789,
+      "grad_norm": 1.1512796878814697,
+      "learning_rate": 1.3378255894584463e-06,
+      "loss": 0.6137411594390869,
+      "step": 1571
+    },
+    {
+      "epoch": 1.5401298860433772,
+      "grad_norm": 0.5368757247924805,
+      "learning_rate": 1.3324289065324608e-06,
+      "loss": 0.35343050956726074,
+      "step": 1572
+    },
+    {
+      "epoch": 1.541110158068864,
+      "grad_norm": 0.5801121592521667,
+      "learning_rate": 1.3270414567123342e-06,
+      "loss": 0.5394392609596252,
+      "step": 1573
+    },
+    {
+      "epoch": 1.5420904300943512,
+      "grad_norm": 1.2381075620651245,
+      "learning_rate": 1.3216632535609664e-06,
+      "loss": 0.3662095367908478,
+      "step": 1574
+    },
+    {
+      "epoch": 1.5430707021198382,
+      "grad_norm": 1.0949516296386719,
+      "learning_rate": 1.3162943106179748e-06,
+      "loss": 0.5478399991989136,
+      "step": 1575
+    },
+    {
+      "epoch": 1.5440509741453252,
+      "grad_norm": 0.600519061088562,
+      "learning_rate": 1.310934641399671e-06,
+      "loss": 0.3420920670032501,
+      "step": 1576
+    },
+    {
+      "epoch": 1.5450312461708124,
+      "grad_norm": 0.5267794132232666,
+      "learning_rate": 1.305584259399013e-06,
+      "loss": 0.3938623368740082,
+      "step": 1577
+    },
+    {
+      "epoch": 1.5460115181962995,
+      "grad_norm": 0.5927364230155945,
+      "learning_rate": 1.3002431780855817e-06,
+      "loss": 0.6037330031394958,
+      "step": 1578
+    },
+    {
+      "epoch": 1.5469917902217865,
+      "grad_norm": 0.5320039987564087,
+      "learning_rate": 1.2949114109055417e-06,
+      "loss": 0.49153977632522583,
+      "step": 1579
+    },
+    {
+      "epoch": 1.5479720622472737,
+      "grad_norm": 0.5074359178543091,
+      "learning_rate": 1.2895889712816129e-06,
+      "loss": 0.49725091457366943,
+      "step": 1580
+    },
+    {
+      "epoch": 1.5489523342727607,
+      "grad_norm": 0.44100022315979004,
+      "learning_rate": 1.2842758726130283e-06,
+      "loss": 0.34668272733688354,
+      "step": 1581
+    },
+    {
+      "epoch": 1.5499326062982477,
+      "grad_norm": 0.6297959089279175,
+      "learning_rate": 1.2789721282755102e-06,
+      "loss": 0.5807473063468933,
+      "step": 1582
+    },
+    {
+      "epoch": 1.550912878323735,
+      "grad_norm": 0.895707905292511,
+      "learning_rate": 1.2736777516212267e-06,
+      "loss": 0.45426493883132935,
+      "step": 1583
+    },
+    {
+      "epoch": 1.551893150349222,
+      "grad_norm": 0.5077658295631409,
+      "learning_rate": 1.2683927559787657e-06,
+      "loss": 0.4310656785964966,
+      "step": 1584
+    },
+    {
+      "epoch": 1.552873422374709,
+      "grad_norm": 0.5879361629486084,
+      "learning_rate": 1.263117154653097e-06,
+      "loss": 0.3991815447807312,
+      "step": 1585
+    },
+    {
+      "epoch": 1.5538536944001962,
+      "grad_norm": 0.6359089612960815,
+      "learning_rate": 1.2578509609255385e-06,
+      "loss": 0.449389785528183,
+      "step": 1586
+    },
+    {
+      "epoch": 1.5548339664256832,
+      "grad_norm": 0.44460681080818176,
+      "learning_rate": 1.2525941880537307e-06,
+      "loss": 0.4537050127983093,
+      "step": 1587
+    },
+    {
+      "epoch": 1.5558142384511702,
+      "grad_norm": 0.5246607065200806,
+      "learning_rate": 1.2473468492715896e-06,
+      "loss": 0.4834633469581604,
+      "step": 1588
+    },
+    {
+      "epoch": 1.5567945104766574,
+      "grad_norm": 0.4515811800956726,
+      "learning_rate": 1.242108957789287e-06,
+      "loss": 0.3937339186668396,
+      "step": 1589
+    },
+    {
+      "epoch": 1.5577747825021442,
+      "grad_norm": 0.46973490715026855,
+      "learning_rate": 1.2368805267932072e-06,
+      "loss": 0.43566322326660156,
+      "step": 1590
+    },
+    {
+      "epoch": 1.5587550545276314,
+      "grad_norm": 0.5791113376617432,
+      "learning_rate": 1.2316615694459188e-06,
+      "loss": 0.40468084812164307,
+      "step": 1591
+    },
+    {
+      "epoch": 1.5597353265531186,
+      "grad_norm": 0.5294288396835327,
+      "learning_rate": 1.22645209888614e-06,
+      "loss": 0.49904078245162964,
+      "step": 1592
+    },
+    {
+      "epoch": 1.5607155985786054,
+      "grad_norm": 0.4481540322303772,
+      "learning_rate": 1.2212521282287093e-06,
+      "loss": 0.40608999133110046,
+      "step": 1593
+    },
+    {
+      "epoch": 1.5616958706040927,
+      "grad_norm": 0.8243662714958191,
+      "learning_rate": 1.2160616705645434e-06,
+      "loss": 0.4111751317977905,
+      "step": 1594
+    },
+    {
+      "epoch": 1.5626761426295797,
+      "grad_norm": 0.44928866624832153,
+      "learning_rate": 1.210880738960616e-06,
+      "loss": 0.4219694137573242,
+      "step": 1595
+    },
+    {
+      "epoch": 1.5636564146550667,
+      "grad_norm": 0.6408407092094421,
+      "learning_rate": 1.2057093464599156e-06,
+      "loss": 0.40732085704803467,
+      "step": 1596
+    },
+    {
+      "epoch": 1.564636686680554,
+      "grad_norm": 0.6224952936172485,
+      "learning_rate": 1.200547506081416e-06,
+      "loss": 0.4951319694519043,
+      "step": 1597
+    },
+    {
+      "epoch": 1.565616958706041,
+      "grad_norm": 0.984839916229248,
+      "learning_rate": 1.1953952308200428e-06,
+      "loss": 0.48763734102249146,
+      "step": 1598
+    },
+    {
+      "epoch": 1.566597230731528,
+      "grad_norm": 0.47617724537849426,
+      "learning_rate": 1.1902525336466465e-06,
+      "loss": 0.3930254876613617,
+      "step": 1599
+    },
+    {
+      "epoch": 1.5675775027570151,
+      "grad_norm": 0.5134947299957275,
+      "learning_rate": 1.1851194275079586e-06,
+      "loss": 0.4373096823692322,
+      "step": 1600
+    },
+    {
+      "epoch": 1.5685577747825021,
+      "grad_norm": 0.5502176284790039,
+      "learning_rate": 1.1799959253265668e-06,
+      "loss": 0.4272688627243042,
+      "step": 1601
+    },
+    {
+      "epoch": 1.5695380468079891,
+      "grad_norm": 0.5675881505012512,
+      "learning_rate": 1.1748820400008843e-06,
+      "loss": 0.5714017152786255,
+      "step": 1602
+    },
+    {
+      "epoch": 1.5705183188334764,
+      "grad_norm": 0.609653651714325,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.4348721206188202,
+      "step": 1603
+    },
+    {
+      "epoch": 1.5714985908589634,
+      "grad_norm": 0.5446346402168274,
+      "learning_rate": 1.164683171389202e-06,
+      "loss": 0.48841187357902527,
+      "step": 1604
+    },
+    {
+      "epoch": 1.5724788628844504,
+      "grad_norm": 0.4983329176902771,
+      "learning_rate": 1.1595982137788403e-06,
+      "loss": 0.5338020324707031,
+      "step": 1605
+    },
+    {
+      "epoch": 1.5734591349099376,
+      "grad_norm": 0.49444273114204407,
+      "learning_rate": 1.154522924375403e-06,
+      "loss": 0.4493732452392578,
+      "step": 1606
+    },
+    {
+      "epoch": 1.5744394069354246,
+      "grad_norm": 0.5851345658302307,
+      "learning_rate": 1.1494573159559214e-06,
+      "loss": 0.4470522403717041,
+      "step": 1607
+    },
+    {
+      "epoch": 1.5754196789609116,
+      "grad_norm": 0.5860568881034851,
+      "learning_rate": 1.144401401273062e-06,
+      "loss": 0.44188910722732544,
+      "step": 1608
+    },
+    {
+      "epoch": 1.5763999509863988,
+      "grad_norm": 0.4428722858428955,
+      "learning_rate": 1.1393551930550828e-06,
+      "loss": 0.45409727096557617,
+      "step": 1609
+    },
+    {
+      "epoch": 1.5773802230118859,
+      "grad_norm": 0.5591416954994202,
+      "learning_rate": 1.1343187040058068e-06,
+      "loss": 0.4411742687225342,
+      "step": 1610
+    },
+    {
+      "epoch": 1.5783604950373729,
+      "grad_norm": 0.5603340864181519,
+      "learning_rate": 1.1292919468045876e-06,
+      "loss": 0.38865429162979126,
+      "step": 1611
+    },
+    {
+      "epoch": 1.57934076706286,
+      "grad_norm": 0.6181705594062805,
+      "learning_rate": 1.1242749341062837e-06,
+      "loss": 0.5278862714767456,
+      "step": 1612
+    },
+    {
+      "epoch": 1.5803210390883469,
+      "grad_norm": 0.639843225479126,
+      "learning_rate": 1.1192676785412154e-06,
+      "loss": 0.6048687100410461,
+      "step": 1613
+    },
+    {
+      "epoch": 1.581301311113834,
+      "grad_norm": 0.5007582306861877,
+      "learning_rate": 1.1142701927151456e-06,
+      "loss": 0.4496884346008301,
+      "step": 1614
+    },
+    {
+      "epoch": 1.5822815831393213,
+      "grad_norm": 0.5776380300521851,
+      "learning_rate": 1.1092824892092375e-06,
+      "loss": 0.5071296095848083,
+      "step": 1615
+    },
+    {
+      "epoch": 1.583261855164808,
+      "grad_norm": 0.49941593408584595,
+      "learning_rate": 1.1043045805800285e-06,
+      "loss": 0.33119434118270874,
+      "step": 1616
+    },
+    {
+      "epoch": 1.5842421271902953,
+      "grad_norm": 1.7760988473892212,
+      "learning_rate": 1.099336479359398e-06,
+      "loss": 0.43989765644073486,
+      "step": 1617
+    },
+    {
+      "epoch": 1.5852223992157823,
+      "grad_norm": 0.5640775561332703,
+      "learning_rate": 1.0943781980545332e-06,
+      "loss": 0.4828563928604126,
+      "step": 1618
+    },
+    {
+      "epoch": 1.5862026712412693,
+      "grad_norm": 0.6271183490753174,
+      "learning_rate": 1.0894297491479044e-06,
+      "loss": 0.4299319386482239,
+      "step": 1619
+    },
+    {
+      "epoch": 1.5871829432667566,
+      "grad_norm": 30.2771053314209,
+      "learning_rate": 1.0844911450972228e-06,
+      "loss": 0.4792664051055908,
+      "step": 1620
+    },
+    {
+      "epoch": 1.5881632152922436,
+      "grad_norm": 0.7280365228652954,
+      "learning_rate": 1.0795623983354214e-06,
+      "loss": 0.482305109500885,
+      "step": 1621
+    },
+    {
+      "epoch": 1.5891434873177306,
+      "grad_norm": 0.6474554538726807,
+      "learning_rate": 1.0746435212706125e-06,
+      "loss": 0.4826841950416565,
+      "step": 1622
+    },
+    {
+      "epoch": 1.5901237593432178,
+      "grad_norm": 0.5415899157524109,
+      "learning_rate": 1.0697345262860638e-06,
+      "loss": 0.5280240178108215,
+      "step": 1623
+    },
+    {
+      "epoch": 1.5911040313687048,
+      "grad_norm": 0.9219223260879517,
+      "learning_rate": 1.0648354257401622e-06,
+      "loss": 0.5524288415908813,
+      "step": 1624
+    },
+    {
+      "epoch": 1.5920843033941918,
+      "grad_norm": 0.5914516448974609,
+      "learning_rate": 1.0599462319663906e-06,
+      "loss": 0.4685550332069397,
+      "step": 1625
+    },
+    {
+      "epoch": 1.593064575419679,
+      "grad_norm": 16.096162796020508,
+      "learning_rate": 1.0550669572732862e-06,
+      "loss": 0.41529643535614014,
+      "step": 1626
+    },
+    {
+      "epoch": 1.594044847445166,
+      "grad_norm": 1.156607747077942,
+      "learning_rate": 1.0501976139444191e-06,
+      "loss": 0.4621277451515198,
+      "step": 1627
+    },
+    {
+      "epoch": 1.595025119470653,
+      "grad_norm": 0.4887218773365021,
+      "learning_rate": 1.0453382142383545e-06,
+      "loss": 0.3737853765487671,
+      "step": 1628
+    },
+    {
+      "epoch": 1.5960053914961403,
+      "grad_norm": 0.46627190709114075,
+      "learning_rate": 1.0404887703886252e-06,
+      "loss": 0.4328254461288452,
+      "step": 1629
+    },
+    {
+      "epoch": 1.5969856635216273,
+      "grad_norm": 0.659539520740509,
+      "learning_rate": 1.0356492946036994e-06,
+      "loss": 0.4790624976158142,
+      "step": 1630
+    },
+    {
+      "epoch": 1.5979659355471143,
+      "grad_norm": 0.45169106125831604,
+      "learning_rate": 1.0308197990669538e-06,
+      "loss": 0.41745269298553467,
+      "step": 1631
+    },
+    {
+      "epoch": 1.5989462075726015,
+      "grad_norm": 2.085352897644043,
+      "learning_rate": 1.026000295936635e-06,
+      "loss": 0.5254413485527039,
+      "step": 1632
+    },
+    {
+      "epoch": 1.5999264795980883,
+      "grad_norm": 0.45684167742729187,
+      "learning_rate": 1.0211907973458391e-06,
+      "loss": 0.562838077545166,
+      "step": 1633
+    },
+    {
+      "epoch": 1.6009067516235755,
+      "grad_norm": 0.44522812962532043,
+      "learning_rate": 1.0163913154024724e-06,
+      "loss": 0.3999680280685425,
+      "step": 1634
+    },
+    {
+      "epoch": 1.6018870236490628,
+      "grad_norm": 1.300023078918457,
+      "learning_rate": 1.0116018621892237e-06,
+      "loss": 0.4016084372997284,
+      "step": 1635
+    },
+    {
+      "epoch": 1.6028672956745496,
+      "grad_norm": 0.45449599623680115,
+      "learning_rate": 1.006822449763537e-06,
+      "loss": 0.38338398933410645,
+      "step": 1636
+    },
+    {
+      "epoch": 1.6038475677000368,
+      "grad_norm": 0.599853515625,
+      "learning_rate": 1.0020530901575754e-06,
+      "loss": 0.40913450717926025,
+      "step": 1637
+    },
+    {
+      "epoch": 1.6048278397255238,
+      "grad_norm": 1.3690499067306519,
+      "learning_rate": 9.972937953781985e-07,
+      "loss": 0.459037184715271,
+      "step": 1638
+    },
+    {
+      "epoch": 1.6058081117510108,
+      "grad_norm": 0.5936954617500305,
+      "learning_rate": 9.925445774069232e-07,
+      "loss": 0.44346359372138977,
+      "step": 1639
+    },
+    {
+      "epoch": 1.606788383776498,
+      "grad_norm": 0.5537636280059814,
+      "learning_rate": 9.878054481999027e-07,
+      "loss": 0.4729841351509094,
+      "step": 1640
+    },
+    {
+      "epoch": 1.607768655801985,
+      "grad_norm": 0.5466724038124084,
+      "learning_rate": 9.830764196878872e-07,
+      "loss": 0.47972679138183594,
+      "step": 1641
+    },
+    {
+      "epoch": 1.608748927827472,
+      "grad_norm": 0.6618354320526123,
+      "learning_rate": 9.783575037762006e-07,
+      "loss": 0.3903306722640991,
+      "step": 1642
+    },
+    {
+      "epoch": 1.6097291998529593,
+      "grad_norm": 0.6704966425895691,
+      "learning_rate": 9.73648712344707e-07,
+      "loss": 0.4191238284111023,
+      "step": 1643
+    },
+    {
+      "epoch": 1.6107094718784463,
+      "grad_norm": 0.7465136647224426,
+      "learning_rate": 9.689500572477857e-07,
+      "loss": 0.4328649640083313,
+      "step": 1644
+    },
+    {
+      "epoch": 1.6116897439039333,
+      "grad_norm": 3.6307878494262695,
+      "learning_rate": 9.642615503142927e-07,
+      "loss": 0.4666591286659241,
+      "step": 1645
+    },
+    {
+      "epoch": 1.6126700159294205,
+      "grad_norm": 0.541968584060669,
+      "learning_rate": 9.595832033475417e-07,
+      "loss": 0.4586426317691803,
+      "step": 1646
+    },
+    {
+      "epoch": 1.6136502879549075,
+      "grad_norm": 0.7496551871299744,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.4554421901702881,
+      "step": 1647
+    },
+    {
+      "epoch": 1.6146305599803945,
+      "grad_norm": 0.5645624995231628,
+      "learning_rate": 9.502570363995839e-07,
+      "loss": 0.44312530755996704,
+      "step": 1648
+    },
+    {
+      "epoch": 1.6156108320058817,
+      "grad_norm": 0.7723249793052673,
+      "learning_rate": 9.456092398969902e-07,
+      "loss": 0.44427910447120667,
+      "step": 1649
+    },
+    {
+      "epoch": 1.6165911040313687,
+      "grad_norm": 0.46417075395584106,
+      "learning_rate": 9.409716503183075e-07,
+      "loss": 0.47685492038726807,
+      "step": 1650
+    },
+    {
+      "epoch": 1.6175713760568557,
+      "grad_norm": 0.8587995767593384,
+      "learning_rate": 9.363442793386606e-07,
+      "loss": 0.5411603450775146,
+      "step": 1651
+    },
+    {
+      "epoch": 1.618551648082343,
+      "grad_norm": 0.48141348361968994,
+      "learning_rate": 9.317271386074489e-07,
+      "loss": 0.4529597759246826,
+      "step": 1652
+    },
+    {
+      "epoch": 1.6195319201078298,
+      "grad_norm": 0.7475388646125793,
+      "learning_rate": 9.271202397483214e-07,
+      "loss": 0.4013572037220001,
+      "step": 1653
+    },
+    {
+      "epoch": 1.620512192133317,
+      "grad_norm": 0.6382339596748352,
+      "learning_rate": 9.225235943591398e-07,
+      "loss": 0.41697216033935547,
+      "step": 1654
+    },
+    {
+      "epoch": 1.6214924641588042,
+      "grad_norm": 0.5399206876754761,
+      "learning_rate": 9.179372140119524e-07,
+      "loss": 0.4487508535385132,
+      "step": 1655
+    },
+    {
+      "epoch": 1.622472736184291,
+      "grad_norm": 0.6563431024551392,
+      "learning_rate": 9.133611102529655e-07,
+      "loss": 0.41831231117248535,
+      "step": 1656
+    },
+    {
+      "epoch": 1.6234530082097782,
+      "grad_norm": 0.5656543970108032,
+      "learning_rate": 9.087952946025175e-07,
+      "loss": 0.37385034561157227,
+      "step": 1657
+    },
+    {
+      "epoch": 1.6244332802352652,
+      "grad_norm": 0.5844793915748596,
+      "learning_rate": 9.042397785550405e-07,
+      "loss": 0.42205750942230225,
+      "step": 1658
+    },
+    {
+      "epoch": 1.6254135522607522,
+      "grad_norm": 1.2819033861160278,
+      "learning_rate": 8.996945735790447e-07,
+      "loss": 0.5121839046478271,
+      "step": 1659
+    },
+    {
+      "epoch": 1.6263938242862395,
+      "grad_norm": 0.6153983473777771,
+      "learning_rate": 8.951596911170756e-07,
+      "loss": 0.48581522703170776,
+      "step": 1660
+    },
+    {
+      "epoch": 1.6273740963117265,
+      "grad_norm": 0.4786304831504822,
+      "learning_rate": 8.906351425856952e-07,
+      "loss": 0.45942214131355286,
+      "step": 1661
+    },
+    {
+      "epoch": 1.6283543683372135,
+      "grad_norm": 0.686417818069458,
+      "learning_rate": 8.861209393754478e-07,
+      "loss": 0.4662189483642578,
+      "step": 1662
+    },
+    {
+      "epoch": 1.6293346403627007,
+      "grad_norm": 0.8793837428092957,
+      "learning_rate": 8.816170928508367e-07,
+      "loss": 0.454819917678833,
+      "step": 1663
+    },
+    {
+      "epoch": 1.6303149123881877,
+      "grad_norm": 0.8305994868278503,
+      "learning_rate": 8.771236143502876e-07,
+      "loss": 0.5060313940048218,
+      "step": 1664
+    },
+    {
+      "epoch": 1.6312951844136747,
+      "grad_norm": 0.5543544292449951,
+      "learning_rate": 8.7264051518613e-07,
+      "loss": 0.5017277598381042,
+      "step": 1665
+    },
+    {
+      "epoch": 1.632275456439162,
+      "grad_norm": 0.6258586645126343,
+      "learning_rate": 8.681678066445598e-07,
+      "loss": 0.49192607402801514,
+      "step": 1666
+    },
+    {
+      "epoch": 1.633255728464649,
+      "grad_norm": 0.5229313969612122,
+      "learning_rate": 8.637054999856148e-07,
+      "loss": 0.41909271478652954,
+      "step": 1667
+    },
+    {
+      "epoch": 1.634236000490136,
+      "grad_norm": 0.5685809850692749,
+      "learning_rate": 8.592536064431467e-07,
+      "loss": 0.4713624119758606,
+      "step": 1668
+    },
+    {
+      "epoch": 1.6352162725156232,
+      "grad_norm": 0.657346785068512,
+      "learning_rate": 8.54812137224792e-07,
+      "loss": 0.4496573805809021,
+      "step": 1669
+    },
+    {
+      "epoch": 1.6361965445411102,
+      "grad_norm": 0.40926048159599304,
+      "learning_rate": 8.503811035119452e-07,
+      "loss": 0.3840830326080322,
+      "step": 1670
+    },
+    {
+      "epoch": 1.6371768165665972,
+      "grad_norm": 0.5244526267051697,
+      "learning_rate": 8.459605164597268e-07,
+      "loss": 0.4947342872619629,
+      "step": 1671
+    },
+    {
+      "epoch": 1.6381570885920844,
+      "grad_norm": 0.5079709887504578,
+      "learning_rate": 8.415503871969616e-07,
+      "loss": 0.3232126235961914,
+      "step": 1672
+    },
+    {
+      "epoch": 1.6391373606175714,
+      "grad_norm": 0.5607351064682007,
+      "learning_rate": 8.371507268261436e-07,
+      "loss": 0.4252610504627228,
+      "step": 1673
+    },
+    {
+      "epoch": 1.6401176326430584,
+      "grad_norm": 0.5929581522941589,
+      "learning_rate": 8.327615464234129e-07,
+      "loss": 0.5064104795455933,
+      "step": 1674
+    },
+    {
+      "epoch": 1.6410979046685457,
+      "grad_norm": 0.4602496922016144,
+      "learning_rate": 8.283828570385239e-07,
+      "loss": 0.44804561138153076,
+      "step": 1675
+    },
+    {
+      "epoch": 1.6420781766940324,
+      "grad_norm": 0.5295189619064331,
+      "learning_rate": 8.240146696948254e-07,
+      "loss": 0.5265501141548157,
+      "step": 1676
+    },
+    {
+      "epoch": 1.6430584487195197,
+      "grad_norm": 0.5260054469108582,
+      "learning_rate": 8.196569953892202e-07,
+      "loss": 0.5025229454040527,
+      "step": 1677
+    },
+    {
+      "epoch": 1.644038720745007,
+      "grad_norm": 0.7239804267883301,
+      "learning_rate": 8.153098450921514e-07,
+      "loss": 0.5124154090881348,
+      "step": 1678
+    },
+    {
+      "epoch": 1.6450189927704937,
+      "grad_norm": 0.4656829833984375,
+      "learning_rate": 8.109732297475637e-07,
+      "loss": 0.43184879422187805,
+      "step": 1679
+    },
+    {
+      "epoch": 1.645999264795981,
+      "grad_norm": 0.6292969584465027,
+      "learning_rate": 8.066471602728804e-07,
+      "loss": 0.40027016401290894,
+      "step": 1680
+    },
+    {
+      "epoch": 1.646979536821468,
+      "grad_norm": 0.5554639101028442,
+      "learning_rate": 8.023316475589754e-07,
+      "loss": 0.3764341473579407,
+      "step": 1681
+    },
+    {
+      "epoch": 1.647959808846955,
+      "grad_norm": 0.671816885471344,
+      "learning_rate": 7.980267024701494e-07,
+      "loss": 0.4938500225543976,
+      "step": 1682
+    },
+    {
+      "epoch": 1.6489400808724421,
+      "grad_norm": 0.8617560863494873,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.46413126587867737,
+      "step": 1683
+    },
+    {
+      "epoch": 1.6499203528979292,
+      "grad_norm": 0.6406113505363464,
+      "learning_rate": 7.894485584918737e-07,
+      "loss": 0.5171228051185608,
+      "step": 1684
+    },
+    {
+      "epoch": 1.6509006249234162,
+      "grad_norm": 1.0307143926620483,
+      "learning_rate": 7.851753811978924e-07,
+      "loss": 0.3773418664932251,
+      "step": 1685
+    },
+    {
+      "epoch": 1.6518808969489034,
+      "grad_norm": 0.9457737803459167,
+      "learning_rate": 7.809128147198692e-07,
+      "loss": 0.35890549421310425,
+      "step": 1686
+    },
+    {
+      "epoch": 1.6528611689743904,
+      "grad_norm": 0.5369381308555603,
+      "learning_rate": 7.766608697888095e-07,
+      "loss": 0.4412789046764374,
+      "step": 1687
+    },
+    {
+      "epoch": 1.6538414409998774,
+      "grad_norm": 2.702284336090088,
+      "learning_rate": 7.724195571089787e-07,
+      "loss": 0.543986439704895,
+      "step": 1688
+    },
+    {
+      "epoch": 1.6548217130253646,
+      "grad_norm": 1.033728003501892,
+      "learning_rate": 7.681888873578786e-07,
+      "loss": 0.5213661789894104,
+      "step": 1689
+    },
+    {
+      "epoch": 1.6558019850508516,
+      "grad_norm": 0.783892810344696,
+      "learning_rate": 7.639688711862131e-07,
+      "loss": 0.5121074914932251,
+      "step": 1690
+    },
+    {
+      "epoch": 1.6567822570763386,
+      "grad_norm": 0.8677245378494263,
+      "learning_rate": 7.597595192178702e-07,
+      "loss": 0.5285683274269104,
+      "step": 1691
+    },
+    {
+      "epoch": 1.6577625291018259,
+      "grad_norm": 1.0014557838439941,
+      "learning_rate": 7.555608420498872e-07,
+      "loss": 0.4320244789123535,
+      "step": 1692
+    },
+    {
+      "epoch": 1.6587428011273129,
+      "grad_norm": 0.4497794806957245,
+      "learning_rate": 7.513728502524286e-07,
+      "loss": 0.3927844762802124,
+      "step": 1693
+    },
+    {
+      "epoch": 1.6597230731527999,
+      "grad_norm": 0.5977275371551514,
+      "learning_rate": 7.471955543687586e-07,
+      "loss": 0.4481373429298401,
+      "step": 1694
+    },
+    {
+      "epoch": 1.660703345178287,
+      "grad_norm": 0.5257479548454285,
+      "learning_rate": 7.430289649152156e-07,
+      "loss": 0.3347582221031189,
+      "step": 1695
+    },
+    {
+      "epoch": 1.6616836172037739,
+      "grad_norm": 0.5114702582359314,
+      "learning_rate": 7.388730923811827e-07,
+      "loss": 0.47129809856414795,
+      "step": 1696
+    },
+    {
+      "epoch": 1.6626638892292611,
+      "grad_norm": 0.47362610697746277,
+      "learning_rate": 7.347279472290647e-07,
+      "loss": 0.4425743520259857,
+      "step": 1697
+    },
+    {
+      "epoch": 1.6636441612547483,
+      "grad_norm": 0.6944350600242615,
+      "learning_rate": 7.305935398942598e-07,
+      "loss": 0.45714670419692993,
+      "step": 1698
+    },
+    {
+      "epoch": 1.6646244332802351,
+      "grad_norm": 0.6112069487571716,
+      "learning_rate": 7.264698807851328e-07,
+      "loss": 0.4847654104232788,
+      "step": 1699
+    },
+    {
+      "epoch": 1.6656047053057224,
+      "grad_norm": 0.47613897919654846,
+      "learning_rate": 7.2235698028299e-07,
+      "loss": 0.40825432538986206,
+      "step": 1700
+    },
+    {
+      "epoch": 1.6665849773312094,
+      "grad_norm": 0.439959317445755,
+      "learning_rate": 7.182548487420555e-07,
+      "loss": 0.400945782661438,
+      "step": 1701
+    },
+    {
+      "epoch": 1.6675652493566964,
+      "grad_norm": 0.5905991792678833,
+      "learning_rate": 7.141634964894389e-07,
+      "loss": 0.37899017333984375,
+      "step": 1702
+    },
+    {
+      "epoch": 1.6685455213821836,
+      "grad_norm": 0.5127450823783875,
+      "learning_rate": 7.100829338251147e-07,
+      "loss": 0.5572842359542847,
+      "step": 1703
+    },
+    {
+      "epoch": 1.6695257934076706,
+      "grad_norm": 0.6347851753234863,
+      "learning_rate": 7.060131710218959e-07,
+      "loss": 0.5185015201568604,
+      "step": 1704
+    },
+    {
+      "epoch": 1.6705060654331576,
+      "grad_norm": 0.5175912976264954,
+      "learning_rate": 7.019542183254047e-07,
+      "loss": 0.3992921710014343,
+      "step": 1705
+    },
+    {
+      "epoch": 1.6714863374586448,
+      "grad_norm": 0.44617775082588196,
+      "learning_rate": 6.979060859540493e-07,
+      "loss": 0.33239397406578064,
+      "step": 1706
+    },
+    {
+      "epoch": 1.6724666094841318,
+      "grad_norm": 0.4812489449977875,
+      "learning_rate": 6.938687840989972e-07,
+      "loss": 0.4037539064884186,
+      "step": 1707
+    },
+    {
+      "epoch": 1.6734468815096188,
+      "grad_norm": 0.4330952763557434,
+      "learning_rate": 6.898423229241535e-07,
+      "loss": 0.44279295206069946,
+      "step": 1708
+    },
+    {
+      "epoch": 1.674427153535106,
+      "grad_norm": 0.459568053483963,
+      "learning_rate": 6.858267125661272e-07,
+      "loss": 0.4847477078437805,
+      "step": 1709
+    },
+    {
+      "epoch": 1.675407425560593,
+      "grad_norm": 0.5936370491981506,
+      "learning_rate": 6.818219631342149e-07,
+      "loss": 0.36596566438674927,
+      "step": 1710
+    },
+    {
+      "epoch": 1.67638769758608,
+      "grad_norm": 0.5384985208511353,
+      "learning_rate": 6.778280847103668e-07,
+      "loss": 0.40649908781051636,
+      "step": 1711
+    },
+    {
+      "epoch": 1.6773679696115673,
+      "grad_norm": 0.5439937710762024,
+      "learning_rate": 6.738450873491675e-07,
+      "loss": 0.38872742652893066,
+      "step": 1712
+    },
+    {
+      "epoch": 1.6783482416370543,
+      "grad_norm": 0.4497339427471161,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.3218197822570801,
+      "step": 1713
+    },
+    {
+      "epoch": 1.6793285136625413,
+      "grad_norm": 2.0870888233184814,
+      "learning_rate": 6.659117758960598e-07,
+      "loss": 0.5049896240234375,
+      "step": 1714
+    },
+    {
+      "epoch": 1.6803087856880285,
+      "grad_norm": 0.6304010152816772,
+      "learning_rate": 6.619614817762537e-07,
+      "loss": 0.47291094064712524,
+      "step": 1715
+    },
+    {
+      "epoch": 1.6812890577135156,
+      "grad_norm": 0.49912068247795105,
+      "learning_rate": 6.580221086632516e-07,
+      "loss": 0.41225558519363403,
+      "step": 1716
+    },
+    {
+      "epoch": 1.6822693297390026,
+      "grad_norm": 0.8838415741920471,
+      "learning_rate": 6.540936664744197e-07,
+      "loss": 0.5509417057037354,
+      "step": 1717
+    },
+    {
+      "epoch": 1.6832496017644898,
+      "grad_norm": 0.7193208336830139,
+      "learning_rate": 6.501761650996052e-07,
+      "loss": 0.49753430485725403,
+      "step": 1718
+    },
+    {
+      "epoch": 1.6842298737899766,
+      "grad_norm": 0.5841873288154602,
+      "learning_rate": 6.462696144011149e-07,
+      "loss": 0.42212575674057007,
+      "step": 1719
+    },
+    {
+      "epoch": 1.6852101458154638,
+      "grad_norm": 0.6215124130249023,
+      "learning_rate": 6.423740242136833e-07,
+      "loss": 0.42674964666366577,
+      "step": 1720
+    },
+    {
+      "epoch": 1.686190417840951,
+      "grad_norm": 1.1202881336212158,
+      "learning_rate": 6.384894043444568e-07,
+      "loss": 0.334888756275177,
+      "step": 1721
+    },
+    {
+      "epoch": 1.6871706898664378,
+      "grad_norm": 0.5193874835968018,
+      "learning_rate": 6.34615764572959e-07,
+      "loss": 0.44838741421699524,
+      "step": 1722
+    },
+    {
+      "epoch": 1.688150961891925,
+      "grad_norm": 0.5526618957519531,
+      "learning_rate": 6.307531146510754e-07,
+      "loss": 0.4563174247741699,
+      "step": 1723
+    },
+    {
+      "epoch": 1.689131233917412,
+      "grad_norm": 1.3023649454116821,
+      "learning_rate": 6.269014643030214e-07,
+      "loss": 0.4477706551551819,
+      "step": 1724
+    },
+    {
+      "epoch": 1.690111505942899,
+      "grad_norm": 0.7621108889579773,
+      "learning_rate": 6.230608232253227e-07,
+      "loss": 0.4551399350166321,
+      "step": 1725
+    },
+    {
+      "epoch": 1.6910917779683863,
+      "grad_norm": 0.5743107199668884,
+      "learning_rate": 6.192312010867868e-07,
+      "loss": 0.41959965229034424,
+      "step": 1726
+    },
+    {
+      "epoch": 1.6920720499938733,
+      "grad_norm": 0.7936354279518127,
+      "learning_rate": 6.154126075284855e-07,
+      "loss": 0.4667820632457733,
+      "step": 1727
+    },
+    {
+      "epoch": 1.6930523220193603,
+      "grad_norm": 0.579326868057251,
+      "learning_rate": 6.116050521637218e-07,
+      "loss": 0.517664909362793,
+      "step": 1728
+    },
+    {
+      "epoch": 1.6940325940448475,
+      "grad_norm": 0.8135713338851929,
+      "learning_rate": 6.07808544578013e-07,
+      "loss": 0.6272374391555786,
+      "step": 1729
+    },
+    {
+      "epoch": 1.6950128660703345,
+      "grad_norm": 0.4043082296848297,
+      "learning_rate": 6.04023094329062e-07,
+      "loss": 0.3940927982330322,
+      "step": 1730
+    },
+    {
+      "epoch": 1.6959931380958215,
+      "grad_norm": 5.948627948760986,
+      "learning_rate": 6.002487109467347e-07,
+      "loss": 0.35454392433166504,
+      "step": 1731
+    },
+    {
+      "epoch": 1.6969734101213088,
+      "grad_norm": 2.5638437271118164,
+      "learning_rate": 5.964854039330365e-07,
+      "loss": 0.3477349579334259,
+      "step": 1732
+    },
+    {
+      "epoch": 1.6979536821467958,
+      "grad_norm": 0.8501605987548828,
+      "learning_rate": 5.927331827620902e-07,
+      "loss": 0.5380195379257202,
+      "step": 1733
+    },
+    {
+      "epoch": 1.6989339541722828,
+      "grad_norm": 0.5704190135002136,
+      "learning_rate": 5.889920568801055e-07,
+      "loss": 0.5046436786651611,
+      "step": 1734
+    },
+    {
+      "epoch": 1.69991422619777,
+      "grad_norm": 0.4976552128791809,
+      "learning_rate": 5.852620357053651e-07,
+      "loss": 0.44678419828414917,
+      "step": 1735
+    },
+    {
+      "epoch": 1.700894498223257,
+      "grad_norm": 0.5562195181846619,
+      "learning_rate": 5.815431286281925e-07,
+      "loss": 0.49987655878067017,
+      "step": 1736
+    },
+    {
+      "epoch": 1.701874770248744,
+      "grad_norm": 0.4525013267993927,
+      "learning_rate": 5.778353450109286e-07,
+      "loss": 0.40802162885665894,
+      "step": 1737
+    },
+    {
+      "epoch": 1.7028550422742312,
+      "grad_norm": 0.5386720895767212,
+      "learning_rate": 5.741386941879179e-07,
+      "loss": 0.41850584745407104,
+      "step": 1738
+    },
+    {
+      "epoch": 1.703835314299718,
+      "grad_norm": 0.556573212146759,
+      "learning_rate": 5.704531854654721e-07,
+      "loss": 0.3448736071586609,
+      "step": 1739
+    },
+    {
+      "epoch": 1.7048155863252052,
+      "grad_norm": 0.6255661845207214,
+      "learning_rate": 5.667788281218567e-07,
+      "loss": 0.4192609190940857,
+      "step": 1740
+    },
+    {
+      "epoch": 1.7057958583506925,
+      "grad_norm": 0.6631842255592346,
+      "learning_rate": 5.631156314072605e-07,
+      "loss": 0.5139786005020142,
+      "step": 1741
+    },
+    {
+      "epoch": 1.7067761303761793,
+      "grad_norm": 0.5389664173126221,
+      "learning_rate": 5.594636045437796e-07,
+      "loss": 0.3771073520183563,
+      "step": 1742
+    },
+    {
+      "epoch": 1.7077564024016665,
+      "grad_norm": 0.4552420377731323,
+      "learning_rate": 5.558227567253832e-07,
+      "loss": 0.42471352219581604,
+      "step": 1743
+    },
+    {
+      "epoch": 1.7087366744271535,
+      "grad_norm": 0.817165195941925,
+      "learning_rate": 5.521930971179041e-07,
+      "loss": 0.4951934218406677,
+      "step": 1744
+    },
+    {
+      "epoch": 1.7097169464526405,
+      "grad_norm": 0.5637513399124146,
+      "learning_rate": 5.485746348590048e-07,
+      "loss": 0.4901920557022095,
+      "step": 1745
+    },
+    {
+      "epoch": 1.7106972184781277,
+      "grad_norm": 0.9385501742362976,
+      "learning_rate": 5.449673790581611e-07,
+      "loss": 0.4220595359802246,
+      "step": 1746
+    },
+    {
+      "epoch": 1.7116774905036147,
+      "grad_norm": 0.5086963772773743,
+      "learning_rate": 5.413713387966329e-07,
+      "loss": 0.4635939598083496,
+      "step": 1747
+    },
+    {
+      "epoch": 1.7126577625291017,
+      "grad_norm": 0.6461208462715149,
+      "learning_rate": 5.377865231274504e-07,
+      "loss": 0.45143288373947144,
+      "step": 1748
+    },
+    {
+      "epoch": 1.713638034554589,
+      "grad_norm": 0.6383857131004333,
+      "learning_rate": 5.34212941075381e-07,
+      "loss": 0.5652772188186646,
+      "step": 1749
+    },
+    {
+      "epoch": 1.714618306580076,
+      "grad_norm": 0.6029959321022034,
+      "learning_rate": 5.306506016369134e-07,
+      "loss": 0.5482159852981567,
+      "step": 1750
+    },
+    {
+      "epoch": 1.715598578605563,
+      "grad_norm": 0.656978189945221,
+      "learning_rate": 5.270995137802315e-07,
+      "loss": 0.4565068483352661,
+      "step": 1751
+    },
+    {
+      "epoch": 1.7165788506310502,
+      "grad_norm": 0.43457481265068054,
+      "learning_rate": 5.235596864451975e-07,
+      "loss": 0.3208303153514862,
+      "step": 1752
+    },
+    {
+      "epoch": 1.7175591226565372,
+      "grad_norm": 0.6544308662414551,
+      "learning_rate": 5.200311285433213e-07,
+      "loss": 0.5067968368530273,
+      "step": 1753
+    },
+    {
+      "epoch": 1.7185393946820242,
+      "grad_norm": 0.7496337294578552,
+      "learning_rate": 5.165138489577415e-07,
+      "loss": 0.4992087483406067,
+      "step": 1754
+    },
+    {
+      "epoch": 1.7195196667075114,
+      "grad_norm": 0.44024550914764404,
+      "learning_rate": 5.130078565432089e-07,
+      "loss": 0.37321993708610535,
+      "step": 1755
+    },
+    {
+      "epoch": 1.7204999387329984,
+      "grad_norm": 0.5472961068153381,
+      "learning_rate": 5.095131601260517e-07,
+      "loss": 0.46043628454208374,
+      "step": 1756
+    },
+    {
+      "epoch": 1.7214802107584855,
+      "grad_norm": 0.5872296690940857,
+      "learning_rate": 5.06029768504166e-07,
+      "loss": 0.4246714115142822,
+      "step": 1757
+    },
+    {
+      "epoch": 1.7224604827839727,
+      "grad_norm": 0.41121867299079895,
+      "learning_rate": 5.025576904469842e-07,
+      "loss": 0.40875205397605896,
+      "step": 1758
+    },
+    {
+      "epoch": 1.7234407548094595,
+      "grad_norm": 0.5615606307983398,
+      "learning_rate": 4.990969346954611e-07,
+      "loss": 0.40258634090423584,
+      "step": 1759
+    },
+    {
+      "epoch": 1.7244210268349467,
+      "grad_norm": 0.40818363428115845,
+      "learning_rate": 4.956475099620433e-07,
+      "loss": 0.3606753349304199,
+      "step": 1760
+    },
+    {
+      "epoch": 1.725401298860434,
+      "grad_norm": 23.671892166137695,
+      "learning_rate": 4.922094249306559e-07,
+      "loss": 0.42479193210601807,
+      "step": 1761
+    },
+    {
+      "epoch": 1.7263815708859207,
+      "grad_norm": 0.7296469211578369,
+      "learning_rate": 4.887826882566699e-07,
+      "loss": 0.46341702342033386,
+      "step": 1762
+    },
+    {
+      "epoch": 1.727361842911408,
+      "grad_norm": 1.304168462753296,
+      "learning_rate": 4.853673085668947e-07,
+      "loss": 0.4242638349533081,
+      "step": 1763
+    },
+    {
+      "epoch": 1.728342114936895,
+      "grad_norm": 0.487373948097229,
+      "learning_rate": 4.819632944595415e-07,
+      "loss": 0.5017347931861877,
+      "step": 1764
+    },
+    {
+      "epoch": 1.729322386962382,
+      "grad_norm": 0.51096111536026,
+      "learning_rate": 4.785706545042141e-07,
+      "loss": 0.4666666090488434,
+      "step": 1765
+    },
+    {
+      "epoch": 1.7303026589878692,
+      "grad_norm": 0.5835421085357666,
+      "learning_rate": 4.751893972418781e-07,
+      "loss": 0.43100428581237793,
+      "step": 1766
+    },
+    {
+      "epoch": 1.7312829310133562,
+      "grad_norm": 0.8265053033828735,
+      "learning_rate": 4.7181953118484556e-07,
+      "loss": 0.5166634321212769,
+      "step": 1767
+    },
+    {
+      "epoch": 1.7322632030388432,
+      "grad_norm": 0.47917771339416504,
+      "learning_rate": 4.6846106481675035e-07,
+      "loss": 0.3971092700958252,
+      "step": 1768
+    },
+    {
+      "epoch": 1.7332434750643304,
+      "grad_norm": 0.5350149273872375,
+      "learning_rate": 4.651140065925269e-07,
+      "loss": 0.413171648979187,
+      "step": 1769
+    },
+    {
+      "epoch": 1.7342237470898174,
+      "grad_norm": 0.4752916097640991,
+      "learning_rate": 4.617783649383906e-07,
+      "loss": 0.4330071210861206,
+      "step": 1770
+    },
+    {
+      "epoch": 1.7352040191153044,
+      "grad_norm": 0.5725945830345154,
+      "learning_rate": 4.58454148251814e-07,
+      "loss": 0.4238564968109131,
+      "step": 1771
+    },
+    {
+      "epoch": 1.7361842911407916,
+      "grad_norm": 0.4779057204723358,
+      "learning_rate": 4.551413649015113e-07,
+      "loss": 0.3402918577194214,
+      "step": 1772
+    },
+    {
+      "epoch": 1.7371645631662787,
+      "grad_norm": 0.6874942779541016,
+      "learning_rate": 4.5184002322740784e-07,
+      "loss": 0.45461729168891907,
+      "step": 1773
+    },
+    {
+      "epoch": 1.7381448351917657,
+      "grad_norm": 0.5682085156440735,
+      "learning_rate": 4.48550131540631e-07,
+      "loss": 0.45160502195358276,
+      "step": 1774
+    },
+    {
+      "epoch": 1.7391251072172529,
+      "grad_norm": 0.5243704915046692,
+      "learning_rate": 4.4527169812347446e-07,
+      "loss": 0.4869271218776703,
+      "step": 1775
+    },
+    {
+      "epoch": 1.74010537924274,
+      "grad_norm": 0.568425714969635,
+      "learning_rate": 4.420047312293946e-07,
+      "loss": 0.5299988389015198,
+      "step": 1776
+    },
+    {
+      "epoch": 1.741085651268227,
+      "grad_norm": 0.47022855281829834,
+      "learning_rate": 4.387492390829734e-07,
+      "loss": 0.44766008853912354,
+      "step": 1777
+    },
+    {
+      "epoch": 1.7420659232937141,
+      "grad_norm": 0.5842577219009399,
+      "learning_rate": 4.355052298799112e-07,
+      "loss": 0.4448380470275879,
+      "step": 1778
+    },
+    {
+      "epoch": 1.7430461953192011,
+      "grad_norm": 0.6317983865737915,
+      "learning_rate": 4.322727117869951e-07,
+      "loss": 0.46058422327041626,
+      "step": 1779
+    },
+    {
+      "epoch": 1.7440264673446881,
+      "grad_norm": 0.6467611193656921,
+      "learning_rate": 4.2905169294208846e-07,
+      "loss": 0.4474758207798004,
+      "step": 1780
+    },
+    {
+      "epoch": 1.7450067393701754,
+      "grad_norm": 0.6873292922973633,
+      "learning_rate": 4.2584218145409916e-07,
+      "loss": 0.38277098536491394,
+      "step": 1781
+    },
+    {
+      "epoch": 1.7459870113956621,
+      "grad_norm": 0.42666471004486084,
+      "learning_rate": 4.2264418540297e-07,
+      "loss": 0.4441336989402771,
+      "step": 1782
+    },
+    {
+      "epoch": 1.7469672834211494,
+      "grad_norm": 0.4923200011253357,
+      "learning_rate": 4.194577128396521e-07,
+      "loss": 0.40558958053588867,
+      "step": 1783
+    },
+    {
+      "epoch": 1.7479475554466366,
+      "grad_norm": 1.9058901071548462,
+      "learning_rate": 4.1628277178608677e-07,
+      "loss": 0.6052329540252686,
+      "step": 1784
+    },
+    {
+      "epoch": 1.7489278274721234,
+      "grad_norm": 0.5461837649345398,
+      "learning_rate": 4.131193702351827e-07,
+      "loss": 0.5211032032966614,
+      "step": 1785
+    },
+    {
+      "epoch": 1.7499080994976106,
+      "grad_norm": 0.442888468503952,
+      "learning_rate": 4.099675161508021e-07,
+      "loss": 0.4277229905128479,
+      "step": 1786
+    },
+    {
+      "epoch": 1.7508883715230976,
+      "grad_norm": 0.5102304816246033,
+      "learning_rate": 4.0682721746773346e-07,
+      "loss": 0.43259677290916443,
+      "step": 1787
+    },
+    {
+      "epoch": 1.7518686435485846,
+      "grad_norm": 0.5137483477592468,
+      "learning_rate": 4.036984820916723e-07,
+      "loss": 0.4784802496433258,
+      "step": 1788
+    },
+    {
+      "epoch": 1.7528489155740719,
+      "grad_norm": 0.6571153998374939,
+      "learning_rate": 4.005813178992091e-07,
+      "loss": 0.4951273798942566,
+      "step": 1789
+    },
+    {
+      "epoch": 1.7538291875995589,
+      "grad_norm": 0.4703921675682068,
+      "learning_rate": 3.9747573273779816e-07,
+      "loss": 0.5274416208267212,
+      "step": 1790
+    },
+    {
+      "epoch": 1.7548094596250459,
+      "grad_norm": 0.47288310527801514,
+      "learning_rate": 3.9438173442575e-07,
+      "loss": 0.40976738929748535,
+      "step": 1791
+    },
+    {
+      "epoch": 1.755789731650533,
+      "grad_norm": 0.54091477394104,
+      "learning_rate": 3.9129933075219837e-07,
+      "loss": 0.4047437608242035,
+      "step": 1792
+    },
+    {
+      "epoch": 1.75677000367602,
+      "grad_norm": 0.41091862320899963,
+      "learning_rate": 3.882285294770938e-07,
+      "loss": 0.47677528858184814,
+      "step": 1793
+    },
+    {
+      "epoch": 1.757750275701507,
+      "grad_norm": 0.5608519315719604,
+      "learning_rate": 3.851693383311722e-07,
+      "loss": 0.3766956925392151,
+      "step": 1794
+    },
+    {
+      "epoch": 1.7587305477269943,
+      "grad_norm": 1.9002785682678223,
+      "learning_rate": 3.821217650159453e-07,
+      "loss": 0.5391687750816345,
+      "step": 1795
+    },
+    {
+      "epoch": 1.7597108197524813,
+      "grad_norm": 0.4525732696056366,
+      "learning_rate": 3.790858172036738e-07,
+      "loss": 0.4377427101135254,
+      "step": 1796
+    },
+    {
+      "epoch": 1.7606910917779683,
+      "grad_norm": 0.6908583045005798,
+      "learning_rate": 3.760615025373543e-07,
+      "loss": 0.46291735768318176,
+      "step": 1797
+    },
+    {
+      "epoch": 1.7616713638034556,
+      "grad_norm": 0.5039394497871399,
+      "learning_rate": 3.730488286306944e-07,
+      "loss": 0.5032774209976196,
+      "step": 1798
+    },
+    {
+      "epoch": 1.7626516358289426,
+      "grad_norm": 0.638607919216156,
+      "learning_rate": 3.7004780306809873e-07,
+      "loss": 0.40094488859176636,
+      "step": 1799
+    },
+    {
+      "epoch": 1.7636319078544296,
+      "grad_norm": 0.8207895159721375,
+      "learning_rate": 3.6705843340464287e-07,
+      "loss": 0.5478153824806213,
+      "step": 1800
+    },
+    {
+      "epoch": 1.7646121798799168,
+      "grad_norm": 0.6444453001022339,
+      "learning_rate": 3.6408072716606346e-07,
+      "loss": 0.44778549671173096,
+      "step": 1801
+    },
+    {
+      "epoch": 1.7655924519054036,
+      "grad_norm": 0.5676345229148865,
+      "learning_rate": 3.611146918487307e-07,
+      "loss": 0.44193071126937866,
+      "step": 1802
+    },
+    {
+      "epoch": 1.7665727239308908,
+      "grad_norm": 0.5010296702384949,
+      "learning_rate": 3.581603349196372e-07,
+      "loss": 0.4195302724838257,
+      "step": 1803
+    },
+    {
+      "epoch": 1.767552995956378,
+      "grad_norm": 0.6718754172325134,
+      "learning_rate": 3.552176638163707e-07,
+      "loss": 0.34599247574806213,
+      "step": 1804
+    },
+    {
+      "epoch": 1.7685332679818648,
+      "grad_norm": 0.6563330292701721,
+      "learning_rate": 3.522866859471047e-07,
+      "loss": 0.5447871685028076,
+      "step": 1805
+    },
+    {
+      "epoch": 1.769513540007352,
+      "grad_norm": 0.5016887784004211,
+      "learning_rate": 3.4936740869057075e-07,
+      "loss": 0.48268669843673706,
+      "step": 1806
+    },
+    {
+      "epoch": 1.770493812032839,
+      "grad_norm": 0.7236999273300171,
+      "learning_rate": 3.46459839396045e-07,
+      "loss": 0.44142189621925354,
+      "step": 1807
+    },
+    {
+      "epoch": 1.771474084058326,
+      "grad_norm": 1.996353030204773,
+      "learning_rate": 3.435639853833317e-07,
+      "loss": 0.42545628547668457,
+      "step": 1808
+    },
+    {
+      "epoch": 1.7724543560838133,
+      "grad_norm": 0.6237255930900574,
+      "learning_rate": 3.406798539427386e-07,
+      "loss": 0.4861873984336853,
+      "step": 1809
+    },
+    {
+      "epoch": 1.7734346281093003,
+      "grad_norm": 0.4381408393383026,
+      "learning_rate": 3.3780745233506384e-07,
+      "loss": 0.4135734438896179,
+      "step": 1810
+    },
+    {
+      "epoch": 1.7744149001347873,
+      "grad_norm": 0.5980325937271118,
+      "learning_rate": 3.3494678779157464e-07,
+      "loss": 0.4729668200016022,
+      "step": 1811
+    },
+    {
+      "epoch": 1.7753951721602745,
+      "grad_norm": 0.5451025366783142,
+      "learning_rate": 3.320978675139919e-07,
+      "loss": 0.4551047086715698,
+      "step": 1812
+    },
+    {
+      "epoch": 1.7763754441857615,
+      "grad_norm": 0.5626180171966553,
+      "learning_rate": 3.2926069867446673e-07,
+      "loss": 0.46117469668388367,
+      "step": 1813
+    },
+    {
+      "epoch": 1.7773557162112485,
+      "grad_norm": 0.5022594332695007,
+      "learning_rate": 3.2643528841556973e-07,
+      "loss": 0.5012738704681396,
+      "step": 1814
+    },
+    {
+      "epoch": 1.7783359882367358,
+      "grad_norm": 0.622039794921875,
+      "learning_rate": 3.2362164385026704e-07,
+      "loss": 0.5177352428436279,
+      "step": 1815
+    },
+    {
+      "epoch": 1.7793162602622228,
+      "grad_norm": 0.5423142910003662,
+      "learning_rate": 3.2081977206190715e-07,
+      "loss": 0.6440224051475525,
+      "step": 1816
+    },
+    {
+      "epoch": 1.7802965322877098,
+      "grad_norm": 0.48715320229530334,
+      "learning_rate": 3.180296801041971e-07,
+      "loss": 0.40212482213974,
+      "step": 1817
+    },
+    {
+      "epoch": 1.781276804313197,
+      "grad_norm": 0.5168865919113159,
+      "learning_rate": 3.1525137500119207e-07,
+      "loss": 0.5141654014587402,
+      "step": 1818
+    },
+    {
+      "epoch": 1.782257076338684,
+      "grad_norm": 0.4482264816761017,
+      "learning_rate": 3.1248486374726884e-07,
+      "loss": 0.36538636684417725,
+      "step": 1819
+    },
+    {
+      "epoch": 1.783237348364171,
+      "grad_norm": 0.8960753679275513,
+      "learning_rate": 3.09730153307119e-07,
+      "loss": 0.5195493698120117,
+      "step": 1820
+    },
+    {
+      "epoch": 1.7842176203896583,
+      "grad_norm": 0.858798623085022,
+      "learning_rate": 3.069872506157212e-07,
+      "loss": 0.3850517272949219,
+      "step": 1821
+    },
+    {
+      "epoch": 1.785197892415145,
+      "grad_norm": 0.8067991137504578,
+      "learning_rate": 3.042561625783308e-07,
+      "loss": 0.35192543268203735,
+      "step": 1822
+    },
+    {
+      "epoch": 1.7861781644406323,
+      "grad_norm": 0.6098344326019287,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.45531773567199707,
+      "step": 1823
+    },
+    {
+      "epoch": 1.7871584364661195,
+      "grad_norm": 0.45518630743026733,
+      "learning_rate": 2.988294579378537e-07,
+      "loss": 0.405795693397522,
+      "step": 1824
+    },
+    {
+      "epoch": 1.7881387084916063,
+      "grad_norm": 1.0298306941986084,
+      "learning_rate": 2.9613385499648926e-07,
+      "loss": 0.4920778274536133,
+      "step": 1825
+    },
+    {
+      "epoch": 1.7891189805170935,
+      "grad_norm": 0.8277662396430969,
+      "learning_rate": 2.934500940325413e-07,
+      "loss": 0.44633984565734863,
+      "step": 1826
+    },
+    {
+      "epoch": 1.7900992525425807,
+      "grad_norm": 0.7844619154930115,
+      "learning_rate": 2.9077818180237693e-07,
+      "loss": 0.4384131133556366,
+      "step": 1827
+    },
+    {
+      "epoch": 1.7910795245680675,
+      "grad_norm": 0.4697599411010742,
+      "learning_rate": 2.881181250325299e-07,
+      "loss": 0.360952764749527,
+      "step": 1828
+    },
+    {
+      "epoch": 1.7920597965935547,
+      "grad_norm": 1.0636643171310425,
+      "learning_rate": 2.8546993041969173e-07,
+      "loss": 0.35457491874694824,
+      "step": 1829
+    },
+    {
+      "epoch": 1.7930400686190417,
+      "grad_norm": 0.4991975426673889,
+      "learning_rate": 2.8283360463068786e-07,
+      "loss": 0.39283955097198486,
+      "step": 1830
+    },
+    {
+      "epoch": 1.7940203406445288,
+      "grad_norm": 0.4297543168067932,
+      "learning_rate": 2.802091543024671e-07,
+      "loss": 0.4083915054798126,
+      "step": 1831
+    },
+    {
+      "epoch": 1.795000612670016,
+      "grad_norm": 0.5845555067062378,
+      "learning_rate": 2.7759658604207705e-07,
+      "loss": 0.5623732805252075,
+      "step": 1832
+    },
+    {
+      "epoch": 1.795980884695503,
+      "grad_norm": 0.6158885359764099,
+      "learning_rate": 2.7499590642665773e-07,
+      "loss": 0.48206067085266113,
+      "step": 1833
+    },
+    {
+      "epoch": 1.79696115672099,
+      "grad_norm": 2.715228319168091,
+      "learning_rate": 2.724071220034158e-07,
+      "loss": 0.45180177688598633,
+      "step": 1834
+    },
+    {
+      "epoch": 1.7979414287464772,
+      "grad_norm": 0.6644836068153381,
+      "learning_rate": 2.6983023928961406e-07,
+      "loss": 0.37124502658843994,
+      "step": 1835
+    },
+    {
+      "epoch": 1.7989217007719642,
+      "grad_norm": 0.4559076428413391,
+      "learning_rate": 2.6726526477254986e-07,
+      "loss": 0.4645237922668457,
+      "step": 1836
+    },
+    {
+      "epoch": 1.7999019727974512,
+      "grad_norm": 0.5588224530220032,
+      "learning_rate": 2.647122049095463e-07,
+      "loss": 0.42912042140960693,
+      "step": 1837
+    },
+    {
+      "epoch": 1.8008822448229385,
+      "grad_norm": 0.4895893931388855,
+      "learning_rate": 2.621710661279253e-07,
+      "loss": 0.4239204525947571,
+      "step": 1838
+    },
+    {
+      "epoch": 1.8018625168484255,
+      "grad_norm": 0.5144309997558594,
+      "learning_rate": 2.596418548250029e-07,
+      "loss": 0.4330456256866455,
+      "step": 1839
+    },
+    {
+      "epoch": 1.8028427888739125,
+      "grad_norm": 0.42308276891708374,
+      "learning_rate": 2.571245773680647e-07,
+      "loss": 0.3533886671066284,
+      "step": 1840
+    },
+    {
+      "epoch": 1.8038230608993997,
+      "grad_norm": 0.5395842790603638,
+      "learning_rate": 2.546192400943537e-07,
+      "loss": 0.5010573267936707,
+      "step": 1841
+    },
+    {
+      "epoch": 1.8048033329248867,
+      "grad_norm": 0.5580001473426819,
+      "learning_rate": 2.52125849311054e-07,
+      "loss": 0.5050355195999146,
+      "step": 1842
+    },
+    {
+      "epoch": 1.8057836049503737,
+      "grad_norm": 0.5212836861610413,
+      "learning_rate": 2.4964441129527337e-07,
+      "loss": 0.6462127566337585,
+      "step": 1843
+    },
+    {
+      "epoch": 1.806763876975861,
+      "grad_norm": 0.38223889470100403,
+      "learning_rate": 2.4717493229402976e-07,
+      "loss": 0.32172685861587524,
+      "step": 1844
+    },
+    {
+      "epoch": 1.8077441490013477,
+      "grad_norm": 0.5720911026000977,
+      "learning_rate": 2.447174185242324e-07,
+      "loss": 0.431645005941391,
+      "step": 1845
+    },
+    {
+      "epoch": 1.808724421026835,
+      "grad_norm": 0.5592072606086731,
+      "learning_rate": 2.4227187617267057e-07,
+      "loss": 0.45566949248313904,
+      "step": 1846
+    },
+    {
+      "epoch": 1.8097046930523222,
+      "grad_norm": 0.3950572907924652,
+      "learning_rate": 2.398383113959929e-07,
+      "loss": 0.39810889959335327,
+      "step": 1847
+    },
+    {
+      "epoch": 1.810684965077809,
+      "grad_norm": 0.4785499572753906,
+      "learning_rate": 2.3741673032069757e-07,
+      "loss": 0.3847421109676361,
+      "step": 1848
+    },
+    {
+      "epoch": 1.8116652371032962,
+      "grad_norm": 1.3525609970092773,
+      "learning_rate": 2.3500713904311023e-07,
+      "loss": 0.4135790467262268,
+      "step": 1849
+    },
+    {
+      "epoch": 1.8126455091287832,
+      "grad_norm": 0.4805648922920227,
+      "learning_rate": 2.3260954362937626e-07,
+      "loss": 0.41442999243736267,
+      "step": 1850
+    },
+    {
+      "epoch": 1.8136257811542702,
+      "grad_norm": 0.6130191683769226,
+      "learning_rate": 2.3022395011543687e-07,
+      "loss": 0.5303109884262085,
+      "step": 1851
+    },
+    {
+      "epoch": 1.8146060531797574,
+      "grad_norm": 0.5477684736251831,
+      "learning_rate": 2.2785036450702237e-07,
+      "loss": 0.42619743943214417,
+      "step": 1852
+    },
+    {
+      "epoch": 1.8155863252052444,
+      "grad_norm": 0.5172080993652344,
+      "learning_rate": 2.2548879277963065e-07,
+      "loss": 0.4615512490272522,
+      "step": 1853
+    },
+    {
+      "epoch": 1.8165665972307314,
+      "grad_norm": 0.47979018092155457,
+      "learning_rate": 2.2313924087851657e-07,
+      "loss": 0.42331331968307495,
+      "step": 1854
+    },
+    {
+      "epoch": 1.8175468692562187,
+      "grad_norm": 0.606453001499176,
+      "learning_rate": 2.2080171471867362e-07,
+      "loss": 0.4292833209037781,
+      "step": 1855
+    },
+    {
+      "epoch": 1.8185271412817057,
+      "grad_norm": 0.5634953379631042,
+      "learning_rate": 2.1847622018482283e-07,
+      "loss": 0.5072528123855591,
+      "step": 1856
+    },
+    {
+      "epoch": 1.8195074133071927,
+      "grad_norm": 0.466601699590683,
+      "learning_rate": 2.161627631313923e-07,
+      "loss": 0.42179855704307556,
+      "step": 1857
+    },
+    {
+      "epoch": 1.82048768533268,
+      "grad_norm": 1.6157302856445312,
+      "learning_rate": 2.1386134938250645e-07,
+      "loss": 0.46553200483322144,
+      "step": 1858
+    },
+    {
+      "epoch": 1.821467957358167,
+      "grad_norm": 0.416071355342865,
+      "learning_rate": 2.1157198473197417e-07,
+      "loss": 0.3804522752761841,
+      "step": 1859
+    },
+    {
+      "epoch": 1.822448229383654,
+      "grad_norm": 0.4381702244281769,
+      "learning_rate": 2.0929467494326616e-07,
+      "loss": 0.3243664801120758,
+      "step": 1860
+    },
+    {
+      "epoch": 1.8234285014091411,
+      "grad_norm": 0.4505363404750824,
+      "learning_rate": 2.0702942574950812e-07,
+      "loss": 0.4905754327774048,
+      "step": 1861
+    },
+    {
+      "epoch": 1.8244087734346282,
+      "grad_norm": 0.5357232093811035,
+      "learning_rate": 2.047762428534622e-07,
+      "loss": 0.5555627346038818,
+      "step": 1862
+    },
+    {
+      "epoch": 1.8253890454601152,
+      "grad_norm": 0.80414217710495,
+      "learning_rate": 2.0253513192751374e-07,
+      "loss": 0.6146029233932495,
+      "step": 1863
+    },
+    {
+      "epoch": 1.8263693174856024,
+      "grad_norm": 0.4719810485839844,
+      "learning_rate": 2.0030609861365568e-07,
+      "loss": 0.43739885091781616,
+      "step": 1864
+    },
+    {
+      "epoch": 1.8273495895110892,
+      "grad_norm": 0.5847609639167786,
+      "learning_rate": 1.9808914852347817e-07,
+      "loss": 0.46729379892349243,
+      "step": 1865
+    },
+    {
+      "epoch": 1.8283298615365764,
+      "grad_norm": 0.46554553508758545,
+      "learning_rate": 1.9588428723814945e-07,
+      "loss": 0.46161988377571106,
+      "step": 1866
+    },
+    {
+      "epoch": 1.8293101335620636,
+      "grad_norm": 0.7154317498207092,
+      "learning_rate": 1.9369152030840553e-07,
+      "loss": 0.4180935025215149,
+      "step": 1867
+    },
+    {
+      "epoch": 1.8302904055875504,
+      "grad_norm": 0.5231561660766602,
+      "learning_rate": 1.9151085325453512e-07,
+      "loss": 0.38244593143463135,
+      "step": 1868
+    },
+    {
+      "epoch": 1.8312706776130376,
+      "grad_norm": 0.4596077799797058,
+      "learning_rate": 1.8934229156636453e-07,
+      "loss": 0.3760368227958679,
+      "step": 1869
+    },
+    {
+      "epoch": 1.8322509496385246,
+      "grad_norm": 2.262033700942993,
+      "learning_rate": 1.8718584070324452e-07,
+      "loss": 0.5451330542564392,
+      "step": 1870
+    },
+    {
+      "epoch": 1.8332312216640116,
+      "grad_norm": 0.4730428159236908,
+      "learning_rate": 1.8504150609403858e-07,
+      "loss": 0.44578754901885986,
+      "step": 1871
+    },
+    {
+      "epoch": 1.8342114936894989,
+      "grad_norm": 0.4594602584838867,
+      "learning_rate": 1.8290929313710514e-07,
+      "loss": 0.460705041885376,
+      "step": 1872
+    },
+    {
+      "epoch": 1.8351917657149859,
+      "grad_norm": 0.4803224503993988,
+      "learning_rate": 1.807892072002898e-07,
+      "loss": 0.3790692389011383,
+      "step": 1873
+    },
+    {
+      "epoch": 1.8361720377404729,
+      "grad_norm": 0.815172553062439,
+      "learning_rate": 1.7868125362090482e-07,
+      "loss": 0.39920926094055176,
+      "step": 1874
+    },
+    {
+      "epoch": 1.8371523097659601,
+      "grad_norm": 0.577421247959137,
+      "learning_rate": 1.765854377057219e-07,
+      "loss": 0.3441392779350281,
+      "step": 1875
+    },
+    {
+      "epoch": 1.8381325817914471,
+      "grad_norm": 0.5438459515571594,
+      "learning_rate": 1.7450176473095436e-07,
+      "loss": 0.37040749192237854,
+      "step": 1876
+    },
+    {
+      "epoch": 1.8391128538169341,
+      "grad_norm": 2.920234203338623,
+      "learning_rate": 1.724302399422456e-07,
+      "loss": 0.6004539132118225,
+      "step": 1877
+    },
+    {
+      "epoch": 1.8400931258424214,
+      "grad_norm": 0.7245340943336487,
+      "learning_rate": 1.7037086855465902e-07,
+      "loss": 0.41499757766723633,
+      "step": 1878
+    },
+    {
+      "epoch": 1.8410733978679084,
+      "grad_norm": 0.4374197721481323,
+      "learning_rate": 1.6832365575265742e-07,
+      "loss": 0.44313156604766846,
+      "step": 1879
+    },
+    {
+      "epoch": 1.8420536698933954,
+      "grad_norm": 0.6867006421089172,
+      "learning_rate": 1.6628860669009816e-07,
+      "loss": 0.45141398906707764,
+      "step": 1880
+    },
+    {
+      "epoch": 1.8430339419188826,
+      "grad_norm": 0.4403974115848541,
+      "learning_rate": 1.6426572649021477e-07,
+      "loss": 0.4557759761810303,
+      "step": 1881
+    },
+    {
+      "epoch": 1.8440142139443696,
+      "grad_norm": 0.7186617255210876,
+      "learning_rate": 1.6225502024560468e-07,
+      "loss": 0.5363942980766296,
+      "step": 1882
+    },
+    {
+      "epoch": 1.8449944859698566,
+      "grad_norm": 2.0424888134002686,
+      "learning_rate": 1.6025649301821877e-07,
+      "loss": 0.4392750859260559,
+      "step": 1883
+    },
+    {
+      "epoch": 1.8459747579953438,
+      "grad_norm": 0.5844815969467163,
+      "learning_rate": 1.582701498393474e-07,
+      "loss": 0.47282278537750244,
+      "step": 1884
+    },
+    {
+      "epoch": 1.8469550300208308,
+      "grad_norm": 0.4067746698856354,
+      "learning_rate": 1.562959957096072e-07,
+      "loss": 0.36924052238464355,
+      "step": 1885
+    },
+    {
+      "epoch": 1.8479353020463178,
+      "grad_norm": 0.5178905129432678,
+      "learning_rate": 1.543340355989287e-07,
+      "loss": 0.48047885298728943,
+      "step": 1886
+    },
+    {
+      "epoch": 1.848915574071805,
+      "grad_norm": 1.0028716325759888,
+      "learning_rate": 1.5238427444654368e-07,
+      "loss": 0.49795693159103394,
+      "step": 1887
+    },
+    {
+      "epoch": 1.8498958460972919,
+      "grad_norm": 0.49718478322029114,
+      "learning_rate": 1.5044671716097414e-07,
+      "loss": 0.4911704361438751,
+      "step": 1888
+    },
+    {
+      "epoch": 1.850876118122779,
+      "grad_norm": 0.5813276171684265,
+      "learning_rate": 1.4852136862001766e-07,
+      "loss": 0.4910121560096741,
+      "step": 1889
+    },
+    {
+      "epoch": 1.8518563901482663,
+      "grad_norm": 0.5754580497741699,
+      "learning_rate": 1.466082336707375e-07,
+      "loss": 0.5036097168922424,
+      "step": 1890
+    },
+    {
+      "epoch": 1.852836662173753,
+      "grad_norm": 0.47939202189445496,
+      "learning_rate": 1.4470731712944885e-07,
+      "loss": 0.481952041387558,
+      "step": 1891
+    },
+    {
+      "epoch": 1.8538169341992403,
+      "grad_norm": 0.7401548027992249,
+      "learning_rate": 1.428186237817053e-07,
+      "loss": 0.3692576289176941,
+      "step": 1892
+    },
+    {
+      "epoch": 1.8547972062247273,
+      "grad_norm": 0.6041969656944275,
+      "learning_rate": 1.4094215838229176e-07,
+      "loss": 0.41709646582603455,
+      "step": 1893
+    },
+    {
+      "epoch": 1.8557774782502143,
+      "grad_norm": 0.4660212993621826,
+      "learning_rate": 1.3907792565520716e-07,
+      "loss": 0.43801939487457275,
+      "step": 1894
+    },
+    {
+      "epoch": 1.8567577502757016,
+      "grad_norm": 0.4720137119293213,
+      "learning_rate": 1.372259302936546e-07,
+      "loss": 0.4119502902030945,
+      "step": 1895
+    },
+    {
+      "epoch": 1.8577380223011886,
+      "grad_norm": 0.6017296314239502,
+      "learning_rate": 1.3538617696003066e-07,
+      "loss": 0.3752719759941101,
+      "step": 1896
+    },
+    {
+      "epoch": 1.8587182943266756,
+      "grad_norm": 0.5979307889938354,
+      "learning_rate": 1.3355867028591209e-07,
+      "loss": 0.3951225280761719,
+      "step": 1897
+    },
+    {
+      "epoch": 1.8596985663521628,
+      "grad_norm": 0.5684795379638672,
+      "learning_rate": 1.3174341487204478e-07,
+      "loss": 0.40364813804626465,
+      "step": 1898
+    },
+    {
+      "epoch": 1.8606788383776498,
+      "grad_norm": 0.47828906774520874,
+      "learning_rate": 1.2994041528833267e-07,
+      "loss": 0.44901227951049805,
+      "step": 1899
+    },
+    {
+      "epoch": 1.8616591104031368,
+      "grad_norm": 0.6910325884819031,
+      "learning_rate": 1.2814967607382433e-07,
+      "loss": 0.44203782081604004,
+      "step": 1900
+    },
+    {
+      "epoch": 1.862639382428624,
+      "grad_norm": 0.5532023310661316,
+      "learning_rate": 1.263712017367036e-07,
+      "loss": 0.38102787733078003,
+      "step": 1901
+    },
+    {
+      "epoch": 1.863619654454111,
+      "grad_norm": 0.5694209337234497,
+      "learning_rate": 1.246049967542773e-07,
+      "loss": 0.46240073442459106,
+      "step": 1902
+    },
+    {
+      "epoch": 1.864599926479598,
+      "grad_norm": 0.751404881477356,
+      "learning_rate": 1.2285106557296479e-07,
+      "loss": 0.43973201513290405,
+      "step": 1903
+    },
+    {
+      "epoch": 1.8655801985050853,
+      "grad_norm": 0.8471018075942993,
+      "learning_rate": 1.211094126082857e-07,
+      "loss": 0.4042569100856781,
+      "step": 1904
+    },
+    {
+      "epoch": 1.8665604705305723,
+      "grad_norm": 0.513670802116394,
+      "learning_rate": 1.193800422448499e-07,
+      "loss": 0.42281681299209595,
+      "step": 1905
+    },
+    {
+      "epoch": 1.8675407425560593,
+      "grad_norm": 0.5502100586891174,
+      "learning_rate": 1.1766295883634427e-07,
+      "loss": 0.4615674614906311,
+      "step": 1906
+    },
+    {
+      "epoch": 1.8685210145815465,
+      "grad_norm": 0.49501538276672363,
+      "learning_rate": 1.1595816670552429e-07,
+      "loss": 0.5072102546691895,
+      "step": 1907
+    },
+    {
+      "epoch": 1.8695012866070333,
+      "grad_norm": 0.44755715131759644,
+      "learning_rate": 1.1426567014420297e-07,
+      "loss": 0.4582875967025757,
+      "step": 1908
+    },
+    {
+      "epoch": 1.8704815586325205,
+      "grad_norm": 0.502082347869873,
+      "learning_rate": 1.12585473413237e-07,
+      "loss": 0.4758835434913635,
+      "step": 1909
+    },
+    {
+      "epoch": 1.8714618306580078,
+      "grad_norm": 0.7396796345710754,
+      "learning_rate": 1.1091758074252112e-07,
+      "loss": 0.45714467763900757,
+      "step": 1910
+    },
+    {
+      "epoch": 1.8724421026834945,
+      "grad_norm": 0.6277248859405518,
+      "learning_rate": 1.0926199633097156e-07,
+      "loss": 0.44521600008010864,
+      "step": 1911
+    },
+    {
+      "epoch": 1.8734223747089818,
+      "grad_norm": 0.48554229736328125,
+      "learning_rate": 1.076187243465221e-07,
+      "loss": 0.355817973613739,
+      "step": 1912
+    },
+    {
+      "epoch": 1.8744026467344688,
+      "grad_norm": 0.44571366906166077,
+      "learning_rate": 1.0598776892610685e-07,
+      "loss": 0.3922297954559326,
+      "step": 1913
+    },
+    {
+      "epoch": 1.8753829187599558,
+      "grad_norm": 0.5140563249588013,
+      "learning_rate": 1.0436913417565364e-07,
+      "loss": 0.473746120929718,
+      "step": 1914
+    },
+    {
+      "epoch": 1.876363190785443,
+      "grad_norm": 0.5642273426055908,
+      "learning_rate": 1.0276282417007399e-07,
+      "loss": 0.2978225350379944,
+      "step": 1915
+    },
+    {
+      "epoch": 1.87734346281093,
+      "grad_norm": 0.9473456740379333,
+      "learning_rate": 1.0116884295325202e-07,
+      "loss": 0.4083951711654663,
+      "step": 1916
+    },
+    {
+      "epoch": 1.878323734836417,
+      "grad_norm": 1.8731892108917236,
+      "learning_rate": 9.958719453803278e-08,
+      "loss": 0.34825029969215393,
+      "step": 1917
+    },
+    {
+      "epoch": 1.8793040068619042,
+      "grad_norm": 0.4795823395252228,
+      "learning_rate": 9.801788290621505e-08,
+      "loss": 0.5096890926361084,
+      "step": 1918
+    },
+    {
+      "epoch": 1.8802842788873912,
+      "grad_norm": 0.802291214466095,
+      "learning_rate": 9.646091200853802e-08,
+      "loss": 0.49362629652023315,
+      "step": 1919
+    },
+    {
+      "epoch": 1.8812645509128783,
+      "grad_norm": 0.4294847249984741,
+      "learning_rate": 9.491628576467516e-08,
+      "loss": 0.342154860496521,
+      "step": 1920
+    },
+    {
+      "epoch": 1.8822448229383655,
+      "grad_norm": 1.4047214984893799,
+      "learning_rate": 9.338400806321979e-08,
+      "loss": 0.581269383430481,
+      "step": 1921
+    },
+    {
+      "epoch": 1.8832250949638525,
+      "grad_norm": 0.5004202723503113,
+      "learning_rate": 9.186408276168012e-08,
+      "loss": 0.4325605630874634,
+      "step": 1922
+    },
+    {
+      "epoch": 1.8842053669893395,
+      "grad_norm": 0.5487098693847656,
+      "learning_rate": 9.035651368646647e-08,
+      "loss": 0.4732184410095215,
+      "step": 1923
+    },
+    {
+      "epoch": 1.8851856390148267,
+      "grad_norm": 0.9227629899978638,
+      "learning_rate": 8.886130463288178e-08,
+      "loss": 0.512689471244812,
+      "step": 1924
+    },
+    {
+      "epoch": 1.8861659110403137,
+      "grad_norm": 0.5078418254852295,
+      "learning_rate": 8.737845936511335e-08,
+      "loss": 0.4965260624885559,
+      "step": 1925
+    },
+    {
+      "epoch": 1.8871461830658007,
+      "grad_norm": 0.8656265139579773,
+      "learning_rate": 8.590798161622227e-08,
+      "loss": 0.5351952314376831,
+      "step": 1926
+    },
+    {
+      "epoch": 1.888126455091288,
+      "grad_norm": 1.0230077505111694,
+      "learning_rate": 8.444987508813451e-08,
+      "loss": 0.5152793526649475,
+      "step": 1927
+    },
+    {
+      "epoch": 1.8891067271167747,
+      "grad_norm": 0.4701693654060364,
+      "learning_rate": 8.300414345163043e-08,
+      "loss": 0.44536110758781433,
+      "step": 1928
+    },
+    {
+      "epoch": 1.890086999142262,
+      "grad_norm": 0.6286927461624146,
+      "learning_rate": 8.157079034633974e-08,
+      "loss": 0.47006624937057495,
+      "step": 1929
+    },
+    {
+      "epoch": 1.8910672711677492,
+      "grad_norm": 0.54852294921875,
+      "learning_rate": 8.014981938072542e-08,
+      "loss": 0.4077466130256653,
+      "step": 1930
+    },
+    {
+      "epoch": 1.892047543193236,
+      "grad_norm": 0.5348506569862366,
+      "learning_rate": 7.874123413208145e-08,
+      "loss": 0.41892942786216736,
+      "step": 1931
+    },
+    {
+      "epoch": 1.8930278152187232,
+      "grad_norm": 0.6889854669570923,
+      "learning_rate": 7.734503814651906e-08,
+      "loss": 0.5781711339950562,
+      "step": 1932
+    },
+    {
+      "epoch": 1.8940080872442102,
+      "grad_norm": 1.1440167427062988,
+      "learning_rate": 7.59612349389599e-08,
+      "loss": 0.4548415541648865,
+      "step": 1933
+    },
+    {
+      "epoch": 1.8949883592696972,
+      "grad_norm": 9.44877815246582,
+      "learning_rate": 7.458982799312676e-08,
+      "loss": 0.39248502254486084,
+      "step": 1934
+    },
+    {
+      "epoch": 1.8959686312951844,
+      "grad_norm": 0.7582590579986572,
+      "learning_rate": 7.32308207615351e-08,
+      "loss": 0.4765550494194031,
+      "step": 1935
+    },
+    {
+      "epoch": 1.8969489033206715,
+      "grad_norm": 0.6219707131385803,
+      "learning_rate": 7.188421666548373e-08,
+      "loss": 0.5452197790145874,
+      "step": 1936
+    },
+    {
+      "epoch": 1.8979291753461585,
+      "grad_norm": 0.4136485457420349,
+      "learning_rate": 7.055001909504755e-08,
+      "loss": 0.34368371963500977,
+      "step": 1937
+    },
+    {
+      "epoch": 1.8989094473716457,
+      "grad_norm": 0.5863478183746338,
+      "learning_rate": 6.922823140906754e-08,
+      "loss": 0.39949125051498413,
+      "step": 1938
+    },
+    {
+      "epoch": 1.8998897193971327,
+      "grad_norm": 0.536173403263092,
+      "learning_rate": 6.791885693514134e-08,
+      "loss": 0.4445520341396332,
+      "step": 1939
+    },
+    {
+      "epoch": 1.9008699914226197,
+      "grad_norm": 0.4607829749584198,
+      "learning_rate": 6.662189896961824e-08,
+      "loss": 0.40381550788879395,
+      "step": 1940
+    },
+    {
+      "epoch": 1.901850263448107,
+      "grad_norm": 0.7370746731758118,
+      "learning_rate": 6.533736077758868e-08,
+      "loss": 0.391458123922348,
+      "step": 1941
+    },
+    {
+      "epoch": 1.902830535473594,
+      "grad_norm": 0.7983855605125427,
+      "learning_rate": 6.406524559287641e-08,
+      "loss": 0.43582460284233093,
+      "step": 1942
+    },
+    {
+      "epoch": 1.903810807499081,
+      "grad_norm": 1.3392868041992188,
+      "learning_rate": 6.280555661802857e-08,
+      "loss": 0.5032989978790283,
+      "step": 1943
+    },
+    {
+      "epoch": 1.9047910795245682,
+      "grad_norm": 0.8695182800292969,
+      "learning_rate": 6.15582970243117e-08,
+      "loss": 0.5524872541427612,
+      "step": 1944
+    },
+    {
+      "epoch": 1.9057713515500552,
+      "grad_norm": 0.6703439354896545,
+      "learning_rate": 6.032346995169968e-08,
+      "loss": 0.6040399670600891,
+      "step": 1945
+    },
+    {
+      "epoch": 1.9067516235755422,
+      "grad_norm": 0.48495686054229736,
+      "learning_rate": 5.9101078508869144e-08,
+      "loss": 0.4884125590324402,
+      "step": 1946
+    },
+    {
+      "epoch": 1.9077318956010294,
+      "grad_norm": 0.4106784462928772,
+      "learning_rate": 5.7891125773187896e-08,
+      "loss": 0.39191728830337524,
+      "step": 1947
+    },
+    {
+      "epoch": 1.9087121676265164,
+      "grad_norm": 0.44426536560058594,
+      "learning_rate": 5.669361479071156e-08,
+      "loss": 0.41355571150779724,
+      "step": 1948
+    },
+    {
+      "epoch": 1.9096924396520034,
+      "grad_norm": 0.4924589991569519,
+      "learning_rate": 5.550854857617194e-08,
+      "loss": 0.47964173555374146,
+      "step": 1949
+    },
+    {
+      "epoch": 1.9106727116774906,
+      "grad_norm": 0.5969910025596619,
+      "learning_rate": 5.4335930112972e-08,
+      "loss": 0.5252376794815063,
+      "step": 1950
+    },
+    {
+      "epoch": 1.9116529837029774,
+      "grad_norm": 0.43021538853645325,
+      "learning_rate": 5.3175762353177563e-08,
+      "loss": 0.3906043767929077,
+      "step": 1951
+    },
+    {
+      "epoch": 1.9126332557284647,
+      "grad_norm": 0.5627011060714722,
+      "learning_rate": 5.202804821750951e-08,
+      "loss": 0.3485060930252075,
+      "step": 1952
+    },
+    {
+      "epoch": 1.9136135277539519,
+      "grad_norm": 0.5871269106864929,
+      "learning_rate": 5.089279059533658e-08,
+      "loss": 0.3971886932849884,
+      "step": 1953
+    },
+    {
+      "epoch": 1.9145937997794387,
+      "grad_norm": 0.6854955554008484,
+      "learning_rate": 4.9769992344668174e-08,
+      "loss": 0.3901163935661316,
+      "step": 1954
+    },
+    {
+      "epoch": 1.915574071804926,
+      "grad_norm": 0.4564671814441681,
+      "learning_rate": 4.865965629214819e-08,
+      "loss": 0.42336922883987427,
+      "step": 1955
+    },
+    {
+      "epoch": 1.916554343830413,
+      "grad_norm": 0.5906099081039429,
+      "learning_rate": 4.756178523304622e-08,
+      "loss": 0.4308120608329773,
+      "step": 1956
+    },
+    {
+      "epoch": 1.9175346158559,
+      "grad_norm": 0.6054607629776001,
+      "learning_rate": 4.6476381931251366e-08,
+      "loss": 0.4885075092315674,
+      "step": 1957
+    },
+    {
+      "epoch": 1.9185148878813871,
+      "grad_norm": 0.6015900373458862,
+      "learning_rate": 4.5403449119265085e-08,
+      "loss": 0.5340933799743652,
+      "step": 1958
+    },
+    {
+      "epoch": 1.9194951599068741,
+      "grad_norm": 0.5624886155128479,
+      "learning_rate": 4.434298949819449e-08,
+      "loss": 0.5054583549499512,
+      "step": 1959
+    },
+    {
+      "epoch": 1.9204754319323611,
+      "grad_norm": 3.656388282775879,
+      "learning_rate": 4.329500573774459e-08,
+      "loss": 0.4391195774078369,
+      "step": 1960
+    },
+    {
+      "epoch": 1.9214557039578484,
+      "grad_norm": 0.7404154539108276,
+      "learning_rate": 4.225950047621441e-08,
+      "loss": 0.39996451139450073,
+      "step": 1961
+    },
+    {
+      "epoch": 1.9224359759833354,
+      "grad_norm": 0.5697031021118164,
+      "learning_rate": 4.123647632048644e-08,
+      "loss": 0.4455189108848572,
+      "step": 1962
+    },
+    {
+      "epoch": 1.9234162480088224,
+      "grad_norm": 0.6449862122535706,
+      "learning_rate": 4.02259358460233e-08,
+      "loss": 0.3961139917373657,
+      "step": 1963
+    },
+    {
+      "epoch": 1.9243965200343096,
+      "grad_norm": 0.5936830639839172,
+      "learning_rate": 3.922788159685942e-08,
+      "loss": 0.581050455570221,
+      "step": 1964
+    },
+    {
+      "epoch": 1.9253767920597966,
+      "grad_norm": 0.70735102891922,
+      "learning_rate": 3.8242316085594923e-08,
+      "loss": 0.4737338423728943,
+      "step": 1965
+    },
+    {
+      "epoch": 1.9263570640852836,
+      "grad_norm": 0.49195075035095215,
+      "learning_rate": 3.726924179339009e-08,
+      "loss": 0.44615018367767334,
+      "step": 1966
+    },
+    {
+      "epoch": 1.9273373361107708,
+      "grad_norm": 0.48140597343444824,
+      "learning_rate": 3.630866116995757e-08,
+      "loss": 0.3997931480407715,
+      "step": 1967
+    },
+    {
+      "epoch": 1.9283176081362579,
+      "grad_norm": 0.554653525352478,
+      "learning_rate": 3.536057663355852e-08,
+      "loss": 0.5077458620071411,
+      "step": 1968
+    },
+    {
+      "epoch": 1.9292978801617449,
+      "grad_norm": 0.4709983170032501,
+      "learning_rate": 3.44249905709948e-08,
+      "loss": 0.4191884398460388,
+      "step": 1969
+    },
+    {
+      "epoch": 1.930278152187232,
+      "grad_norm": 0.6067174077033997,
+      "learning_rate": 3.350190533760289e-08,
+      "loss": 0.5175186395645142,
+      "step": 1970
+    },
+    {
+      "epoch": 1.9312584242127189,
+      "grad_norm": 0.46343591809272766,
+      "learning_rate": 3.25913232572489e-08,
+      "loss": 0.45350590348243713,
+      "step": 1971
+    },
+    {
+      "epoch": 1.932238696238206,
+      "grad_norm": 0.5357552766799927,
+      "learning_rate": 3.169324662232187e-08,
+      "loss": 0.45746350288391113,
+      "step": 1972
+    },
+    {
+      "epoch": 1.9332189682636933,
+      "grad_norm": 0.5588996410369873,
+      "learning_rate": 3.080767769372939e-08,
+      "loss": 0.38622236251831055,
+      "step": 1973
+    },
+    {
+      "epoch": 1.93419924028918,
+      "grad_norm": 0.4605850577354431,
+      "learning_rate": 2.9934618700889205e-08,
+      "loss": 0.34298330545425415,
+      "step": 1974
+    },
+    {
+      "epoch": 1.9351795123146673,
+      "grad_norm": 0.5302512645721436,
+      "learning_rate": 2.907407184172706e-08,
+      "loss": 0.47001948952674866,
+      "step": 1975
+    },
+    {
+      "epoch": 1.9361597843401543,
+      "grad_norm": 0.6759960055351257,
+      "learning_rate": 2.8226039282667762e-08,
+      "loss": 0.46635645627975464,
+      "step": 1976
+    },
+    {
+      "epoch": 1.9371400563656413,
+      "grad_norm": 0.44337689876556396,
+      "learning_rate": 2.7390523158633552e-08,
+      "loss": 0.4438217878341675,
+      "step": 1977
+    },
+    {
+      "epoch": 1.9381203283911286,
+      "grad_norm": 0.6156742572784424,
+      "learning_rate": 2.6567525573034102e-08,
+      "loss": 0.4663146138191223,
+      "step": 1978
+    },
+    {
+      "epoch": 1.9391006004166156,
+      "grad_norm": 0.7220153212547302,
+      "learning_rate": 2.57570485977654e-08,
+      "loss": 0.4533311724662781,
+      "step": 1979
+    },
+    {
+      "epoch": 1.9400808724421026,
+      "grad_norm": 0.4567999243736267,
+      "learning_rate": 2.495909427320198e-08,
+      "loss": 0.3989451825618744,
+      "step": 1980
+    },
+    {
+      "epoch": 1.9410611444675898,
+      "grad_norm": 0.6177561283111572,
+      "learning_rate": 2.4173664608193592e-08,
+      "loss": 0.5735886096954346,
+      "step": 1981
+    },
+    {
+      "epoch": 1.9420414164930768,
+      "grad_norm": 0.6381194591522217,
+      "learning_rate": 2.3400761580057442e-08,
+      "loss": 0.5144109129905701,
+      "step": 1982
+    },
+    {
+      "epoch": 1.9430216885185638,
+      "grad_norm": 0.5260413289070129,
+      "learning_rate": 2.264038713457706e-08,
+      "loss": 0.40311023592948914,
+      "step": 1983
+    },
+    {
+      "epoch": 1.944001960544051,
+      "grad_norm": 0.6878109574317932,
+      "learning_rate": 2.189254318599343e-08,
+      "loss": 0.368513286113739,
+      "step": 1984
+    },
+    {
+      "epoch": 1.944982232569538,
+      "grad_norm": 0.4822657108306885,
+      "learning_rate": 2.1157231617002783e-08,
+      "loss": 0.3771039843559265,
+      "step": 1985
+    },
+    {
+      "epoch": 1.945962504595025,
+      "grad_norm": 0.5470884442329407,
+      "learning_rate": 2.0434454278752126e-08,
+      "loss": 0.4660874605178833,
+      "step": 1986
+    },
+    {
+      "epoch": 1.9469427766205123,
+      "grad_norm": 0.4203990697860718,
+      "learning_rate": 1.9724212990830938e-08,
+      "loss": 0.32562848925590515,
+      "step": 1987
+    },
+    {
+      "epoch": 1.9479230486459993,
+      "grad_norm": 0.553207278251648,
+      "learning_rate": 1.9026509541272276e-08,
+      "loss": 0.4342890977859497,
+      "step": 1988
+    },
+    {
+      "epoch": 1.9489033206714863,
+      "grad_norm": 0.5573920011520386,
+      "learning_rate": 1.834134568654333e-08,
+      "loss": 0.5062669515609741,
+      "step": 1989
+    },
+    {
+      "epoch": 1.9498835926969735,
+      "grad_norm": 0.5155161619186401,
+      "learning_rate": 1.7668723151542667e-08,
+      "loss": 0.4343412518501282,
+      "step": 1990
+    },
+    {
+      "epoch": 1.9508638647224603,
+      "grad_norm": 0.6901623606681824,
+      "learning_rate": 1.7008643629596866e-08,
+      "loss": 0.6550469398498535,
+      "step": 1991
+    },
+    {
+      "epoch": 1.9518441367479475,
+      "grad_norm": 0.5579574704170227,
+      "learning_rate": 1.6361108782456113e-08,
+      "loss": 0.44486093521118164,
+      "step": 1992
+    },
+    {
+      "epoch": 1.9528244087734348,
+      "grad_norm": 0.5197591781616211,
+      "learning_rate": 1.5726120240288632e-08,
+      "loss": 0.4059602618217468,
+      "step": 1993
+    },
+    {
+      "epoch": 1.9538046807989216,
+      "grad_norm": 0.36868494749069214,
+      "learning_rate": 1.510367960167569e-08,
+      "loss": 0.32157665491104126,
+      "step": 1994
+    },
+    {
+      "epoch": 1.9547849528244088,
+      "grad_norm": 0.6862021684646606,
+      "learning_rate": 1.449378843361271e-08,
+      "loss": 0.40972402691841125,
+      "step": 1995
+    },
+    {
+      "epoch": 1.955765224849896,
+      "grad_norm": 0.7209951281547546,
+      "learning_rate": 1.3896448271499275e-08,
+      "loss": 0.4901972711086273,
+      "step": 1996
+    },
+    {
+      "epoch": 1.9567454968753828,
+      "grad_norm": 0.4967319369316101,
+      "learning_rate": 1.3311660619138578e-08,
+      "loss": 0.46000999212265015,
+      "step": 1997
+    },
+    {
+      "epoch": 1.95772576890087,
+      "grad_norm": 0.4791561961174011,
+      "learning_rate": 1.2739426948732426e-08,
+      "loss": 0.36261147260665894,
+      "step": 1998
+    },
+    {
+      "epoch": 1.958706040926357,
+      "grad_norm": 0.5726127624511719,
+      "learning_rate": 1.2179748700879013e-08,
+      "loss": 0.4656495451927185,
+      "step": 1999
+    },
+    {
+      "epoch": 1.959686312951844,
+      "grad_norm": 0.6408140659332275,
+      "learning_rate": 1.1632627284566822e-08,
+      "loss": 0.4069877564907074,
+      "step": 2000
+    },
+    {
+      "epoch": 1.9606665849773313,
+      "grad_norm": 0.5542259216308594,
+      "learning_rate": 1.109806407717462e-08,
+      "loss": 0.47622790932655334,
+      "step": 2001
+    },
+    {
+      "epoch": 1.9616468570028183,
+      "grad_norm": 0.6945074200630188,
+      "learning_rate": 1.0576060424463686e-08,
+      "loss": 0.5247766971588135,
+      "step": 2002
+    },
+    {
+      "epoch": 1.9626271290283053,
+      "grad_norm": 0.8039277791976929,
+      "learning_rate": 1.006661764057837e-08,
+      "loss": 0.3815796971321106,
+      "step": 2003
+    },
+    {
+      "epoch": 1.9636074010537925,
+      "grad_norm": 0.5804888010025024,
+      "learning_rate": 9.56973700803887e-09,
+      "loss": 0.4335627853870392,
+      "step": 2004
+    },
+    {
+      "epoch": 1.9645876730792795,
+      "grad_norm": 0.6415610909461975,
+      "learning_rate": 9.085419777743465e-09,
+      "loss": 0.6014455556869507,
+      "step": 2005
+    },
+    {
+      "epoch": 1.9655679451047665,
+      "grad_norm": 0.5289047956466675,
+      "learning_rate": 8.613667168958506e-09,
+      "loss": 0.4468574523925781,
+      "step": 2006
+    },
+    {
+      "epoch": 1.9665482171302537,
+      "grad_norm": 0.7096878886222839,
+      "learning_rate": 8.15448036932176e-09,
+      "loss": 0.49082621932029724,
+      "step": 2007
+    },
+    {
+      "epoch": 1.9675284891557407,
+      "grad_norm": 0.5897271037101746,
+      "learning_rate": 7.707860534834632e-09,
+      "loss": 0.45144224166870117,
+      "step": 2008
+    },
+    {
+      "epoch": 1.9685087611812278,
+      "grad_norm": 0.5561026334762573,
+      "learning_rate": 7.273808789862724e-09,
+      "loss": 0.3502047061920166,
+      "step": 2009
+    },
+    {
+      "epoch": 1.969489033206715,
+      "grad_norm": 2.1456868648529053,
+      "learning_rate": 6.852326227130835e-09,
+      "loss": 0.5044934153556824,
+      "step": 2010
+    },
+    {
+      "epoch": 1.970469305232202,
+      "grad_norm": 0.6080367565155029,
+      "learning_rate": 6.4434139077201865e-09,
+      "loss": 0.4637754261493683,
+      "step": 2011
+    },
+    {
+      "epoch": 1.971449577257689,
+      "grad_norm": 0.48481234908103943,
+      "learning_rate": 6.0470728610673155e-09,
+      "loss": 0.44471535086631775,
+      "step": 2012
+    },
+    {
+      "epoch": 1.9724298492831762,
+      "grad_norm": 0.5679246783256531,
+      "learning_rate": 5.6633040849601865e-09,
+      "loss": 0.5313770174980164,
+      "step": 2013
+    },
+    {
+      "epoch": 1.973410121308663,
+      "grad_norm": 0.4563775062561035,
+      "learning_rate": 5.292108545536523e-09,
+      "loss": 0.39058300852775574,
+      "step": 2014
+    },
+    {
+      "epoch": 1.9743903933341502,
+      "grad_norm": 0.6261054873466492,
+      "learning_rate": 4.933487177280483e-09,
+      "loss": 0.5773292779922485,
+      "step": 2015
+    },
+    {
+      "epoch": 1.9753706653596375,
+      "grad_norm": 0.48486658930778503,
+      "learning_rate": 4.587440883021543e-09,
+      "loss": 0.42626726627349854,
+      "step": 2016
+    },
+    {
+      "epoch": 1.9763509373851242,
+      "grad_norm": 0.4783303439617157,
+      "learning_rate": 4.253970533929508e-09,
+      "loss": 0.41985732316970825,
+      "step": 2017
+    },
+    {
+      "epoch": 1.9773312094106115,
+      "grad_norm": 0.5955907702445984,
+      "learning_rate": 3.9330769695167245e-09,
+      "loss": 0.39971280097961426,
+      "step": 2018
+    },
+    {
+      "epoch": 1.9783114814360985,
+      "grad_norm": 0.8066526651382446,
+      "learning_rate": 3.6247609976319818e-09,
+      "loss": 0.4745522141456604,
+      "step": 2019
+    },
+    {
+      "epoch": 1.9792917534615855,
+      "grad_norm": 0.5380304455757141,
+      "learning_rate": 3.3290233944605067e-09,
+      "loss": 0.43057122826576233,
+      "step": 2020
+    },
+    {
+      "epoch": 1.9802720254870727,
+      "grad_norm": 1.305221438407898,
+      "learning_rate": 3.0458649045211897e-09,
+      "loss": 0.42735588550567627,
+      "step": 2021
+    },
+    {
+      "epoch": 1.9812522975125597,
+      "grad_norm": 0.5447163581848145,
+      "learning_rate": 2.7752862406654755e-09,
+      "loss": 0.4578840136528015,
+      "step": 2022
+    },
+    {
+      "epoch": 1.9822325695380467,
+      "grad_norm": 0.422486275434494,
+      "learning_rate": 2.5172880840745873e-09,
+      "loss": 0.42595726251602173,
+      "step": 2023
+    },
+    {
+      "epoch": 1.983212841563534,
+      "grad_norm": 0.5234951972961426,
+      "learning_rate": 2.2718710842584145e-09,
+      "loss": 0.4243493676185608,
+      "step": 2024
+    },
+    {
+      "epoch": 1.984193113589021,
+      "grad_norm": 0.5258898735046387,
+      "learning_rate": 2.0390358590538507e-09,
+      "loss": 0.4304959177970886,
+      "step": 2025
+    },
+    {
+      "epoch": 1.985173385614508,
+      "grad_norm": 0.6476520895957947,
+      "learning_rate": 1.8187829946236802e-09,
+      "loss": 0.4684514105319977,
+      "step": 2026
+    },
+    {
+      "epoch": 1.9861536576399952,
+      "grad_norm": 0.4058888554573059,
+      "learning_rate": 1.61111304545436e-09,
+      "loss": 0.31605929136276245,
+      "step": 2027
+    },
+    {
+      "epoch": 1.9871339296654822,
+      "grad_norm": 0.40702277421951294,
+      "learning_rate": 1.4160265343549084e-09,
+      "loss": 0.4311891198158264,
+      "step": 2028
+    },
+    {
+      "epoch": 1.9881142016909692,
+      "grad_norm": 0.5184072852134705,
+      "learning_rate": 1.2335239524541298e-09,
+      "loss": 0.45399418473243713,
+      "step": 2029
+    },
+    {
+      "epoch": 1.9890944737164564,
+      "grad_norm": 0.4722048342227936,
+      "learning_rate": 1.0636057592033899e-09,
+      "loss": 0.4487224221229553,
+      "step": 2030
+    },
+    {
+      "epoch": 1.9900747457419434,
+      "grad_norm": 0.46525847911834717,
+      "learning_rate": 9.062723823710651e-10,
+      "loss": 0.37769845128059387,
+      "step": 2031
+    },
+    {
+      "epoch": 1.9910550177674304,
+      "grad_norm": 0.7857242822647095,
+      "learning_rate": 7.615242180436521e-10,
+      "loss": 0.5099886655807495,
+      "step": 2032
+    },
+    {
+      "epoch": 1.9920352897929177,
+      "grad_norm": 0.5199786424636841,
+      "learning_rate": 6.293616306246586e-10,
+      "loss": 0.5014153718948364,
+      "step": 2033
+    },
+    {
+      "epoch": 1.9930155618184044,
+      "grad_norm": 0.6570830345153809,
+      "learning_rate": 5.09784952833492e-10,
+      "loss": 0.3988020420074463,
+      "step": 2034
+    },
+    {
+      "epoch": 1.9939958338438917,
+      "grad_norm": 0.4883534908294678,
+      "learning_rate": 4.027944857032395e-10,
+      "loss": 0.5205972194671631,
+      "step": 2035
+    },
+    {
+      "epoch": 1.994976105869379,
+      "grad_norm": 1.1817388534545898,
+      "learning_rate": 3.0839049858344406e-10,
+      "loss": 0.4344862103462219,
+      "step": 2036
+    },
+    {
+      "epoch": 1.9959563778948657,
+      "grad_norm": 0.6104335188865662,
+      "learning_rate": 2.265732291356626e-10,
+      "loss": 0.46816766262054443,
+      "step": 2037
+    },
+    {
+      "epoch": 1.996936649920353,
+      "grad_norm": 0.5169486403465271,
+      "learning_rate": 1.5734288333457692e-10,
+      "loss": 0.5320672392845154,
+      "step": 2038
+    },
+    {
+      "epoch": 1.99791692194584,
+      "grad_norm": 0.6468532085418701,
+      "learning_rate": 1.0069963546743833e-10,
+      "loss": 0.46354761719703674,
+      "step": 2039
+    },
+    {
+      "epoch": 1.998897193971327,
+      "grad_norm": 0.628599226474762,
+      "learning_rate": 5.6643628134067655e-11,
+      "loss": 0.47993552684783936,
+      "step": 2040
+    },
+    {
+      "epoch": 1.9998774659968142,
+      "grad_norm": 0.41873377561569214,
+      "learning_rate": 2.5174972244634834e-11,
+      "loss": 0.3825181722640991,
+      "step": 2041
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.2331719398498535,
+      "learning_rate": 6.2937470224344596e-12,
+      "loss": 0.7386448979377747,
+      "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": 9.875924520890532e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}