diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12831 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 1827,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005473453749315818,
+      "grad_norm": 6.498785495758057,
+      "learning_rate": 9.999992607989888e-06,
+      "loss": 3.6359,
+      "step": 1
+    },
+    {
+      "epoch": 0.0010946907498631637,
+      "grad_norm": 6.897223472595215,
+      "learning_rate": 9.999970431981408e-06,
+      "loss": 4.6947,
+      "step": 2
+    },
+    {
+      "epoch": 0.0016420361247947454,
+      "grad_norm": 7.065066337585449,
+      "learning_rate": 9.999933472040129e-06,
+      "loss": 4.3915,
+      "step": 3
+    },
+    {
+      "epoch": 0.0021893814997263274,
+      "grad_norm": 6.855522155761719,
+      "learning_rate": 9.999881728275334e-06,
+      "loss": 3.9864,
+      "step": 4
+    },
+    {
+      "epoch": 0.002736726874657909,
+      "grad_norm": 9.54251480102539,
+      "learning_rate": 9.99981520084002e-06,
+      "loss": 3.7584,
+      "step": 5
+    },
+    {
+      "epoch": 0.003284072249589491,
+      "grad_norm": 9.166783332824707,
+      "learning_rate": 9.999733889930897e-06,
+      "loss": 4.3282,
+      "step": 6
+    },
+    {
+      "epoch": 0.0038314176245210726,
+      "grad_norm": 9.869710922241211,
+      "learning_rate": 9.999637795788383e-06,
+      "loss": 3.745,
+      "step": 7
+    },
+    {
+      "epoch": 0.004378762999452655,
+      "grad_norm": 10.634393692016602,
+      "learning_rate": 9.999526918696613e-06,
+      "loss": 3.2058,
+      "step": 8
+    },
+    {
+      "epoch": 0.0049261083743842365,
+      "grad_norm": 11.015477180480957,
+      "learning_rate": 9.999401258983426e-06,
+      "loss": 4.0171,
+      "step": 9
+    },
+    {
+      "epoch": 0.005473453749315818,
+      "grad_norm": 13.21458911895752,
+      "learning_rate": 9.999260817020373e-06,
+      "loss": 2.6568,
+      "step": 10
+    },
+    {
+      "epoch": 0.0060207991242474,
+      "grad_norm": 11.63002872467041,
+      "learning_rate": 9.999105593222714e-06,
+      "loss": 2.5763,
+      "step": 11
+    },
+    {
+      "epoch": 0.006568144499178982,
+      "grad_norm": 6.239895820617676,
+      "learning_rate": 9.998935588049414e-06,
+      "loss": 2.3638,
+      "step": 12
+    },
+    {
+      "epoch": 0.0071154898741105635,
+      "grad_norm": 16.099512100219727,
+      "learning_rate": 9.998750802003148e-06,
+      "loss": 3.0956,
+      "step": 13
+    },
+    {
+      "epoch": 0.007662835249042145,
+      "grad_norm": 9.356679916381836,
+      "learning_rate": 9.99855123563029e-06,
+      "loss": 2.1751,
+      "step": 14
+    },
+    {
+      "epoch": 0.008210180623973728,
+      "grad_norm": 7.839269638061523,
+      "learning_rate": 9.99833688952092e-06,
+      "loss": 2.5912,
+      "step": 15
+    },
+    {
+      "epoch": 0.00875752599890531,
+      "grad_norm": 10.665407180786133,
+      "learning_rate": 9.998107764308815e-06,
+      "loss": 2.2888,
+      "step": 16
+    },
+    {
+      "epoch": 0.009304871373836891,
+      "grad_norm": 8.104930877685547,
+      "learning_rate": 9.997863860671457e-06,
+      "loss": 1.4194,
+      "step": 17
+    },
+    {
+      "epoch": 0.009852216748768473,
+      "grad_norm": 5.245537757873535,
+      "learning_rate": 9.997605179330018e-06,
+      "loss": 2.3864,
+      "step": 18
+    },
+    {
+      "epoch": 0.010399562123700055,
+      "grad_norm": 9.132122039794922,
+      "learning_rate": 9.99733172104937e-06,
+      "loss": 2.0268,
+      "step": 19
+    },
+    {
+      "epoch": 0.010946907498631636,
+      "grad_norm": 6.973398208618164,
+      "learning_rate": 9.997043486638076e-06,
+      "loss": 1.4236,
+      "step": 20
+    },
+    {
+      "epoch": 0.011494252873563218,
+      "grad_norm": 4.4784159660339355,
+      "learning_rate": 9.996740476948386e-06,
+      "loss": 1.0039,
+      "step": 21
+    },
+    {
+      "epoch": 0.0120415982484948,
+      "grad_norm": 4.953526020050049,
+      "learning_rate": 9.996422692876242e-06,
+      "loss": 1.2214,
+      "step": 22
+    },
+    {
+      "epoch": 0.012588943623426382,
+      "grad_norm": 9.643818855285645,
+      "learning_rate": 9.996090135361269e-06,
+      "loss": 1.1022,
+      "step": 23
+    },
+    {
+      "epoch": 0.013136288998357963,
+      "grad_norm": 7.458937644958496,
+      "learning_rate": 9.995742805386775e-06,
+      "loss": 1.861,
+      "step": 24
+    },
+    {
+      "epoch": 0.013683634373289545,
+      "grad_norm": 3.3844003677368164,
+      "learning_rate": 9.995380703979744e-06,
+      "loss": 0.3331,
+      "step": 25
+    },
+    {
+      "epoch": 0.014230979748221127,
+      "grad_norm": 4.211463928222656,
+      "learning_rate": 9.995003832210843e-06,
+      "loss": 0.3263,
+      "step": 26
+    },
+    {
+      "epoch": 0.014778325123152709,
+      "grad_norm": 5.280601978302002,
+      "learning_rate": 9.994612191194407e-06,
+      "loss": 0.6696,
+      "step": 27
+    },
+    {
+      "epoch": 0.01532567049808429,
+      "grad_norm": 1.4705464839935303,
+      "learning_rate": 9.994205782088438e-06,
+      "loss": 0.1472,
+      "step": 28
+    },
+    {
+      "epoch": 0.015873015873015872,
+      "grad_norm": 4.857443332672119,
+      "learning_rate": 9.993784606094612e-06,
+      "loss": 0.3761,
+      "step": 29
+    },
+    {
+      "epoch": 0.016420361247947456,
+      "grad_norm": 1.6145434379577637,
+      "learning_rate": 9.993348664458263e-06,
+      "loss": 0.0976,
+      "step": 30
+    },
+    {
+      "epoch": 0.016967706622879036,
+      "grad_norm": 0.5500306487083435,
+      "learning_rate": 9.992897958468386e-06,
+      "loss": 0.0244,
+      "step": 31
+    },
+    {
+      "epoch": 0.01751505199781062,
+      "grad_norm": 1.4634612798690796,
+      "learning_rate": 9.992432489457626e-06,
+      "loss": 0.0901,
+      "step": 32
+    },
+    {
+      "epoch": 0.0180623973727422,
+      "grad_norm": 0.18377567827701569,
+      "learning_rate": 9.991952258802288e-06,
+      "loss": 0.0113,
+      "step": 33
+    },
+    {
+      "epoch": 0.018609742747673783,
+      "grad_norm": 0.2462746500968933,
+      "learning_rate": 9.99145726792232e-06,
+      "loss": 0.0164,
+      "step": 34
+    },
+    {
+      "epoch": 0.019157088122605363,
+      "grad_norm": 0.3507516384124756,
+      "learning_rate": 9.990947518281312e-06,
+      "loss": 0.0084,
+      "step": 35
+    },
+    {
+      "epoch": 0.019704433497536946,
+      "grad_norm": 0.05946524441242218,
+      "learning_rate": 9.990423011386489e-06,
+      "loss": 0.0025,
+      "step": 36
+    },
+    {
+      "epoch": 0.020251778872468526,
+      "grad_norm": 0.29986003041267395,
+      "learning_rate": 9.989883748788724e-06,
+      "loss": 0.0155,
+      "step": 37
+    },
+    {
+      "epoch": 0.02079912424740011,
+      "grad_norm": 0.036302998661994934,
+      "learning_rate": 9.989329732082504e-06,
+      "loss": 0.0018,
+      "step": 38
+    },
+    {
+      "epoch": 0.021346469622331693,
+      "grad_norm": 0.18141533434391022,
+      "learning_rate": 9.98876096290595e-06,
+      "loss": 0.0096,
+      "step": 39
+    },
+    {
+      "epoch": 0.021893814997263273,
+      "grad_norm": 3.4637537002563477,
+      "learning_rate": 9.988177442940803e-06,
+      "loss": 1.3062,
+      "step": 40
+    },
+    {
+      "epoch": 0.022441160372194856,
+      "grad_norm": 0.18254823982715607,
+      "learning_rate": 9.987579173912413e-06,
+      "loss": 0.0111,
+      "step": 41
+    },
+    {
+      "epoch": 0.022988505747126436,
+      "grad_norm": 5.862061023712158,
+      "learning_rate": 9.986966157589751e-06,
+      "loss": 1.1001,
+      "step": 42
+    },
+    {
+      "epoch": 0.02353585112205802,
+      "grad_norm": 0.026641806587576866,
+      "learning_rate": 9.986338395785377e-06,
+      "loss": 0.0012,
+      "step": 43
+    },
+    {
+      "epoch": 0.0240831964969896,
+      "grad_norm": 1.614923119544983,
+      "learning_rate": 9.985695890355467e-06,
+      "loss": 0.0334,
+      "step": 44
+    },
+    {
+      "epoch": 0.024630541871921183,
+      "grad_norm": 0.09787950664758682,
+      "learning_rate": 9.98503864319978e-06,
+      "loss": 0.0042,
+      "step": 45
+    },
+    {
+      "epoch": 0.025177887246852763,
+      "grad_norm": 0.04275774955749512,
+      "learning_rate": 9.98436665626167e-06,
+      "loss": 0.0019,
+      "step": 46
+    },
+    {
+      "epoch": 0.025725232621784347,
+      "grad_norm": 0.03735092282295227,
+      "learning_rate": 9.983679931528068e-06,
+      "loss": 0.0018,
+      "step": 47
+    },
+    {
+      "epoch": 0.026272577996715927,
+      "grad_norm": 0.07040659338235855,
+      "learning_rate": 9.982978471029485e-06,
+      "loss": 0.0039,
+      "step": 48
+    },
+    {
+      "epoch": 0.02681992337164751,
+      "grad_norm": 0.015653884038329124,
+      "learning_rate": 9.982262276840002e-06,
+      "loss": 0.0007,
+      "step": 49
+    },
+    {
+      "epoch": 0.02736726874657909,
+      "grad_norm": 1.0839862823486328,
+      "learning_rate": 9.981531351077266e-06,
+      "loss": 0.0126,
+      "step": 50
+    },
+    {
+      "epoch": 0.027914614121510674,
+      "grad_norm": 0.02633080445230007,
+      "learning_rate": 9.980785695902481e-06,
+      "loss": 0.0009,
+      "step": 51
+    },
+    {
+      "epoch": 0.028461959496442254,
+      "grad_norm": 7.923893928527832,
+      "learning_rate": 9.980025313520403e-06,
+      "loss": 0.2134,
+      "step": 52
+    },
+    {
+      "epoch": 0.029009304871373837,
+      "grad_norm": 6.742312431335449,
+      "learning_rate": 9.979250206179333e-06,
+      "loss": 0.4451,
+      "step": 53
+    },
+    {
+      "epoch": 0.029556650246305417,
+      "grad_norm": 0.008141374215483665,
+      "learning_rate": 9.978460376171113e-06,
+      "loss": 0.0004,
+      "step": 54
+    },
+    {
+      "epoch": 0.030103995621237,
+      "grad_norm": 0.006501413881778717,
+      "learning_rate": 9.977655825831114e-06,
+      "loss": 0.0003,
+      "step": 55
+    },
+    {
+      "epoch": 0.03065134099616858,
+      "grad_norm": 0.04851650074124336,
+      "learning_rate": 9.976836557538234e-06,
+      "loss": 0.0023,
+      "step": 56
+    },
+    {
+      "epoch": 0.031198686371100164,
+      "grad_norm": 0.008002854883670807,
+      "learning_rate": 9.97600257371489e-06,
+      "loss": 0.0003,
+      "step": 57
+    },
+    {
+      "epoch": 0.031746031746031744,
+      "grad_norm": 0.0038702317979186773,
+      "learning_rate": 9.975153876827008e-06,
+      "loss": 0.0002,
+      "step": 58
+    },
+    {
+      "epoch": 0.03229337712096333,
+      "grad_norm": 0.006718830205500126,
+      "learning_rate": 9.974290469384019e-06,
+      "loss": 0.0003,
+      "step": 59
+    },
+    {
+      "epoch": 0.03284072249589491,
+      "grad_norm": 0.011266892775893211,
+      "learning_rate": 9.973412353938847e-06,
+      "loss": 0.0005,
+      "step": 60
+    },
+    {
+      "epoch": 0.033388067870826495,
+      "grad_norm": 0.07028752565383911,
+      "learning_rate": 9.97251953308791e-06,
+      "loss": 0.0024,
+      "step": 61
+    },
+    {
+      "epoch": 0.03393541324575807,
+      "grad_norm": 0.022092929109930992,
+      "learning_rate": 9.971612009471105e-06,
+      "loss": 0.0008,
+      "step": 62
+    },
+    {
+      "epoch": 0.034482758620689655,
+      "grad_norm": 0.024217141792178154,
+      "learning_rate": 9.970689785771798e-06,
+      "loss": 0.0011,
+      "step": 63
+    },
+    {
+      "epoch": 0.03503010399562124,
+      "grad_norm": 0.021001331508159637,
+      "learning_rate": 9.969752864716828e-06,
+      "loss": 0.0009,
+      "step": 64
+    },
+    {
+      "epoch": 0.03557744937055282,
+      "grad_norm": 0.01548935379832983,
+      "learning_rate": 9.968801249076484e-06,
+      "loss": 0.0007,
+      "step": 65
+    },
+    {
+      "epoch": 0.0361247947454844,
+      "grad_norm": 5.745925426483154,
+      "learning_rate": 9.967834941664508e-06,
+      "loss": 1.477,
+      "step": 66
+    },
+    {
+      "epoch": 0.03667214012041598,
+      "grad_norm": 0.02599332295358181,
+      "learning_rate": 9.96685394533808e-06,
+      "loss": 0.0009,
+      "step": 67
+    },
+    {
+      "epoch": 0.037219485495347565,
+      "grad_norm": 7.921395778656006,
+      "learning_rate": 9.965858262997817e-06,
+      "loss": 1.0394,
+      "step": 68
+    },
+    {
+      "epoch": 0.03776683087027915,
+      "grad_norm": 1.6005064249038696,
+      "learning_rate": 9.964847897587753e-06,
+      "loss": 0.0348,
+      "step": 69
+    },
+    {
+      "epoch": 0.038314176245210725,
+      "grad_norm": 0.01647241599857807,
+      "learning_rate": 9.963822852095344e-06,
+      "loss": 0.0007,
+      "step": 70
+    },
+    {
+      "epoch": 0.03886152162014231,
+      "grad_norm": 0.01896149478852749,
+      "learning_rate": 9.962783129551447e-06,
+      "loss": 0.0007,
+      "step": 71
+    },
+    {
+      "epoch": 0.03940886699507389,
+      "grad_norm": 3.117213487625122,
+      "learning_rate": 9.961728733030318e-06,
+      "loss": 0.1582,
+      "step": 72
+    },
+    {
+      "epoch": 0.039956212370005476,
+      "grad_norm": 0.005239859223365784,
+      "learning_rate": 9.9606596656496e-06,
+      "loss": 0.0002,
+      "step": 73
+    },
+    {
+      "epoch": 0.04050355774493705,
+      "grad_norm": 0.010306102223694324,
+      "learning_rate": 9.959575930570318e-06,
+      "loss": 0.0005,
+      "step": 74
+    },
+    {
+      "epoch": 0.041050903119868636,
+      "grad_norm": 0.005997025407850742,
+      "learning_rate": 9.958477530996862e-06,
+      "loss": 0.0002,
+      "step": 75
+    },
+    {
+      "epoch": 0.04159824849480022,
+      "grad_norm": 0.00372750754468143,
+      "learning_rate": 9.957364470176986e-06,
+      "loss": 0.0002,
+      "step": 76
+    },
+    {
+      "epoch": 0.0421455938697318,
+      "grad_norm": 0.021045928820967674,
+      "learning_rate": 9.95623675140179e-06,
+      "loss": 0.0009,
+      "step": 77
+    },
+    {
+      "epoch": 0.042692939244663386,
+      "grad_norm": 6.066615581512451,
+      "learning_rate": 9.955094378005723e-06,
+      "loss": 0.912,
+      "step": 78
+    },
+    {
+      "epoch": 0.04324028461959496,
+      "grad_norm": 0.002792640123516321,
+      "learning_rate": 9.953937353366551e-06,
+      "loss": 0.0001,
+      "step": 79
+    },
+    {
+      "epoch": 0.043787629994526546,
+      "grad_norm": 0.03558952733874321,
+      "learning_rate": 9.952765680905378e-06,
+      "loss": 0.0017,
+      "step": 80
+    },
+    {
+      "epoch": 0.04433497536945813,
+      "grad_norm": 0.23785309493541718,
+      "learning_rate": 9.951579364086603e-06,
+      "loss": 0.0061,
+      "step": 81
+    },
+    {
+      "epoch": 0.04488232074438971,
+      "grad_norm": 0.005056190770119429,
+      "learning_rate": 9.950378406417935e-06,
+      "loss": 0.0002,
+      "step": 82
+    },
+    {
+      "epoch": 0.04542966611932129,
+      "grad_norm": 0.005788102280348539,
+      "learning_rate": 9.949162811450373e-06,
+      "loss": 0.0003,
+      "step": 83
+    },
+    {
+      "epoch": 0.04597701149425287,
+      "grad_norm": 0.027935905382037163,
+      "learning_rate": 9.947932582778188e-06,
+      "loss": 0.0011,
+      "step": 84
+    },
+    {
+      "epoch": 0.046524356869184456,
+      "grad_norm": 0.10212934762239456,
+      "learning_rate": 9.946687724038929e-06,
+      "loss": 0.003,
+      "step": 85
+    },
+    {
+      "epoch": 0.04707170224411604,
+      "grad_norm": 0.0034679218661040068,
+      "learning_rate": 9.945428238913396e-06,
+      "loss": 0.0002,
+      "step": 86
+    },
+    {
+      "epoch": 0.047619047619047616,
+      "grad_norm": 0.006359061226248741,
+      "learning_rate": 9.944154131125643e-06,
+      "loss": 0.0003,
+      "step": 87
+    },
+    {
+      "epoch": 0.0481663929939792,
+      "grad_norm": 0.011999201029539108,
+      "learning_rate": 9.942865404442955e-06,
+      "loss": 0.0006,
+      "step": 88
+    },
+    {
+      "epoch": 0.04871373836891078,
+      "grad_norm": 5.890190601348877,
+      "learning_rate": 9.941562062675848e-06,
+      "loss": 0.572,
+      "step": 89
+    },
+    {
+      "epoch": 0.04926108374384237,
+      "grad_norm": 3.3754000663757324,
+      "learning_rate": 9.940244109678043e-06,
+      "loss": 0.2577,
+      "step": 90
+    },
+    {
+      "epoch": 0.04980842911877394,
+      "grad_norm": 0.003820559708401561,
+      "learning_rate": 9.938911549346473e-06,
+      "loss": 0.0002,
+      "step": 91
+    },
+    {
+      "epoch": 0.05035577449370553,
+      "grad_norm": 0.18967939913272858,
+      "learning_rate": 9.937564385621254e-06,
+      "loss": 0.0082,
+      "step": 92
+    },
+    {
+      "epoch": 0.05090311986863711,
+      "grad_norm": 0.007278892211616039,
+      "learning_rate": 9.936202622485687e-06,
+      "loss": 0.0004,
+      "step": 93
+    },
+    {
+      "epoch": 0.051450465243568694,
+      "grad_norm": 6.680490016937256,
+      "learning_rate": 9.93482626396624e-06,
+      "loss": 0.8037,
+      "step": 94
+    },
+    {
+      "epoch": 0.05199781061850027,
+      "grad_norm": 0.09887053072452545,
+      "learning_rate": 9.933435314132534e-06,
+      "loss": 0.0048,
+      "step": 95
+    },
+    {
+      "epoch": 0.052545155993431854,
+      "grad_norm": 3.9046106338500977,
+      "learning_rate": 9.932029777097333e-06,
+      "loss": 0.3514,
+      "step": 96
+    },
+    {
+      "epoch": 0.05309250136836344,
+      "grad_norm": 0.015004507265985012,
+      "learning_rate": 9.93060965701654e-06,
+      "loss": 0.0007,
+      "step": 97
+    },
+    {
+      "epoch": 0.05363984674329502,
+      "grad_norm": 0.1083778440952301,
+      "learning_rate": 9.929174958089167e-06,
+      "loss": 0.0059,
+      "step": 98
+    },
+    {
+      "epoch": 0.054187192118226604,
+      "grad_norm": 0.0027883078437298536,
+      "learning_rate": 9.927725684557339e-06,
+      "loss": 0.0001,
+      "step": 99
+    },
+    {
+      "epoch": 0.05473453749315818,
+      "grad_norm": 0.03726564347743988,
+      "learning_rate": 9.926261840706275e-06,
+      "loss": 0.0019,
+      "step": 100
+    },
+    {
+      "epoch": 0.055281882868089764,
+      "grad_norm": 0.016526643186807632,
+      "learning_rate": 9.924783430864273e-06,
+      "loss": 0.0008,
+      "step": 101
+    },
+    {
+      "epoch": 0.05582922824302135,
+      "grad_norm": 0.009402341209352016,
+      "learning_rate": 9.923290459402701e-06,
+      "loss": 0.0004,
+      "step": 102
+    },
+    {
+      "epoch": 0.05637657361795293,
+      "grad_norm": 0.012751290574669838,
+      "learning_rate": 9.921782930735985e-06,
+      "loss": 0.0006,
+      "step": 103
+    },
+    {
+      "epoch": 0.05692391899288451,
+      "grad_norm": 0.0394417904317379,
+      "learning_rate": 9.92026084932159e-06,
+      "loss": 0.0019,
+      "step": 104
+    },
+    {
+      "epoch": 0.05747126436781609,
+      "grad_norm": 0.02528909221291542,
+      "learning_rate": 9.918724219660013e-06,
+      "loss": 0.0013,
+      "step": 105
+    },
+    {
+      "epoch": 0.058018609742747675,
+      "grad_norm": 0.0044928742572665215,
+      "learning_rate": 9.917173046294769e-06,
+      "loss": 0.0002,
+      "step": 106
+    },
+    {
+      "epoch": 0.05856595511767926,
+      "grad_norm": 0.011569995433092117,
+      "learning_rate": 9.91560733381237e-06,
+      "loss": 0.0005,
+      "step": 107
+    },
+    {
+      "epoch": 0.059113300492610835,
+      "grad_norm": 0.005935823544859886,
+      "learning_rate": 9.914027086842323e-06,
+      "loss": 0.0003,
+      "step": 108
+    },
+    {
+      "epoch": 0.05966064586754242,
+      "grad_norm": 0.0639987364411354,
+      "learning_rate": 9.912432310057108e-06,
+      "loss": 0.0031,
+      "step": 109
+    },
+    {
+      "epoch": 0.060207991242474,
+      "grad_norm": 0.2608802318572998,
+      "learning_rate": 9.910823008172168e-06,
+      "loss": 0.011,
+      "step": 110
+    },
+    {
+      "epoch": 0.060755336617405585,
+      "grad_norm": 0.0046829953789711,
+      "learning_rate": 9.909199185945893e-06,
+      "loss": 0.0002,
+      "step": 111
+    },
+    {
+      "epoch": 0.06130268199233716,
+      "grad_norm": 0.008316273801028728,
+      "learning_rate": 9.907560848179607e-06,
+      "loss": 0.0003,
+      "step": 112
+    },
+    {
+      "epoch": 0.061850027367268745,
+      "grad_norm": 0.012634074315428734,
+      "learning_rate": 9.905907999717551e-06,
+      "loss": 0.0004,
+      "step": 113
+    },
+    {
+      "epoch": 0.06239737274220033,
+      "grad_norm": 0.2069849669933319,
+      "learning_rate": 9.90424064544688e-06,
+      "loss": 0.0092,
+      "step": 114
+    },
+    {
+      "epoch": 0.06294471811713191,
+      "grad_norm": 4.1578803062438965,
+      "learning_rate": 9.902558790297631e-06,
+      "loss": 0.3153,
+      "step": 115
+    },
+    {
+      "epoch": 0.06349206349206349,
+      "grad_norm": 0.3431181013584137,
+      "learning_rate": 9.900862439242719e-06,
+      "loss": 0.0144,
+      "step": 116
+    },
+    {
+      "epoch": 0.06403940886699508,
+      "grad_norm": 0.18196187913417816,
+      "learning_rate": 9.899151597297923e-06,
+      "loss": 0.0085,
+      "step": 117
+    },
+    {
+      "epoch": 0.06458675424192666,
+      "grad_norm": 0.011420561000704765,
+      "learning_rate": 9.897426269521868e-06,
+      "loss": 0.0005,
+      "step": 118
+    },
+    {
+      "epoch": 0.06513409961685823,
+      "grad_norm": 0.20257817208766937,
+      "learning_rate": 9.895686461016007e-06,
+      "loss": 0.0096,
+      "step": 119
+    },
+    {
+      "epoch": 0.06568144499178982,
+      "grad_norm": 0.0031977526377886534,
+      "learning_rate": 9.893932176924616e-06,
+      "loss": 0.0002,
+      "step": 120
+    },
+    {
+      "epoch": 0.0662287903667214,
+      "grad_norm": 0.0007180224638432264,
+      "learning_rate": 9.892163422434767e-06,
+      "loss": 0.0001,
+      "step": 121
+    },
+    {
+      "epoch": 0.06677613574165299,
+      "grad_norm": 5.340695381164551,
+      "learning_rate": 9.890380202776323e-06,
+      "loss": 0.503,
+      "step": 122
+    },
+    {
+      "epoch": 0.06732348111658457,
+      "grad_norm": 0.0025924837682396173,
+      "learning_rate": 9.888582523221912e-06,
+      "loss": 0.0001,
+      "step": 123
+    },
+    {
+      "epoch": 0.06787082649151614,
+      "grad_norm": 2.3252947330474854,
+      "learning_rate": 9.886770389086923e-06,
+      "loss": 0.2639,
+      "step": 124
+    },
+    {
+      "epoch": 0.06841817186644773,
+      "grad_norm": 0.004119732417166233,
+      "learning_rate": 9.884943805729481e-06,
+      "loss": 0.0002,
+      "step": 125
+    },
+    {
+      "epoch": 0.06896551724137931,
+      "grad_norm": 0.06781020015478134,
+      "learning_rate": 9.883102778550434e-06,
+      "loss": 0.0022,
+      "step": 126
+    },
+    {
+      "epoch": 0.06951286261631089,
+      "grad_norm": 4.811086654663086,
+      "learning_rate": 9.88124731299334e-06,
+      "loss": 0.4408,
+      "step": 127
+    },
+    {
+      "epoch": 0.07006020799124248,
+      "grad_norm": 0.008488141000270844,
+      "learning_rate": 9.879377414544444e-06,
+      "loss": 0.0003,
+      "step": 128
+    },
+    {
+      "epoch": 0.07060755336617405,
+      "grad_norm": 0.007522579748183489,
+      "learning_rate": 9.877493088732672e-06,
+      "loss": 0.0004,
+      "step": 129
+    },
+    {
+      "epoch": 0.07115489874110564,
+      "grad_norm": 0.011271845549345016,
+      "learning_rate": 9.875594341129607e-06,
+      "loss": 0.0005,
+      "step": 130
+    },
+    {
+      "epoch": 0.07170224411603722,
+      "grad_norm": 0.87716144323349,
+      "learning_rate": 9.873681177349473e-06,
+      "loss": 0.0288,
+      "step": 131
+    },
+    {
+      "epoch": 0.0722495894909688,
+      "grad_norm": 0.021911056712269783,
+      "learning_rate": 9.871753603049117e-06,
+      "loss": 0.0009,
+      "step": 132
+    },
+    {
+      "epoch": 0.07279693486590039,
+      "grad_norm": 0.0031497348099946976,
+      "learning_rate": 9.869811623928001e-06,
+      "loss": 0.0001,
+      "step": 133
+    },
+    {
+      "epoch": 0.07334428024083196,
+      "grad_norm": 1.4936490058898926,
+      "learning_rate": 9.86785524572818e-06,
+      "loss": 0.1167,
+      "step": 134
+    },
+    {
+      "epoch": 0.07389162561576355,
+      "grad_norm": 0.11308709532022476,
+      "learning_rate": 9.865884474234275e-06,
+      "loss": 0.0048,
+      "step": 135
+    },
+    {
+      "epoch": 0.07443897099069513,
+      "grad_norm": 0.008609531447291374,
+      "learning_rate": 9.863899315273475e-06,
+      "loss": 0.0004,
+      "step": 136
+    },
+    {
+      "epoch": 0.0749863163656267,
+      "grad_norm": 0.35647737979888916,
+      "learning_rate": 9.861899774715504e-06,
+      "loss": 0.0147,
+      "step": 137
+    },
+    {
+      "epoch": 0.0755336617405583,
+      "grad_norm": 6.980875492095947,
+      "learning_rate": 9.859885858472614e-06,
+      "loss": 0.8246,
+      "step": 138
+    },
+    {
+      "epoch": 0.07608100711548987,
+      "grad_norm": 0.002529384568333626,
+      "learning_rate": 9.857857572499559e-06,
+      "loss": 0.0001,
+      "step": 139
+    },
+    {
+      "epoch": 0.07662835249042145,
+      "grad_norm": 7.442034721374512,
+      "learning_rate": 9.855814922793583e-06,
+      "loss": 0.3612,
+      "step": 140
+    },
+    {
+      "epoch": 0.07717569786535304,
+      "grad_norm": 3.36582350730896,
+      "learning_rate": 9.853757915394403e-06,
+      "loss": 0.2024,
+      "step": 141
+    },
+    {
+      "epoch": 0.07772304324028462,
+      "grad_norm": 0.004160053096711636,
+      "learning_rate": 9.851686556384182e-06,
+      "loss": 0.0002,
+      "step": 142
+    },
+    {
+      "epoch": 0.07827038861521621,
+      "grad_norm": 0.0015179223846644163,
+      "learning_rate": 9.849600851887528e-06,
+      "loss": 0.0001,
+      "step": 143
+    },
+    {
+      "epoch": 0.07881773399014778,
+      "grad_norm": 5.545576095581055,
+      "learning_rate": 9.847500808071458e-06,
+      "loss": 0.596,
+      "step": 144
+    },
+    {
+      "epoch": 0.07936507936507936,
+      "grad_norm": 0.01277522835880518,
+      "learning_rate": 9.84538643114539e-06,
+      "loss": 0.0007,
+      "step": 145
+    },
+    {
+      "epoch": 0.07991242474001095,
+      "grad_norm": 0.0032218769192695618,
+      "learning_rate": 9.843257727361124e-06,
+      "loss": 0.0002,
+      "step": 146
+    },
+    {
+      "epoch": 0.08045977011494253,
+      "grad_norm": 0.009027930907905102,
+      "learning_rate": 9.841114703012817e-06,
+      "loss": 0.0005,
+      "step": 147
+    },
+    {
+      "epoch": 0.0810071154898741,
+      "grad_norm": 3.7516531944274902,
+      "learning_rate": 9.838957364436973e-06,
+      "loss": 0.551,
+      "step": 148
+    },
+    {
+      "epoch": 0.0815544608648057,
+      "grad_norm": 0.01132990699261427,
+      "learning_rate": 9.836785718012422e-06,
+      "loss": 0.0005,
+      "step": 149
+    },
+    {
+      "epoch": 0.08210180623973727,
+      "grad_norm": 0.011869224719703197,
+      "learning_rate": 9.834599770160296e-06,
+      "loss": 0.0005,
+      "step": 150
+    },
+    {
+      "epoch": 0.08264915161466886,
+      "grad_norm": 0.011585843749344349,
+      "learning_rate": 9.832399527344012e-06,
+      "loss": 0.0005,
+      "step": 151
+    },
+    {
+      "epoch": 0.08319649698960044,
+      "grad_norm": 0.10732828080654144,
+      "learning_rate": 9.830184996069259e-06,
+      "loss": 0.0033,
+      "step": 152
+    },
+    {
+      "epoch": 0.08374384236453201,
+      "grad_norm": 4.160403251647949,
+      "learning_rate": 9.82795618288397e-06,
+      "loss": 0.4624,
+      "step": 153
+    },
+    {
+      "epoch": 0.0842911877394636,
+      "grad_norm": 0.5677698254585266,
+      "learning_rate": 9.82571309437831e-06,
+      "loss": 0.0172,
+      "step": 154
+    },
+    {
+      "epoch": 0.08483853311439518,
+      "grad_norm": 0.013016915880143642,
+      "learning_rate": 9.823455737184655e-06,
+      "loss": 0.0004,
+      "step": 155
+    },
+    {
+      "epoch": 0.08538587848932677,
+      "grad_norm": 0.02737693302333355,
+      "learning_rate": 9.821184117977564e-06,
+      "loss": 0.0014,
+      "step": 156
+    },
+    {
+      "epoch": 0.08593322386425835,
+      "grad_norm": 0.002458269475027919,
+      "learning_rate": 9.81889824347377e-06,
+      "loss": 0.0001,
+      "step": 157
+    },
+    {
+      "epoch": 0.08648056923918993,
+      "grad_norm": 0.21147428452968597,
+      "learning_rate": 9.816598120432159e-06,
+      "loss": 0.0073,
+      "step": 158
+    },
+    {
+      "epoch": 0.08702791461412152,
+      "grad_norm": 0.007792654912918806,
+      "learning_rate": 9.81428375565374e-06,
+      "loss": 0.0004,
+      "step": 159
+    },
+    {
+      "epoch": 0.08757525998905309,
+      "grad_norm": 10.54150390625,
+      "learning_rate": 9.811955155981641e-06,
+      "loss": 0.621,
+      "step": 160
+    },
+    {
+      "epoch": 0.08812260536398467,
+      "grad_norm": 0.33467328548431396,
+      "learning_rate": 9.809612328301071e-06,
+      "loss": 0.0132,
+      "step": 161
+    },
+    {
+      "epoch": 0.08866995073891626,
+      "grad_norm": 0.025340264663100243,
+      "learning_rate": 9.807255279539313e-06,
+      "loss": 0.0009,
+      "step": 162
+    },
+    {
+      "epoch": 0.08921729611384784,
+      "grad_norm": 2.977665901184082,
+      "learning_rate": 9.8048840166657e-06,
+      "loss": 0.0356,
+      "step": 163
+    },
+    {
+      "epoch": 0.08976464148877943,
+      "grad_norm": 0.02018612250685692,
+      "learning_rate": 9.80249854669159e-06,
+      "loss": 0.001,
+      "step": 164
+    },
+    {
+      "epoch": 0.090311986863711,
+      "grad_norm": 0.024333668872714043,
+      "learning_rate": 9.80009887667035e-06,
+      "loss": 0.0011,
+      "step": 165
+    },
+    {
+      "epoch": 0.09085933223864258,
+      "grad_norm": 0.02486356534063816,
+      "learning_rate": 9.797685013697336e-06,
+      "loss": 0.0008,
+      "step": 166
+    },
+    {
+      "epoch": 0.09140667761357417,
+      "grad_norm": 0.009108465164899826,
+      "learning_rate": 9.795256964909868e-06,
+      "loss": 0.0004,
+      "step": 167
+    },
+    {
+      "epoch": 0.09195402298850575,
+      "grad_norm": 0.004747547209262848,
+      "learning_rate": 9.792814737487207e-06,
+      "loss": 0.0002,
+      "step": 168
+    },
+    {
+      "epoch": 0.09250136836343732,
+      "grad_norm": 2.016352653503418,
+      "learning_rate": 9.790358338650546e-06,
+      "loss": 0.1343,
+      "step": 169
+    },
+    {
+      "epoch": 0.09304871373836891,
+      "grad_norm": 0.003539568977430463,
+      "learning_rate": 9.787887775662969e-06,
+      "loss": 0.0002,
+      "step": 170
+    },
+    {
+      "epoch": 0.09359605911330049,
+      "grad_norm": 0.01274325605481863,
+      "learning_rate": 9.78540305582945e-06,
+      "loss": 0.0005,
+      "step": 171
+    },
+    {
+      "epoch": 0.09414340448823208,
+      "grad_norm": 0.8760842680931091,
+      "learning_rate": 9.78290418649682e-06,
+      "loss": 0.0427,
+      "step": 172
+    },
+    {
+      "epoch": 0.09469074986316366,
+      "grad_norm": 0.008066395297646523,
+      "learning_rate": 9.780391175053744e-06,
+      "loss": 0.0003,
+      "step": 173
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 3.9684042930603027,
+      "learning_rate": 9.777864028930705e-06,
+      "loss": 0.5087,
+      "step": 174
+    },
+    {
+      "epoch": 0.09578544061302682,
+      "grad_norm": 1.021899938583374,
+      "learning_rate": 9.775322755599979e-06,
+      "loss": 0.0392,
+      "step": 175
+    },
+    {
+      "epoch": 0.0963327859879584,
+      "grad_norm": 0.6588534712791443,
+      "learning_rate": 9.77276736257561e-06,
+      "loss": 0.0338,
+      "step": 176
+    },
+    {
+      "epoch": 0.09688013136288999,
+      "grad_norm": 0.017111433669924736,
+      "learning_rate": 9.7701978574134e-06,
+      "loss": 0.0005,
+      "step": 177
+    },
+    {
+      "epoch": 0.09742747673782157,
+      "grad_norm": 0.010617710649967194,
+      "learning_rate": 9.76761424771087e-06,
+      "loss": 0.0004,
+      "step": 178
+    },
+    {
+      "epoch": 0.09797482211275314,
+      "grad_norm": 0.0035194838419556618,
+      "learning_rate": 9.765016541107247e-06,
+      "loss": 0.0002,
+      "step": 179
+    },
+    {
+      "epoch": 0.09852216748768473,
+      "grad_norm": 0.0017975402297452092,
+      "learning_rate": 9.762404745283439e-06,
+      "loss": 0.0001,
+      "step": 180
+    },
+    {
+      "epoch": 0.09906951286261631,
+      "grad_norm": 0.007557627744972706,
+      "learning_rate": 9.759778867962017e-06,
+      "loss": 0.0002,
+      "step": 181
+    },
+    {
+      "epoch": 0.09961685823754789,
+      "grad_norm": 1.118253469467163,
+      "learning_rate": 9.757138916907184e-06,
+      "loss": 0.0372,
+      "step": 182
+    },
+    {
+      "epoch": 0.10016420361247948,
+      "grad_norm": 2.939692258834839,
+      "learning_rate": 9.754484899924762e-06,
+      "loss": 0.3909,
+      "step": 183
+    },
+    {
+      "epoch": 0.10071154898741105,
+      "grad_norm": 0.13486160337924957,
+      "learning_rate": 9.751816824862152e-06,
+      "loss": 0.0047,
+      "step": 184
+    },
+    {
+      "epoch": 0.10125889436234264,
+      "grad_norm": 5.695981025695801,
+      "learning_rate": 9.749134699608336e-06,
+      "loss": 0.1484,
+      "step": 185
+    },
+    {
+      "epoch": 0.10180623973727422,
+      "grad_norm": 0.004681828897446394,
+      "learning_rate": 9.746438532093827e-06,
+      "loss": 0.0002,
+      "step": 186
+    },
+    {
+      "epoch": 0.1023535851122058,
+      "grad_norm": 0.007142484653741121,
+      "learning_rate": 9.74372833029067e-06,
+      "loss": 0.0003,
+      "step": 187
+    },
+    {
+      "epoch": 0.10290093048713739,
+      "grad_norm": 0.016086289659142494,
+      "learning_rate": 9.741004102212395e-06,
+      "loss": 0.0008,
+      "step": 188
+    },
+    {
+      "epoch": 0.10344827586206896,
+      "grad_norm": 0.0052831051871180534,
+      "learning_rate": 9.738265855914014e-06,
+      "loss": 0.0003,
+      "step": 189
+    },
+    {
+      "epoch": 0.10399562123700054,
+      "grad_norm": 0.4634156823158264,
+      "learning_rate": 9.735513599491982e-06,
+      "loss": 0.0108,
+      "step": 190
+    },
+    {
+      "epoch": 0.10454296661193213,
+      "grad_norm": 0.004725305829197168,
+      "learning_rate": 9.732747341084185e-06,
+      "loss": 0.0002,
+      "step": 191
+    },
+    {
+      "epoch": 0.10509031198686371,
+      "grad_norm": 0.0022925010416656733,
+      "learning_rate": 9.729967088869907e-06,
+      "loss": 0.0001,
+      "step": 192
+    },
+    {
+      "epoch": 0.1056376573617953,
+      "grad_norm": 0.0022578220814466476,
+      "learning_rate": 9.727172851069807e-06,
+      "loss": 0.0001,
+      "step": 193
+    },
+    {
+      "epoch": 0.10618500273672687,
+      "grad_norm": 0.005604981444776058,
+      "learning_rate": 9.7243646359459e-06,
+      "loss": 0.0002,
+      "step": 194
+    },
+    {
+      "epoch": 0.10673234811165845,
+      "grad_norm": 0.0032916353084146976,
+      "learning_rate": 9.721542451801526e-06,
+      "loss": 0.0001,
+      "step": 195
+    },
+    {
+      "epoch": 0.10727969348659004,
+      "grad_norm": 0.009358298033475876,
+      "learning_rate": 9.718706306981332e-06,
+      "loss": 0.0004,
+      "step": 196
+    },
+    {
+      "epoch": 0.10782703886152162,
+      "grad_norm": 0.0041911546140909195,
+      "learning_rate": 9.715856209871243e-06,
+      "loss": 0.0002,
+      "step": 197
+    },
+    {
+      "epoch": 0.10837438423645321,
+      "grad_norm": 0.003506641834974289,
+      "learning_rate": 9.712992168898436e-06,
+      "loss": 0.0002,
+      "step": 198
+    },
+    {
+      "epoch": 0.10892172961138478,
+      "grad_norm": 4.677643775939941,
+      "learning_rate": 9.71011419253132e-06,
+      "loss": 0.6282,
+      "step": 199
+    },
+    {
+      "epoch": 0.10946907498631636,
+      "grad_norm": 0.02117490954697132,
+      "learning_rate": 9.707222289279508e-06,
+      "loss": 0.0008,
+      "step": 200
+    },
+    {
+      "epoch": 0.11001642036124795,
+      "grad_norm": 0.003004522994160652,
+      "learning_rate": 9.704316467693789e-06,
+      "loss": 0.0001,
+      "step": 201
+    },
+    {
+      "epoch": 0.11056376573617953,
+      "grad_norm": 0.003951009828597307,
+      "learning_rate": 9.701396736366108e-06,
+      "loss": 0.0001,
+      "step": 202
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 0.005045830737799406,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.0003,
+      "step": 203
+    },
+    {
+      "epoch": 0.1116584564860427,
+      "grad_norm": 0.001341570750810206,
+      "learning_rate": 9.695515579058265e-06,
+      "loss": 0.0001,
+      "step": 204
+    },
+    {
+      "epoch": 0.11220580186097427,
+      "grad_norm": 0.007657123729586601,
+      "learning_rate": 9.692554170467529e-06,
+      "loss": 0.0002,
+      "step": 205
+    },
+    {
+      "epoch": 0.11275314723590586,
+      "grad_norm": 0.10520672798156738,
+      "learning_rate": 9.689578886913641e-06,
+      "loss": 0.0032,
+      "step": 206
+    },
+    {
+      "epoch": 0.11330049261083744,
+      "grad_norm": 0.001422167639248073,
+      "learning_rate": 9.686589737193929e-06,
+      "loss": 0.0001,
+      "step": 207
+    },
+    {
+      "epoch": 0.11384783798576902,
+      "grad_norm": 0.00470214756205678,
+      "learning_rate": 9.683586730146727e-06,
+      "loss": 0.0001,
+      "step": 208
+    },
+    {
+      "epoch": 0.1143951833607006,
+      "grad_norm": 4.35874080657959,
+      "learning_rate": 9.680569874651336e-06,
+      "loss": 0.0846,
+      "step": 209
+    },
+    {
+      "epoch": 0.11494252873563218,
+      "grad_norm": 0.006586556322872639,
+      "learning_rate": 9.677539179628005e-06,
+      "loss": 0.0004,
+      "step": 210
+    },
+    {
+      "epoch": 0.11548987411056376,
+      "grad_norm": 0.004729663487523794,
+      "learning_rate": 9.674494654037909e-06,
+      "loss": 0.0001,
+      "step": 211
+    },
+    {
+      "epoch": 0.11603721948549535,
+      "grad_norm": 0.015061924234032631,
+      "learning_rate": 9.67143630688311e-06,
+      "loss": 0.0006,
+      "step": 212
+    },
+    {
+      "epoch": 0.11658456486042693,
+      "grad_norm": 0.009670287370681763,
+      "learning_rate": 9.668364147206542e-06,
+      "loss": 0.0004,
+      "step": 213
+    },
+    {
+      "epoch": 0.11713191023535852,
+      "grad_norm": 7.630548477172852,
+      "learning_rate": 9.665278184091981e-06,
+      "loss": 0.928,
+      "step": 214
+    },
+    {
+      "epoch": 0.11767925561029009,
+      "grad_norm": 0.011364479549229145,
+      "learning_rate": 9.662178426664014e-06,
+      "loss": 0.0004,
+      "step": 215
+    },
+    {
+      "epoch": 0.11822660098522167,
+      "grad_norm": 0.042831990867853165,
+      "learning_rate": 9.659064884088017e-06,
+      "loss": 0.0014,
+      "step": 216
+    },
+    {
+      "epoch": 0.11877394636015326,
+      "grad_norm": 0.004559504333883524,
+      "learning_rate": 9.655937565570124e-06,
+      "loss": 0.0002,
+      "step": 217
+    },
+    {
+      "epoch": 0.11932129173508484,
+      "grad_norm": 0.0060465335845947266,
+      "learning_rate": 9.652796480357203e-06,
+      "loss": 0.0002,
+      "step": 218
+    },
+    {
+      "epoch": 0.11986863711001643,
+      "grad_norm": 0.002930757123976946,
+      "learning_rate": 9.649641637736829e-06,
+      "loss": 0.0002,
+      "step": 219
+    },
+    {
+      "epoch": 0.120415982484948,
+      "grad_norm": 4.215134143829346,
+      "learning_rate": 9.646473047037252e-06,
+      "loss": 0.5299,
+      "step": 220
+    },
+    {
+      "epoch": 0.12096332785987958,
+      "grad_norm": 7.453273296356201,
+      "learning_rate": 9.643290717627376e-06,
+      "loss": 0.6441,
+      "step": 221
+    },
+    {
+      "epoch": 0.12151067323481117,
+      "grad_norm": 0.01627381704747677,
+      "learning_rate": 9.640094658916723e-06,
+      "loss": 0.0009,
+      "step": 222
+    },
+    {
+      "epoch": 0.12205801860974275,
+      "grad_norm": 0.003751779207959771,
+      "learning_rate": 9.636884880355412e-06,
+      "loss": 0.0001,
+      "step": 223
+    },
+    {
+      "epoch": 0.12260536398467432,
+      "grad_norm": 0.0018137397710233927,
+      "learning_rate": 9.63366139143413e-06,
+      "loss": 0.0001,
+      "step": 224
+    },
+    {
+      "epoch": 0.12315270935960591,
+      "grad_norm": 0.9278050661087036,
+      "learning_rate": 9.630424201684105e-06,
+      "loss": 0.0315,
+      "step": 225
+    },
+    {
+      "epoch": 0.12370005473453749,
+      "grad_norm": 0.004502156283706427,
+      "learning_rate": 9.62717332067707e-06,
+      "loss": 0.0002,
+      "step": 226
+    },
+    {
+      "epoch": 0.12424740010946908,
+      "grad_norm": 0.047333408147096634,
+      "learning_rate": 9.623908758025243e-06,
+      "loss": 0.0007,
+      "step": 227
+    },
+    {
+      "epoch": 0.12479474548440066,
+      "grad_norm": 0.08214745670557022,
+      "learning_rate": 9.620630523381295e-06,
+      "loss": 0.0029,
+      "step": 228
+    },
+    {
+      "epoch": 0.12534209085933223,
+      "grad_norm": 0.00138851348310709,
+      "learning_rate": 9.617338626438326e-06,
+      "loss": 0.0001,
+      "step": 229
+    },
+    {
+      "epoch": 0.12588943623426382,
+      "grad_norm": 4.332123279571533,
+      "learning_rate": 9.61403307692983e-06,
+      "loss": 0.772,
+      "step": 230
+    },
+    {
+      "epoch": 0.12643678160919541,
+      "grad_norm": 6.4362874031066895,
+      "learning_rate": 9.610713884629667e-06,
+      "loss": 0.9625,
+      "step": 231
+    },
+    {
+      "epoch": 0.12698412698412698,
+      "grad_norm": 0.004271378740668297,
+      "learning_rate": 9.60738105935204e-06,
+      "loss": 0.0002,
+      "step": 232
+    },
+    {
+      "epoch": 0.12753147235905857,
+      "grad_norm": 0.005139256827533245,
+      "learning_rate": 9.604034610951458e-06,
+      "loss": 0.0002,
+      "step": 233
+    },
+    {
+      "epoch": 0.12807881773399016,
+      "grad_norm": 0.005095008760690689,
+      "learning_rate": 9.600674549322716e-06,
+      "loss": 0.0002,
+      "step": 234
+    },
+    {
+      "epoch": 0.12862616310892172,
+      "grad_norm": 5.136434078216553,
+      "learning_rate": 9.597300884400858e-06,
+      "loss": 0.943,
+      "step": 235
+    },
+    {
+      "epoch": 0.1291735084838533,
+      "grad_norm": 0.04719545692205429,
+      "learning_rate": 9.593913626161148e-06,
+      "loss": 0.002,
+      "step": 236
+    },
+    {
+      "epoch": 0.1297208538587849,
+      "grad_norm": 0.013157535344362259,
+      "learning_rate": 9.590512784619045e-06,
+      "loss": 0.0004,
+      "step": 237
+    },
+    {
+      "epoch": 0.13026819923371646,
+      "grad_norm": 4.134466648101807,
+      "learning_rate": 9.587098369830171e-06,
+      "loss": 0.9442,
+      "step": 238
+    },
+    {
+      "epoch": 0.13081554460864805,
+      "grad_norm": 0.0218354444950819,
+      "learning_rate": 9.583670391890285e-06,
+      "loss": 0.0013,
+      "step": 239
+    },
+    {
+      "epoch": 0.13136288998357964,
+      "grad_norm": 0.025551216676831245,
+      "learning_rate": 9.580228860935242e-06,
+      "loss": 0.001,
+      "step": 240
+    },
+    {
+      "epoch": 0.1319102353585112,
+      "grad_norm": 0.9895619750022888,
+      "learning_rate": 9.576773787140974e-06,
+      "loss": 0.0341,
+      "step": 241
+    },
+    {
+      "epoch": 0.1324575807334428,
+      "grad_norm": 4.647812366485596,
+      "learning_rate": 9.57330518072346e-06,
+      "loss": 0.2535,
+      "step": 242
+    },
+    {
+      "epoch": 0.1330049261083744,
+      "grad_norm": 0.027296705171465874,
+      "learning_rate": 9.569823051938689e-06,
+      "loss": 0.0012,
+      "step": 243
+    },
+    {
+      "epoch": 0.13355227148330598,
+      "grad_norm": 0.030924847349524498,
+      "learning_rate": 9.566327411082634e-06,
+      "loss": 0.001,
+      "step": 244
+    },
+    {
+      "epoch": 0.13409961685823754,
+      "grad_norm": 0.00540255568921566,
+      "learning_rate": 9.562818268491216e-06,
+      "loss": 0.0003,
+      "step": 245
+    },
+    {
+      "epoch": 0.13464696223316913,
+      "grad_norm": 0.014651118777692318,
+      "learning_rate": 9.559295634540287e-06,
+      "loss": 0.0006,
+      "step": 246
+    },
+    {
+      "epoch": 0.13519430760810072,
+      "grad_norm": 0.0771181657910347,
+      "learning_rate": 9.555759519645584e-06,
+      "loss": 0.0025,
+      "step": 247
+    },
+    {
+      "epoch": 0.13574165298303228,
+      "grad_norm": 0.036993276327848434,
+      "learning_rate": 9.552209934262703e-06,
+      "loss": 0.0015,
+      "step": 248
+    },
+    {
+      "epoch": 0.13628899835796388,
+      "grad_norm": 0.014038922265172005,
+      "learning_rate": 9.548646888887076e-06,
+      "loss": 0.0007,
+      "step": 249
+    },
+    {
+      "epoch": 0.13683634373289547,
+      "grad_norm": 0.045163851231336594,
+      "learning_rate": 9.54507039405393e-06,
+      "loss": 0.0018,
+      "step": 250
+    },
+    {
+      "epoch": 0.13738368910782703,
+      "grad_norm": 4.474141597747803,
+      "learning_rate": 9.541480460338255e-06,
+      "loss": 0.7153,
+      "step": 251
+    },
+    {
+      "epoch": 0.13793103448275862,
+      "grad_norm": 2.494176149368286,
+      "learning_rate": 9.537877098354787e-06,
+      "loss": 0.1233,
+      "step": 252
+    },
+    {
+      "epoch": 0.1384783798576902,
+      "grad_norm": 0.03589564189314842,
+      "learning_rate": 9.534260318757956e-06,
+      "loss": 0.0011,
+      "step": 253
+    },
+    {
+      "epoch": 0.13902572523262177,
+      "grad_norm": 0.1724754422903061,
+      "learning_rate": 9.530630132241876e-06,
+      "loss": 0.0075,
+      "step": 254
+    },
+    {
+      "epoch": 0.13957307060755336,
+      "grad_norm": 0.8134304881095886,
+      "learning_rate": 9.526986549540292e-06,
+      "loss": 0.1045,
+      "step": 255
+    },
+    {
+      "epoch": 0.14012041598248495,
+      "grad_norm": 0.19864898920059204,
+      "learning_rate": 9.523329581426568e-06,
+      "loss": 0.0081,
+      "step": 256
+    },
+    {
+      "epoch": 0.14066776135741654,
+      "grad_norm": 0.04995502904057503,
+      "learning_rate": 9.519659238713642e-06,
+      "loss": 0.0017,
+      "step": 257
+    },
+    {
+      "epoch": 0.1412151067323481,
+      "grad_norm": 1.2576309442520142,
+      "learning_rate": 9.515975532253994e-06,
+      "loss": 0.2187,
+      "step": 258
+    },
+    {
+      "epoch": 0.1417624521072797,
+      "grad_norm": 0.04681723564863205,
+      "learning_rate": 9.512278472939627e-06,
+      "loss": 0.0018,
+      "step": 259
+    },
+    {
+      "epoch": 0.1423097974822113,
+      "grad_norm": 0.8970088958740234,
+      "learning_rate": 9.508568071702016e-06,
+      "loss": 0.1183,
+      "step": 260
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.037138741463422775,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.0013,
+      "step": 261
+    },
+    {
+      "epoch": 0.14340448823207444,
+      "grad_norm": 0.30721408128738403,
+      "learning_rate": 9.50110728738021e-06,
+      "loss": 0.0112,
+      "step": 262
+    },
+    {
+      "epoch": 0.14395183360700603,
+      "grad_norm": 0.038838911801576614,
+      "learning_rate": 9.49735692635609e-06,
+      "loss": 0.0013,
+      "step": 263
+    },
+    {
+      "epoch": 0.1444991789819376,
+      "grad_norm": 8.811029434204102,
+      "learning_rate": 9.493593267528818e-06,
+      "loss": 1.0208,
+      "step": 264
+    },
+    {
+      "epoch": 0.14504652435686918,
+      "grad_norm": 0.03491875156760216,
+      "learning_rate": 9.489816322026796e-06,
+      "loss": 0.0013,
+      "step": 265
+    },
+    {
+      "epoch": 0.14559386973180077,
+      "grad_norm": 0.3128603398799896,
+      "learning_rate": 9.486026101017711e-06,
+      "loss": 0.01,
+      "step": 266
+    },
+    {
+      "epoch": 0.14614121510673234,
+      "grad_norm": 0.012525409460067749,
+      "learning_rate": 9.482222615708506e-06,
+      "loss": 0.0006,
+      "step": 267
+    },
+    {
+      "epoch": 0.14668856048166393,
+      "grad_norm": 3.2567107677459717,
+      "learning_rate": 9.478405877345339e-06,
+      "loss": 0.2549,
+      "step": 268
+    },
+    {
+      "epoch": 0.14723590585659552,
+      "grad_norm": 0.11480681598186493,
+      "learning_rate": 9.474575897213558e-06,
+      "loss": 0.0059,
+      "step": 269
+    },
+    {
+      "epoch": 0.1477832512315271,
+      "grad_norm": 0.0677303895354271,
+      "learning_rate": 9.470732686637665e-06,
+      "loss": 0.0028,
+      "step": 270
+    },
+    {
+      "epoch": 0.14833059660645867,
+      "grad_norm": 0.010763351805508137,
+      "learning_rate": 9.466876256981279e-06,
+      "loss": 0.0004,
+      "step": 271
+    },
+    {
+      "epoch": 0.14887794198139026,
+      "grad_norm": 0.17489972710609436,
+      "learning_rate": 9.463006619647109e-06,
+      "loss": 0.0061,
+      "step": 272
+    },
+    {
+      "epoch": 0.14942528735632185,
+      "grad_norm": 0.33539700508117676,
+      "learning_rate": 9.459123786076911e-06,
+      "loss": 0.0116,
+      "step": 273
+    },
+    {
+      "epoch": 0.1499726327312534,
+      "grad_norm": 0.004740248434245586,
+      "learning_rate": 9.455227767751467e-06,
+      "loss": 0.0002,
+      "step": 274
+    },
+    {
+      "epoch": 0.150519978106185,
+      "grad_norm": 0.0359928198158741,
+      "learning_rate": 9.451318576190538e-06,
+      "loss": 0.0015,
+      "step": 275
+    },
+    {
+      "epoch": 0.1510673234811166,
+      "grad_norm": 0.03405028209090233,
+      "learning_rate": 9.447396222952837e-06,
+      "loss": 0.0012,
+      "step": 276
+    },
+    {
+      "epoch": 0.15161466885604816,
+      "grad_norm": 0.010478261858224869,
+      "learning_rate": 9.443460719635993e-06,
+      "loss": 0.0004,
+      "step": 277
+    },
+    {
+      "epoch": 0.15216201423097975,
+      "grad_norm": 0.10749869793653488,
+      "learning_rate": 9.43951207787652e-06,
+      "loss": 0.0048,
+      "step": 278
+    },
+    {
+      "epoch": 0.15270935960591134,
+      "grad_norm": 0.021480495110154152,
+      "learning_rate": 9.435550309349776e-06,
+      "loss": 0.001,
+      "step": 279
+    },
+    {
+      "epoch": 0.1532567049808429,
+      "grad_norm": 0.012531301937997341,
+      "learning_rate": 9.431575425769938e-06,
+      "loss": 0.0005,
+      "step": 280
+    },
+    {
+      "epoch": 0.1538040503557745,
+      "grad_norm": 3.6380410194396973,
+      "learning_rate": 9.427587438889954e-06,
+      "loss": 0.4189,
+      "step": 281
+    },
+    {
+      "epoch": 0.15435139573070608,
+      "grad_norm": 3.237821340560913,
+      "learning_rate": 9.423586360501521e-06,
+      "loss": 0.2345,
+      "step": 282
+    },
+    {
+      "epoch": 0.15489874110563764,
+      "grad_norm": 0.009717877954244614,
+      "learning_rate": 9.419572202435044e-06,
+      "loss": 0.0004,
+      "step": 283
+    },
+    {
+      "epoch": 0.15544608648056923,
+      "grad_norm": 0.018061315640807152,
+      "learning_rate": 9.415544976559601e-06,
+      "loss": 0.0007,
+      "step": 284
+    },
+    {
+      "epoch": 0.15599343185550082,
+      "grad_norm": 3.929448366165161,
+      "learning_rate": 9.411504694782909e-06,
+      "loss": 0.5633,
+      "step": 285
+    },
+    {
+      "epoch": 0.15654077723043242,
+      "grad_norm": 0.016803627833724022,
+      "learning_rate": 9.407451369051293e-06,
+      "loss": 0.0006,
+      "step": 286
+    },
+    {
+      "epoch": 0.15708812260536398,
+      "grad_norm": 0.011305440217256546,
+      "learning_rate": 9.40338501134964e-06,
+      "loss": 0.0005,
+      "step": 287
+    },
+    {
+      "epoch": 0.15763546798029557,
+      "grad_norm": 0.8303712606430054,
+      "learning_rate": 9.399305633701372e-06,
+      "loss": 0.0285,
+      "step": 288
+    },
+    {
+      "epoch": 0.15818281335522716,
+      "grad_norm": 0.009512335993349552,
+      "learning_rate": 9.395213248168414e-06,
+      "loss": 0.0004,
+      "step": 289
+    },
+    {
+      "epoch": 0.15873015873015872,
+      "grad_norm": 0.004181810654699802,
+      "learning_rate": 9.391107866851143e-06,
+      "loss": 0.0002,
+      "step": 290
+    },
+    {
+      "epoch": 0.1592775041050903,
+      "grad_norm": 0.12350267171859741,
+      "learning_rate": 9.38698950188837e-06,
+      "loss": 0.004,
+      "step": 291
+    },
+    {
+      "epoch": 0.1598248494800219,
+      "grad_norm": 0.04803336039185524,
+      "learning_rate": 9.382858165457291e-06,
+      "loss": 0.0014,
+      "step": 292
+    },
+    {
+      "epoch": 0.16037219485495346,
+      "grad_norm": 2.1996352672576904,
+      "learning_rate": 9.378713869773462e-06,
+      "loss": 0.1333,
+      "step": 293
+    },
+    {
+      "epoch": 0.16091954022988506,
+      "grad_norm": 0.014819197356700897,
+      "learning_rate": 9.374556627090749e-06,
+      "loss": 0.0008,
+      "step": 294
+    },
+    {
+      "epoch": 0.16146688560481665,
+      "grad_norm": 0.012402137741446495,
+      "learning_rate": 9.370386449701306e-06,
+      "loss": 0.0006,
+      "step": 295
+    },
+    {
+      "epoch": 0.1620142309797482,
+      "grad_norm": 0.05959853157401085,
+      "learning_rate": 9.366203349935531e-06,
+      "loss": 0.0025,
+      "step": 296
+    },
+    {
+      "epoch": 0.1625615763546798,
+      "grad_norm": 0.006656737066805363,
+      "learning_rate": 9.36200734016203e-06,
+      "loss": 0.0003,
+      "step": 297
+    },
+    {
+      "epoch": 0.1631089217296114,
+      "grad_norm": 0.019398879259824753,
+      "learning_rate": 9.35779843278758e-06,
+      "loss": 0.0005,
+      "step": 298
+    },
+    {
+      "epoch": 0.16365626710454298,
+      "grad_norm": 0.022218007594347,
+      "learning_rate": 9.353576640257096e-06,
+      "loss": 0.0008,
+      "step": 299
+    },
+    {
+      "epoch": 0.16420361247947454,
+      "grad_norm": 0.023649122565984726,
+      "learning_rate": 9.349341975053593e-06,
+      "loss": 0.001,
+      "step": 300
+    },
+    {
+      "epoch": 0.16475095785440613,
+      "grad_norm": 1.304732084274292,
+      "learning_rate": 9.345094449698143e-06,
+      "loss": 0.0465,
+      "step": 301
+    },
+    {
+      "epoch": 0.16529830322933772,
+      "grad_norm": 0.006432794965803623,
+      "learning_rate": 9.34083407674985e-06,
+      "loss": 0.0003,
+      "step": 302
+    },
+    {
+      "epoch": 0.16584564860426929,
+      "grad_norm": 0.4880763590335846,
+      "learning_rate": 9.336560868805799e-06,
+      "loss": 0.0188,
+      "step": 303
+    },
+    {
+      "epoch": 0.16639299397920088,
+      "grad_norm": 0.007425636053085327,
+      "learning_rate": 9.33227483850103e-06,
+      "loss": 0.0002,
+      "step": 304
+    },
+    {
+      "epoch": 0.16694033935413247,
+      "grad_norm": 0.057750653475522995,
+      "learning_rate": 9.327975998508496e-06,
+      "loss": 0.0008,
+      "step": 305
+    },
+    {
+      "epoch": 0.16748768472906403,
+      "grad_norm": 0.035874757915735245,
+      "learning_rate": 9.32366436153902e-06,
+      "loss": 0.0017,
+      "step": 306
+    },
+    {
+      "epoch": 0.16803503010399562,
+      "grad_norm": 0.006478885188698769,
+      "learning_rate": 9.319339940341272e-06,
+      "loss": 0.0003,
+      "step": 307
+    },
+    {
+      "epoch": 0.1685823754789272,
+      "grad_norm": 0.05246639624238014,
+      "learning_rate": 9.315002747701716e-06,
+      "loss": 0.0021,
+      "step": 308
+    },
+    {
+      "epoch": 0.16912972085385877,
+      "grad_norm": 0.019346576184034348,
+      "learning_rate": 9.310652796444581e-06,
+      "loss": 0.001,
+      "step": 309
+    },
+    {
+      "epoch": 0.16967706622879036,
+      "grad_norm": 0.0032381361816078424,
+      "learning_rate": 9.306290099431822e-06,
+      "loss": 0.0001,
+      "step": 310
+    },
+    {
+      "epoch": 0.17022441160372195,
+      "grad_norm": 0.002722548320889473,
+      "learning_rate": 9.301914669563077e-06,
+      "loss": 0.0002,
+      "step": 311
+    },
+    {
+      "epoch": 0.17077175697865354,
+      "grad_norm": 0.13813874125480652,
+      "learning_rate": 9.297526519775637e-06,
+      "loss": 0.0055,
+      "step": 312
+    },
+    {
+      "epoch": 0.1713191023535851,
+      "grad_norm": 0.014929019846022129,
+      "learning_rate": 9.293125663044399e-06,
+      "loss": 0.0002,
+      "step": 313
+    },
+    {
+      "epoch": 0.1718664477285167,
+      "grad_norm": 0.016266122460365295,
+      "learning_rate": 9.288712112381834e-06,
+      "loss": 0.0008,
+      "step": 314
+    },
+    {
+      "epoch": 0.1724137931034483,
+      "grad_norm": 0.002234517829492688,
+      "learning_rate": 9.284285880837947e-06,
+      "loss": 0.0001,
+      "step": 315
+    },
+    {
+      "epoch": 0.17296113847837985,
+      "grad_norm": 0.054850004613399506,
+      "learning_rate": 9.279846981500237e-06,
+      "loss": 0.0029,
+      "step": 316
+    },
+    {
+      "epoch": 0.17350848385331144,
+      "grad_norm": 0.0061068604700267315,
+      "learning_rate": 9.275395427493662e-06,
+      "loss": 0.0003,
+      "step": 317
+    },
+    {
+      "epoch": 0.17405582922824303,
+      "grad_norm": 0.005567502696067095,
+      "learning_rate": 9.27093123198059e-06,
+      "loss": 0.0002,
+      "step": 318
+    },
+    {
+      "epoch": 0.1746031746031746,
+      "grad_norm": 2.0154500007629395,
+      "learning_rate": 9.266454408160779e-06,
+      "loss": 0.2103,
+      "step": 319
+    },
+    {
+      "epoch": 0.17515051997810618,
+      "grad_norm": 0.006063805893063545,
+      "learning_rate": 9.261964969271315e-06,
+      "loss": 0.0002,
+      "step": 320
+    },
+    {
+      "epoch": 0.17569786535303777,
+      "grad_norm": 0.007771521806716919,
+      "learning_rate": 9.257462928586589e-06,
+      "loss": 0.0004,
+      "step": 321
+    },
+    {
+      "epoch": 0.17624521072796934,
+      "grad_norm": 0.0036675152368843555,
+      "learning_rate": 9.252948299418255e-06,
+      "loss": 0.0001,
+      "step": 322
+    },
+    {
+      "epoch": 0.17679255610290093,
+      "grad_norm": 0.003957046661525965,
+      "learning_rate": 9.248421095115185e-06,
+      "loss": 0.0002,
+      "step": 323
+    },
+    {
+      "epoch": 0.17733990147783252,
+      "grad_norm": 0.008962135761976242,
+      "learning_rate": 9.243881329063436e-06,
+      "loss": 0.0003,
+      "step": 324
+    },
+    {
+      "epoch": 0.17788724685276408,
+      "grad_norm": 0.4392240047454834,
+      "learning_rate": 9.239329014686207e-06,
+      "loss": 0.0103,
+      "step": 325
+    },
+    {
+      "epoch": 0.17843459222769567,
+      "grad_norm": 0.003559105796739459,
+      "learning_rate": 9.2347641654438e-06,
+      "loss": 0.0001,
+      "step": 326
+    },
+    {
+      "epoch": 0.17898193760262726,
+      "grad_norm": 0.009468616917729378,
+      "learning_rate": 9.230186794833578e-06,
+      "loss": 0.0005,
+      "step": 327
+    },
+    {
+      "epoch": 0.17952928297755885,
+      "grad_norm": 0.025194939225912094,
+      "learning_rate": 9.225596916389929e-06,
+      "loss": 0.0012,
+      "step": 328
+    },
+    {
+      "epoch": 0.18007662835249041,
+      "grad_norm": 0.03295588493347168,
+      "learning_rate": 9.220994543684225e-06,
+      "loss": 0.0013,
+      "step": 329
+    },
+    {
+      "epoch": 0.180623973727422,
+      "grad_norm": 0.025521008297801018,
+      "learning_rate": 9.216379690324782e-06,
+      "loss": 0.0012,
+      "step": 330
+    },
+    {
+      "epoch": 0.1811713191023536,
+      "grad_norm": 0.004285611677914858,
+      "learning_rate": 9.211752369956814e-06,
+      "loss": 0.0002,
+      "step": 331
+    },
+    {
+      "epoch": 0.18171866447728516,
+      "grad_norm": 0.010934761725366116,
+      "learning_rate": 9.207112596262404e-06,
+      "loss": 0.0005,
+      "step": 332
+    },
+    {
+      "epoch": 0.18226600985221675,
+      "grad_norm": 4.3676371574401855,
+      "learning_rate": 9.202460382960449e-06,
+      "loss": 1.2057,
+      "step": 333
+    },
+    {
+      "epoch": 0.18281335522714834,
+      "grad_norm": 0.023604271933436394,
+      "learning_rate": 9.197795743806634e-06,
+      "loss": 0.0009,
+      "step": 334
+    },
+    {
+      "epoch": 0.1833607006020799,
+      "grad_norm": 0.09049314260482788,
+      "learning_rate": 9.193118692593385e-06,
+      "loss": 0.0027,
+      "step": 335
+    },
+    {
+      "epoch": 0.1839080459770115,
+      "grad_norm": 0.05073768272995949,
+      "learning_rate": 9.188429243149824e-06,
+      "loss": 0.002,
+      "step": 336
+    },
+    {
+      "epoch": 0.18445539135194308,
+      "grad_norm": 0.010885998606681824,
+      "learning_rate": 9.183727409341737e-06,
+      "loss": 0.0002,
+      "step": 337
+    },
+    {
+      "epoch": 0.18500273672687464,
+      "grad_norm": 0.0029311482794582844,
+      "learning_rate": 9.179013205071518e-06,
+      "loss": 0.0001,
+      "step": 338
+    },
+    {
+      "epoch": 0.18555008210180624,
+      "grad_norm": 4.0211181640625,
+      "learning_rate": 9.174286644278154e-06,
+      "loss": 0.849,
+      "step": 339
+    },
+    {
+      "epoch": 0.18609742747673783,
+      "grad_norm": 0.001774416770786047,
+      "learning_rate": 9.169547740937152e-06,
+      "loss": 0.0001,
+      "step": 340
+    },
+    {
+      "epoch": 0.18664477285166942,
+      "grad_norm": 0.0026573315262794495,
+      "learning_rate": 9.164796509060526e-06,
+      "loss": 0.0001,
+      "step": 341
+    },
+    {
+      "epoch": 0.18719211822660098,
+      "grad_norm": 5.1158905029296875,
+      "learning_rate": 9.160032962696734e-06,
+      "loss": 0.8304,
+      "step": 342
+    },
+    {
+      "epoch": 0.18773946360153257,
+      "grad_norm": 1.3284435272216797,
+      "learning_rate": 9.155257115930651e-06,
+      "loss": 0.0391,
+      "step": 343
+    },
+    {
+      "epoch": 0.18828680897646416,
+      "grad_norm": 0.004486436489969492,
+      "learning_rate": 9.15046898288352e-06,
+      "loss": 0.0002,
+      "step": 344
+    },
+    {
+      "epoch": 0.18883415435139572,
+      "grad_norm": 3.2804343700408936,
+      "learning_rate": 9.145668577712911e-06,
+      "loss": 0.2478,
+      "step": 345
+    },
+    {
+      "epoch": 0.1893814997263273,
+      "grad_norm": 0.017663557082414627,
+      "learning_rate": 9.140855914612683e-06,
+      "loss": 0.0008,
+      "step": 346
+    },
+    {
+      "epoch": 0.1899288451012589,
+      "grad_norm": 0.003916335292160511,
+      "learning_rate": 9.136031007812937e-06,
+      "loss": 0.0002,
+      "step": 347
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 0.0020537979435175657,
+      "learning_rate": 9.131193871579975e-06,
+      "loss": 0.0001,
+      "step": 348
+    },
+    {
+      "epoch": 0.19102353585112206,
+      "grad_norm": 0.01471023727208376,
+      "learning_rate": 9.126344520216264e-06,
+      "loss": 0.0007,
+      "step": 349
+    },
+    {
+      "epoch": 0.19157088122605365,
+      "grad_norm": 0.00609373115003109,
+      "learning_rate": 9.121482968060384e-06,
+      "loss": 0.0002,
+      "step": 350
+    },
+    {
+      "epoch": 0.1921182266009852,
+      "grad_norm": 0.004218249581754208,
+      "learning_rate": 9.116609229486992e-06,
+      "loss": 0.0002,
+      "step": 351
+    },
+    {
+      "epoch": 0.1926655719759168,
+      "grad_norm": 0.17256547510623932,
+      "learning_rate": 9.11172331890678e-06,
+      "loss": 0.0087,
+      "step": 352
+    },
+    {
+      "epoch": 0.1932129173508484,
+      "grad_norm": 0.33836865425109863,
+      "learning_rate": 9.106825250766424e-06,
+      "loss": 0.0185,
+      "step": 353
+    },
+    {
+      "epoch": 0.19376026272577998,
+      "grad_norm": 0.016286412253975868,
+      "learning_rate": 9.101915039548557e-06,
+      "loss": 0.0008,
+      "step": 354
+    },
+    {
+      "epoch": 0.19430760810071154,
+      "grad_norm": 2.213080883026123,
+      "learning_rate": 9.096992699771707e-06,
+      "loss": 0.3761,
+      "step": 355
+    },
+    {
+      "epoch": 0.19485495347564313,
+      "grad_norm": 0.03972277790307999,
+      "learning_rate": 9.092058245990271e-06,
+      "loss": 0.0017,
+      "step": 356
+    },
+    {
+      "epoch": 0.19540229885057472,
+      "grad_norm": 0.028462251648306847,
+      "learning_rate": 9.08711169279446e-06,
+      "loss": 0.0013,
+      "step": 357
+    },
+    {
+      "epoch": 0.1959496442255063,
+      "grad_norm": 0.04045112803578377,
+      "learning_rate": 9.082153054810263e-06,
+      "loss": 0.0016,
+      "step": 358
+    },
+    {
+      "epoch": 0.19649698960043788,
+      "grad_norm": 3.064483642578125,
+      "learning_rate": 9.077182346699402e-06,
+      "loss": 0.1838,
+      "step": 359
+    },
+    {
+      "epoch": 0.19704433497536947,
+      "grad_norm": 0.020319310948252678,
+      "learning_rate": 9.072199583159285e-06,
+      "loss": 0.0008,
+      "step": 360
+    },
+    {
+      "epoch": 0.19759168035030103,
+      "grad_norm": 0.004529138095676899,
+      "learning_rate": 9.067204778922968e-06,
+      "loss": 0.0002,
+      "step": 361
+    },
+    {
+      "epoch": 0.19813902572523262,
+      "grad_norm": 0.010784142650663853,
+      "learning_rate": 9.062197948759112e-06,
+      "loss": 0.0004,
+      "step": 362
+    },
+    {
+      "epoch": 0.1986863711001642,
+      "grad_norm": 0.02503327652812004,
+      "learning_rate": 9.057179107471926e-06,
+      "loss": 0.001,
+      "step": 363
+    },
+    {
+      "epoch": 0.19923371647509577,
+      "grad_norm": 0.024941373616456985,
+      "learning_rate": 9.052148269901145e-06,
+      "loss": 0.0011,
+      "step": 364
+    },
+    {
+      "epoch": 0.19978106185002736,
+      "grad_norm": 2.407067060470581,
+      "learning_rate": 9.047105450921968e-06,
+      "loss": 0.3134,
+      "step": 365
+    },
+    {
+      "epoch": 0.20032840722495895,
+      "grad_norm": 3.505232810974121,
+      "learning_rate": 9.042050665445024e-06,
+      "loss": 0.1596,
+      "step": 366
+    },
+    {
+      "epoch": 0.20087575259989054,
+      "grad_norm": 0.01830684021115303,
+      "learning_rate": 9.03698392841632e-06,
+      "loss": 0.0009,
+      "step": 367
+    },
+    {
+      "epoch": 0.2014230979748221,
+      "grad_norm": 0.03134550899267197,
+      "learning_rate": 9.031905254817209e-06,
+      "loss": 0.0011,
+      "step": 368
+    },
+    {
+      "epoch": 0.2019704433497537,
+      "grad_norm": 0.03162311762571335,
+      "learning_rate": 9.026814659664331e-06,
+      "loss": 0.0013,
+      "step": 369
+    },
+    {
+      "epoch": 0.2025177887246853,
+      "grad_norm": 0.05484980717301369,
+      "learning_rate": 9.021712158009578e-06,
+      "loss": 0.0025,
+      "step": 370
+    },
+    {
+      "epoch": 0.20306513409961685,
+      "grad_norm": 0.14212408661842346,
+      "learning_rate": 9.01659776494005e-06,
+      "loss": 0.0053,
+      "step": 371
+    },
+    {
+      "epoch": 0.20361247947454844,
+      "grad_norm": 9.681675910949707,
+      "learning_rate": 9.011471495578e-06,
+      "loss": 0.4362,
+      "step": 372
+    },
+    {
+      "epoch": 0.20415982484948003,
+      "grad_norm": 0.023493144661188126,
+      "learning_rate": 9.006333365080808e-06,
+      "loss": 0.0013,
+      "step": 373
+    },
+    {
+      "epoch": 0.2047071702244116,
+      "grad_norm": 0.14610296487808228,
+      "learning_rate": 9.001183388640915e-06,
+      "loss": 0.004,
+      "step": 374
+    },
+    {
+      "epoch": 0.20525451559934318,
+      "grad_norm": 0.017246192321181297,
+      "learning_rate": 8.996021581485795e-06,
+      "loss": 0.0006,
+      "step": 375
+    },
+    {
+      "epoch": 0.20580186097427478,
+      "grad_norm": 0.023940537124872208,
+      "learning_rate": 8.990847958877897e-06,
+      "loss": 0.0009,
+      "step": 376
+    },
+    {
+      "epoch": 0.20634920634920634,
+      "grad_norm": 0.1029309406876564,
+      "learning_rate": 8.985662536114614e-06,
+      "loss": 0.0049,
+      "step": 377
+    },
+    {
+      "epoch": 0.20689655172413793,
+      "grad_norm": 0.00324469106271863,
+      "learning_rate": 8.98046532852822e-06,
+      "loss": 0.0001,
+      "step": 378
+    },
+    {
+      "epoch": 0.20744389709906952,
+      "grad_norm": 0.04180926829576492,
+      "learning_rate": 8.975256351485842e-06,
+      "loss": 0.0015,
+      "step": 379
+    },
+    {
+      "epoch": 0.20799124247400108,
+      "grad_norm": 1.3598195314407349,
+      "learning_rate": 8.970035620389404e-06,
+      "loss": 0.083,
+      "step": 380
+    },
+    {
+      "epoch": 0.20853858784893267,
+      "grad_norm": 0.015766430646181107,
+      "learning_rate": 8.964803150675583e-06,
+      "loss": 0.0006,
+      "step": 381
+    },
+    {
+      "epoch": 0.20908593322386426,
+      "grad_norm": 0.024671118706464767,
+      "learning_rate": 8.95955895781577e-06,
+      "loss": 0.0008,
+      "step": 382
+    },
+    {
+      "epoch": 0.20963327859879585,
+      "grad_norm": 0.07474718987941742,
+      "learning_rate": 8.954303057316014e-06,
+      "loss": 0.004,
+      "step": 383
+    },
+    {
+      "epoch": 0.21018062397372742,
+      "grad_norm": 0.26028382778167725,
+      "learning_rate": 8.949035464716984e-06,
+      "loss": 0.011,
+      "step": 384
+    },
+    {
+      "epoch": 0.210727969348659,
+      "grad_norm": 0.024138959124684334,
+      "learning_rate": 8.943756195593916e-06,
+      "loss": 0.001,
+      "step": 385
+    },
+    {
+      "epoch": 0.2112753147235906,
+      "grad_norm": 0.01153239980340004,
+      "learning_rate": 8.938465265556576e-06,
+      "loss": 0.0004,
+      "step": 386
+    },
+    {
+      "epoch": 0.21182266009852216,
+      "grad_norm": 0.006579091772437096,
+      "learning_rate": 8.93316269024921e-06,
+      "loss": 0.0003,
+      "step": 387
+    },
+    {
+      "epoch": 0.21237000547345375,
+      "grad_norm": 0.023460250347852707,
+      "learning_rate": 8.92784848535049e-06,
+      "loss": 0.0009,
+      "step": 388
+    },
+    {
+      "epoch": 0.21291735084838534,
+      "grad_norm": 3.0115833282470703,
+      "learning_rate": 8.92252266657348e-06,
+      "loss": 0.4978,
+      "step": 389
+    },
+    {
+      "epoch": 0.2134646962233169,
+      "grad_norm": 1.2171604633331299,
+      "learning_rate": 8.917185249665583e-06,
+      "loss": 0.0114,
+      "step": 390
+    },
+    {
+      "epoch": 0.2140120415982485,
+      "grad_norm": 0.004426921717822552,
+      "learning_rate": 8.911836250408494e-06,
+      "loss": 0.0002,
+      "step": 391
+    },
+    {
+      "epoch": 0.21455938697318008,
+      "grad_norm": 0.022343961521983147,
+      "learning_rate": 8.90647568461816e-06,
+      "loss": 0.0009,
+      "step": 392
+    },
+    {
+      "epoch": 0.21510673234811165,
+      "grad_norm": 0.03706027567386627,
+      "learning_rate": 8.901103568144715e-06,
+      "loss": 0.0013,
+      "step": 393
+    },
+    {
+      "epoch": 0.21565407772304324,
+      "grad_norm": 0.10373899340629578,
+      "learning_rate": 8.895719916872463e-06,
+      "loss": 0.0031,
+      "step": 394
+    },
+    {
+      "epoch": 0.21620142309797483,
+      "grad_norm": 3.4080610275268555,
+      "learning_rate": 8.8903247467198e-06,
+      "loss": 0.5519,
+      "step": 395
+    },
+    {
+      "epoch": 0.21674876847290642,
+      "grad_norm": 0.0201762355864048,
+      "learning_rate": 8.88491807363919e-06,
+      "loss": 0.0007,
+      "step": 396
+    },
+    {
+      "epoch": 0.21729611384783798,
+      "grad_norm": 0.04070815071463585,
+      "learning_rate": 8.879499913617107e-06,
+      "loss": 0.0017,
+      "step": 397
+    },
+    {
+      "epoch": 0.21784345922276957,
+      "grad_norm": 0.9010310173034668,
+      "learning_rate": 8.874070282673985e-06,
+      "loss": 0.0271,
+      "step": 398
+    },
+    {
+      "epoch": 0.21839080459770116,
+      "grad_norm": 0.16603036224842072,
+      "learning_rate": 8.868629196864182e-06,
+      "loss": 0.0057,
+      "step": 399
+    },
+    {
+      "epoch": 0.21893814997263272,
+      "grad_norm": 0.010245956480503082,
+      "learning_rate": 8.863176672275921e-06,
+      "loss": 0.0005,
+      "step": 400
+    },
+    {
+      "epoch": 0.2194854953475643,
+      "grad_norm": 0.02312450110912323,
+      "learning_rate": 8.857712725031247e-06,
+      "loss": 0.0011,
+      "step": 401
+    },
+    {
+      "epoch": 0.2200328407224959,
+      "grad_norm": 5.873071193695068,
+      "learning_rate": 8.852237371285984e-06,
+      "loss": 0.2469,
+      "step": 402
+    },
+    {
+      "epoch": 0.22058018609742747,
+      "grad_norm": 0.7911509871482849,
+      "learning_rate": 8.84675062722968e-06,
+      "loss": 0.0259,
+      "step": 403
+    },
+    {
+      "epoch": 0.22112753147235906,
+      "grad_norm": 0.06114649772644043,
+      "learning_rate": 8.841252509085561e-06,
+      "loss": 0.0026,
+      "step": 404
+    },
+    {
+      "epoch": 0.22167487684729065,
+      "grad_norm": 0.01913605071604252,
+      "learning_rate": 8.835743033110482e-06,
+      "loss": 0.0007,
+      "step": 405
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 2.3251266479492188,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.309,
+      "step": 406
+    },
+    {
+      "epoch": 0.2227695675971538,
+      "grad_norm": 0.04632899910211563,
+      "learning_rate": 8.824690072862758e-06,
+      "loss": 0.002,
+      "step": 407
+    },
+    {
+      "epoch": 0.2233169129720854,
+      "grad_norm": 0.23529572784900665,
+      "learning_rate": 8.819146621271546e-06,
+      "loss": 0.0107,
+      "step": 408
+    },
+    {
+      "epoch": 0.22386425834701698,
+      "grad_norm": 0.005736156366765499,
+      "learning_rate": 8.813591877212157e-06,
+      "loss": 0.0003,
+      "step": 409
+    },
+    {
+      "epoch": 0.22441160372194854,
+      "grad_norm": 1.7037173509597778,
+      "learning_rate": 8.80802585710888e-06,
+      "loss": 0.0472,
+      "step": 410
+    },
+    {
+      "epoch": 0.22495894909688013,
+      "grad_norm": 0.1383262276649475,
+      "learning_rate": 8.802448577419343e-06,
+      "loss": 0.0057,
+      "step": 411
+    },
+    {
+      "epoch": 0.22550629447181172,
+      "grad_norm": 0.003472214797511697,
+      "learning_rate": 8.796860054634471e-06,
+      "loss": 0.0002,
+      "step": 412
+    },
+    {
+      "epoch": 0.2260536398467433,
+      "grad_norm": 0.006429862696677446,
+      "learning_rate": 8.791260305278434e-06,
+      "loss": 0.0003,
+      "step": 413
+    },
+    {
+      "epoch": 0.22660098522167488,
+      "grad_norm": 0.2246072143316269,
+      "learning_rate": 8.78564934590859e-06,
+      "loss": 0.0117,
+      "step": 414
+    },
+    {
+      "epoch": 0.22714833059660647,
+      "grad_norm": 0.005228335503488779,
+      "learning_rate": 8.780027193115444e-06,
+      "loss": 0.0002,
+      "step": 415
+    },
+    {
+      "epoch": 0.22769567597153803,
+      "grad_norm": 0.6419472098350525,
+      "learning_rate": 8.774393863522606e-06,
+      "loss": 0.0406,
+      "step": 416
+    },
+    {
+      "epoch": 0.22824302134646962,
+      "grad_norm": 0.9705145955085754,
+      "learning_rate": 8.768749373786722e-06,
+      "loss": 0.0587,
+      "step": 417
+    },
+    {
+      "epoch": 0.2287903667214012,
+      "grad_norm": 0.0039025377482175827,
+      "learning_rate": 8.763093740597447e-06,
+      "loss": 0.0002,
+      "step": 418
+    },
+    {
+      "epoch": 0.22933771209633277,
+      "grad_norm": 0.08845080435276031,
+      "learning_rate": 8.757426980677377e-06,
+      "loss": 0.0032,
+      "step": 419
+    },
+    {
+      "epoch": 0.22988505747126436,
+      "grad_norm": 0.005025999154895544,
+      "learning_rate": 8.751749110782013e-06,
+      "loss": 0.0002,
+      "step": 420
+    },
+    {
+      "epoch": 0.23043240284619596,
+      "grad_norm": 0.0018684992101043463,
+      "learning_rate": 8.746060147699701e-06,
+      "loss": 0.0001,
+      "step": 421
+    },
+    {
+      "epoch": 0.23097974822112752,
+      "grad_norm": 0.0019199148518964648,
+      "learning_rate": 8.740360108251592e-06,
+      "loss": 0.0001,
+      "step": 422
+    },
+    {
+      "epoch": 0.2315270935960591,
+      "grad_norm": 13.037283897399902,
+      "learning_rate": 8.734649009291586e-06,
+      "loss": 1.4691,
+      "step": 423
+    },
+    {
+      "epoch": 0.2320744389709907,
+      "grad_norm": 0.003358412766829133,
+      "learning_rate": 8.72892686770628e-06,
+      "loss": 0.0002,
+      "step": 424
+    },
+    {
+      "epoch": 0.2326217843459223,
+      "grad_norm": 0.014812062494456768,
+      "learning_rate": 8.72319370041493e-06,
+      "loss": 0.0007,
+      "step": 425
+    },
+    {
+      "epoch": 0.23316912972085385,
+      "grad_norm": 0.0031867721118032932,
+      "learning_rate": 8.717449524369386e-06,
+      "loss": 0.0002,
+      "step": 426
+    },
+    {
+      "epoch": 0.23371647509578544,
+      "grad_norm": 0.6433345079421997,
+      "learning_rate": 8.71169435655405e-06,
+      "loss": 0.0296,
+      "step": 427
+    },
+    {
+      "epoch": 0.23426382047071703,
+      "grad_norm": 0.0018559806048870087,
+      "learning_rate": 8.705928213985827e-06,
+      "loss": 0.0001,
+      "step": 428
+    },
+    {
+      "epoch": 0.2348111658456486,
+      "grad_norm": 0.02164350636303425,
+      "learning_rate": 8.700151113714071e-06,
+      "loss": 0.001,
+      "step": 429
+    },
+    {
+      "epoch": 0.23535851122058019,
+      "grad_norm": 0.0030800742097198963,
+      "learning_rate": 8.694363072820535e-06,
+      "loss": 0.0002,
+      "step": 430
+    },
+    {
+      "epoch": 0.23590585659551178,
+      "grad_norm": 0.11861609667539597,
+      "learning_rate": 8.688564108419321e-06,
+      "loss": 0.0041,
+      "step": 431
+    },
+    {
+      "epoch": 0.23645320197044334,
+      "grad_norm": 0.01565435156226158,
+      "learning_rate": 8.68275423765683e-06,
+      "loss": 0.0009,
+      "step": 432
+    },
+    {
+      "epoch": 0.23700054734537493,
+      "grad_norm": 0.024173466488718987,
+      "learning_rate": 8.676933477711714e-06,
+      "loss": 0.0012,
+      "step": 433
+    },
+    {
+      "epoch": 0.23754789272030652,
+      "grad_norm": 0.003673731815069914,
+      "learning_rate": 8.671101845794816e-06,
+      "loss": 0.0001,
+      "step": 434
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 2.293191909790039,
+      "learning_rate": 8.665259359149132e-06,
+      "loss": 0.0733,
+      "step": 435
+    },
+    {
+      "epoch": 0.23864258347016967,
+      "grad_norm": 0.02679836004972458,
+      "learning_rate": 8.65940603504975e-06,
+      "loss": 0.0016,
+      "step": 436
+    },
+    {
+      "epoch": 0.23918992884510126,
+      "grad_norm": 0.006045842077583075,
+      "learning_rate": 8.653541890803798e-06,
+      "loss": 0.0003,
+      "step": 437
+    },
+    {
+      "epoch": 0.23973727422003285,
+      "grad_norm": 0.009248864836990833,
+      "learning_rate": 8.647666943750405e-06,
+      "loss": 0.0005,
+      "step": 438
+    },
+    {
+      "epoch": 0.24028461959496442,
+      "grad_norm": 0.001006833277642727,
+      "learning_rate": 8.641781211260641e-06,
+      "loss": 0.0001,
+      "step": 439
+    },
+    {
+      "epoch": 0.240831964969896,
+      "grad_norm": 3.960789680480957,
+      "learning_rate": 8.635884710737458e-06,
+      "loss": 0.1464,
+      "step": 440
+    },
+    {
+      "epoch": 0.2413793103448276,
+      "grad_norm": 0.0017514158971607685,
+      "learning_rate": 8.629977459615655e-06,
+      "loss": 0.0001,
+      "step": 441
+    },
+    {
+      "epoch": 0.24192665571975916,
+      "grad_norm": 0.0009652414591982961,
+      "learning_rate": 8.624059475361818e-06,
+      "loss": 0.0001,
+      "step": 442
+    },
+    {
+      "epoch": 0.24247400109469075,
+      "grad_norm": 0.0015257166232913733,
+      "learning_rate": 8.618130775474262e-06,
+      "loss": 0.0001,
+      "step": 443
+    },
+    {
+      "epoch": 0.24302134646962234,
+      "grad_norm": 0.0018407206516712904,
+      "learning_rate": 8.612191377482995e-06,
+      "loss": 0.0001,
+      "step": 444
+    },
+    {
+      "epoch": 0.2435686918445539,
+      "grad_norm": 0.002263927599415183,
+      "learning_rate": 8.606241298949651e-06,
+      "loss": 0.0001,
+      "step": 445
+    },
+    {
+      "epoch": 0.2441160372194855,
+      "grad_norm": 0.008505883626639843,
+      "learning_rate": 8.600280557467448e-06,
+      "loss": 0.0004,
+      "step": 446
+    },
+    {
+      "epoch": 0.24466338259441708,
+      "grad_norm": 0.0033861438278108835,
+      "learning_rate": 8.594309170661128e-06,
+      "loss": 0.0002,
+      "step": 447
+    },
+    {
+      "epoch": 0.24521072796934865,
+      "grad_norm": 8.31734561920166,
+      "learning_rate": 8.588327156186915e-06,
+      "loss": 0.2658,
+      "step": 448
+    },
+    {
+      "epoch": 0.24575807334428024,
+      "grad_norm": 0.001098312553949654,
+      "learning_rate": 8.58233453173245e-06,
+      "loss": 0.0001,
+      "step": 449
+    },
+    {
+      "epoch": 0.24630541871921183,
+      "grad_norm": 6.409268379211426,
+      "learning_rate": 8.576331315016753e-06,
+      "loss": 0.4498,
+      "step": 450
+    },
+    {
+      "epoch": 0.24685276409414342,
+      "grad_norm": 0.0007233834476210177,
+      "learning_rate": 8.570317523790155e-06,
+      "loss": 0.0,
+      "step": 451
+    },
+    {
+      "epoch": 0.24740010946907498,
+      "grad_norm": 0.003745045978575945,
+      "learning_rate": 8.564293175834261e-06,
+      "loss": 0.0001,
+      "step": 452
+    },
+    {
+      "epoch": 0.24794745484400657,
+      "grad_norm": 0.0014460004167631269,
+      "learning_rate": 8.558258288961887e-06,
+      "loss": 0.0001,
+      "step": 453
+    },
+    {
+      "epoch": 0.24849480021893816,
+      "grad_norm": 0.2529180347919464,
+      "learning_rate": 8.552212881017012e-06,
+      "loss": 0.0083,
+      "step": 454
+    },
+    {
+      "epoch": 0.24904214559386972,
+      "grad_norm": 0.01214388944208622,
+      "learning_rate": 8.546156969874723e-06,
+      "loss": 0.0007,
+      "step": 455
+    },
+    {
+      "epoch": 0.24958949096880131,
+      "grad_norm": 0.0014981742715463042,
+      "learning_rate": 8.540090573441159e-06,
+      "loss": 0.0001,
+      "step": 456
+    },
+    {
+      "epoch": 0.2501368363437329,
+      "grad_norm": 0.001638900488615036,
+      "learning_rate": 8.534013709653469e-06,
+      "loss": 0.0001,
+      "step": 457
+    },
+    {
+      "epoch": 0.25068418171866447,
+      "grad_norm": 0.05050064995884895,
+      "learning_rate": 8.527926396479746e-06,
+      "loss": 0.002,
+      "step": 458
+    },
+    {
+      "epoch": 0.2512315270935961,
+      "grad_norm": 0.0042986744083464146,
+      "learning_rate": 8.521828651918983e-06,
+      "loss": 0.0002,
+      "step": 459
+    },
+    {
+      "epoch": 0.25177887246852765,
+      "grad_norm": 0.012750094756484032,
+      "learning_rate": 8.515720494001016e-06,
+      "loss": 0.0006,
+      "step": 460
+    },
+    {
+      "epoch": 0.2523262178434592,
+      "grad_norm": 0.008734261617064476,
+      "learning_rate": 8.509601940786472e-06,
+      "loss": 0.0004,
+      "step": 461
+    },
+    {
+      "epoch": 0.25287356321839083,
+      "grad_norm": 0.0008263569907285273,
+      "learning_rate": 8.503473010366713e-06,
+      "loss": 0.0001,
+      "step": 462
+    },
+    {
+      "epoch": 0.2534209085933224,
+      "grad_norm": 0.0020867646671831608,
+      "learning_rate": 8.497333720863786e-06,
+      "loss": 0.0001,
+      "step": 463
+    },
+    {
+      "epoch": 0.25396825396825395,
+      "grad_norm": 0.0010886021191254258,
+      "learning_rate": 8.491184090430365e-06,
+      "loss": 0.0001,
+      "step": 464
+    },
+    {
+      "epoch": 0.2545155993431856,
+      "grad_norm": 0.6500905752182007,
+      "learning_rate": 8.485024137249705e-06,
+      "loss": 0.0172,
+      "step": 465
+    },
+    {
+      "epoch": 0.25506294471811713,
+      "grad_norm": 0.26900380849838257,
+      "learning_rate": 8.478853879535578e-06,
+      "loss": 0.0092,
+      "step": 466
+    },
+    {
+      "epoch": 0.2556102900930487,
+      "grad_norm": 0.09617394208908081,
+      "learning_rate": 8.472673335532226e-06,
+      "loss": 0.0046,
+      "step": 467
+    },
+    {
+      "epoch": 0.2561576354679803,
+      "grad_norm": 3.6172397136688232,
+      "learning_rate": 8.46648252351431e-06,
+      "loss": 0.9964,
+      "step": 468
+    },
+    {
+      "epoch": 0.2567049808429119,
+      "grad_norm": 0.005054984707385302,
+      "learning_rate": 8.460281461786848e-06,
+      "loss": 0.0002,
+      "step": 469
+    },
+    {
+      "epoch": 0.25725232621784344,
+      "grad_norm": 3.7851755619049072,
+      "learning_rate": 8.454070168685162e-06,
+      "loss": 0.4502,
+      "step": 470
+    },
+    {
+      "epoch": 0.25779967159277506,
+      "grad_norm": 0.00290496414527297,
+      "learning_rate": 8.447848662574828e-06,
+      "loss": 0.0002,
+      "step": 471
+    },
+    {
+      "epoch": 0.2583470169677066,
+      "grad_norm": 0.42932426929473877,
+      "learning_rate": 8.441616961851624e-06,
+      "loss": 0.0123,
+      "step": 472
+    },
+    {
+      "epoch": 0.2588943623426382,
+      "grad_norm": 3.719259738922119,
+      "learning_rate": 8.435375084941464e-06,
+      "loss": 0.4588,
+      "step": 473
+    },
+    {
+      "epoch": 0.2594417077175698,
+      "grad_norm": 0.07546942681074142,
+      "learning_rate": 8.429123050300357e-06,
+      "loss": 0.0038,
+      "step": 474
+    },
+    {
+      "epoch": 0.25998905309250137,
+      "grad_norm": 0.01607462391257286,
+      "learning_rate": 8.422860876414344e-06,
+      "loss": 0.0009,
+      "step": 475
+    },
+    {
+      "epoch": 0.26053639846743293,
+      "grad_norm": 0.0010948505951091647,
+      "learning_rate": 8.416588581799447e-06,
+      "loss": 0.0001,
+      "step": 476
+    },
+    {
+      "epoch": 0.26108374384236455,
+      "grad_norm": 3.172180652618408,
+      "learning_rate": 8.41030618500161e-06,
+      "loss": 0.0499,
+      "step": 477
+    },
+    {
+      "epoch": 0.2616310892172961,
+      "grad_norm": 0.002868334762752056,
+      "learning_rate": 8.404013704596653e-06,
+      "loss": 0.0001,
+      "step": 478
+    },
+    {
+      "epoch": 0.26217843459222767,
+      "grad_norm": 0.0034138199407607317,
+      "learning_rate": 8.3977111591902e-06,
+      "loss": 0.0002,
+      "step": 479
+    },
+    {
+      "epoch": 0.2627257799671593,
+      "grad_norm": 0.11267642676830292,
+      "learning_rate": 8.391398567417653e-06,
+      "loss": 0.0031,
+      "step": 480
+    },
+    {
+      "epoch": 0.26327312534209085,
+      "grad_norm": 0.06314973533153534,
+      "learning_rate": 8.385075947944101e-06,
+      "loss": 0.0035,
+      "step": 481
+    },
+    {
+      "epoch": 0.2638204707170224,
+      "grad_norm": 0.0866016298532486,
+      "learning_rate": 8.378743319464293e-06,
+      "loss": 0.004,
+      "step": 482
+    },
+    {
+      "epoch": 0.26436781609195403,
+      "grad_norm": 4.425952434539795,
+      "learning_rate": 8.372400700702569e-06,
+      "loss": 1.0502,
+      "step": 483
+    },
+    {
+      "epoch": 0.2649151614668856,
+      "grad_norm": 0.011416326276957989,
+      "learning_rate": 8.366048110412817e-06,
+      "loss": 0.0006,
+      "step": 484
+    },
+    {
+      "epoch": 0.2654625068418172,
+      "grad_norm": 0.006862805690616369,
+      "learning_rate": 8.359685567378392e-06,
+      "loss": 0.0004,
+      "step": 485
+    },
+    {
+      "epoch": 0.2660098522167488,
+      "grad_norm": 0.13846512138843536,
+      "learning_rate": 8.353313090412093e-06,
+      "loss": 0.0073,
+      "step": 486
+    },
+    {
+      "epoch": 0.26655719759168034,
+      "grad_norm": 0.003636781359091401,
+      "learning_rate": 8.346930698356083e-06,
+      "loss": 0.0002,
+      "step": 487
+    },
+    {
+      "epoch": 0.26710454296661196,
+      "grad_norm": 0.03555435314774513,
+      "learning_rate": 8.340538410081846e-06,
+      "loss": 0.0017,
+      "step": 488
+    },
+    {
+      "epoch": 0.2676518883415435,
+      "grad_norm": 0.004512585233896971,
+      "learning_rate": 8.334136244490128e-06,
+      "loss": 0.0002,
+      "step": 489
+    },
+    {
+      "epoch": 0.2681992337164751,
+      "grad_norm": 0.3295449912548065,
+      "learning_rate": 8.327724220510873e-06,
+      "loss": 0.02,
+      "step": 490
+    },
+    {
+      "epoch": 0.2687465790914067,
+      "grad_norm": 0.06443799287080765,
+      "learning_rate": 8.321302357103183e-06,
+      "loss": 0.0036,
+      "step": 491
+    },
+    {
+      "epoch": 0.26929392446633826,
+      "grad_norm": 0.01787388324737549,
+      "learning_rate": 8.314870673255248e-06,
+      "loss": 0.0009,
+      "step": 492
+    },
+    {
+      "epoch": 0.2698412698412698,
+      "grad_norm": 0.018945252522826195,
+      "learning_rate": 8.308429187984298e-06,
+      "loss": 0.001,
+      "step": 493
+    },
+    {
+      "epoch": 0.27038861521620144,
+      "grad_norm": 0.006173065863549709,
+      "learning_rate": 8.301977920336542e-06,
+      "loss": 0.0002,
+      "step": 494
+    },
+    {
+      "epoch": 0.270935960591133,
+      "grad_norm": 2.493547201156616,
+      "learning_rate": 8.295516889387115e-06,
+      "loss": 0.0723,
+      "step": 495
+    },
+    {
+      "epoch": 0.27148330596606457,
+      "grad_norm": 2.351107120513916,
+      "learning_rate": 8.289046114240019e-06,
+      "loss": 0.4947,
+      "step": 496
+    },
+    {
+      "epoch": 0.2720306513409962,
+      "grad_norm": 0.0063329474069178104,
+      "learning_rate": 8.282565614028068e-06,
+      "loss": 0.0003,
+      "step": 497
+    },
+    {
+      "epoch": 0.27257799671592775,
+      "grad_norm": 0.00849709752947092,
+      "learning_rate": 8.276075407912831e-06,
+      "loss": 0.0003,
+      "step": 498
+    },
+    {
+      "epoch": 0.2731253420908593,
+      "grad_norm": 0.01383709441870451,
+      "learning_rate": 8.269575515084577e-06,
+      "loss": 0.0007,
+      "step": 499
+    },
+    {
+      "epoch": 0.27367268746579093,
+      "grad_norm": 0.006640062667429447,
+      "learning_rate": 8.263065954762212e-06,
+      "loss": 0.0003,
+      "step": 500
+    },
+    {
+      "epoch": 0.2742200328407225,
+      "grad_norm": 0.008006825111806393,
+      "learning_rate": 8.256546746193237e-06,
+      "loss": 0.0004,
+      "step": 501
+    },
+    {
+      "epoch": 0.27476737821565406,
+      "grad_norm": 0.03927391767501831,
+      "learning_rate": 8.250017908653666e-06,
+      "loss": 0.0022,
+      "step": 502
+    },
+    {
+      "epoch": 0.2753147235905857,
+      "grad_norm": 4.720467567443848,
+      "learning_rate": 8.243479461447999e-06,
+      "loss": 0.2078,
+      "step": 503
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 4.6646552085876465,
+      "learning_rate": 8.23693142390914e-06,
+      "loss": 0.8793,
+      "step": 504
+    },
+    {
+      "epoch": 0.2764094143404488,
+      "grad_norm": 4.503306865692139,
+      "learning_rate": 8.230373815398352e-06,
+      "loss": 0.4268,
+      "step": 505
+    },
+    {
+      "epoch": 0.2769567597153804,
+      "grad_norm": 0.004571705125272274,
+      "learning_rate": 8.2238066553052e-06,
+      "loss": 0.0002,
+      "step": 506
+    },
+    {
+      "epoch": 0.277504105090312,
+      "grad_norm": 0.039737485349178314,
+      "learning_rate": 8.21722996304749e-06,
+      "loss": 0.0016,
+      "step": 507
+    },
+    {
+      "epoch": 0.27805145046524354,
+      "grad_norm": 0.02338779717683792,
+      "learning_rate": 8.210643758071211e-06,
+      "loss": 0.0009,
+      "step": 508
+    },
+    {
+      "epoch": 0.27859879584017516,
+      "grad_norm": 0.36878153681755066,
+      "learning_rate": 8.20404805985048e-06,
+      "loss": 0.0202,
+      "step": 509
+    },
+    {
+      "epoch": 0.2791461412151067,
+      "grad_norm": 0.009232649579644203,
+      "learning_rate": 8.197442887887488e-06,
+      "loss": 0.0005,
+      "step": 510
+    },
+    {
+      "epoch": 0.2796934865900383,
+      "grad_norm": 0.2624075710773468,
+      "learning_rate": 8.19082826171243e-06,
+      "loss": 0.0159,
+      "step": 511
+    },
+    {
+      "epoch": 0.2802408319649699,
+      "grad_norm": 2.7321317195892334,
+      "learning_rate": 8.184204200883458e-06,
+      "loss": 0.4077,
+      "step": 512
+    },
+    {
+      "epoch": 0.28078817733990147,
+      "grad_norm": 0.011593791656196117,
+      "learning_rate": 8.177570724986627e-06,
+      "loss": 0.0006,
+      "step": 513
+    },
+    {
+      "epoch": 0.2813355227148331,
+      "grad_norm": 0.05139797925949097,
+      "learning_rate": 8.170927853635824e-06,
+      "loss": 0.0024,
+      "step": 514
+    },
+    {
+      "epoch": 0.28188286808976465,
+      "grad_norm": 0.06633631885051727,
+      "learning_rate": 8.164275606472716e-06,
+      "loss": 0.004,
+      "step": 515
+    },
+    {
+      "epoch": 0.2824302134646962,
+      "grad_norm": 0.12292248010635376,
+      "learning_rate": 8.157614003166695e-06,
+      "loss": 0.0054,
+      "step": 516
+    },
+    {
+      "epoch": 0.28297755883962783,
+      "grad_norm": 0.01715407706797123,
+      "learning_rate": 8.150943063414815e-06,
+      "loss": 0.0008,
+      "step": 517
+    },
+    {
+      "epoch": 0.2835249042145594,
+      "grad_norm": 1.500406265258789,
+      "learning_rate": 8.144262806941743e-06,
+      "loss": 0.1147,
+      "step": 518
+    },
+    {
+      "epoch": 0.28407224958949095,
+      "grad_norm": 0.012590247206389904,
+      "learning_rate": 8.137573253499683e-06,
+      "loss": 0.0006,
+      "step": 519
+    },
+    {
+      "epoch": 0.2846195949644226,
+      "grad_norm": 0.009325980208814144,
+      "learning_rate": 8.130874422868335e-06,
+      "loss": 0.0004,
+      "step": 520
+    },
+    {
+      "epoch": 0.28516694033935414,
+      "grad_norm": 0.009420981630682945,
+      "learning_rate": 8.124166334854831e-06,
+      "loss": 0.0004,
+      "step": 521
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.010938627645373344,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.0005,
+      "step": 522
+    },
+    {
+      "epoch": 0.2862616310892173,
+      "grad_norm": 2.296308755874634,
+      "learning_rate": 8.110722466046666e-06,
+      "loss": 0.4474,
+      "step": 523
+    },
+    {
+      "epoch": 0.2868089764641489,
+      "grad_norm": 0.030201993882656097,
+      "learning_rate": 8.103986725002893e-06,
+      "loss": 0.0014,
+      "step": 524
+    },
+    {
+      "epoch": 0.28735632183908044,
+      "grad_norm": 0.0328528992831707,
+      "learning_rate": 8.097241806078616e-06,
+      "loss": 0.0017,
+      "step": 525
+    },
+    {
+      "epoch": 0.28790366721401206,
+      "grad_norm": 0.06072849780321121,
+      "learning_rate": 8.090487729217238e-06,
+      "loss": 0.0035,
+      "step": 526
+    },
+    {
+      "epoch": 0.2884510125889436,
+      "grad_norm": 0.040568139404058456,
+      "learning_rate": 8.083724514389242e-06,
+      "loss": 0.0022,
+      "step": 527
+    },
+    {
+      "epoch": 0.2889983579638752,
+      "grad_norm": 0.03415596857666969,
+      "learning_rate": 8.076952181592125e-06,
+      "loss": 0.0016,
+      "step": 528
+    },
+    {
+      "epoch": 0.2895457033388068,
+      "grad_norm": 0.04164343699812889,
+      "learning_rate": 8.070170750850354e-06,
+      "loss": 0.0018,
+      "step": 529
+    },
+    {
+      "epoch": 0.29009304871373837,
+      "grad_norm": 0.1452062726020813,
+      "learning_rate": 8.063380242215289e-06,
+      "loss": 0.0096,
+      "step": 530
+    },
+    {
+      "epoch": 0.29064039408866993,
+      "grad_norm": 0.057472992688417435,
+      "learning_rate": 8.05658067576513e-06,
+      "loss": 0.0027,
+      "step": 531
+    },
+    {
+      "epoch": 0.29118773946360155,
+      "grad_norm": 0.0152335399761796,
+      "learning_rate": 8.049772071604864e-06,
+      "loss": 0.0007,
+      "step": 532
+    },
+    {
+      "epoch": 0.2917350848385331,
+      "grad_norm": 0.09490533173084259,
+      "learning_rate": 8.042954449866203e-06,
+      "loss": 0.0056,
+      "step": 533
+    },
+    {
+      "epoch": 0.2922824302134647,
+      "grad_norm": 0.025839099660515785,
+      "learning_rate": 8.036127830707515e-06,
+      "loss": 0.0011,
+      "step": 534
+    },
+    {
+      "epoch": 0.2928297755883963,
+      "grad_norm": 0.02462073415517807,
+      "learning_rate": 8.029292234313777e-06,
+      "loss": 0.001,
+      "step": 535
+    },
+    {
+      "epoch": 0.29337712096332785,
+      "grad_norm": 1.0036952495574951,
+      "learning_rate": 8.022447680896505e-06,
+      "loss": 0.0427,
+      "step": 536
+    },
+    {
+      "epoch": 0.2939244663382594,
+      "grad_norm": 0.014462477527558804,
+      "learning_rate": 8.015594190693705e-06,
+      "loss": 0.0007,
+      "step": 537
+    },
+    {
+      "epoch": 0.29447181171319103,
+      "grad_norm": 0.03679874539375305,
+      "learning_rate": 8.008731783969803e-06,
+      "loss": 0.0016,
+      "step": 538
+    },
+    {
+      "epoch": 0.2950191570881226,
+      "grad_norm": 0.036358315497636795,
+      "learning_rate": 8.001860481015594e-06,
+      "loss": 0.0018,
+      "step": 539
+    },
+    {
+      "epoch": 0.2955665024630542,
+      "grad_norm": 2.7873520851135254,
+      "learning_rate": 7.99498030214817e-06,
+      "loss": 0.447,
+      "step": 540
+    },
+    {
+      "epoch": 0.2961138478379858,
+      "grad_norm": 0.1810341328382492,
+      "learning_rate": 7.988091267710873e-06,
+      "loss": 0.0074,
+      "step": 541
+    },
+    {
+      "epoch": 0.29666119321291734,
+      "grad_norm": 0.1556759476661682,
+      "learning_rate": 7.981193398073232e-06,
+      "loss": 0.0121,
+      "step": 542
+    },
+    {
+      "epoch": 0.29720853858784896,
+      "grad_norm": 3.2317357063293457,
+      "learning_rate": 7.97428671363089e-06,
+      "loss": 0.6869,
+      "step": 543
+    },
+    {
+      "epoch": 0.2977558839627805,
+      "grad_norm": 0.01287668477743864,
+      "learning_rate": 7.967371234805563e-06,
+      "loss": 0.0006,
+      "step": 544
+    },
+    {
+      "epoch": 0.2983032293377121,
+      "grad_norm": 0.13862210512161255,
+      "learning_rate": 7.960446982044964e-06,
+      "loss": 0.0088,
+      "step": 545
+    },
+    {
+      "epoch": 0.2988505747126437,
+      "grad_norm": 0.16142381727695465,
+      "learning_rate": 7.953513975822755e-06,
+      "loss": 0.012,
+      "step": 546
+    },
+    {
+      "epoch": 0.29939792008757526,
+      "grad_norm": 0.12508529424667358,
+      "learning_rate": 7.946572236638477e-06,
+      "loss": 0.0083,
+      "step": 547
+    },
+    {
+      "epoch": 0.2999452654625068,
+      "grad_norm": 0.02337013930082321,
+      "learning_rate": 7.939621785017488e-06,
+      "loss": 0.0012,
+      "step": 548
+    },
+    {
+      "epoch": 0.30049261083743845,
+      "grad_norm": 0.03486739471554756,
+      "learning_rate": 7.932662641510915e-06,
+      "loss": 0.0017,
+      "step": 549
+    },
+    {
+      "epoch": 0.30103995621237,
+      "grad_norm": 0.131191223859787,
+      "learning_rate": 7.925694826695582e-06,
+      "loss": 0.0056,
+      "step": 550
+    },
+    {
+      "epoch": 0.30158730158730157,
+      "grad_norm": 0.058438822627067566,
+      "learning_rate": 7.918718361173951e-06,
+      "loss": 0.004,
+      "step": 551
+    },
+    {
+      "epoch": 0.3021346469622332,
+      "grad_norm": 0.16554298996925354,
+      "learning_rate": 7.911733265574061e-06,
+      "loss": 0.0104,
+      "step": 552
+    },
+    {
+      "epoch": 0.30268199233716475,
+      "grad_norm": 0.017998240888118744,
+      "learning_rate": 7.904739560549475e-06,
+      "loss": 0.0009,
+      "step": 553
+    },
+    {
+      "epoch": 0.3032293377120963,
+      "grad_norm": 0.03947608172893524,
+      "learning_rate": 7.897737266779207e-06,
+      "loss": 0.0023,
+      "step": 554
+    },
+    {
+      "epoch": 0.30377668308702793,
+      "grad_norm": 0.07796397805213928,
+      "learning_rate": 7.890726404967665e-06,
+      "loss": 0.0037,
+      "step": 555
+    },
+    {
+      "epoch": 0.3043240284619595,
+      "grad_norm": 4.119320869445801,
+      "learning_rate": 7.883706995844598e-06,
+      "loss": 0.9868,
+      "step": 556
+    },
+    {
+      "epoch": 0.30487137383689106,
+      "grad_norm": 0.01604565419256687,
+      "learning_rate": 7.87667906016502e-06,
+      "loss": 0.0007,
+      "step": 557
+    },
+    {
+      "epoch": 0.3054187192118227,
+      "grad_norm": 1.9092696905136108,
+      "learning_rate": 7.869642618709162e-06,
+      "loss": 0.4158,
+      "step": 558
+    },
+    {
+      "epoch": 0.30596606458675424,
+      "grad_norm": 0.2341250330209732,
+      "learning_rate": 7.8625976922824e-06,
+      "loss": 0.0119,
+      "step": 559
+    },
+    {
+      "epoch": 0.3065134099616858,
+      "grad_norm": 0.07946468889713287,
+      "learning_rate": 7.855544301715203e-06,
+      "loss": 0.0048,
+      "step": 560
+    },
+    {
+      "epoch": 0.3070607553366174,
+      "grad_norm": 0.12750709056854248,
+      "learning_rate": 7.848482467863062e-06,
+      "loss": 0.0057,
+      "step": 561
+    },
+    {
+      "epoch": 0.307608100711549,
+      "grad_norm": 0.2701328992843628,
+      "learning_rate": 7.841412211606439e-06,
+      "loss": 0.0146,
+      "step": 562
+    },
+    {
+      "epoch": 0.30815544608648054,
+      "grad_norm": 4.954006195068359,
+      "learning_rate": 7.834333553850694e-06,
+      "loss": 0.7856,
+      "step": 563
+    },
+    {
+      "epoch": 0.30870279146141216,
+      "grad_norm": 0.25672805309295654,
+      "learning_rate": 7.827246515526035e-06,
+      "loss": 0.017,
+      "step": 564
+    },
+    {
+      "epoch": 0.3092501368363437,
+      "grad_norm": 0.02571881376206875,
+      "learning_rate": 7.82015111758744e-06,
+      "loss": 0.0013,
+      "step": 565
+    },
+    {
+      "epoch": 0.3097974822112753,
+      "grad_norm": 0.05526258423924446,
+      "learning_rate": 7.813047381014613e-06,
+      "loss": 0.0026,
+      "step": 566
+    },
+    {
+      "epoch": 0.3103448275862069,
+      "grad_norm": 1.4756897687911987,
+      "learning_rate": 7.805935326811913e-06,
+      "loss": 0.048,
+      "step": 567
+    },
+    {
+      "epoch": 0.31089217296113847,
+      "grad_norm": 0.09228195250034332,
+      "learning_rate": 7.798814976008286e-06,
+      "loss": 0.0043,
+      "step": 568
+    },
+    {
+      "epoch": 0.3114395183360701,
+      "grad_norm": 0.4741865396499634,
+      "learning_rate": 7.791686349657219e-06,
+      "loss": 0.0377,
+      "step": 569
+    },
+    {
+      "epoch": 0.31198686371100165,
+      "grad_norm": 0.24840384721755981,
+      "learning_rate": 7.78454946883666e-06,
+      "loss": 0.0219,
+      "step": 570
+    },
+    {
+      "epoch": 0.3125342090859332,
+      "grad_norm": 0.13321872055530548,
+      "learning_rate": 7.777404354648967e-06,
+      "loss": 0.0069,
+      "step": 571
+    },
+    {
+      "epoch": 0.31308155446086483,
+      "grad_norm": 2.217121124267578,
+      "learning_rate": 7.770251028220844e-06,
+      "loss": 0.6064,
+      "step": 572
+    },
+    {
+      "epoch": 0.3136288998357964,
+      "grad_norm": 0.048713941127061844,
+      "learning_rate": 7.763089510703276e-06,
+      "loss": 0.0025,
+      "step": 573
+    },
+    {
+      "epoch": 0.31417624521072796,
+      "grad_norm": 0.128813236951828,
+      "learning_rate": 7.755919823271466e-06,
+      "loss": 0.0072,
+      "step": 574
+    },
+    {
+      "epoch": 0.3147235905856596,
+      "grad_norm": 0.03228011354804039,
+      "learning_rate": 7.748741987124773e-06,
+      "loss": 0.0017,
+      "step": 575
+    },
+    {
+      "epoch": 0.31527093596059114,
+      "grad_norm": 0.12434003502130508,
+      "learning_rate": 7.741556023486655e-06,
+      "loss": 0.0071,
+      "step": 576
+    },
+    {
+      "epoch": 0.3158182813355227,
+      "grad_norm": 0.05032031983137131,
+      "learning_rate": 7.734361953604596e-06,
+      "loss": 0.0021,
+      "step": 577
+    },
+    {
+      "epoch": 0.3163656267104543,
+      "grad_norm": 0.03245704621076584,
+      "learning_rate": 7.727159798750054e-06,
+      "loss": 0.0015,
+      "step": 578
+    },
+    {
+      "epoch": 0.3169129720853859,
+      "grad_norm": 0.09331010282039642,
+      "learning_rate": 7.719949580218387e-06,
+      "loss": 0.0056,
+      "step": 579
+    },
+    {
+      "epoch": 0.31746031746031744,
+      "grad_norm": 0.04344509541988373,
+      "learning_rate": 7.712731319328798e-06,
+      "loss": 0.002,
+      "step": 580
+    },
+    {
+      "epoch": 0.31800766283524906,
+      "grad_norm": 0.016618408262729645,
+      "learning_rate": 7.70550503742427e-06,
+      "loss": 0.0008,
+      "step": 581
+    },
+    {
+      "epoch": 0.3185550082101806,
+      "grad_norm": 0.018471628427505493,
+      "learning_rate": 7.698270755871506e-06,
+      "loss": 0.0008,
+      "step": 582
+    },
+    {
+      "epoch": 0.3191023535851122,
+      "grad_norm": 0.05424068868160248,
+      "learning_rate": 7.691028496060856e-06,
+      "loss": 0.0026,
+      "step": 583
+    },
+    {
+      "epoch": 0.3196496989600438,
+      "grad_norm": 0.28279751539230347,
+      "learning_rate": 7.683778279406261e-06,
+      "loss": 0.0188,
+      "step": 584
+    },
+    {
+      "epoch": 0.32019704433497537,
+      "grad_norm": 0.041227713227272034,
+      "learning_rate": 7.676520127345198e-06,
+      "loss": 0.0019,
+      "step": 585
+    },
+    {
+      "epoch": 0.32074438970990693,
+      "grad_norm": 2.211728096008301,
+      "learning_rate": 7.669254061338591e-06,
+      "loss": 0.4059,
+      "step": 586
+    },
+    {
+      "epoch": 0.32129173508483855,
+      "grad_norm": 2.1793460845947266,
+      "learning_rate": 7.66198010287078e-06,
+      "loss": 0.3825,
+      "step": 587
+    },
+    {
+      "epoch": 0.3218390804597701,
+      "grad_norm": 0.633358359336853,
+      "learning_rate": 7.654698273449435e-06,
+      "loss": 0.0113,
+      "step": 588
+    },
+    {
+      "epoch": 0.3223864258347017,
+      "grad_norm": 0.018418803811073303,
+      "learning_rate": 7.647408594605495e-06,
+      "loss": 0.001,
+      "step": 589
+    },
+    {
+      "epoch": 0.3229337712096333,
+      "grad_norm": 0.019276276230812073,
+      "learning_rate": 7.640111087893114e-06,
+      "loss": 0.001,
+      "step": 590
+    },
+    {
+      "epoch": 0.32348111658456485,
+      "grad_norm": 0.017273103818297386,
+      "learning_rate": 7.632805774889589e-06,
+      "loss": 0.0009,
+      "step": 591
+    },
+    {
+      "epoch": 0.3240284619594964,
+      "grad_norm": 0.0777779296040535,
+      "learning_rate": 7.625492677195298e-06,
+      "loss": 0.0038,
+      "step": 592
+    },
+    {
+      "epoch": 0.32457580733442803,
+      "grad_norm": 0.1303318440914154,
+      "learning_rate": 7.6181718164336415e-06,
+      "loss": 0.0071,
+      "step": 593
+    },
+    {
+      "epoch": 0.3251231527093596,
+      "grad_norm": 0.18606723845005035,
+      "learning_rate": 7.610843214250964e-06,
+      "loss": 0.011,
+      "step": 594
+    },
+    {
+      "epoch": 0.32567049808429116,
+      "grad_norm": 1.8400012254714966,
+      "learning_rate": 7.603506892316513e-06,
+      "loss": 0.1371,
+      "step": 595
+    },
+    {
+      "epoch": 0.3262178434592228,
+      "grad_norm": 0.019331350922584534,
+      "learning_rate": 7.5961628723223505e-06,
+      "loss": 0.0008,
+      "step": 596
+    },
+    {
+      "epoch": 0.32676518883415434,
+      "grad_norm": 0.02180156670510769,
+      "learning_rate": 7.588811175983305e-06,
+      "loss": 0.0011,
+      "step": 597
+    },
+    {
+      "epoch": 0.32731253420908596,
+      "grad_norm": 0.49529480934143066,
+      "learning_rate": 7.581451825036903e-06,
+      "loss": 0.0257,
+      "step": 598
+    },
+    {
+      "epoch": 0.3278598795840175,
+      "grad_norm": 0.08971985429525375,
+      "learning_rate": 7.574084841243302e-06,
+      "loss": 0.0044,
+      "step": 599
+    },
+    {
+      "epoch": 0.3284072249589491,
+      "grad_norm": 1.5169864892959595,
+      "learning_rate": 7.5667102463852314e-06,
+      "loss": 0.4389,
+      "step": 600
+    },
+    {
+      "epoch": 0.3289545703338807,
+      "grad_norm": 0.16540849208831787,
+      "learning_rate": 7.55932806226792e-06,
+      "loss": 0.0108,
+      "step": 601
+    },
+    {
+      "epoch": 0.32950191570881227,
+      "grad_norm": 0.01784505322575569,
+      "learning_rate": 7.551938310719043e-06,
+      "loss": 0.0009,
+      "step": 602
+    },
+    {
+      "epoch": 0.33004926108374383,
+      "grad_norm": 0.02484745904803276,
+      "learning_rate": 7.5445410135886455e-06,
+      "loss": 0.0013,
+      "step": 603
+    },
+    {
+      "epoch": 0.33059660645867545,
+      "grad_norm": 0.00939080398529768,
+      "learning_rate": 7.537136192749086e-06,
+      "loss": 0.0005,
+      "step": 604
+    },
+    {
+      "epoch": 0.331143951833607,
+      "grad_norm": 0.006826468743383884,
+      "learning_rate": 7.529723870094969e-06,
+      "loss": 0.0004,
+      "step": 605
+    },
+    {
+      "epoch": 0.33169129720853857,
+      "grad_norm": 0.5507104396820068,
+      "learning_rate": 7.522304067543082e-06,
+      "loss": 0.0289,
+      "step": 606
+    },
+    {
+      "epoch": 0.3322386425834702,
+      "grad_norm": 0.03984001278877258,
+      "learning_rate": 7.514876807032323e-06,
+      "loss": 0.0015,
+      "step": 607
+    },
+    {
+      "epoch": 0.33278598795840175,
+      "grad_norm": 0.010095106437802315,
+      "learning_rate": 7.507442110523649e-06,
+      "loss": 0.0005,
+      "step": 608
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.014431447722017765,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.0006,
+      "step": 609
+    },
+    {
+      "epoch": 0.33388067870826493,
+      "grad_norm": 2.57116436958313,
+      "learning_rate": 7.492550497466239e-06,
+      "loss": 0.5546,
+      "step": 610
+    },
+    {
+      "epoch": 0.3344280240831965,
+      "grad_norm": 1.8662731647491455,
+      "learning_rate": 7.485093624949085e-06,
+      "loss": 0.2615,
+      "step": 611
+    },
+    {
+      "epoch": 0.33497536945812806,
+      "grad_norm": 0.3963170051574707,
+      "learning_rate": 7.477629404497048e-06,
+      "loss": 0.0106,
+      "step": 612
+    },
+    {
+      "epoch": 0.3355227148330597,
+      "grad_norm": 0.02827736735343933,
+      "learning_rate": 7.470157858180365e-06,
+      "loss": 0.0016,
+      "step": 613
+    },
+    {
+      "epoch": 0.33607006020799124,
+      "grad_norm": 0.1261405646800995,
+      "learning_rate": 7.462679008090935e-06,
+      "loss": 0.0064,
+      "step": 614
+    },
+    {
+      "epoch": 0.3366174055829228,
+      "grad_norm": 0.2770984470844269,
+      "learning_rate": 7.455192876342253e-06,
+      "loss": 0.0154,
+      "step": 615
+    },
+    {
+      "epoch": 0.3371647509578544,
+      "grad_norm": 0.019770491868257523,
+      "learning_rate": 7.447699485069342e-06,
+      "loss": 0.0009,
+      "step": 616
+    },
+    {
+      "epoch": 0.337712096332786,
+      "grad_norm": 0.01447894237935543,
+      "learning_rate": 7.440198856428693e-06,
+      "loss": 0.0007,
+      "step": 617
+    },
+    {
+      "epoch": 0.33825944170771755,
+      "grad_norm": 0.048531387001276016,
+      "learning_rate": 7.432691012598196e-06,
+      "loss": 0.0016,
+      "step": 618
+    },
+    {
+      "epoch": 0.33880678708264916,
+      "grad_norm": 2.619145393371582,
+      "learning_rate": 7.42517597577707e-06,
+      "loss": 0.5279,
+      "step": 619
+    },
+    {
+      "epoch": 0.3393541324575807,
+      "grad_norm": 0.17820040881633759,
+      "learning_rate": 7.41765376818581e-06,
+      "loss": 0.0093,
+      "step": 620
+    },
+    {
+      "epoch": 0.3399014778325123,
+      "grad_norm": 0.08188218623399734,
+      "learning_rate": 7.4101244120661105e-06,
+      "loss": 0.0038,
+      "step": 621
+    },
+    {
+      "epoch": 0.3404488232074439,
+      "grad_norm": 2.2620508670806885,
+      "learning_rate": 7.4025879296807975e-06,
+      "loss": 0.2475,
+      "step": 622
+    },
+    {
+      "epoch": 0.34099616858237547,
+      "grad_norm": 0.013281342573463917,
+      "learning_rate": 7.395044343313777e-06,
+      "loss": 0.0006,
+      "step": 623
+    },
+    {
+      "epoch": 0.3415435139573071,
+      "grad_norm": 0.21208718419075012,
+      "learning_rate": 7.387493675269955e-06,
+      "loss": 0.0121,
+      "step": 624
+    },
+    {
+      "epoch": 0.34209085933223865,
+      "grad_norm": 0.11422328650951385,
+      "learning_rate": 7.379935947875177e-06,
+      "loss": 0.0067,
+      "step": 625
+    },
+    {
+      "epoch": 0.3426382047071702,
+      "grad_norm": 0.015390058048069477,
+      "learning_rate": 7.372371183476159e-06,
+      "loss": 0.0006,
+      "step": 626
+    },
+    {
+      "epoch": 0.34318555008210183,
+      "grad_norm": 0.022556424140930176,
+      "learning_rate": 7.36479940444043e-06,
+      "loss": 0.0009,
+      "step": 627
+    },
+    {
+      "epoch": 0.3437328954570334,
+      "grad_norm": 1.4255520105361938,
+      "learning_rate": 7.3572206331562575e-06,
+      "loss": 0.1016,
+      "step": 628
+    },
+    {
+      "epoch": 0.34428024083196496,
+      "grad_norm": 0.017720935866236687,
+      "learning_rate": 7.349634892032582e-06,
+      "loss": 0.0008,
+      "step": 629
+    },
+    {
+      "epoch": 0.3448275862068966,
+      "grad_norm": 0.021948307752609253,
+      "learning_rate": 7.342042203498952e-06,
+      "loss": 0.001,
+      "step": 630
+    },
+    {
+      "epoch": 0.34537493158182814,
+      "grad_norm": 0.22511789202690125,
+      "learning_rate": 7.334442590005462e-06,
+      "loss": 0.0121,
+      "step": 631
+    },
+    {
+      "epoch": 0.3459222769567597,
+      "grad_norm": 0.012582589872181416,
+      "learning_rate": 7.3268360740226785e-06,
+      "loss": 0.0005,
+      "step": 632
+    },
+    {
+      "epoch": 0.3464696223316913,
+      "grad_norm": 0.0653594508767128,
+      "learning_rate": 7.319222678041578e-06,
+      "loss": 0.0024,
+      "step": 633
+    },
+    {
+      "epoch": 0.3470169677066229,
+      "grad_norm": 0.005534766241908073,
+      "learning_rate": 7.311602424573483e-06,
+      "loss": 0.0002,
+      "step": 634
+    },
+    {
+      "epoch": 0.34756431308155444,
+      "grad_norm": 0.023336416110396385,
+      "learning_rate": 7.3039753361499885e-06,
+      "loss": 0.0012,
+      "step": 635
+    },
+    {
+      "epoch": 0.34811165845648606,
+      "grad_norm": 0.08834053575992584,
+      "learning_rate": 7.2963414353229e-06,
+      "loss": 0.0037,
+      "step": 636
+    },
+    {
+      "epoch": 0.3486590038314176,
+      "grad_norm": 0.028155624866485596,
+      "learning_rate": 7.288700744664167e-06,
+      "loss": 0.0007,
+      "step": 637
+    },
+    {
+      "epoch": 0.3492063492063492,
+      "grad_norm": 1.5263876914978027,
+      "learning_rate": 7.281053286765816e-06,
+      "loss": 0.0783,
+      "step": 638
+    },
+    {
+      "epoch": 0.3497536945812808,
+      "grad_norm": 0.007886398583650589,
+      "learning_rate": 7.273399084239878e-06,
+      "loss": 0.0004,
+      "step": 639
+    },
+    {
+      "epoch": 0.35030103995621237,
+      "grad_norm": 0.08708926290273666,
+      "learning_rate": 7.265738159718332e-06,
+      "loss": 0.0051,
+      "step": 640
+    },
+    {
+      "epoch": 0.35084838533114393,
+      "grad_norm": 0.005664496682584286,
+      "learning_rate": 7.258070535853031e-06,
+      "loss": 0.0002,
+      "step": 641
+    },
+    {
+      "epoch": 0.35139573070607555,
+      "grad_norm": 0.06383395940065384,
+      "learning_rate": 7.250396235315634e-06,
+      "loss": 0.0022,
+      "step": 642
+    },
+    {
+      "epoch": 0.3519430760810071,
+      "grad_norm": 1.0289912223815918,
+      "learning_rate": 7.242715280797547e-06,
+      "loss": 0.0553,
+      "step": 643
+    },
+    {
+      "epoch": 0.3524904214559387,
+      "grad_norm": 0.017405716702342033,
+      "learning_rate": 7.235027695009846e-06,
+      "loss": 0.0006,
+      "step": 644
+    },
+    {
+      "epoch": 0.3530377668308703,
+      "grad_norm": 0.2064036875963211,
+      "learning_rate": 7.2273335006832144e-06,
+      "loss": 0.0124,
+      "step": 645
+    },
+    {
+      "epoch": 0.35358511220580185,
+      "grad_norm": 0.0043057892471551895,
+      "learning_rate": 7.219632720567879e-06,
+      "loss": 0.0002,
+      "step": 646
+    },
+    {
+      "epoch": 0.3541324575807334,
+      "grad_norm": 0.015848618000745773,
+      "learning_rate": 7.211925377433537e-06,
+      "loss": 0.0008,
+      "step": 647
+    },
+    {
+      "epoch": 0.35467980295566504,
+      "grad_norm": 0.006341234780848026,
+      "learning_rate": 7.204211494069292e-06,
+      "loss": 0.0003,
+      "step": 648
+    },
+    {
+      "epoch": 0.3552271483305966,
+      "grad_norm": 0.2039993554353714,
+      "learning_rate": 7.196491093283585e-06,
+      "loss": 0.012,
+      "step": 649
+    },
+    {
+      "epoch": 0.35577449370552816,
+      "grad_norm": 0.01146707870066166,
+      "learning_rate": 7.188764197904129e-06,
+      "loss": 0.0005,
+      "step": 650
+    },
+    {
+      "epoch": 0.3563218390804598,
+      "grad_norm": 1.2888134717941284,
+      "learning_rate": 7.181030830777838e-06,
+      "loss": 0.0737,
+      "step": 651
+    },
+    {
+      "epoch": 0.35686918445539134,
+      "grad_norm": 0.007257033605128527,
+      "learning_rate": 7.173291014770765e-06,
+      "loss": 0.0003,
+      "step": 652
+    },
+    {
+      "epoch": 0.35741652983032296,
+      "grad_norm": 0.09801796823740005,
+      "learning_rate": 7.165544772768027e-06,
+      "loss": 0.0047,
+      "step": 653
+    },
+    {
+      "epoch": 0.3579638752052545,
+      "grad_norm": 0.0024445722810924053,
+      "learning_rate": 7.157792127673747e-06,
+      "loss": 0.0001,
+      "step": 654
+    },
+    {
+      "epoch": 0.3585112205801861,
+      "grad_norm": 0.02576364018023014,
+      "learning_rate": 7.150033102410975e-06,
+      "loss": 0.0013,
+      "step": 655
+    },
+    {
+      "epoch": 0.3590585659551177,
+      "grad_norm": 5.51544713973999,
+      "learning_rate": 7.142267719921629e-06,
+      "loss": 0.6532,
+      "step": 656
+    },
+    {
+      "epoch": 0.35960591133004927,
+      "grad_norm": 0.09530629217624664,
+      "learning_rate": 7.134496003166423e-06,
+      "loss": 0.0057,
+      "step": 657
+    },
+    {
+      "epoch": 0.36015325670498083,
+      "grad_norm": 0.03840158134698868,
+      "learning_rate": 7.1267179751248005e-06,
+      "loss": 0.0012,
+      "step": 658
+    },
+    {
+      "epoch": 0.36070060207991245,
+      "grad_norm": 0.00522483279928565,
+      "learning_rate": 7.118933658794868e-06,
+      "loss": 0.0002,
+      "step": 659
+    },
+    {
+      "epoch": 0.361247947454844,
+      "grad_norm": 0.2433670610189438,
+      "learning_rate": 7.111143077193321e-06,
+      "loss": 0.0168,
+      "step": 660
+    },
+    {
+      "epoch": 0.36179529282977557,
+      "grad_norm": 0.08343702554702759,
+      "learning_rate": 7.103346253355383e-06,
+      "loss": 0.0018,
+      "step": 661
+    },
+    {
+      "epoch": 0.3623426382047072,
+      "grad_norm": 0.006387744098901749,
+      "learning_rate": 7.0955432103347355e-06,
+      "loss": 0.0003,
+      "step": 662
+    },
+    {
+      "epoch": 0.36288998357963875,
+      "grad_norm": 0.0036262532230466604,
+      "learning_rate": 7.087733971203448e-06,
+      "loss": 0.0002,
+      "step": 663
+    },
+    {
+      "epoch": 0.3634373289545703,
+      "grad_norm": 0.10606978833675385,
+      "learning_rate": 7.0799185590519086e-06,
+      "loss": 0.006,
+      "step": 664
+    },
+    {
+      "epoch": 0.36398467432950193,
+      "grad_norm": 0.13019196689128876,
+      "learning_rate": 7.0720969969887595e-06,
+      "loss": 0.0045,
+      "step": 665
+    },
+    {
+      "epoch": 0.3645320197044335,
+      "grad_norm": 0.10207099467515945,
+      "learning_rate": 7.06426930814083e-06,
+      "loss": 0.0027,
+      "step": 666
+    },
+    {
+      "epoch": 0.36507936507936506,
+      "grad_norm": 0.00420819316059351,
+      "learning_rate": 7.056435515653059e-06,
+      "loss": 0.0002,
+      "step": 667
+    },
+    {
+      "epoch": 0.3656267104542967,
+      "grad_norm": 0.09126473218202591,
+      "learning_rate": 7.048595642688436e-06,
+      "loss": 0.0053,
+      "step": 668
+    },
+    {
+      "epoch": 0.36617405582922824,
+      "grad_norm": 0.004398214165121317,
+      "learning_rate": 7.040749712427932e-06,
+      "loss": 0.0002,
+      "step": 669
+    },
+    {
+      "epoch": 0.3667214012041598,
+      "grad_norm": 0.00617770291864872,
+      "learning_rate": 7.032897748070423e-06,
+      "loss": 0.0003,
+      "step": 670
+    },
+    {
+      "epoch": 0.3672687465790914,
+      "grad_norm": 0.10934311896562576,
+      "learning_rate": 7.0250397728326295e-06,
+      "loss": 0.0063,
+      "step": 671
+    },
+    {
+      "epoch": 0.367816091954023,
+      "grad_norm": 0.004969688132405281,
+      "learning_rate": 7.017175809949044e-06,
+      "loss": 0.0002,
+      "step": 672
+    },
+    {
+      "epoch": 0.36836343732895455,
+      "grad_norm": 2.228437900543213,
+      "learning_rate": 7.009305882671864e-06,
+      "loss": 0.2851,
+      "step": 673
+    },
+    {
+      "epoch": 0.36891078270388616,
+      "grad_norm": 0.007352550979703665,
+      "learning_rate": 7.001430014270921e-06,
+      "loss": 0.0003,
+      "step": 674
+    },
+    {
+      "epoch": 0.3694581280788177,
+      "grad_norm": 0.00255901413038373,
+      "learning_rate": 6.993548228033618e-06,
+      "loss": 0.0001,
+      "step": 675
+    },
+    {
+      "epoch": 0.3700054734537493,
+      "grad_norm": 3.149127960205078,
+      "learning_rate": 6.9856605472648494e-06,
+      "loss": 0.5427,
+      "step": 676
+    },
+    {
+      "epoch": 0.3705528188286809,
+      "grad_norm": 0.5940586924552917,
+      "learning_rate": 6.977766995286943e-06,
+      "loss": 0.0356,
+      "step": 677
+    },
+    {
+      "epoch": 0.37110016420361247,
+      "grad_norm": 2.30926775932312,
+      "learning_rate": 6.969867595439586e-06,
+      "loss": 0.4709,
+      "step": 678
+    },
+    {
+      "epoch": 0.3716475095785441,
+      "grad_norm": 0.16510634124279022,
+      "learning_rate": 6.961962371079752e-06,
+      "loss": 0.0081,
+      "step": 679
+    },
+    {
+      "epoch": 0.37219485495347565,
+      "grad_norm": 0.05402740463614464,
+      "learning_rate": 6.954051345581645e-06,
+      "loss": 0.0016,
+      "step": 680
+    },
+    {
+      "epoch": 0.3727422003284072,
+      "grad_norm": 0.004461527802050114,
+      "learning_rate": 6.946134542336615e-06,
+      "loss": 0.0003,
+      "step": 681
+    },
+    {
+      "epoch": 0.37328954570333883,
+      "grad_norm": 3.6112473011016846,
+      "learning_rate": 6.938211984753097e-06,
+      "loss": 0.3841,
+      "step": 682
+    },
+    {
+      "epoch": 0.3738368910782704,
+      "grad_norm": 0.008113270625472069,
+      "learning_rate": 6.930283696256543e-06,
+      "loss": 0.0003,
+      "step": 683
+    },
+    {
+      "epoch": 0.37438423645320196,
+      "grad_norm": 0.15984749794006348,
+      "learning_rate": 6.922349700289348e-06,
+      "loss": 0.0107,
+      "step": 684
+    },
+    {
+      "epoch": 0.3749315818281336,
+      "grad_norm": 0.006165068130940199,
+      "learning_rate": 6.914410020310782e-06,
+      "loss": 0.0003,
+      "step": 685
+    },
+    {
+      "epoch": 0.37547892720306514,
+      "grad_norm": 0.030400443822145462,
+      "learning_rate": 6.906464679796927e-06,
+      "loss": 0.0015,
+      "step": 686
+    },
+    {
+      "epoch": 0.3760262725779967,
+      "grad_norm": 0.005032232962548733,
+      "learning_rate": 6.898513702240592e-06,
+      "loss": 0.0002,
+      "step": 687
+    },
+    {
+      "epoch": 0.3765736179529283,
+      "grad_norm": 0.19334912300109863,
+      "learning_rate": 6.890557111151266e-06,
+      "loss": 0.0082,
+      "step": 688
+    },
+    {
+      "epoch": 0.3771209633278599,
+      "grad_norm": 0.022890301421284676,
+      "learning_rate": 6.882594930055024e-06,
+      "loss": 0.0015,
+      "step": 689
+    },
+    {
+      "epoch": 0.37766830870279144,
+      "grad_norm": 2.6308631896972656,
+      "learning_rate": 6.8746271824944774e-06,
+      "loss": 0.6082,
+      "step": 690
+    },
+    {
+      "epoch": 0.37821565407772306,
+      "grad_norm": 0.004297096747905016,
+      "learning_rate": 6.8666538920286965e-06,
+      "loss": 0.0002,
+      "step": 691
+    },
+    {
+      "epoch": 0.3787629994526546,
+      "grad_norm": 4.101519584655762,
+      "learning_rate": 6.858675082233135e-06,
+      "loss": 0.2531,
+      "step": 692
+    },
+    {
+      "epoch": 0.3793103448275862,
+      "grad_norm": 2.1500627994537354,
+      "learning_rate": 6.850690776699574e-06,
+      "loss": 0.3575,
+      "step": 693
+    },
+    {
+      "epoch": 0.3798576902025178,
+      "grad_norm": 0.4142707586288452,
+      "learning_rate": 6.842700999036036e-06,
+      "loss": 0.0254,
+      "step": 694
+    },
+    {
+      "epoch": 0.38040503557744937,
+      "grad_norm": 0.1441921442747116,
+      "learning_rate": 6.834705772866732e-06,
+      "loss": 0.0101,
+      "step": 695
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 0.016346001997590065,
+      "learning_rate": 6.8267051218319766e-06,
+      "loss": 0.0006,
+      "step": 696
+    },
+    {
+      "epoch": 0.38149972632731255,
+      "grad_norm": 0.017199577763676643,
+      "learning_rate": 6.8186990695881275e-06,
+      "loss": 0.0008,
+      "step": 697
+    },
+    {
+      "epoch": 0.3820470717022441,
+      "grad_norm": 0.009311121888458729,
+      "learning_rate": 6.810687639807514e-06,
+      "loss": 0.0004,
+      "step": 698
+    },
+    {
+      "epoch": 0.3825944170771757,
+      "grad_norm": 1.2476736307144165,
+      "learning_rate": 6.802670856178362e-06,
+      "loss": 0.1364,
+      "step": 699
+    },
+    {
+      "epoch": 0.3831417624521073,
+      "grad_norm": 1.8799997568130493,
+      "learning_rate": 6.79464874240473e-06,
+      "loss": 0.3712,
+      "step": 700
+    },
+    {
+      "epoch": 0.38368910782703886,
+      "grad_norm": 1.3058347702026367,
+      "learning_rate": 6.7866213222064385e-06,
+      "loss": 0.2762,
+      "step": 701
+    },
+    {
+      "epoch": 0.3842364532019704,
+      "grad_norm": 0.01594836823642254,
+      "learning_rate": 6.7785886193189936e-06,
+      "loss": 0.0007,
+      "step": 702
+    },
+    {
+      "epoch": 0.38478379857690204,
+      "grad_norm": 0.012764069251716137,
+      "learning_rate": 6.770550657493525e-06,
+      "loss": 0.0005,
+      "step": 703
+    },
+    {
+      "epoch": 0.3853311439518336,
+      "grad_norm": 1.4421093463897705,
+      "learning_rate": 6.76250746049671e-06,
+      "loss": 0.0832,
+      "step": 704
+    },
+    {
+      "epoch": 0.38587848932676516,
+      "grad_norm": 0.002905939007177949,
+      "learning_rate": 6.754459052110707e-06,
+      "loss": 0.0001,
+      "step": 705
+    },
+    {
+      "epoch": 0.3864258347016968,
+      "grad_norm": 2.3534743785858154,
+      "learning_rate": 6.7464054561330805e-06,
+      "loss": 0.4796,
+      "step": 706
+    },
+    {
+      "epoch": 0.38697318007662834,
+      "grad_norm": 0.04106984660029411,
+      "learning_rate": 6.7383466963767386e-06,
+      "loss": 0.0018,
+      "step": 707
+    },
+    {
+      "epoch": 0.38752052545155996,
+      "grad_norm": 0.007894037291407585,
+      "learning_rate": 6.730282796669853e-06,
+      "loss": 0.0004,
+      "step": 708
+    },
+    {
+      "epoch": 0.3880678708264915,
+      "grad_norm": 0.36676910519599915,
+      "learning_rate": 6.722213780855795e-06,
+      "loss": 0.0271,
+      "step": 709
+    },
+    {
+      "epoch": 0.3886152162014231,
+      "grad_norm": 0.5855551958084106,
+      "learning_rate": 6.714139672793063e-06,
+      "loss": 0.0555,
+      "step": 710
+    },
+    {
+      "epoch": 0.3891625615763547,
+      "grad_norm": 0.7090444564819336,
+      "learning_rate": 6.7060604963552125e-06,
+      "loss": 0.0533,
+      "step": 711
+    },
+    {
+      "epoch": 0.38970990695128627,
+      "grad_norm": 0.22933818399906158,
+      "learning_rate": 6.697976275430786e-06,
+      "loss": 0.0136,
+      "step": 712
+    },
+    {
+      "epoch": 0.39025725232621783,
+      "grad_norm": 0.344751238822937,
+      "learning_rate": 6.6898870339232405e-06,
+      "loss": 0.0281,
+      "step": 713
+    },
+    {
+      "epoch": 0.39080459770114945,
+      "grad_norm": 0.16770361363887787,
+      "learning_rate": 6.681792795750876e-06,
+      "loss": 0.008,
+      "step": 714
+    },
+    {
+      "epoch": 0.391351943076081,
+      "grad_norm": 0.016307057812809944,
+      "learning_rate": 6.673693584846771e-06,
+      "loss": 0.0006,
+      "step": 715
+    },
+    {
+      "epoch": 0.3918992884510126,
+      "grad_norm": 0.3089595139026642,
+      "learning_rate": 6.665589425158705e-06,
+      "loss": 0.0221,
+      "step": 716
+    },
+    {
+      "epoch": 0.3924466338259442,
+      "grad_norm": 0.05962289124727249,
+      "learning_rate": 6.657480340649088e-06,
+      "loss": 0.0036,
+      "step": 717
+    },
+    {
+      "epoch": 0.39299397920087575,
+      "grad_norm": 0.13849852979183197,
+      "learning_rate": 6.649366355294895e-06,
+      "loss": 0.0091,
+      "step": 718
+    },
+    {
+      "epoch": 0.3935413245758073,
+      "grad_norm": 0.2673451602458954,
+      "learning_rate": 6.641247493087591e-06,
+      "loss": 0.0216,
+      "step": 719
+    },
+    {
+      "epoch": 0.39408866995073893,
+      "grad_norm": 0.02874746359884739,
+      "learning_rate": 6.633123778033061e-06,
+      "loss": 0.0012,
+      "step": 720
+    },
+    {
+      "epoch": 0.3946360153256705,
+      "grad_norm": 0.04847461357712746,
+      "learning_rate": 6.624995234151539e-06,
+      "loss": 0.0025,
+      "step": 721
+    },
+    {
+      "epoch": 0.39518336070060206,
+      "grad_norm": 4.155821800231934,
+      "learning_rate": 6.616861885477535e-06,
+      "loss": 0.2413,
+      "step": 722
+    },
+    {
+      "epoch": 0.3957307060755337,
+      "grad_norm": 0.025641655549407005,
+      "learning_rate": 6.608723756059768e-06,
+      "loss": 0.0011,
+      "step": 723
+    },
+    {
+      "epoch": 0.39627805145046524,
+      "grad_norm": 0.10781197994947433,
+      "learning_rate": 6.600580869961091e-06,
+      "loss": 0.0063,
+      "step": 724
+    },
+    {
+      "epoch": 0.3968253968253968,
+      "grad_norm": 0.01060369610786438,
+      "learning_rate": 6.592433251258423e-06,
+      "loss": 0.0005,
+      "step": 725
+    },
+    {
+      "epoch": 0.3973727422003284,
+      "grad_norm": 0.153262197971344,
+      "learning_rate": 6.5842809240426765e-06,
+      "loss": 0.0087,
+      "step": 726
+    },
+    {
+      "epoch": 0.39792008757526,
+      "grad_norm": 0.017993014305830002,
+      "learning_rate": 6.576123912418686e-06,
+      "loss": 0.0008,
+      "step": 727
+    },
+    {
+      "epoch": 0.39846743295019155,
+      "grad_norm": 0.828525722026825,
+      "learning_rate": 6.567962240505136e-06,
+      "loss": 0.0422,
+      "step": 728
+    },
+    {
+      "epoch": 0.39901477832512317,
+      "grad_norm": 2.484757900238037,
+      "learning_rate": 6.559795932434489e-06,
+      "loss": 0.4287,
+      "step": 729
+    },
+    {
+      "epoch": 0.3995621237000547,
+      "grad_norm": 0.007983885705471039,
+      "learning_rate": 6.551625012352921e-06,
+      "loss": 0.0003,
+      "step": 730
+    },
+    {
+      "epoch": 0.4001094690749863,
+      "grad_norm": 0.3982315957546234,
+      "learning_rate": 6.543449504420241e-06,
+      "loss": 0.0341,
+      "step": 731
+    },
+    {
+      "epoch": 0.4006568144499179,
+      "grad_norm": 0.08536089211702347,
+      "learning_rate": 6.535269432809821e-06,
+      "loss": 0.0056,
+      "step": 732
+    },
+    {
+      "epoch": 0.40120415982484947,
+      "grad_norm": 1.6352207660675049,
+      "learning_rate": 6.5270848217085325e-06,
+      "loss": 0.1665,
+      "step": 733
+    },
+    {
+      "epoch": 0.4017515051997811,
+      "grad_norm": 0.01377387810498476,
+      "learning_rate": 6.518895695316666e-06,
+      "loss": 0.0007,
+      "step": 734
+    },
+    {
+      "epoch": 0.40229885057471265,
+      "grad_norm": 0.30763447284698486,
+      "learning_rate": 6.510702077847864e-06,
+      "loss": 0.0218,
+      "step": 735
+    },
+    {
+      "epoch": 0.4028461959496442,
+      "grad_norm": 0.09024977684020996,
+      "learning_rate": 6.502503993529048e-06,
+      "loss": 0.0051,
+      "step": 736
+    },
+    {
+      "epoch": 0.40339354132457583,
+      "grad_norm": 0.08540990948677063,
+      "learning_rate": 6.494301466600345e-06,
+      "loss": 0.0054,
+      "step": 737
+    },
+    {
+      "epoch": 0.4039408866995074,
+      "grad_norm": 0.011326467618346214,
+      "learning_rate": 6.486094521315022e-06,
+      "loss": 0.0005,
+      "step": 738
+    },
+    {
+      "epoch": 0.40448823207443896,
+      "grad_norm": 0.0065598557703197,
+      "learning_rate": 6.477883181939406e-06,
+      "loss": 0.0003,
+      "step": 739
+    },
+    {
+      "epoch": 0.4050355774493706,
+      "grad_norm": 0.012220407836139202,
+      "learning_rate": 6.469667472752821e-06,
+      "loss": 0.0005,
+      "step": 740
+    },
+    {
+      "epoch": 0.40558292282430214,
+      "grad_norm": 0.004128745291382074,
+      "learning_rate": 6.461447418047506e-06,
+      "loss": 0.0002,
+      "step": 741
+    },
+    {
+      "epoch": 0.4061302681992337,
+      "grad_norm": 0.039554912596940994,
+      "learning_rate": 6.453223042128556e-06,
+      "loss": 0.0018,
+      "step": 742
+    },
+    {
+      "epoch": 0.4066776135741653,
+      "grad_norm": 0.009743032976984978,
+      "learning_rate": 6.444994369313835e-06,
+      "loss": 0.0004,
+      "step": 743
+    },
+    {
+      "epoch": 0.4072249589490969,
+      "grad_norm": 0.0073027294129133224,
+      "learning_rate": 6.4367614239339185e-06,
+      "loss": 0.0003,
+      "step": 744
+    },
+    {
+      "epoch": 0.40777230432402845,
+      "grad_norm": 0.01663777604699135,
+      "learning_rate": 6.428524230332012e-06,
+      "loss": 0.0007,
+      "step": 745
+    },
+    {
+      "epoch": 0.40831964969896006,
+      "grad_norm": 3.9854557514190674,
+      "learning_rate": 6.420282812863881e-06,
+      "loss": 0.2017,
+      "step": 746
+    },
+    {
+      "epoch": 0.4088669950738916,
+      "grad_norm": 0.20734180510044098,
+      "learning_rate": 6.412037195897786e-06,
+      "loss": 0.011,
+      "step": 747
+    },
+    {
+      "epoch": 0.4094143404488232,
+      "grad_norm": 0.004745072685182095,
+      "learning_rate": 6.403787403814399e-06,
+      "loss": 0.0003,
+      "step": 748
+    },
+    {
+      "epoch": 0.4099616858237548,
+      "grad_norm": 2.455504894256592,
+      "learning_rate": 6.395533461006736e-06,
+      "loss": 0.3553,
+      "step": 749
+    },
+    {
+      "epoch": 0.41050903119868637,
+      "grad_norm": 0.005224741529673338,
+      "learning_rate": 6.387275391880091e-06,
+      "loss": 0.0002,
+      "step": 750
+    },
+    {
+      "epoch": 0.41105637657361793,
+      "grad_norm": 0.002598387422040105,
+      "learning_rate": 6.379013220851956e-06,
+      "loss": 0.0001,
+      "step": 751
+    },
+    {
+      "epoch": 0.41160372194854955,
+      "grad_norm": 0.00500458711758256,
+      "learning_rate": 6.370746972351952e-06,
+      "loss": 0.0002,
+      "step": 752
+    },
+    {
+      "epoch": 0.4121510673234811,
+      "grad_norm": 0.055540215224027634,
+      "learning_rate": 6.362476670821755e-06,
+      "loss": 0.0032,
+      "step": 753
+    },
+    {
+      "epoch": 0.4126984126984127,
+      "grad_norm": 0.009147963486611843,
+      "learning_rate": 6.354202340715027e-06,
+      "loss": 0.0003,
+      "step": 754
+    },
+    {
+      "epoch": 0.4132457580733443,
+      "grad_norm": 0.008779282681643963,
+      "learning_rate": 6.345924006497339e-06,
+      "loss": 0.0004,
+      "step": 755
+    },
+    {
+      "epoch": 0.41379310344827586,
+      "grad_norm": 0.08578510582447052,
+      "learning_rate": 6.337641692646106e-06,
+      "loss": 0.005,
+      "step": 756
+    },
+    {
+      "epoch": 0.4143404488232074,
+      "grad_norm": 2.1286394596099854,
+      "learning_rate": 6.329355423650504e-06,
+      "loss": 0.4173,
+      "step": 757
+    },
+    {
+      "epoch": 0.41488779419813904,
+      "grad_norm": 0.005920462775975466,
+      "learning_rate": 6.321065224011408e-06,
+      "loss": 0.0003,
+      "step": 758
+    },
+    {
+      "epoch": 0.4154351395730706,
+      "grad_norm": 0.0021460726857185364,
+      "learning_rate": 6.312771118241314e-06,
+      "loss": 0.0001,
+      "step": 759
+    },
+    {
+      "epoch": 0.41598248494800216,
+      "grad_norm": 0.02204807847738266,
+      "learning_rate": 6.3044731308642685e-06,
+      "loss": 0.0012,
+      "step": 760
+    },
+    {
+      "epoch": 0.4165298303229338,
+      "grad_norm": 0.010705336928367615,
+      "learning_rate": 6.296171286415791e-06,
+      "loss": 0.0005,
+      "step": 761
+    },
+    {
+      "epoch": 0.41707717569786534,
+      "grad_norm": 0.004768910817801952,
+      "learning_rate": 6.287865609442812e-06,
+      "loss": 0.0002,
+      "step": 762
+    },
+    {
+      "epoch": 0.41762452107279696,
+      "grad_norm": 2.3559985160827637,
+      "learning_rate": 6.2795561245035895e-06,
+      "loss": 0.1317,
+      "step": 763
+    },
+    {
+      "epoch": 0.4181718664477285,
+      "grad_norm": 0.27280962467193604,
+      "learning_rate": 6.271242856167642e-06,
+      "loss": 0.013,
+      "step": 764
+    },
+    {
+      "epoch": 0.4187192118226601,
+      "grad_norm": 0.030251115560531616,
+      "learning_rate": 6.262925829015675e-06,
+      "loss": 0.0017,
+      "step": 765
+    },
+    {
+      "epoch": 0.4192665571975917,
+      "grad_norm": 0.08326046913862228,
+      "learning_rate": 6.254605067639509e-06,
+      "loss": 0.002,
+      "step": 766
+    },
+    {
+      "epoch": 0.41981390257252327,
+      "grad_norm": 0.0023234374821186066,
+      "learning_rate": 6.246280596642004e-06,
+      "loss": 0.0001,
+      "step": 767
+    },
+    {
+      "epoch": 0.42036124794745483,
+      "grad_norm": 0.01413232646882534,
+      "learning_rate": 6.23795244063699e-06,
+      "loss": 0.0005,
+      "step": 768
+    },
+    {
+      "epoch": 0.42090859332238645,
+      "grad_norm": 0.00254431227222085,
+      "learning_rate": 6.229620624249189e-06,
+      "loss": 0.0001,
+      "step": 769
+    },
+    {
+      "epoch": 0.421455938697318,
+      "grad_norm": 0.11062794923782349,
+      "learning_rate": 6.221285172114156e-06,
+      "loss": 0.0069,
+      "step": 770
+    },
+    {
+      "epoch": 0.4220032840722496,
+      "grad_norm": 0.07323046773672104,
+      "learning_rate": 6.212946108878185e-06,
+      "loss": 0.0045,
+      "step": 771
+    },
+    {
+      "epoch": 0.4225506294471812,
+      "grad_norm": 0.0021918388083577156,
+      "learning_rate": 6.204603459198252e-06,
+      "loss": 0.0001,
+      "step": 772
+    },
+    {
+      "epoch": 0.42309797482211275,
+      "grad_norm": 2.624161720275879,
+      "learning_rate": 6.196257247741939e-06,
+      "loss": 0.5911,
+      "step": 773
+    },
+    {
+      "epoch": 0.4236453201970443,
+      "grad_norm": 0.00811794027686119,
+      "learning_rate": 6.187907499187357e-06,
+      "loss": 0.0004,
+      "step": 774
+    },
+    {
+      "epoch": 0.42419266557197594,
+      "grad_norm": 0.041902635246515274,
+      "learning_rate": 6.179554238223076e-06,
+      "loss": 0.0018,
+      "step": 775
+    },
+    {
+      "epoch": 0.4247400109469075,
+      "grad_norm": 0.05677172169089317,
+      "learning_rate": 6.171197489548051e-06,
+      "loss": 0.0025,
+      "step": 776
+    },
+    {
+      "epoch": 0.42528735632183906,
+      "grad_norm": 0.007656686939299107,
+      "learning_rate": 6.162837277871553e-06,
+      "loss": 0.0003,
+      "step": 777
+    },
+    {
+      "epoch": 0.4258347016967707,
+      "grad_norm": 0.21811577677726746,
+      "learning_rate": 6.1544736279130865e-06,
+      "loss": 0.0146,
+      "step": 778
+    },
+    {
+      "epoch": 0.42638204707170224,
+      "grad_norm": 0.0371868871152401,
+      "learning_rate": 6.146106564402329e-06,
+      "loss": 0.002,
+      "step": 779
+    },
+    {
+      "epoch": 0.4269293924466338,
+      "grad_norm": 0.5981050729751587,
+      "learning_rate": 6.1377361120790445e-06,
+      "loss": 0.0398,
+      "step": 780
+    },
+    {
+      "epoch": 0.4274767378215654,
+      "grad_norm": 0.005229018162935972,
+      "learning_rate": 6.129362295693022e-06,
+      "loss": 0.0002,
+      "step": 781
+    },
+    {
+      "epoch": 0.428024083196497,
+      "grad_norm": 0.00975254736840725,
+      "learning_rate": 6.120985140003996e-06,
+      "loss": 0.0005,
+      "step": 782
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 2.9124345779418945,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.6284,
+      "step": 783
+    },
+    {
+      "epoch": 0.42911877394636017,
+      "grad_norm": 0.0032201369758695364,
+      "learning_rate": 6.104220909805162e-06,
+      "loss": 0.0002,
+      "step": 784
+    },
+    {
+      "epoch": 0.42966611932129173,
+      "grad_norm": 0.007221321575343609,
+      "learning_rate": 6.095833884863897e-06,
+      "loss": 0.0003,
+      "step": 785
+    },
+    {
+      "epoch": 0.4302134646962233,
+      "grad_norm": 0.06140093877911568,
+      "learning_rate": 6.08744361975657e-06,
+      "loss": 0.0029,
+      "step": 786
+    },
+    {
+      "epoch": 0.4307608100711549,
+      "grad_norm": 0.006075866986066103,
+      "learning_rate": 6.07905013929155e-06,
+      "loss": 0.0003,
+      "step": 787
+    },
+    {
+      "epoch": 0.43130815544608647,
+      "grad_norm": 0.10445165634155273,
+      "learning_rate": 6.0706534682867125e-06,
+      "loss": 0.0059,
+      "step": 788
+    },
+    {
+      "epoch": 0.4318555008210181,
+      "grad_norm": 0.09483800083398819,
+      "learning_rate": 6.062253631569368e-06,
+      "loss": 0.0064,
+      "step": 789
+    },
+    {
+      "epoch": 0.43240284619594965,
+      "grad_norm": 0.22675608098506927,
+      "learning_rate": 6.053850653976191e-06,
+      "loss": 0.0166,
+      "step": 790
+    },
+    {
+      "epoch": 0.4329501915708812,
+      "grad_norm": 0.08051242679357529,
+      "learning_rate": 6.045444560353136e-06,
+      "loss": 0.0055,
+      "step": 791
+    },
+    {
+      "epoch": 0.43349753694581283,
+      "grad_norm": 0.1378559172153473,
+      "learning_rate": 6.037035375555376e-06,
+      "loss": 0.0096,
+      "step": 792
+    },
+    {
+      "epoch": 0.4340448823207444,
+      "grad_norm": 0.004898442886769772,
+      "learning_rate": 6.028623124447224e-06,
+      "loss": 0.0002,
+      "step": 793
+    },
+    {
+      "epoch": 0.43459222769567596,
+      "grad_norm": 0.07252166420221329,
+      "learning_rate": 6.020207831902056e-06,
+      "loss": 0.0041,
+      "step": 794
+    },
+    {
+      "epoch": 0.4351395730706076,
+      "grad_norm": 0.011364106088876724,
+      "learning_rate": 6.011789522802242e-06,
+      "loss": 0.0005,
+      "step": 795
+    },
+    {
+      "epoch": 0.43568691844553914,
+      "grad_norm": 0.03275460749864578,
+      "learning_rate": 6.003368222039078e-06,
+      "loss": 0.0016,
+      "step": 796
+    },
+    {
+      "epoch": 0.4362342638204707,
+      "grad_norm": 0.6348550319671631,
+      "learning_rate": 5.994943954512694e-06,
+      "loss": 0.0458,
+      "step": 797
+    },
+    {
+      "epoch": 0.4367816091954023,
+      "grad_norm": 1.7934825420379639,
+      "learning_rate": 5.986516745132e-06,
+      "loss": 0.2173,
+      "step": 798
+    },
+    {
+      "epoch": 0.4373289545703339,
+      "grad_norm": 0.004570923279970884,
+      "learning_rate": 5.978086618814606e-06,
+      "loss": 0.0002,
+      "step": 799
+    },
+    {
+      "epoch": 0.43787629994526545,
+      "grad_norm": 0.013217059895396233,
+      "learning_rate": 5.96965360048674e-06,
+      "loss": 0.0006,
+      "step": 800
+    },
+    {
+      "epoch": 0.43842364532019706,
+      "grad_norm": 0.8358885049819946,
+      "learning_rate": 5.961217715083185e-06,
+      "loss": 0.0486,
+      "step": 801
+    },
+    {
+      "epoch": 0.4389709906951286,
+      "grad_norm": 0.08642356842756271,
+      "learning_rate": 5.952778987547203e-06,
+      "loss": 0.0054,
+      "step": 802
+    },
+    {
+      "epoch": 0.4395183360700602,
+      "grad_norm": 0.09967074543237686,
+      "learning_rate": 5.944337442830457e-06,
+      "loss": 0.0039,
+      "step": 803
+    },
+    {
+      "epoch": 0.4400656814449918,
+      "grad_norm": 2.0377254486083984,
+      "learning_rate": 5.935893105892938e-06,
+      "loss": 0.2651,
+      "step": 804
+    },
+    {
+      "epoch": 0.44061302681992337,
+      "grad_norm": 0.0057347621768713,
+      "learning_rate": 5.927446001702899e-06,
+      "loss": 0.0003,
+      "step": 805
+    },
+    {
+      "epoch": 0.44116037219485493,
+      "grad_norm": 0.003230379894375801,
+      "learning_rate": 5.918996155236771e-06,
+      "loss": 0.0002,
+      "step": 806
+    },
+    {
+      "epoch": 0.44170771756978655,
+      "grad_norm": 2.145993709564209,
+      "learning_rate": 5.9105435914790935e-06,
+      "loss": 0.4657,
+      "step": 807
+    },
+    {
+      "epoch": 0.4422550629447181,
+      "grad_norm": 0.038496311753988266,
+      "learning_rate": 5.902088335422442e-06,
+      "loss": 0.0021,
+      "step": 808
+    },
+    {
+      "epoch": 0.4428024083196497,
+      "grad_norm": 0.006777400616556406,
+      "learning_rate": 5.893630412067351e-06,
+      "loss": 0.0003,
+      "step": 809
+    },
+    {
+      "epoch": 0.4433497536945813,
+      "grad_norm": 0.04935964569449425,
+      "learning_rate": 5.885169846422242e-06,
+      "loss": 0.0018,
+      "step": 810
+    },
+    {
+      "epoch": 0.44389709906951286,
+      "grad_norm": 0.011123371310532093,
+      "learning_rate": 5.876706663503352e-06,
+      "loss": 0.0005,
+      "step": 811
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.03125881776213646,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.0015,
+      "step": 812
+    },
+    {
+      "epoch": 0.44499178981937604,
+      "grad_norm": 0.004140671342611313,
+      "learning_rate": 5.859772545947782e-06,
+      "loss": 0.0002,
+      "step": 813
+    },
+    {
+      "epoch": 0.4455391351943076,
+      "grad_norm": 0.002423380734398961,
+      "learning_rate": 5.85130166138197e-06,
+      "loss": 0.0001,
+      "step": 814
+    },
+    {
+      "epoch": 0.44608648056923916,
+      "grad_norm": 0.004158463794738054,
+      "learning_rate": 5.8428282596839625e-06,
+      "loss": 0.0002,
+      "step": 815
+    },
+    {
+      "epoch": 0.4466338259441708,
+      "grad_norm": 0.08258446305990219,
+      "learning_rate": 5.834352365907946e-06,
+      "loss": 0.003,
+      "step": 816
+    },
+    {
+      "epoch": 0.44718117131910234,
+      "grad_norm": 0.004286042880266905,
+      "learning_rate": 5.82587400511548e-06,
+      "loss": 0.0002,
+      "step": 817
+    },
+    {
+      "epoch": 0.44772851669403396,
+      "grad_norm": 0.006518741603940725,
+      "learning_rate": 5.817393202375416e-06,
+      "loss": 0.0003,
+      "step": 818
+    },
+    {
+      "epoch": 0.4482758620689655,
+      "grad_norm": 0.33282148838043213,
+      "learning_rate": 5.808909982763825e-06,
+      "loss": 0.0305,
+      "step": 819
+    },
+    {
+      "epoch": 0.4488232074438971,
+      "grad_norm": 0.04650212079286575,
+      "learning_rate": 5.800424371363924e-06,
+      "loss": 0.0022,
+      "step": 820
+    },
+    {
+      "epoch": 0.4493705528188287,
+      "grad_norm": 0.0025998507626354694,
+      "learning_rate": 5.791936393266004e-06,
+      "loss": 0.0002,
+      "step": 821
+    },
+    {
+      "epoch": 0.44991789819376027,
+      "grad_norm": 0.4078565239906311,
+      "learning_rate": 5.783446073567353e-06,
+      "loss": 0.0282,
+      "step": 822
+    },
+    {
+      "epoch": 0.45046524356869183,
+      "grad_norm": 0.03496750444173813,
+      "learning_rate": 5.774953437372181e-06,
+      "loss": 0.0011,
+      "step": 823
+    },
+    {
+      "epoch": 0.45101258894362345,
+      "grad_norm": 0.15228895843029022,
+      "learning_rate": 5.766458509791553e-06,
+      "loss": 0.0082,
+      "step": 824
+    },
+    {
+      "epoch": 0.451559934318555,
+      "grad_norm": 0.026087448000907898,
+      "learning_rate": 5.757961315943303e-06,
+      "loss": 0.0013,
+      "step": 825
+    },
+    {
+      "epoch": 0.4521072796934866,
+      "grad_norm": 0.1130935400724411,
+      "learning_rate": 5.749461880951966e-06,
+      "loss": 0.0067,
+      "step": 826
+    },
+    {
+      "epoch": 0.4526546250684182,
+      "grad_norm": 0.06168792396783829,
+      "learning_rate": 5.7409602299487085e-06,
+      "loss": 0.0035,
+      "step": 827
+    },
+    {
+      "epoch": 0.45320197044334976,
+      "grad_norm": 2.115023612976074,
+      "learning_rate": 5.732456388071247e-06,
+      "loss": 0.4833,
+      "step": 828
+    },
+    {
+      "epoch": 0.4537493158182813,
+      "grad_norm": 0.0042996820993721485,
+      "learning_rate": 5.723950380463774e-06,
+      "loss": 0.0002,
+      "step": 829
+    },
+    {
+      "epoch": 0.45429666119321294,
+      "grad_norm": 0.003995430190116167,
+      "learning_rate": 5.715442232276887e-06,
+      "loss": 0.0002,
+      "step": 830
+    },
+    {
+      "epoch": 0.4548440065681445,
+      "grad_norm": 0.006226207595318556,
+      "learning_rate": 5.706931968667514e-06,
+      "loss": 0.0003,
+      "step": 831
+    },
+    {
+      "epoch": 0.45539135194307606,
+      "grad_norm": 0.2677818238735199,
+      "learning_rate": 5.6984196147988365e-06,
+      "loss": 0.017,
+      "step": 832
+    },
+    {
+      "epoch": 0.4559386973180077,
+      "grad_norm": 0.025894341990351677,
+      "learning_rate": 5.689905195840216e-06,
+      "loss": 0.001,
+      "step": 833
+    },
+    {
+      "epoch": 0.45648604269293924,
+      "grad_norm": 0.11859169602394104,
+      "learning_rate": 5.681388736967124e-06,
+      "loss": 0.007,
+      "step": 834
+    },
+    {
+      "epoch": 0.4570333880678708,
+      "grad_norm": 0.002087466651573777,
+      "learning_rate": 5.672870263361057e-06,
+      "loss": 0.0001,
+      "step": 835
+    },
+    {
+      "epoch": 0.4575807334428024,
+      "grad_norm": 1.965535283088684,
+      "learning_rate": 5.6643498002094725e-06,
+      "loss": 0.1036,
+      "step": 836
+    },
+    {
+      "epoch": 0.458128078817734,
+      "grad_norm": 0.26401904225349426,
+      "learning_rate": 5.655827372705712e-06,
+      "loss": 0.0155,
+      "step": 837
+    },
+    {
+      "epoch": 0.45867542419266555,
+      "grad_norm": 1.1218135356903076,
+      "learning_rate": 5.647303006048924e-06,
+      "loss": 0.0511,
+      "step": 838
+    },
+    {
+      "epoch": 0.45922276956759717,
+      "grad_norm": 0.003350113518536091,
+      "learning_rate": 5.638776725443989e-06,
+      "loss": 0.0002,
+      "step": 839
+    },
+    {
+      "epoch": 0.45977011494252873,
+      "grad_norm": 0.15477736294269562,
+      "learning_rate": 5.630248556101448e-06,
+      "loss": 0.0096,
+      "step": 840
+    },
+    {
+      "epoch": 0.4603174603174603,
+      "grad_norm": 1.8222780227661133,
+      "learning_rate": 5.621718523237427e-06,
+      "loss": 0.1304,
+      "step": 841
+    },
+    {
+      "epoch": 0.4608648056923919,
+      "grad_norm": 0.47362422943115234,
+      "learning_rate": 5.613186652073561e-06,
+      "loss": 0.0269,
+      "step": 842
+    },
+    {
+      "epoch": 0.4614121510673235,
+      "grad_norm": 0.0013261528220027685,
+      "learning_rate": 5.604652967836922e-06,
+      "loss": 0.0001,
+      "step": 843
+    },
+    {
+      "epoch": 0.46195949644225504,
+      "grad_norm": 2.21683669090271,
+      "learning_rate": 5.596117495759943e-06,
+      "loss": 0.4404,
+      "step": 844
+    },
+    {
+      "epoch": 0.46250684181718665,
+      "grad_norm": 0.5288764834403992,
+      "learning_rate": 5.58758026108034e-06,
+      "loss": 0.0157,
+      "step": 845
+    },
+    {
+      "epoch": 0.4630541871921182,
+      "grad_norm": 0.0019743768498301506,
+      "learning_rate": 5.579041289041045e-06,
+      "loss": 0.0001,
+      "step": 846
+    },
+    {
+      "epoch": 0.46360153256704983,
+      "grad_norm": 0.005323327146470547,
+      "learning_rate": 5.570500604890124e-06,
+      "loss": 0.0002,
+      "step": 847
+    },
+    {
+      "epoch": 0.4641488779419814,
+      "grad_norm": 5.332734107971191,
+      "learning_rate": 5.561958233880707e-06,
+      "loss": 0.1358,
+      "step": 848
+    },
+    {
+      "epoch": 0.46469622331691296,
+      "grad_norm": 0.08172398060560226,
+      "learning_rate": 5.55341420127091e-06,
+      "loss": 0.0038,
+      "step": 849
+    },
+    {
+      "epoch": 0.4652435686918446,
+      "grad_norm": 0.009597660973668098,
+      "learning_rate": 5.544868532323766e-06,
+      "loss": 0.0004,
+      "step": 850
+    },
+    {
+      "epoch": 0.46579091406677614,
+      "grad_norm": 0.0016919386107474566,
+      "learning_rate": 5.536321252307141e-06,
+      "loss": 0.0001,
+      "step": 851
+    },
+    {
+      "epoch": 0.4663382594417077,
+      "grad_norm": 0.0031833848915994167,
+      "learning_rate": 5.527772386493667e-06,
+      "loss": 0.0002,
+      "step": 852
+    },
+    {
+      "epoch": 0.4668856048166393,
+      "grad_norm": 0.0018079314613714814,
+      "learning_rate": 5.519221960160666e-06,
+      "loss": 0.0001,
+      "step": 853
+    },
+    {
+      "epoch": 0.4674329501915709,
+      "grad_norm": 0.7406489253044128,
+      "learning_rate": 5.510669998590074e-06,
+      "loss": 0.0509,
+      "step": 854
+    },
+    {
+      "epoch": 0.46798029556650245,
+      "grad_norm": 0.05459873005747795,
+      "learning_rate": 5.502116527068363e-06,
+      "loss": 0.0019,
+      "step": 855
+    },
+    {
+      "epoch": 0.46852764094143406,
+      "grad_norm": 0.00270917359739542,
+      "learning_rate": 5.493561570886473e-06,
+      "loss": 0.0001,
+      "step": 856
+    },
+    {
+      "epoch": 0.4690749863163656,
+      "grad_norm": 4.040920734405518,
+      "learning_rate": 5.485005155339736e-06,
+      "loss": 0.1073,
+      "step": 857
+    },
+    {
+      "epoch": 0.4696223316912972,
+      "grad_norm": 0.7534052133560181,
+      "learning_rate": 5.4764473057277925e-06,
+      "loss": 0.0456,
+      "step": 858
+    },
+    {
+      "epoch": 0.4701696770662288,
+      "grad_norm": 0.030180972069501877,
+      "learning_rate": 5.467888047354528e-06,
+      "loss": 0.0013,
+      "step": 859
+    },
+    {
+      "epoch": 0.47071702244116037,
+      "grad_norm": 0.021435558795928955,
+      "learning_rate": 5.4593274055279935e-06,
+      "loss": 0.0008,
+      "step": 860
+    },
+    {
+      "epoch": 0.47126436781609193,
+      "grad_norm": 0.0014348177937790751,
+      "learning_rate": 5.450765405560328e-06,
+      "loss": 0.0001,
+      "step": 861
+    },
+    {
+      "epoch": 0.47181171319102355,
+      "grad_norm": 0.0019303731387481093,
+      "learning_rate": 5.442202072767686e-06,
+      "loss": 0.0001,
+      "step": 862
+    },
+    {
+      "epoch": 0.4723590585659551,
+      "grad_norm": 3.485708236694336,
+      "learning_rate": 5.433637432470169e-06,
+      "loss": 0.5884,
+      "step": 863
+    },
+    {
+      "epoch": 0.4729064039408867,
+      "grad_norm": 0.06138644367456436,
+      "learning_rate": 5.425071509991737e-06,
+      "loss": 0.0042,
+      "step": 864
+    },
+    {
+      "epoch": 0.4734537493158183,
+      "grad_norm": 0.009944554418325424,
+      "learning_rate": 5.4165043306601436e-06,
+      "loss": 0.0004,
+      "step": 865
+    },
+    {
+      "epoch": 0.47400109469074986,
+      "grad_norm": 0.0024830952752381563,
+      "learning_rate": 5.407935919806862e-06,
+      "loss": 0.0001,
+      "step": 866
+    },
+    {
+      "epoch": 0.4745484400656814,
+      "grad_norm": 0.001542185782454908,
+      "learning_rate": 5.399366302767003e-06,
+      "loss": 0.0001,
+      "step": 867
+    },
+    {
+      "epoch": 0.47509578544061304,
+      "grad_norm": 1.6952769756317139,
+      "learning_rate": 5.390795504879243e-06,
+      "loss": 0.1442,
+      "step": 868
+    },
+    {
+      "epoch": 0.4756431308155446,
+      "grad_norm": 0.005125279538333416,
+      "learning_rate": 5.382223551485754e-06,
+      "loss": 0.0003,
+      "step": 869
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 0.0007472229772247374,
+      "learning_rate": 5.373650467932122e-06,
+      "loss": 0.0001,
+      "step": 870
+    },
+    {
+      "epoch": 0.4767378215654078,
+      "grad_norm": 1.5565142631530762,
+      "learning_rate": 5.3650762795672755e-06,
+      "loss": 0.0455,
+      "step": 871
+    },
+    {
+      "epoch": 0.47728516694033934,
+      "grad_norm": 1.660844087600708,
+      "learning_rate": 5.356501011743408e-06,
+      "loss": 0.0606,
+      "step": 872
+    },
+    {
+      "epoch": 0.47783251231527096,
+      "grad_norm": 0.0034840109292417765,
+      "learning_rate": 5.347924689815906e-06,
+      "loss": 0.0002,
+      "step": 873
+    },
+    {
+      "epoch": 0.4783798576902025,
+      "grad_norm": 0.9910545945167542,
+      "learning_rate": 5.3393473391432745e-06,
+      "loss": 0.0963,
+      "step": 874
+    },
+    {
+      "epoch": 0.4789272030651341,
+      "grad_norm": 1.6280092000961304,
+      "learning_rate": 5.330768985087059e-06,
+      "loss": 0.1113,
+      "step": 875
+    },
+    {
+      "epoch": 0.4794745484400657,
+      "grad_norm": 0.340638130903244,
+      "learning_rate": 5.32218965301177e-06,
+      "loss": 0.0135,
+      "step": 876
+    },
+    {
+      "epoch": 0.48002189381499727,
+      "grad_norm": 0.04405590519309044,
+      "learning_rate": 5.313609368284813e-06,
+      "loss": 0.0019,
+      "step": 877
+    },
+    {
+      "epoch": 0.48056923918992883,
+      "grad_norm": 0.05662679672241211,
+      "learning_rate": 5.305028156276405e-06,
+      "loss": 0.0024,
+      "step": 878
+    },
+    {
+      "epoch": 0.48111658456486045,
+      "grad_norm": 0.001325729419477284,
+      "learning_rate": 5.296446042359512e-06,
+      "loss": 0.0001,
+      "step": 879
+    },
+    {
+      "epoch": 0.481663929939792,
+      "grad_norm": 1.2463085651397705,
+      "learning_rate": 5.2878630519097605e-06,
+      "loss": 0.0509,
+      "step": 880
+    },
+    {
+      "epoch": 0.4822112753147236,
+      "grad_norm": 0.0031781319994479418,
+      "learning_rate": 5.279279210305373e-06,
+      "loss": 0.0002,
+      "step": 881
+    },
+    {
+      "epoch": 0.4827586206896552,
+      "grad_norm": 0.024987971410155296,
+      "learning_rate": 5.270694542927089e-06,
+      "loss": 0.0012,
+      "step": 882
+    },
+    {
+      "epoch": 0.48330596606458676,
+      "grad_norm": 0.005412014201283455,
+      "learning_rate": 5.262109075158084e-06,
+      "loss": 0.0003,
+      "step": 883
+    },
+    {
+      "epoch": 0.4838533114395183,
+      "grad_norm": 0.05908210948109627,
+      "learning_rate": 5.2535228323839046e-06,
+      "loss": 0.0035,
+      "step": 884
+    },
+    {
+      "epoch": 0.48440065681444994,
+      "grad_norm": 0.04235808923840523,
+      "learning_rate": 5.2449358399923885e-06,
+      "loss": 0.0017,
+      "step": 885
+    },
+    {
+      "epoch": 0.4849480021893815,
+      "grad_norm": 0.00043109082616865635,
+      "learning_rate": 5.236348123373593e-06,
+      "loss": 0.0001,
+      "step": 886
+    },
+    {
+      "epoch": 0.48549534756431306,
+      "grad_norm": 0.5738411545753479,
+      "learning_rate": 5.227759707919707e-06,
+      "loss": 0.0308,
+      "step": 887
+    },
+    {
+      "epoch": 0.4860426929392447,
+      "grad_norm": 0.04854976013302803,
+      "learning_rate": 5.219170619024996e-06,
+      "loss": 0.0028,
+      "step": 888
+    },
+    {
+      "epoch": 0.48659003831417624,
+      "grad_norm": 0.0018312755273655057,
+      "learning_rate": 5.2105808820857126e-06,
+      "loss": 0.0001,
+      "step": 889
+    },
+    {
+      "epoch": 0.4871373836891078,
+      "grad_norm": 1.7619683742523193,
+      "learning_rate": 5.201990522500027e-06,
+      "loss": 0.4159,
+      "step": 890
+    },
+    {
+      "epoch": 0.4876847290640394,
+      "grad_norm": 0.019682124257087708,
+      "learning_rate": 5.193399565667945e-06,
+      "loss": 0.0008,
+      "step": 891
+    },
+    {
+      "epoch": 0.488232074438971,
+      "grad_norm": 0.0009164040675386786,
+      "learning_rate": 5.184808036991246e-06,
+      "loss": 0.0001,
+      "step": 892
+    },
+    {
+      "epoch": 0.48877941981390255,
+      "grad_norm": 0.0016248204046860337,
+      "learning_rate": 5.1762159618733954e-06,
+      "loss": 0.0001,
+      "step": 893
+    },
+    {
+      "epoch": 0.48932676518883417,
+      "grad_norm": 0.015387938357889652,
+      "learning_rate": 5.167623365719474e-06,
+      "loss": 0.0004,
+      "step": 894
+    },
+    {
+      "epoch": 0.48987411056376573,
+      "grad_norm": 0.0004146917490288615,
+      "learning_rate": 5.1590302739361096e-06,
+      "loss": 0.0,
+      "step": 895
+    },
+    {
+      "epoch": 0.4904214559386973,
+      "grad_norm": 0.0017720222240313888,
+      "learning_rate": 5.150436711931387e-06,
+      "loss": 0.0001,
+      "step": 896
+    },
+    {
+      "epoch": 0.4909688013136289,
+      "grad_norm": 0.0020141142886132,
+      "learning_rate": 5.1418427051147855e-06,
+      "loss": 0.0001,
+      "step": 897
+    },
+    {
+      "epoch": 0.4915161466885605,
+      "grad_norm": 0.0037431721575558186,
+      "learning_rate": 5.1332482788971005e-06,
+      "loss": 0.0002,
+      "step": 898
+    },
+    {
+      "epoch": 0.49206349206349204,
+      "grad_norm": 0.05117342248558998,
+      "learning_rate": 5.1246534586903655e-06,
+      "loss": 0.0015,
+      "step": 899
+    },
+    {
+      "epoch": 0.49261083743842365,
+      "grad_norm": 0.006282655987888575,
+      "learning_rate": 5.116058269907779e-06,
+      "loss": 0.0003,
+      "step": 900
+    },
+    {
+      "epoch": 0.4931581828133552,
+      "grad_norm": 0.001593268709257245,
+      "learning_rate": 5.107462737963631e-06,
+      "loss": 0.0001,
+      "step": 901
+    },
+    {
+      "epoch": 0.49370552818828684,
+      "grad_norm": 0.0004761675954796374,
+      "learning_rate": 5.098866888273224e-06,
+      "loss": 0.0001,
+      "step": 902
+    },
+    {
+      "epoch": 0.4942528735632184,
+      "grad_norm": 0.0016890882980078459,
+      "learning_rate": 5.090270746252803e-06,
+      "loss": 0.0001,
+      "step": 903
+    },
+    {
+      "epoch": 0.49480021893814996,
+      "grad_norm": 0.04701818525791168,
+      "learning_rate": 5.081674337319473e-06,
+      "loss": 0.0024,
+      "step": 904
+    },
+    {
+      "epoch": 0.4953475643130816,
+      "grad_norm": 1.8550165891647339,
+      "learning_rate": 5.073077686891132e-06,
+      "loss": 0.5051,
+      "step": 905
+    },
+    {
+      "epoch": 0.49589490968801314,
+      "grad_norm": 2.175128936767578,
+      "learning_rate": 5.0644808203863926e-06,
+      "loss": 0.448,
+      "step": 906
+    },
+    {
+      "epoch": 0.4964422550629447,
+      "grad_norm": 0.008438384160399437,
+      "learning_rate": 5.055883763224502e-06,
+      "loss": 0.0005,
+      "step": 907
+    },
+    {
+      "epoch": 0.4969896004378763,
+      "grad_norm": 0.0014406866393983364,
+      "learning_rate": 5.047286540825273e-06,
+      "loss": 0.0001,
+      "step": 908
+    },
+    {
+      "epoch": 0.4975369458128079,
+      "grad_norm": 0.02875029854476452,
+      "learning_rate": 5.038689178609011e-06,
+      "loss": 0.0009,
+      "step": 909
+    },
+    {
+      "epoch": 0.49808429118773945,
+      "grad_norm": 0.0011718091554939747,
+      "learning_rate": 5.030091701996428e-06,
+      "loss": 0.0001,
+      "step": 910
+    },
+    {
+      "epoch": 0.49863163656267107,
+      "grad_norm": 0.037694063037633896,
+      "learning_rate": 5.021494136408578e-06,
+      "loss": 0.0019,
+      "step": 911
+    },
+    {
+      "epoch": 0.49917898193760263,
+      "grad_norm": 0.0005312002613209188,
+      "learning_rate": 5.012896507266779e-06,
+      "loss": 0.0,
+      "step": 912
+    },
+    {
+      "epoch": 0.4997263273125342,
+      "grad_norm": 3.6381168365478516,
+      "learning_rate": 5.0042988399925365e-06,
+      "loss": 1.3302,
+      "step": 913
+    },
+    {
+      "epoch": 0.5002736726874658,
+      "grad_norm": 0.2435426563024521,
+      "learning_rate": 4.995701160007466e-06,
+      "loss": 0.0075,
+      "step": 914
+    },
+    {
+      "epoch": 0.5008210180623974,
+      "grad_norm": 0.001816119416616857,
+      "learning_rate": 4.987103492733221e-06,
+      "loss": 0.0001,
+      "step": 915
+    },
+    {
+      "epoch": 0.5013683634373289,
+      "grad_norm": 0.0019303924636915326,
+      "learning_rate": 4.9785058635914234e-06,
+      "loss": 0.0001,
+      "step": 916
+    },
+    {
+      "epoch": 0.5019157088122606,
+      "grad_norm": 0.16450561583042145,
+      "learning_rate": 4.9699082980035735e-06,
+      "loss": 0.0094,
+      "step": 917
+    },
+    {
+      "epoch": 0.5024630541871922,
+      "grad_norm": 0.005837967619299889,
+      "learning_rate": 4.96131082139099e-06,
+      "loss": 0.0003,
+      "step": 918
+    },
+    {
+      "epoch": 0.5030103995621237,
+      "grad_norm": 0.2567005157470703,
+      "learning_rate": 4.952713459174728e-06,
+      "loss": 0.01,
+      "step": 919
+    },
+    {
+      "epoch": 0.5035577449370553,
+      "grad_norm": 0.0013026647502556443,
+      "learning_rate": 4.944116236775499e-06,
+      "loss": 0.0001,
+      "step": 920
+    },
+    {
+      "epoch": 0.5041050903119869,
+      "grad_norm": 0.0013909480767324567,
+      "learning_rate": 4.935519179613607e-06,
+      "loss": 0.0001,
+      "step": 921
+    },
+    {
+      "epoch": 0.5046524356869184,
+      "grad_norm": 0.044553812593221664,
+      "learning_rate": 4.9269223131088685e-06,
+      "loss": 0.0019,
+      "step": 922
+    },
+    {
+      "epoch": 0.50519978106185,
+      "grad_norm": 0.034393515437841415,
+      "learning_rate": 4.9183256626805276e-06,
+      "loss": 0.0018,
+      "step": 923
+    },
+    {
+      "epoch": 0.5057471264367817,
+      "grad_norm": 0.04988914355635643,
+      "learning_rate": 4.909729253747197e-06,
+      "loss": 0.0024,
+      "step": 924
+    },
+    {
+      "epoch": 0.5062944718117132,
+      "grad_norm": 0.30242329835891724,
+      "learning_rate": 4.901133111726777e-06,
+      "loss": 0.0205,
+      "step": 925
+    },
+    {
+      "epoch": 0.5068418171866448,
+      "grad_norm": 2.0898563861846924,
+      "learning_rate": 4.892537262036371e-06,
+      "loss": 0.1189,
+      "step": 926
+    },
+    {
+      "epoch": 0.5073891625615764,
+      "grad_norm": 0.27229753136634827,
+      "learning_rate": 4.883941730092222e-06,
+      "loss": 0.0146,
+      "step": 927
+    },
+    {
+      "epoch": 0.5079365079365079,
+      "grad_norm": 0.033172935247421265,
+      "learning_rate": 4.875346541309637e-06,
+      "loss": 0.0017,
+      "step": 928
+    },
+    {
+      "epoch": 0.5084838533114395,
+      "grad_norm": 3.3485257625579834,
+      "learning_rate": 4.866751721102901e-06,
+      "loss": 0.7556,
+      "step": 929
+    },
+    {
+      "epoch": 0.5090311986863711,
+      "grad_norm": 0.032600533217191696,
+      "learning_rate": 4.858157294885215e-06,
+      "loss": 0.0015,
+      "step": 930
+    },
+    {
+      "epoch": 0.5095785440613027,
+      "grad_norm": 0.017700565978884697,
+      "learning_rate": 4.8495632880686155e-06,
+      "loss": 0.0009,
+      "step": 931
+    },
+    {
+      "epoch": 0.5101258894362343,
+      "grad_norm": 0.0006235113833099604,
+      "learning_rate": 4.840969726063892e-06,
+      "loss": 0.0001,
+      "step": 932
+    },
+    {
+      "epoch": 0.5106732348111659,
+      "grad_norm": 0.005358157679438591,
+      "learning_rate": 4.832376634280526e-06,
+      "loss": 0.0003,
+      "step": 933
+    },
+    {
+      "epoch": 0.5112205801860974,
+      "grad_norm": 0.0027646853122860193,
+      "learning_rate": 4.823784038126608e-06,
+      "loss": 0.0002,
+      "step": 934
+    },
+    {
+      "epoch": 0.511767925561029,
+      "grad_norm": 0.012461655773222446,
+      "learning_rate": 4.8151919630087565e-06,
+      "loss": 0.0005,
+      "step": 935
+    },
+    {
+      "epoch": 0.5123152709359606,
+      "grad_norm": 0.0028007435612380505,
+      "learning_rate": 4.806600434332056e-06,
+      "loss": 0.0001,
+      "step": 936
+    },
+    {
+      "epoch": 0.5128626163108921,
+      "grad_norm": 0.02188277803361416,
+      "learning_rate": 4.7980094774999765e-06,
+      "loss": 0.0009,
+      "step": 937
+    },
+    {
+      "epoch": 0.5134099616858238,
+      "grad_norm": 0.003332045627757907,
+      "learning_rate": 4.789419117914288e-06,
+      "loss": 0.0002,
+      "step": 938
+    },
+    {
+      "epoch": 0.5139573070607554,
+      "grad_norm": 0.0016168846050277352,
+      "learning_rate": 4.780829380975004e-06,
+      "loss": 0.0001,
+      "step": 939
+    },
+    {
+      "epoch": 0.5145046524356869,
+      "grad_norm": 0.0034641437232494354,
+      "learning_rate": 4.772240292080295e-06,
+      "loss": 0.0001,
+      "step": 940
+    },
+    {
+      "epoch": 0.5150519978106185,
+      "grad_norm": 0.043745800852775574,
+      "learning_rate": 4.76365187662641e-06,
+      "loss": 0.0026,
+      "step": 941
+    },
+    {
+      "epoch": 0.5155993431855501,
+      "grad_norm": 0.00854698196053505,
+      "learning_rate": 4.755064160007612e-06,
+      "loss": 0.0005,
+      "step": 942
+    },
+    {
+      "epoch": 0.5161466885604816,
+      "grad_norm": 0.0028649435844272375,
+      "learning_rate": 4.746477167616098e-06,
+      "loss": 0.0001,
+      "step": 943
+    },
+    {
+      "epoch": 0.5166940339354132,
+      "grad_norm": 2.511202812194824,
+      "learning_rate": 4.737890924841918e-06,
+      "loss": 0.4582,
+      "step": 944
+    },
+    {
+      "epoch": 0.5172413793103449,
+      "grad_norm": 0.02500765584409237,
+      "learning_rate": 4.729305457072913e-06,
+      "loss": 0.0009,
+      "step": 945
+    },
+    {
+      "epoch": 0.5177887246852764,
+      "grad_norm": 0.021063437685370445,
+      "learning_rate": 4.7207207896946275e-06,
+      "loss": 0.001,
+      "step": 946
+    },
+    {
+      "epoch": 0.518336070060208,
+      "grad_norm": 0.004225156735628843,
+      "learning_rate": 4.712136948090241e-06,
+      "loss": 0.0002,
+      "step": 947
+    },
+    {
+      "epoch": 0.5188834154351396,
+      "grad_norm": 0.004292478784918785,
+      "learning_rate": 4.70355395764049e-06,
+      "loss": 0.0002,
+      "step": 948
+    },
+    {
+      "epoch": 0.5194307608100711,
+      "grad_norm": 0.003189288079738617,
+      "learning_rate": 4.694971843723596e-06,
+      "loss": 0.0002,
+      "step": 949
+    },
+    {
+      "epoch": 0.5199781061850027,
+      "grad_norm": 0.19809499382972717,
+      "learning_rate": 4.68639063171519e-06,
+      "loss": 0.0099,
+      "step": 950
+    },
+    {
+      "epoch": 0.5205254515599343,
+      "grad_norm": 0.07779690623283386,
+      "learning_rate": 4.677810346988231e-06,
+      "loss": 0.0042,
+      "step": 951
+    },
+    {
+      "epoch": 0.5210727969348659,
+      "grad_norm": 0.0021239686757326126,
+      "learning_rate": 4.6692310149129425e-06,
+      "loss": 0.0001,
+      "step": 952
+    },
+    {
+      "epoch": 0.5216201423097975,
+      "grad_norm": 0.0036316164769232273,
+      "learning_rate": 4.660652660856726e-06,
+      "loss": 0.0002,
+      "step": 953
+    },
+    {
+      "epoch": 0.5221674876847291,
+      "grad_norm": 0.06568455696105957,
+      "learning_rate": 4.6520753101840945e-06,
+      "loss": 0.0039,
+      "step": 954
+    },
+    {
+      "epoch": 0.5227148330596606,
+      "grad_norm": 0.0028860997408628464,
+      "learning_rate": 4.643498988256595e-06,
+      "loss": 0.0001,
+      "step": 955
+    },
+    {
+      "epoch": 0.5232621784345922,
+      "grad_norm": 1.3472000360488892,
+      "learning_rate": 4.634923720432727e-06,
+      "loss": 0.0462,
+      "step": 956
+    },
+    {
+      "epoch": 0.5238095238095238,
+      "grad_norm": 0.023126907646656036,
+      "learning_rate": 4.626349532067879e-06,
+      "loss": 0.0012,
+      "step": 957
+    },
+    {
+      "epoch": 0.5243568691844553,
+      "grad_norm": 2.742182970046997,
+      "learning_rate": 4.617776448514248e-06,
+      "loss": 0.3063,
+      "step": 958
+    },
+    {
+      "epoch": 0.524904214559387,
+      "grad_norm": 0.01860973611474037,
+      "learning_rate": 4.609204495120759e-06,
+      "loss": 0.0009,
+      "step": 959
+    },
+    {
+      "epoch": 0.5254515599343186,
+      "grad_norm": 0.00370593904517591,
+      "learning_rate": 4.600633697232999e-06,
+      "loss": 0.0002,
+      "step": 960
+    },
+    {
+      "epoch": 0.5259989053092501,
+      "grad_norm": 0.002041914500296116,
+      "learning_rate": 4.59206408019314e-06,
+      "loss": 0.0001,
+      "step": 961
+    },
+    {
+      "epoch": 0.5265462506841817,
+      "grad_norm": 3.0877633094787598,
+      "learning_rate": 4.583495669339857e-06,
+      "loss": 0.2035,
+      "step": 962
+    },
+    {
+      "epoch": 0.5270935960591133,
+      "grad_norm": 2.1611199378967285,
+      "learning_rate": 4.574928490008264e-06,
+      "loss": 0.4401,
+      "step": 963
+    },
+    {
+      "epoch": 0.5276409414340448,
+      "grad_norm": 0.003639899892732501,
+      "learning_rate": 4.566362567529834e-06,
+      "loss": 0.0002,
+      "step": 964
+    },
+    {
+      "epoch": 0.5281882868089764,
+      "grad_norm": 0.005092285107821226,
+      "learning_rate": 4.557797927232315e-06,
+      "loss": 0.0002,
+      "step": 965
+    },
+    {
+      "epoch": 0.5287356321839081,
+      "grad_norm": 0.007063568569719791,
+      "learning_rate": 4.549234594439674e-06,
+      "loss": 0.0003,
+      "step": 966
+    },
+    {
+      "epoch": 0.5292829775588396,
+      "grad_norm": 0.029361741617321968,
+      "learning_rate": 4.54067259447201e-06,
+      "loss": 0.0014,
+      "step": 967
+    },
+    {
+      "epoch": 0.5298303229337712,
+      "grad_norm": 0.0071528819389641285,
+      "learning_rate": 4.532111952645474e-06,
+      "loss": 0.0003,
+      "step": 968
+    },
+    {
+      "epoch": 0.5303776683087028,
+      "grad_norm": 0.0060792481526732445,
+      "learning_rate": 4.523552694272208e-06,
+      "loss": 0.0003,
+      "step": 969
+    },
+    {
+      "epoch": 0.5309250136836344,
+      "grad_norm": 0.03650781884789467,
+      "learning_rate": 4.514994844660265e-06,
+      "loss": 0.0023,
+      "step": 970
+    },
+    {
+      "epoch": 0.5314723590585659,
+      "grad_norm": 0.018084436655044556,
+      "learning_rate": 4.506438429113528e-06,
+      "loss": 0.0009,
+      "step": 971
+    },
+    {
+      "epoch": 0.5320197044334976,
+      "grad_norm": 0.007830976508557796,
+      "learning_rate": 4.497883472931639e-06,
+      "loss": 0.0003,
+      "step": 972
+    },
+    {
+      "epoch": 0.5325670498084292,
+      "grad_norm": 2.4423887729644775,
+      "learning_rate": 4.489330001409929e-06,
+      "loss": 0.268,
+      "step": 973
+    },
+    {
+      "epoch": 0.5331143951833607,
+      "grad_norm": 0.015997666865587234,
+      "learning_rate": 4.480778039839336e-06,
+      "loss": 0.0007,
+      "step": 974
+    },
+    {
+      "epoch": 0.5336617405582923,
+      "grad_norm": 0.00490145618095994,
+      "learning_rate": 4.472227613506334e-06,
+      "loss": 0.0003,
+      "step": 975
+    },
+    {
+      "epoch": 0.5342090859332239,
+      "grad_norm": 0.03548009321093559,
+      "learning_rate": 4.4636787476928605e-06,
+      "loss": 0.0013,
+      "step": 976
+    },
+    {
+      "epoch": 0.5347564313081554,
+      "grad_norm": 0.0016715583624318242,
+      "learning_rate": 4.455131467676235e-06,
+      "loss": 0.0001,
+      "step": 977
+    },
+    {
+      "epoch": 0.535303776683087,
+      "grad_norm": 3.094775915145874,
+      "learning_rate": 4.446585798729091e-06,
+      "loss": 0.1449,
+      "step": 978
+    },
+    {
+      "epoch": 0.5358511220580187,
+      "grad_norm": 0.08493595570325851,
+      "learning_rate": 4.438041766119293e-06,
+      "loss": 0.0045,
+      "step": 979
+    },
+    {
+      "epoch": 0.5363984674329502,
+      "grad_norm": 0.6037715673446655,
+      "learning_rate": 4.429499395109877e-06,
+      "loss": 0.0485,
+      "step": 980
+    },
+    {
+      "epoch": 0.5369458128078818,
+      "grad_norm": 1.0072492361068726,
+      "learning_rate": 4.4209587109589565e-06,
+      "loss": 0.1076,
+      "step": 981
+    },
+    {
+      "epoch": 0.5374931581828134,
+      "grad_norm": 0.00744901318103075,
+      "learning_rate": 4.412419738919661e-06,
+      "loss": 0.0003,
+      "step": 982
+    },
+    {
+      "epoch": 0.5380405035577449,
+      "grad_norm": 0.002214734675362706,
+      "learning_rate": 4.40388250424006e-06,
+      "loss": 0.0001,
+      "step": 983
+    },
+    {
+      "epoch": 0.5385878489326765,
+      "grad_norm": 0.004368369467556477,
+      "learning_rate": 4.395347032163079e-06,
+      "loss": 0.0003,
+      "step": 984
+    },
+    {
+      "epoch": 0.5391351943076081,
+      "grad_norm": 2.0607409477233887,
+      "learning_rate": 4.38681334792644e-06,
+      "loss": 0.394,
+      "step": 985
+    },
+    {
+      "epoch": 0.5396825396825397,
+      "grad_norm": 5.10407829284668,
+      "learning_rate": 4.3782814767625755e-06,
+      "loss": 0.2882,
+      "step": 986
+    },
+    {
+      "epoch": 0.5402298850574713,
+      "grad_norm": 0.00618614861741662,
+      "learning_rate": 4.369751443898554e-06,
+      "loss": 0.0003,
+      "step": 987
+    },
+    {
+      "epoch": 0.5407772304324029,
+      "grad_norm": 3.2158572673797607,
+      "learning_rate": 4.361223274556012e-06,
+      "loss": 0.6837,
+      "step": 988
+    },
+    {
+      "epoch": 0.5413245758073344,
+      "grad_norm": 0.01705230213701725,
+      "learning_rate": 4.3526969939510785e-06,
+      "loss": 0.0007,
+      "step": 989
+    },
+    {
+      "epoch": 0.541871921182266,
+      "grad_norm": 0.0015571071999147534,
+      "learning_rate": 4.3441726272942895e-06,
+      "loss": 0.0001,
+      "step": 990
+    },
+    {
+      "epoch": 0.5424192665571976,
+      "grad_norm": 0.010929671116173267,
+      "learning_rate": 4.335650199790528e-06,
+      "loss": 0.0004,
+      "step": 991
+    },
+    {
+      "epoch": 0.5429666119321291,
+      "grad_norm": 2.3123574256896973,
+      "learning_rate": 4.327129736638946e-06,
+      "loss": 0.1457,
+      "step": 992
+    },
+    {
+      "epoch": 0.5435139573070608,
+      "grad_norm": 2.3527815341949463,
+      "learning_rate": 4.318611263032878e-06,
+      "loss": 0.4302,
+      "step": 993
+    },
+    {
+      "epoch": 0.5440613026819924,
+      "grad_norm": 0.42728134989738464,
+      "learning_rate": 4.310094804159784e-06,
+      "loss": 0.0214,
+      "step": 994
+    },
+    {
+      "epoch": 0.5446086480569239,
+      "grad_norm": 0.01541733555495739,
+      "learning_rate": 4.301580385201166e-06,
+      "loss": 0.0008,
+      "step": 995
+    },
+    {
+      "epoch": 0.5451559934318555,
+      "grad_norm": 0.02181682363152504,
+      "learning_rate": 4.293068031332488e-06,
+      "loss": 0.0011,
+      "step": 996
+    },
+    {
+      "epoch": 0.5457033388067871,
+      "grad_norm": 0.0581437423825264,
+      "learning_rate": 4.284557767723114e-06,
+      "loss": 0.0028,
+      "step": 997
+    },
+    {
+      "epoch": 0.5462506841817186,
+      "grad_norm": 0.013395448215305805,
+      "learning_rate": 4.2760496195362285e-06,
+      "loss": 0.0006,
+      "step": 998
+    },
+    {
+      "epoch": 0.5467980295566502,
+      "grad_norm": 1.165745735168457,
+      "learning_rate": 4.267543611928755e-06,
+      "loss": 0.0997,
+      "step": 999
+    },
+    {
+      "epoch": 0.5473453749315819,
+      "grad_norm": 0.03456006571650505,
+      "learning_rate": 4.259039770051292e-06,
+      "loss": 0.0021,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5478927203065134,
+      "grad_norm": 0.2218150794506073,
+      "learning_rate": 4.250538119048036e-06,
+      "loss": 0.018,
+      "step": 1001
+    },
+    {
+      "epoch": 0.548440065681445,
+      "grad_norm": 0.00396326370537281,
+      "learning_rate": 4.2420386840567e-06,
+      "loss": 0.0002,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5489874110563766,
+      "grad_norm": 0.19975824654102325,
+      "learning_rate": 4.233541490208448e-06,
+      "loss": 0.0134,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5495347564313081,
+      "grad_norm": 0.012520854361355305,
+      "learning_rate": 4.22504656262782e-06,
+      "loss": 0.0004,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5500821018062397,
+      "grad_norm": 2.170820951461792,
+      "learning_rate": 4.2165539264326495e-06,
+      "loss": 0.5925,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5506294471811713,
+      "grad_norm": 0.004577749874442816,
+      "learning_rate": 4.208063606733999e-06,
+      "loss": 0.0002,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5511767925561029,
+      "grad_norm": 0.103189617395401,
+      "learning_rate": 4.199575628636078e-06,
+      "loss": 0.006,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 0.004631855525076389,
+      "learning_rate": 4.191090017236177e-06,
+      "loss": 0.0002,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5522714833059661,
+      "grad_norm": 1.1117832660675049,
+      "learning_rate": 4.182606797624585e-06,
+      "loss": 0.1485,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5528188286808976,
+      "grad_norm": 0.0021044062450528145,
+      "learning_rate": 4.1741259948845206e-06,
+      "loss": 0.0001,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5533661740558292,
+      "grad_norm": 1.9612685441970825,
+      "learning_rate": 4.165647634092055e-06,
+      "loss": 0.3809,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5539135194307608,
+      "grad_norm": 0.0078058550134301186,
+      "learning_rate": 4.157171740316039e-06,
+      "loss": 0.0003,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5544608648056923,
+      "grad_norm": 0.09045789390802383,
+      "learning_rate": 4.148698338618031e-06,
+      "loss": 0.0039,
+      "step": 1013
+    },
+    {
+      "epoch": 0.555008210180624,
+      "grad_norm": 0.003742016153410077,
+      "learning_rate": 4.14022745405222e-06,
+      "loss": 0.0002,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 0.004482665564864874,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.0002,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5561029009304871,
+      "grad_norm": 1.2101866006851196,
+      "learning_rate": 4.123293336496651e-06,
+      "loss": 0.0996,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5566502463054187,
+      "grad_norm": 0.001449257368221879,
+      "learning_rate": 4.114830153577759e-06,
+      "loss": 0.0001,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5571975916803503,
+      "grad_norm": 0.12253908812999725,
+      "learning_rate": 4.10636958793265e-06,
+      "loss": 0.0076,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5577449370552818,
+      "grad_norm": 0.010266436263918877,
+      "learning_rate": 4.0979116645775606e-06,
+      "loss": 0.0005,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5582922824302134,
+      "grad_norm": 0.06671036034822464,
+      "learning_rate": 4.089456408520908e-06,
+      "loss": 0.0034,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5588396278051451,
+      "grad_norm": 3.1648943424224854,
+      "learning_rate": 4.0810038447632296e-06,
+      "loss": 0.5862,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5593869731800766,
+      "grad_norm": 0.08388476818799973,
+      "learning_rate": 4.072553998297103e-06,
+      "loss": 0.0056,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5599343185550082,
+      "grad_norm": 0.07822927832603455,
+      "learning_rate": 4.064106894107064e-06,
+      "loss": 0.0041,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5604816639299398,
+      "grad_norm": 0.06111420318484306,
+      "learning_rate": 4.055662557169545e-06,
+      "loss": 0.0031,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5610290093048714,
+      "grad_norm": 0.21099752187728882,
+      "learning_rate": 4.047221012452798e-06,
+      "loss": 0.0146,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5615763546798029,
+      "grad_norm": 0.0014589588390663266,
+      "learning_rate": 4.0387822849168165e-06,
+      "loss": 0.0001,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5621237000547346,
+      "grad_norm": 0.00508915726095438,
+      "learning_rate": 4.030346399513261e-06,
+      "loss": 0.0002,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5626710454296662,
+      "grad_norm": 0.0018473148811608553,
+      "learning_rate": 4.021913381185394e-06,
+      "loss": 0.0001,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5632183908045977,
+      "grad_norm": 0.12212100625038147,
+      "learning_rate": 4.013483254868001e-06,
+      "loss": 0.0089,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5637657361795293,
+      "grad_norm": 3.567409038543701,
+      "learning_rate": 4.005056045487307e-06,
+      "loss": 0.2606,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5643130815544609,
+      "grad_norm": 0.11344505101442337,
+      "learning_rate": 3.996631777960923e-06,
+      "loss": 0.0075,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5648604269293924,
+      "grad_norm": 0.007242574356496334,
+      "learning_rate": 3.9882104771977585e-06,
+      "loss": 0.0004,
+      "step": 1032
+    },
+    {
+      "epoch": 0.565407772304324,
+      "grad_norm": 0.20782946050167084,
+      "learning_rate": 3.979792168097946e-06,
+      "loss": 0.0083,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5659551176792557,
+      "grad_norm": 0.049997005611658096,
+      "learning_rate": 3.971376875552777e-06,
+      "loss": 0.0024,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5665024630541872,
+      "grad_norm": 0.010352439247071743,
+      "learning_rate": 3.962964624444625e-06,
+      "loss": 0.0005,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5670498084291188,
+      "grad_norm": 0.12570813298225403,
+      "learning_rate": 3.9545554396468655e-06,
+      "loss": 0.0063,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5675971538040504,
+      "grad_norm": 0.005166537594050169,
+      "learning_rate": 3.946149346023811e-06,
+      "loss": 0.0002,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5681444991789819,
+      "grad_norm": 0.004181237425655127,
+      "learning_rate": 3.937746368430633e-06,
+      "loss": 0.0002,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5686918445539135,
+      "grad_norm": 0.14119568467140198,
+      "learning_rate": 3.929346531713289e-06,
+      "loss": 0.0089,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5692391899288451,
+      "grad_norm": 0.03071342408657074,
+      "learning_rate": 3.920949860708452e-06,
+      "loss": 0.0015,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5697865353037767,
+      "grad_norm": 0.003992474637925625,
+      "learning_rate": 3.912556380243431e-06,
+      "loss": 0.0002,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5703338806787083,
+      "grad_norm": 0.011039703153073788,
+      "learning_rate": 3.9041661151361045e-06,
+      "loss": 0.0006,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5708812260536399,
+      "grad_norm": 0.01362689770758152,
+      "learning_rate": 3.89577909019484e-06,
+      "loss": 0.0006,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.016619103029370308,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.0007,
+      "step": 1044
+    },
+    {
+      "epoch": 0.571975916803503,
+      "grad_norm": 0.0034893490374088287,
+      "learning_rate": 3.879014859996006e-06,
+      "loss": 0.0002,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5725232621784346,
+      "grad_norm": 2.8486711978912354,
+      "learning_rate": 3.8706377043069785e-06,
+      "loss": 0.1652,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5730706075533661,
+      "grad_norm": 0.027762796729803085,
+      "learning_rate": 3.862263887920957e-06,
+      "loss": 0.0013,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5736179529282978,
+      "grad_norm": 0.09105691313743591,
+      "learning_rate": 3.853893435597673e-06,
+      "loss": 0.0061,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5741652983032294,
+      "grad_norm": 0.00692678801715374,
+      "learning_rate": 3.8455263720869134e-06,
+      "loss": 0.0003,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5747126436781609,
+      "grad_norm": 1.1762075424194336,
+      "learning_rate": 3.8371627221284495e-06,
+      "loss": 0.0687,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5752599890530925,
+      "grad_norm": 0.14164984226226807,
+      "learning_rate": 3.82880251045195e-06,
+      "loss": 0.0088,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5758073344280241,
+      "grad_norm": 0.00480201980099082,
+      "learning_rate": 3.820445761776925e-06,
+      "loss": 0.0002,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5763546798029556,
+      "grad_norm": 0.3967006802558899,
+      "learning_rate": 3.8120925008126457e-06,
+      "loss": 0.0231,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5769020251778872,
+      "grad_norm": 0.01882929727435112,
+      "learning_rate": 3.8037427522580627e-06,
+      "loss": 0.0009,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5774493705528189,
+      "grad_norm": 0.30294129252433777,
+      "learning_rate": 3.7953965408017483e-06,
+      "loss": 0.0205,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5779967159277504,
+      "grad_norm": 0.07470423728227615,
+      "learning_rate": 3.7870538911218176e-06,
+      "loss": 0.0045,
+      "step": 1056
+    },
+    {
+      "epoch": 0.578544061302682,
+      "grad_norm": 1.4331607818603516,
+      "learning_rate": 3.7787148278858453e-06,
+      "loss": 0.1268,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5790914066776136,
+      "grad_norm": 0.0020402672234922647,
+      "learning_rate": 3.77037937575081e-06,
+      "loss": 0.0001,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5796387520525451,
+      "grad_norm": 1.464131236076355,
+      "learning_rate": 3.762047559363013e-06,
+      "loss": 0.0974,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5801860974274767,
+      "grad_norm": 0.7270787358283997,
+      "learning_rate": 3.753719403357997e-06,
+      "loss": 0.0442,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5807334428024084,
+      "grad_norm": 0.006362659856677055,
+      "learning_rate": 3.745394932360491e-06,
+      "loss": 0.0003,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5812807881773399,
+      "grad_norm": 0.0032875300385057926,
+      "learning_rate": 3.7370741709843263e-06,
+      "loss": 0.0002,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5818281335522715,
+      "grad_norm": 0.0008078487007878721,
+      "learning_rate": 3.728757143832359e-06,
+      "loss": 0.0001,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5823754789272031,
+      "grad_norm": 0.03340085595846176,
+      "learning_rate": 3.7204438754964113e-06,
+      "loss": 0.0015,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5829228243021346,
+      "grad_norm": 3.268207550048828,
+      "learning_rate": 3.7121343905571897e-06,
+      "loss": 0.5775,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5834701696770662,
+      "grad_norm": 0.003012469271197915,
+      "learning_rate": 3.70382871358421e-06,
+      "loss": 0.0002,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5840175150519978,
+      "grad_norm": 0.0514325350522995,
+      "learning_rate": 3.695526869135733e-06,
+      "loss": 0.0029,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5845648604269293,
+      "grad_norm": 0.0025664675049483776,
+      "learning_rate": 3.6872288817586883e-06,
+      "loss": 0.0001,
+      "step": 1068
+    },
+    {
+      "epoch": 0.585112205801861,
+      "grad_norm": 0.051885541528463364,
+      "learning_rate": 3.678934775988594e-06,
+      "loss": 0.0031,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5856595511767926,
+      "grad_norm": 0.11281336098909378,
+      "learning_rate": 3.6706445763494976e-06,
+      "loss": 0.0066,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5862068965517241,
+      "grad_norm": 0.03128824383020401,
+      "learning_rate": 3.662358307353897e-06,
+      "loss": 0.0016,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5867542419266557,
+      "grad_norm": 0.04773644730448723,
+      "learning_rate": 3.6540759935026627e-06,
+      "loss": 0.0028,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5873015873015873,
+      "grad_norm": 0.023716391995549202,
+      "learning_rate": 3.6457976592849753e-06,
+      "loss": 0.0012,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5878489326765188,
+      "grad_norm": 0.10552944242954254,
+      "learning_rate": 3.637523329178247e-06,
+      "loss": 0.0061,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5883962780514504,
+      "grad_norm": 0.005593111272901297,
+      "learning_rate": 3.6292530276480493e-06,
+      "loss": 0.0003,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5889436234263821,
+      "grad_norm": 0.054677512496709824,
+      "learning_rate": 3.6209867791480446e-06,
+      "loss": 0.0035,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5894909688013136,
+      "grad_norm": 0.1693568378686905,
+      "learning_rate": 3.6127246081199107e-06,
+      "loss": 0.0135,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5900383141762452,
+      "grad_norm": 0.04031054675579071,
+      "learning_rate": 3.6044665389932663e-06,
+      "loss": 0.0022,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5905856595511768,
+      "grad_norm": 0.0014612111262977123,
+      "learning_rate": 3.596212596185603e-06,
+      "loss": 0.0001,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5911330049261084,
+      "grad_norm": 3.7056195735931396,
+      "learning_rate": 3.587962804102214e-06,
+      "loss": 0.7452,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5916803503010399,
+      "grad_norm": 0.028046006336808205,
+      "learning_rate": 3.5797171871361203e-06,
+      "loss": 0.0013,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5922276956759716,
+      "grad_norm": 0.16697245836257935,
+      "learning_rate": 3.57147576966799e-06,
+      "loss": 0.0103,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5927750410509032,
+      "grad_norm": 0.16656920313835144,
+      "learning_rate": 3.5632385760660828e-06,
+      "loss": 0.0086,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5933223864258347,
+      "grad_norm": 0.0012677456252276897,
+      "learning_rate": 3.5550056306861667e-06,
+      "loss": 0.0001,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5938697318007663,
+      "grad_norm": 0.023179659619927406,
+      "learning_rate": 3.5467769578714455e-06,
+      "loss": 0.0012,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5944170771756979,
+      "grad_norm": 0.007856342010200024,
+      "learning_rate": 3.5385525819524933e-06,
+      "loss": 0.0004,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5949644225506294,
+      "grad_norm": 0.0038795997388660908,
+      "learning_rate": 3.530332527247181e-06,
+      "loss": 0.0002,
+      "step": 1087
+    },
+    {
+      "epoch": 0.595511767925561,
+      "grad_norm": 0.057348210364580154,
+      "learning_rate": 3.5221168180605946e-06,
+      "loss": 0.0036,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5960591133004927,
+      "grad_norm": 0.0044167679734528065,
+      "learning_rate": 3.5139054786849787e-06,
+      "loss": 0.0002,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5966064586754242,
+      "grad_norm": 0.0067407819442451,
+      "learning_rate": 3.5056985333996566e-06,
+      "loss": 0.0003,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5971538040503558,
+      "grad_norm": 0.006905603222548962,
+      "learning_rate": 3.4974960064709534e-06,
+      "loss": 0.0003,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5977011494252874,
+      "grad_norm": 1.9404516220092773,
+      "learning_rate": 3.489297922152136e-06,
+      "loss": 0.2303,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5982484948002189,
+      "grad_norm": 0.05944973602890968,
+      "learning_rate": 3.4811043046833353e-06,
+      "loss": 0.0033,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5987958401751505,
+      "grad_norm": 0.1706078201532364,
+      "learning_rate": 3.4729151782914683e-06,
+      "loss": 0.0122,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5993431855500821,
+      "grad_norm": 0.0021984418854117393,
+      "learning_rate": 3.4647305671901797e-06,
+      "loss": 0.0001,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5998905309250137,
+      "grad_norm": 0.017191147431731224,
+      "learning_rate": 3.456550495579762e-06,
+      "loss": 0.0007,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6004378762999453,
+      "grad_norm": 2.794296979904175,
+      "learning_rate": 3.44837498764708e-06,
+      "loss": 0.4829,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6009852216748769,
+      "grad_norm": 0.3021494448184967,
+      "learning_rate": 3.440204067565511e-06,
+      "loss": 0.0199,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6015325670498084,
+      "grad_norm": 0.0023824695963412523,
+      "learning_rate": 3.432037759494867e-06,
+      "loss": 0.0002,
+      "step": 1099
+    },
+    {
+      "epoch": 0.60207991242474,
+      "grad_norm": 0.0029570909682661295,
+      "learning_rate": 3.4238760875813155e-06,
+      "loss": 0.0002,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6026272577996716,
+      "grad_norm": 2.326396942138672,
+      "learning_rate": 3.4157190759573243e-06,
+      "loss": 0.5833,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6031746031746031,
+      "grad_norm": 0.21300382912158966,
+      "learning_rate": 3.4075667487415785e-06,
+      "loss": 0.0169,
+      "step": 1102
+    },
+    {
+      "epoch": 0.6037219485495348,
+      "grad_norm": 0.054834965616464615,
+      "learning_rate": 3.3994191300389103e-06,
+      "loss": 0.0028,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6042692939244664,
+      "grad_norm": 0.0037344787269830704,
+      "learning_rate": 3.391276243940234e-06,
+      "loss": 0.0002,
+      "step": 1104
+    },
+    {
+      "epoch": 0.6048166392993979,
+      "grad_norm": 0.0019440932665020227,
+      "learning_rate": 3.3831381145224667e-06,
+      "loss": 0.0001,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6053639846743295,
+      "grad_norm": 0.0031255807261914015,
+      "learning_rate": 3.375004765848463e-06,
+      "loss": 0.0002,
+      "step": 1106
+    },
+    {
+      "epoch": 0.6059113300492611,
+      "grad_norm": 0.04441095516085625,
+      "learning_rate": 3.3668762219669393e-06,
+      "loss": 0.0024,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6064586754241926,
+      "grad_norm": 0.5900030732154846,
+      "learning_rate": 3.3587525069124093e-06,
+      "loss": 0.0258,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6070060207991242,
+      "grad_norm": 0.0020720800384879112,
+      "learning_rate": 3.350633644705107e-06,
+      "loss": 0.0001,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6075533661740559,
+      "grad_norm": 0.04347815364599228,
+      "learning_rate": 3.3425196593509135e-06,
+      "loss": 0.0021,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6081007115489874,
+      "grad_norm": 1.996151089668274,
+      "learning_rate": 3.334410574841298e-06,
+      "loss": 0.0879,
+      "step": 1111
+    },
+    {
+      "epoch": 0.608648056923919,
+      "grad_norm": 0.0028991817962378263,
+      "learning_rate": 3.3263064151532303e-06,
+      "loss": 0.0002,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6091954022988506,
+      "grad_norm": 0.14750750362873077,
+      "learning_rate": 3.3182072042491244e-06,
+      "loss": 0.009,
+      "step": 1113
+    },
+    {
+      "epoch": 0.6097427476737821,
+      "grad_norm": 0.17952144145965576,
+      "learning_rate": 3.310112966076762e-06,
+      "loss": 0.0127,
+      "step": 1114
+    },
+    {
+      "epoch": 0.6102900930487137,
+      "grad_norm": 0.0014247623039409518,
+      "learning_rate": 3.3020237245692154e-06,
+      "loss": 0.0001,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6108374384236454,
+      "grad_norm": 0.0048622991889715195,
+      "learning_rate": 3.293939503644788e-06,
+      "loss": 0.0003,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6113847837985769,
+      "grad_norm": 0.004406985826790333,
+      "learning_rate": 3.285860327206939e-06,
+      "loss": 0.0002,
+      "step": 1117
+    },
+    {
+      "epoch": 0.6119321291735085,
+      "grad_norm": 0.0034555234014987946,
+      "learning_rate": 3.277786219144207e-06,
+      "loss": 0.0002,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6124794745484401,
+      "grad_norm": 0.21532057225704193,
+      "learning_rate": 3.2697172033301485e-06,
+      "loss": 0.0077,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6130268199233716,
+      "grad_norm": 0.05346650257706642,
+      "learning_rate": 3.2616533036232635e-06,
+      "loss": 0.003,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6135741652983032,
+      "grad_norm": 0.001968423603102565,
+      "learning_rate": 3.2535945438669203e-06,
+      "loss": 0.0001,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6141215106732348,
+      "grad_norm": 0.0016493768198415637,
+      "learning_rate": 3.245540947889294e-06,
+      "loss": 0.0001,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6146688560481663,
+      "grad_norm": 0.06369902193546295,
+      "learning_rate": 3.2374925395032926e-06,
+      "loss": 0.0034,
+      "step": 1123
+    },
+    {
+      "epoch": 0.615216201423098,
+      "grad_norm": 0.0024321821983903646,
+      "learning_rate": 3.229449342506477e-06,
+      "loss": 0.0001,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6157635467980296,
+      "grad_norm": 0.014392325654625893,
+      "learning_rate": 3.2214113806810077e-06,
+      "loss": 0.0008,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6163108921729611,
+      "grad_norm": 0.0016964362002909184,
+      "learning_rate": 3.2133786777935645e-06,
+      "loss": 0.0001,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6168582375478927,
+      "grad_norm": 0.01708538644015789,
+      "learning_rate": 3.205351257595272e-06,
+      "loss": 0.0009,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6174055829228243,
+      "grad_norm": 0.0007502536755055189,
+      "learning_rate": 3.197329143821639e-06,
+      "loss": 0.0001,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6179529282977558,
+      "grad_norm": 2.2683279514312744,
+      "learning_rate": 3.189312360192489e-06,
+      "loss": 0.2333,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6185002736726875,
+      "grad_norm": 0.001995793776586652,
+      "learning_rate": 3.181300930411874e-06,
+      "loss": 0.0001,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6190476190476191,
+      "grad_norm": 0.02659580484032631,
+      "learning_rate": 3.173294878168025e-06,
+      "loss": 0.0017,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6195949644225506,
+      "grad_norm": 1.8708947896957397,
+      "learning_rate": 3.165294227133271e-06,
+      "loss": 0.1571,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6201423097974822,
+      "grad_norm": 0.0021453495137393475,
+      "learning_rate": 3.157299000963966e-06,
+      "loss": 0.0001,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6206896551724138,
+      "grad_norm": 0.011976787820458412,
+      "learning_rate": 3.149309223300428e-06,
+      "loss": 0.0006,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6212370005473454,
+      "grad_norm": 0.5774064064025879,
+      "learning_rate": 3.141324917766866e-06,
+      "loss": 0.0201,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6217843459222769,
+      "grad_norm": 0.023726046085357666,
+      "learning_rate": 3.1333461079713056e-06,
+      "loss": 0.0006,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6223316912972086,
+      "grad_norm": 0.23722946643829346,
+      "learning_rate": 3.1253728175055242e-06,
+      "loss": 0.0162,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6228790366721402,
+      "grad_norm": 0.0039423611015081406,
+      "learning_rate": 3.1174050699449776e-06,
+      "loss": 0.0002,
+      "step": 1138
+    },
+    {
+      "epoch": 0.6234263820470717,
+      "grad_norm": 0.012701333500444889,
+      "learning_rate": 3.109442888848736e-06,
+      "loss": 0.0006,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6239737274220033,
+      "grad_norm": 0.01759052835404873,
+      "learning_rate": 3.1014862977594083e-06,
+      "loss": 0.0007,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6245210727969349,
+      "grad_norm": 0.0015605682274326682,
+      "learning_rate": 3.093535320203074e-06,
+      "loss": 0.0001,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6250684181718664,
+      "grad_norm": 2.1218316555023193,
+      "learning_rate": 3.0855899796892188e-06,
+      "loss": 0.0915,
+      "step": 1142
+    },
+    {
+      "epoch": 0.625615763546798,
+      "grad_norm": 0.024543100968003273,
+      "learning_rate": 3.0776502997106526e-06,
+      "loss": 0.0011,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6261631089217297,
+      "grad_norm": 0.0035148372408002615,
+      "learning_rate": 3.0697163037434573e-06,
+      "loss": 0.0002,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6267104542966612,
+      "grad_norm": 0.00434495834633708,
+      "learning_rate": 3.061788015246905e-06,
+      "loss": 0.0002,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6272577996715928,
+      "grad_norm": 0.030992476269602776,
+      "learning_rate": 3.0538654576633865e-06,
+      "loss": 0.0015,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6278051450465244,
+      "grad_norm": 0.006613498087972403,
+      "learning_rate": 3.045948654418356e-06,
+      "loss": 0.0004,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6283524904214559,
+      "grad_norm": 0.0013742909068241715,
+      "learning_rate": 3.0380376289202497e-06,
+      "loss": 0.0001,
+      "step": 1148
+    },
+    {
+      "epoch": 0.6288998357963875,
+      "grad_norm": 0.006594480946660042,
+      "learning_rate": 3.0301324045604163e-06,
+      "loss": 0.0003,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6294471811713191,
+      "grad_norm": 0.0012503663310781121,
+      "learning_rate": 3.0222330047130572e-06,
+      "loss": 0.0001,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6299945265462507,
+      "grad_norm": 0.024310488253831863,
+      "learning_rate": 3.0143394527351522e-06,
+      "loss": 0.0014,
+      "step": 1151
+    },
+    {
+      "epoch": 0.6305418719211823,
+      "grad_norm": 0.001007006736472249,
+      "learning_rate": 3.0064517719663833e-06,
+      "loss": 0.0001,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6310892172961139,
+      "grad_norm": 2.5153512954711914,
+      "learning_rate": 2.9985699857290788e-06,
+      "loss": 0.5184,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6316365626710454,
+      "grad_norm": 0.002309981267899275,
+      "learning_rate": 2.990694117328139e-06,
+      "loss": 0.0001,
+      "step": 1154
+    },
+    {
+      "epoch": 0.632183908045977,
+      "grad_norm": 2.245021104812622,
+      "learning_rate": 2.982824190050958e-06,
+      "loss": 0.5133,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6327312534209086,
+      "grad_norm": 0.000964795530308038,
+      "learning_rate": 2.9749602271673717e-06,
+      "loss": 0.0001,
+      "step": 1156
+    },
+    {
+      "epoch": 0.6332785987958401,
+      "grad_norm": 2.1958580017089844,
+      "learning_rate": 2.967102251929579e-06,
+      "loss": 0.28,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6338259441707718,
+      "grad_norm": 0.7682985067367554,
+      "learning_rate": 2.959250287572069e-06,
+      "loss": 0.0413,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6343732895457034,
+      "grad_norm": 0.0010651570046320558,
+      "learning_rate": 2.9514043573115635e-06,
+      "loss": 0.0001,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6349206349206349,
+      "grad_norm": 0.04936596006155014,
+      "learning_rate": 2.9435644843469434e-06,
+      "loss": 0.0021,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6354679802955665,
+      "grad_norm": 0.0013516498729586601,
+      "learning_rate": 2.935730691859172e-06,
+      "loss": 0.0001,
+      "step": 1161
+    },
+    {
+      "epoch": 0.6360153256704981,
+      "grad_norm": 0.0008485048892907798,
+      "learning_rate": 2.927903003011241e-06,
+      "loss": 0.0001,
+      "step": 1162
+    },
+    {
+      "epoch": 0.6365626710454296,
+      "grad_norm": 0.0014783508377149701,
+      "learning_rate": 2.920081440948094e-06,
+      "loss": 0.0001,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6371100164203612,
+      "grad_norm": 3.501877546310425,
+      "learning_rate": 2.912266028796554e-06,
+      "loss": 0.4404,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6376573617952929,
+      "grad_norm": 2.515601634979248,
+      "learning_rate": 2.9044567896652666e-06,
+      "loss": 0.1112,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6382047071702244,
+      "grad_norm": 0.296254962682724,
+      "learning_rate": 2.8966537466446186e-06,
+      "loss": 0.0086,
+      "step": 1166
+    },
+    {
+      "epoch": 0.638752052545156,
+      "grad_norm": 0.001710059237666428,
+      "learning_rate": 2.888856922806682e-06,
+      "loss": 0.0001,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6392993979200876,
+      "grad_norm": 1.275455355644226,
+      "learning_rate": 2.881066341205133e-06,
+      "loss": 0.0864,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6398467432950191,
+      "grad_norm": 0.019870832562446594,
+      "learning_rate": 2.8732820248752016e-06,
+      "loss": 0.001,
+      "step": 1169
+    },
+    {
+      "epoch": 0.6403940886699507,
+      "grad_norm": 1.7314554452896118,
+      "learning_rate": 2.8655039968335774e-06,
+      "loss": 0.2768,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6409414340448824,
+      "grad_norm": 2.4131124019622803,
+      "learning_rate": 2.8577322800783717e-06,
+      "loss": 0.6951,
+      "step": 1171
+    },
+    {
+      "epoch": 0.6414887794198139,
+      "grad_norm": 0.004058561287820339,
+      "learning_rate": 2.849966897589026e-06,
+      "loss": 0.0002,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6420361247947455,
+      "grad_norm": 0.12226421386003494,
+      "learning_rate": 2.842207872326255e-06,
+      "loss": 0.0061,
+      "step": 1173
+    },
+    {
+      "epoch": 0.6425834701696771,
+      "grad_norm": 0.0881904885172844,
+      "learning_rate": 2.8344552272319727e-06,
+      "loss": 0.0048,
+      "step": 1174
+    },
+    {
+      "epoch": 0.6431308155446086,
+      "grad_norm": 0.9335182905197144,
+      "learning_rate": 2.826708985229238e-06,
+      "loss": 0.2188,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6436781609195402,
+      "grad_norm": 0.035821665078401566,
+      "learning_rate": 2.8189691692221627e-06,
+      "loss": 0.0014,
+      "step": 1176
+    },
+    {
+      "epoch": 0.6442255062944718,
+      "grad_norm": 1.6445341110229492,
+      "learning_rate": 2.811235802095873e-06,
+      "loss": 0.1193,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6447728516694033,
+      "grad_norm": 0.01749396324157715,
+      "learning_rate": 2.803508906716417e-06,
+      "loss": 0.0009,
+      "step": 1178
+    },
+    {
+      "epoch": 0.645320197044335,
+      "grad_norm": 1.3020265102386475,
+      "learning_rate": 2.7957885059307097e-06,
+      "loss": 0.0759,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6458675424192666,
+      "grad_norm": 0.004729298409074545,
+      "learning_rate": 2.7880746225664623e-06,
+      "loss": 0.0002,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6464148877941981,
+      "grad_norm": 0.27152588963508606,
+      "learning_rate": 2.780367279432123e-06,
+      "loss": 0.0131,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6469622331691297,
+      "grad_norm": 0.004669299814850092,
+      "learning_rate": 2.7726664993167864e-06,
+      "loss": 0.0002,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6475095785440613,
+      "grad_norm": 0.0026488250587135553,
+      "learning_rate": 2.7649723049901554e-06,
+      "loss": 0.0001,
+      "step": 1183
+    },
+    {
+      "epoch": 0.6480569239189928,
+      "grad_norm": 0.004148549400269985,
+      "learning_rate": 2.7572847192024544e-06,
+      "loss": 0.0003,
+      "step": 1184
+    },
+    {
+      "epoch": 0.6486042692939245,
+      "grad_norm": 0.08109666407108307,
+      "learning_rate": 2.749603764684367e-06,
+      "loss": 0.0042,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6491516146688561,
+      "grad_norm": 0.011724872514605522,
+      "learning_rate": 2.7419294641469718e-06,
+      "loss": 0.0006,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6496989600437876,
+      "grad_norm": 0.004057840444147587,
+      "learning_rate": 2.73426184028167e-06,
+      "loss": 0.0002,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6502463054187192,
+      "grad_norm": 0.2738402485847473,
+      "learning_rate": 2.7266009157601226e-06,
+      "loss": 0.0188,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6507936507936508,
+      "grad_norm": 0.0022387555800378323,
+      "learning_rate": 2.718946713234185e-06,
+      "loss": 0.0001,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6513409961685823,
+      "grad_norm": 1.4459933042526245,
+      "learning_rate": 2.711299255335833e-06,
+      "loss": 0.1288,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6518883415435139,
+      "grad_norm": 0.24324724078178406,
+      "learning_rate": 2.703658564677101e-06,
+      "loss": 0.0167,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6524356869184456,
+      "grad_norm": 0.0050564659759402275,
+      "learning_rate": 2.696024663850013e-06,
+      "loss": 0.0003,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6529830322933772,
+      "grad_norm": 0.014988393522799015,
+      "learning_rate": 2.688397575426517e-06,
+      "loss": 0.0007,
+      "step": 1193
+    },
+    {
+      "epoch": 0.6535303776683087,
+      "grad_norm": 0.13157279789447784,
+      "learning_rate": 2.680777321958424e-06,
+      "loss": 0.0071,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6540777230432403,
+      "grad_norm": 5.4224419593811035,
+      "learning_rate": 2.6731639259773235e-06,
+      "loss": 0.8259,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6546250684181719,
+      "grad_norm": 0.04328963905572891,
+      "learning_rate": 2.6655574099945403e-06,
+      "loss": 0.0024,
+      "step": 1196
+    },
+    {
+      "epoch": 0.6551724137931034,
+      "grad_norm": 0.005359618458896875,
+      "learning_rate": 2.65795779650105e-06,
+      "loss": 0.0003,
+      "step": 1197
+    },
+    {
+      "epoch": 0.655719759168035,
+      "grad_norm": 0.07889597117900848,
+      "learning_rate": 2.6503651079674207e-06,
+      "loss": 0.0045,
+      "step": 1198
+    },
+    {
+      "epoch": 0.6562671045429667,
+      "grad_norm": 0.09890451282262802,
+      "learning_rate": 2.642779366843743e-06,
+      "loss": 0.0067,
+      "step": 1199
+    },
+    {
+      "epoch": 0.6568144499178982,
+      "grad_norm": 0.0020398315973579884,
+      "learning_rate": 2.6352005955595715e-06,
+      "loss": 0.0001,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6573617952928298,
+      "grad_norm": 0.032349683344364166,
+      "learning_rate": 2.6276288165238416e-06,
+      "loss": 0.002,
+      "step": 1201
+    },
+    {
+      "epoch": 0.6579091406677614,
+      "grad_norm": 0.049246739596128464,
+      "learning_rate": 2.620064052124825e-06,
+      "loss": 0.0029,
+      "step": 1202
+    },
+    {
+      "epoch": 0.6584564860426929,
+      "grad_norm": 0.04094109684228897,
+      "learning_rate": 2.612506324730046e-06,
+      "loss": 0.0021,
+      "step": 1203
+    },
+    {
+      "epoch": 0.6590038314176245,
+      "grad_norm": 0.013727393932640553,
+      "learning_rate": 2.6049556566862234e-06,
+      "loss": 0.0006,
+      "step": 1204
+    },
+    {
+      "epoch": 0.6595511767925561,
+      "grad_norm": 0.11889801919460297,
+      "learning_rate": 2.597412070319201e-06,
+      "loss": 0.0068,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6600985221674877,
+      "grad_norm": 2.300886631011963,
+      "learning_rate": 2.589875587933892e-06,
+      "loss": 0.2447,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6606458675424193,
+      "grad_norm": 0.0018244871171191335,
+      "learning_rate": 2.582346231814189e-06,
+      "loss": 0.0001,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6611932129173509,
+      "grad_norm": 0.28391480445861816,
+      "learning_rate": 2.57482402422293e-06,
+      "loss": 0.0223,
+      "step": 1208
+    },
+    {
+      "epoch": 0.6617405582922824,
+      "grad_norm": 0.03892725333571434,
+      "learning_rate": 2.567308987401806e-06,
+      "loss": 0.0016,
+      "step": 1209
+    },
+    {
+      "epoch": 0.662287903667214,
+      "grad_norm": 0.029897578060626984,
+      "learning_rate": 2.5598011435713077e-06,
+      "loss": 0.0018,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6628352490421456,
+      "grad_norm": 0.11231425404548645,
+      "learning_rate": 2.552300514930657e-06,
+      "loss": 0.0071,
+      "step": 1211
+    },
+    {
+      "epoch": 0.6633825944170771,
+      "grad_norm": 0.6822946667671204,
+      "learning_rate": 2.5448071236577493e-06,
+      "loss": 0.0485,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6639299397920088,
+      "grad_norm": 0.011305405758321285,
+      "learning_rate": 2.5373209919090657e-06,
+      "loss": 0.0006,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6644772851669404,
+      "grad_norm": 0.011048182845115662,
+      "learning_rate": 2.5298421418196363e-06,
+      "loss": 0.0005,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6650246305418719,
+      "grad_norm": 0.01335719134658575,
+      "learning_rate": 2.522370595502954e-06,
+      "loss": 0.0006,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6655719759168035,
+      "grad_norm": 0.004644239321351051,
+      "learning_rate": 2.5149063750509166e-06,
+      "loss": 0.0002,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6661193212917351,
+      "grad_norm": 0.06332351267337799,
+      "learning_rate": 2.507449502533762e-06,
+      "loss": 0.0033,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.057875651866197586,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.0026,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6672140120415982,
+      "grad_norm": 0.0002122679288731888,
+      "learning_rate": 2.4925578894763524e-06,
+      "loss": 0.0,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6677613574165299,
+      "grad_norm": 0.01880548521876335,
+      "learning_rate": 2.485123192967677e-06,
+      "loss": 0.001,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6683087027914614,
+      "grad_norm": 0.0019574714824557304,
+      "learning_rate": 2.4776959324569193e-06,
+      "loss": 0.0001,
+      "step": 1221
+    },
+    {
+      "epoch": 0.668856048166393,
+      "grad_norm": 0.028929315507411957,
+      "learning_rate": 2.4702761299050314e-06,
+      "loss": 0.0012,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6694033935413246,
+      "grad_norm": 0.04426882788538933,
+      "learning_rate": 2.462863807250915e-06,
+      "loss": 0.0027,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6699507389162561,
+      "grad_norm": 0.005717442370951176,
+      "learning_rate": 2.4554589864113566e-06,
+      "loss": 0.0003,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6704980842911877,
+      "grad_norm": 1.8422279357910156,
+      "learning_rate": 2.4480616892809593e-06,
+      "loss": 0.2233,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6710454296661194,
+      "grad_norm": 0.23395343124866486,
+      "learning_rate": 2.4406719377320808e-06,
+      "loss": 0.0146,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6715927750410509,
+      "grad_norm": 0.001357329892925918,
+      "learning_rate": 2.4332897536147728e-06,
+      "loss": 0.0001,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6721401204159825,
+      "grad_norm": 0.07431194186210632,
+      "learning_rate": 2.425915158756699e-06,
+      "loss": 0.0045,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6726874657909141,
+      "grad_norm": 2.2066705226898193,
+      "learning_rate": 2.418548174963099e-06,
+      "loss": 0.5243,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6732348111658456,
+      "grad_norm": 0.0052291397005319595,
+      "learning_rate": 2.411188824016697e-06,
+      "loss": 0.0003,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6737821565407772,
+      "grad_norm": 0.0007067082915455103,
+      "learning_rate": 2.4038371276776525e-06,
+      "loss": 0.0001,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6743295019157088,
+      "grad_norm": 3.0892605781555176,
+      "learning_rate": 2.396493107683488e-06,
+      "loss": 0.8147,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6748768472906403,
+      "grad_norm": 0.0053380937315523624,
+      "learning_rate": 2.3891567857490373e-06,
+      "loss": 0.0003,
+      "step": 1233
+    },
+    {
+      "epoch": 0.675424192665572,
+      "grad_norm": 2.618467092514038,
+      "learning_rate": 2.38182818356636e-06,
+      "loss": 0.5732,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6759715380405036,
+      "grad_norm": 0.14668996632099152,
+      "learning_rate": 2.374507322804702e-06,
+      "loss": 0.0059,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6765188834154351,
+      "grad_norm": 0.495018869638443,
+      "learning_rate": 2.3671942251104125e-06,
+      "loss": 0.0381,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6770662287903667,
+      "grad_norm": 0.0009198206826113164,
+      "learning_rate": 2.359888912106888e-06,
+      "loss": 0.0001,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6776135741652983,
+      "grad_norm": 0.005943139083683491,
+      "learning_rate": 2.3525914053945054e-06,
+      "loss": 0.0003,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6781609195402298,
+      "grad_norm": 0.1299518346786499,
+      "learning_rate": 2.345301726550567e-06,
+      "loss": 0.0094,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6787082649151615,
+      "grad_norm": 0.013089645653963089,
+      "learning_rate": 2.3380198971292195e-06,
+      "loss": 0.0006,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6792556102900931,
+      "grad_norm": 2.11917781829834,
+      "learning_rate": 2.3307459386614095e-06,
+      "loss": 0.3652,
+      "step": 1241
+    },
+    {
+      "epoch": 0.6798029556650246,
+      "grad_norm": 0.020909443497657776,
+      "learning_rate": 2.323479872654805e-06,
+      "loss": 0.0011,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6803503010399562,
+      "grad_norm": 0.004706669598817825,
+      "learning_rate": 2.316221720593739e-06,
+      "loss": 0.0003,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6808976464148878,
+      "grad_norm": 0.010048504918813705,
+      "learning_rate": 2.3089715039391447e-06,
+      "loss": 0.0005,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6814449917898193,
+      "grad_norm": 0.32672691345214844,
+      "learning_rate": 2.301729244128496e-06,
+      "loss": 0.0203,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6819923371647509,
+      "grad_norm": 0.37050729990005493,
+      "learning_rate": 2.2944949625757295e-06,
+      "loss": 0.0143,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6825396825396826,
+      "grad_norm": 0.07902567833662033,
+      "learning_rate": 2.2872686806712037e-06,
+      "loss": 0.0046,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6830870279146142,
+      "grad_norm": 0.0054732682183384895,
+      "learning_rate": 2.2800504197816147e-06,
+      "loss": 0.0003,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6836343732895457,
+      "grad_norm": 0.002054156269878149,
+      "learning_rate": 2.2728402012499477e-06,
+      "loss": 0.0001,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6841817186644773,
+      "grad_norm": 2.3521831035614014,
+      "learning_rate": 2.265638046395405e-06,
+      "loss": 0.297,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6847290640394089,
+      "grad_norm": 4.510025501251221,
+      "learning_rate": 2.2584439765133453e-06,
+      "loss": 0.15,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6852764094143404,
+      "grad_norm": 0.0020269020460546017,
+      "learning_rate": 2.251258012875228e-06,
+      "loss": 0.0001,
+      "step": 1252
+    },
+    {
+      "epoch": 0.685823754789272,
+      "grad_norm": 0.1867758184671402,
+      "learning_rate": 2.244080176728536e-06,
+      "loss": 0.0132,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6863711001642037,
+      "grad_norm": 0.021715736016631126,
+      "learning_rate": 2.2369104892967253e-06,
+      "loss": 0.001,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6869184455391352,
+      "grad_norm": 0.9083892107009888,
+      "learning_rate": 2.229748971779157e-06,
+      "loss": 0.0824,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6874657909140668,
+      "grad_norm": 0.001186563284136355,
+      "learning_rate": 2.2225956453510345e-06,
+      "loss": 0.0001,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6880131362889984,
+      "grad_norm": 2.4314286708831787,
+      "learning_rate": 2.2154505311633406e-06,
+      "loss": 0.3183,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6885604816639299,
+      "grad_norm": 0.2594597041606903,
+      "learning_rate": 2.208313650342784e-06,
+      "loss": 0.0179,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6891078270388615,
+      "grad_norm": 0.1705789715051651,
+      "learning_rate": 2.2011850239917136e-06,
+      "loss": 0.0117,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6896551724137931,
+      "grad_norm": 0.010192793793976307,
+      "learning_rate": 2.1940646731880887e-06,
+      "loss": 0.0004,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6902025177887247,
+      "grad_norm": 0.008953562937676907,
+      "learning_rate": 2.186952618985387e-06,
+      "loss": 0.0003,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6907498631636563,
+      "grad_norm": 0.022463621571660042,
+      "learning_rate": 2.1798488824125613e-06,
+      "loss": 0.0013,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6912972085385879,
+      "grad_norm": 0.06705313920974731,
+      "learning_rate": 2.1727534844739658e-06,
+      "loss": 0.0032,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6918445539135194,
+      "grad_norm": 0.2966002821922302,
+      "learning_rate": 2.1656664461493073e-06,
+      "loss": 0.0223,
+      "step": 1264
+    },
+    {
+      "epoch": 0.692391899288451,
+      "grad_norm": 0.06438577175140381,
+      "learning_rate": 2.1585877883935617e-06,
+      "loss": 0.0037,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6929392446633826,
+      "grad_norm": 0.01615588553249836,
+      "learning_rate": 2.151517532136939e-06,
+      "loss": 0.0008,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6934865900383141,
+      "grad_norm": 0.004233787767589092,
+      "learning_rate": 2.1444556982847996e-06,
+      "loss": 0.0002,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6940339354132458,
+      "grad_norm": 0.6990953683853149,
+      "learning_rate": 2.137402307717602e-06,
+      "loss": 0.0346,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6945812807881774,
+      "grad_norm": 2.0177769660949707,
+      "learning_rate": 2.1303573812908383e-06,
+      "loss": 0.2455,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6951286261631089,
+      "grad_norm": 0.14173780381679535,
+      "learning_rate": 2.1233209398349817e-06,
+      "loss": 0.0088,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6956759715380405,
+      "grad_norm": 0.018130486831068993,
+      "learning_rate": 2.1162930041554026e-06,
+      "loss": 0.001,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6962233169129721,
+      "grad_norm": 0.023228706791996956,
+      "learning_rate": 2.109273595032335e-06,
+      "loss": 0.0009,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6967706622879036,
+      "grad_norm": 0.06960176676511765,
+      "learning_rate": 2.1022627332207944e-06,
+      "loss": 0.0034,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6973180076628352,
+      "grad_norm": 0.01118812058120966,
+      "learning_rate": 2.095260439450526e-06,
+      "loss": 0.0005,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6978653530377669,
+      "grad_norm": 0.0013710305793210864,
+      "learning_rate": 2.0882667344259384e-06,
+      "loss": 0.0001,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6984126984126984,
+      "grad_norm": 0.009152219630777836,
+      "learning_rate": 2.081281638826052e-06,
+      "loss": 0.0005,
+      "step": 1276
+    },
+    {
+      "epoch": 0.69896004378763,
+      "grad_norm": 1.5331132411956787,
+      "learning_rate": 2.0743051733044184e-06,
+      "loss": 0.2018,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6995073891625616,
+      "grad_norm": 0.006271702703088522,
+      "learning_rate": 2.0673373584890847e-06,
+      "loss": 0.0003,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7000547345374931,
+      "grad_norm": 0.005529694724828005,
+      "learning_rate": 2.0603782149825126e-06,
+      "loss": 0.0003,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7006020799124247,
+      "grad_norm": 0.3179608881473541,
+      "learning_rate": 2.053427763361525e-06,
+      "loss": 0.0212,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7011494252873564,
+      "grad_norm": 0.007049943320453167,
+      "learning_rate": 2.0464860241772454e-06,
+      "loss": 0.0003,
+      "step": 1281
+    },
+    {
+      "epoch": 0.7016967706622879,
+      "grad_norm": 0.42016157507896423,
+      "learning_rate": 2.0395530179550365e-06,
+      "loss": 0.0337,
+      "step": 1282
+    },
+    {
+      "epoch": 0.7022441160372195,
+      "grad_norm": 0.049554865807294846,
+      "learning_rate": 2.0326287651944392e-06,
+      "loss": 0.003,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7027914614121511,
+      "grad_norm": 2.7460994720458984,
+      "learning_rate": 2.0257132863691108e-06,
+      "loss": 0.5833,
+      "step": 1284
+    },
+    {
+      "epoch": 0.7033388067870826,
+      "grad_norm": 0.8982927799224854,
+      "learning_rate": 2.01880660192677e-06,
+      "loss": 0.0693,
+      "step": 1285
+    },
+    {
+      "epoch": 0.7038861521620142,
+      "grad_norm": 0.12810498476028442,
+      "learning_rate": 2.011908732289127e-06,
+      "loss": 0.0062,
+      "step": 1286
+    },
+    {
+      "epoch": 0.7044334975369458,
+      "grad_norm": 0.001164785004220903,
+      "learning_rate": 2.0050196978518323e-06,
+      "loss": 0.0001,
+      "step": 1287
+    },
+    {
+      "epoch": 0.7049808429118773,
+      "grad_norm": 2.3705291748046875,
+      "learning_rate": 1.998139518984409e-06,
+      "loss": 0.5842,
+      "step": 1288
+    },
+    {
+      "epoch": 0.705528188286809,
+      "grad_norm": 1.3433531522750854,
+      "learning_rate": 1.9912682160301986e-06,
+      "loss": 0.2473,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7060755336617406,
+      "grad_norm": 1.682137131690979,
+      "learning_rate": 1.9844058093062962e-06,
+      "loss": 0.3179,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7066228790366721,
+      "grad_norm": 0.007880594581365585,
+      "learning_rate": 1.977552319103498e-06,
+      "loss": 0.0004,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7071702244116037,
+      "grad_norm": 0.005036715883761644,
+      "learning_rate": 1.970707765686225e-06,
+      "loss": 0.0002,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7077175697865353,
+      "grad_norm": 0.1129072830080986,
+      "learning_rate": 1.963872169292486e-06,
+      "loss": 0.0074,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7082649151614668,
+      "grad_norm": 0.06477373838424683,
+      "learning_rate": 1.957045550133798e-06,
+      "loss": 0.0037,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7088122605363985,
+      "grad_norm": 0.0029708503279834986,
+      "learning_rate": 1.9502279283951363e-06,
+      "loss": 0.0002,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7093596059113301,
+      "grad_norm": 0.043561290949583054,
+      "learning_rate": 1.943419324234871e-06,
+      "loss": 0.0025,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7099069512862616,
+      "grad_norm": 0.03608259931206703,
+      "learning_rate": 1.9366197577847144e-06,
+      "loss": 0.0017,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7104542966611932,
+      "grad_norm": 0.017917588353157043,
+      "learning_rate": 1.929829249149646e-06,
+      "loss": 0.0009,
+      "step": 1298
+    },
+    {
+      "epoch": 0.7110016420361248,
+      "grad_norm": 0.0025389937218278646,
+      "learning_rate": 1.923047818407875e-06,
+      "loss": 0.0002,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7115489874110563,
+      "grad_norm": 0.004917440470308065,
+      "learning_rate": 1.916275485610761e-06,
+      "loss": 0.0002,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7120963327859879,
+      "grad_norm": 0.46073776483535767,
+      "learning_rate": 1.909512270782764e-06,
+      "loss": 0.0267,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7126436781609196,
+      "grad_norm": 0.06749262660741806,
+      "learning_rate": 1.9027581939213852e-06,
+      "loss": 0.0034,
+      "step": 1302
+    },
+    {
+      "epoch": 0.7131910235358512,
+      "grad_norm": 0.1648429036140442,
+      "learning_rate": 1.8960132749971077e-06,
+      "loss": 0.0105,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7137383689107827,
+      "grad_norm": 0.003856829833239317,
+      "learning_rate": 1.8892775339533354e-06,
+      "loss": 0.0002,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.19861356914043427,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.0131,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7148330596606459,
+      "grad_norm": 0.003126499243080616,
+      "learning_rate": 1.8758336651451697e-06,
+      "loss": 0.0002,
+      "step": 1306
+    },
+    {
+      "epoch": 0.7153804050355774,
+      "grad_norm": 0.003605087986215949,
+      "learning_rate": 1.8691255771316664e-06,
+      "loss": 0.0002,
+      "step": 1307
+    },
+    {
+      "epoch": 0.715927750410509,
+      "grad_norm": 0.2548067271709442,
+      "learning_rate": 1.8624267465003176e-06,
+      "loss": 0.0152,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7164750957854407,
+      "grad_norm": 0.00206976430490613,
+      "learning_rate": 1.8557371930582579e-06,
+      "loss": 0.0001,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7170224411603722,
+      "grad_norm": 0.002555719343945384,
+      "learning_rate": 1.8490569365851846e-06,
+      "loss": 0.0001,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7175697865353038,
+      "grad_norm": 0.00730897206813097,
+      "learning_rate": 1.8423859968333063e-06,
+      "loss": 0.0003,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7181171319102354,
+      "grad_norm": 0.09046211838722229,
+      "learning_rate": 1.8357243935272856e-06,
+      "loss": 0.0052,
+      "step": 1312
+    },
+    {
+      "epoch": 0.7186644772851669,
+      "grad_norm": 0.5083016157150269,
+      "learning_rate": 1.8290721463641782e-06,
+      "loss": 0.0485,
+      "step": 1313
+    },
+    {
+      "epoch": 0.7192118226600985,
+      "grad_norm": 0.18554258346557617,
+      "learning_rate": 1.8224292750133743e-06,
+      "loss": 0.0146,
+      "step": 1314
+    },
+    {
+      "epoch": 0.7197591680350302,
+      "grad_norm": 0.005190784577280283,
+      "learning_rate": 1.8157957991165415e-06,
+      "loss": 0.0002,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7203065134099617,
+      "grad_norm": 0.06379152089357376,
+      "learning_rate": 1.8091717382875723e-06,
+      "loss": 0.0035,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7208538587848933,
+      "grad_norm": 0.04362674430012703,
+      "learning_rate": 1.8025571121125141e-06,
+      "loss": 0.0024,
+      "step": 1317
+    },
+    {
+      "epoch": 0.7214012041598249,
+      "grad_norm": 0.0061056301929056644,
+      "learning_rate": 1.7959519401495208e-06,
+      "loss": 0.0003,
+      "step": 1318
+    },
+    {
+      "epoch": 0.7219485495347564,
+      "grad_norm": 0.0022033448331058025,
+      "learning_rate": 1.7893562419287908e-06,
+      "loss": 0.0001,
+      "step": 1319
+    },
+    {
+      "epoch": 0.722495894909688,
+      "grad_norm": 0.0009400215349160135,
+      "learning_rate": 1.7827700369525125e-06,
+      "loss": 0.0001,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7230432402846196,
+      "grad_norm": 2.527238368988037,
+      "learning_rate": 1.7761933446948004e-06,
+      "loss": 0.1914,
+      "step": 1321
+    },
+    {
+      "epoch": 0.7235905856595511,
+      "grad_norm": 0.1244870126247406,
+      "learning_rate": 1.7696261846016505e-06,
+      "loss": 0.0085,
+      "step": 1322
+    },
+    {
+      "epoch": 0.7241379310344828,
+      "grad_norm": 0.04909505695104599,
+      "learning_rate": 1.7630685760908623e-06,
+      "loss": 0.0022,
+      "step": 1323
+    },
+    {
+      "epoch": 0.7246852764094144,
+      "grad_norm": 2.485495090484619,
+      "learning_rate": 1.756520538552003e-06,
+      "loss": 0.1524,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7252326217843459,
+      "grad_norm": 0.0722983255982399,
+      "learning_rate": 1.749982091346335e-06,
+      "loss": 0.0033,
+      "step": 1325
+    },
+    {
+      "epoch": 0.7257799671592775,
+      "grad_norm": 0.003003154881298542,
+      "learning_rate": 1.7434532538067655e-06,
+      "loss": 0.0002,
+      "step": 1326
+    },
+    {
+      "epoch": 0.7263273125342091,
+      "grad_norm": 0.0019068497931584716,
+      "learning_rate": 1.736934045237787e-06,
+      "loss": 0.0001,
+      "step": 1327
+    },
+    {
+      "epoch": 0.7268746579091406,
+      "grad_norm": 0.004010593984276056,
+      "learning_rate": 1.7304244849154256e-06,
+      "loss": 0.0002,
+      "step": 1328
+    },
+    {
+      "epoch": 0.7274220032840722,
+      "grad_norm": 0.15621091425418854,
+      "learning_rate": 1.72392459208717e-06,
+      "loss": 0.0071,
+      "step": 1329
+    },
+    {
+      "epoch": 0.7279693486590039,
+      "grad_norm": 0.39699360728263855,
+      "learning_rate": 1.7174343859719334e-06,
+      "loss": 0.0214,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7285166940339354,
+      "grad_norm": 0.01176412496715784,
+      "learning_rate": 1.7109538857599829e-06,
+      "loss": 0.0006,
+      "step": 1331
+    },
+    {
+      "epoch": 0.729064039408867,
+      "grad_norm": 0.0023852819576859474,
+      "learning_rate": 1.7044831106128867e-06,
+      "loss": 0.0001,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7296113847837986,
+      "grad_norm": 0.011824233457446098,
+      "learning_rate": 1.6980220796634583e-06,
+      "loss": 0.0005,
+      "step": 1333
+    },
+    {
+      "epoch": 0.7301587301587301,
+      "grad_norm": 0.0041591511107981205,
+      "learning_rate": 1.6915708120157042e-06,
+      "loss": 0.0002,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7307060755336617,
+      "grad_norm": 1.6555798053741455,
+      "learning_rate": 1.6851293267447527e-06,
+      "loss": 0.2759,
+      "step": 1335
+    },
+    {
+      "epoch": 0.7312534209085934,
+      "grad_norm": 0.056029561907052994,
+      "learning_rate": 1.6786976428968188e-06,
+      "loss": 0.003,
+      "step": 1336
+    },
+    {
+      "epoch": 0.7318007662835249,
+      "grad_norm": 0.07709532231092453,
+      "learning_rate": 1.6722757794891287e-06,
+      "loss": 0.0043,
+      "step": 1337
+    },
+    {
+      "epoch": 0.7323481116584565,
+      "grad_norm": 0.049780573695898056,
+      "learning_rate": 1.6658637555098744e-06,
+      "loss": 0.0029,
+      "step": 1338
+    },
+    {
+      "epoch": 0.7328954570333881,
+      "grad_norm": 1.2633681297302246,
+      "learning_rate": 1.6594615899181526e-06,
+      "loss": 0.1035,
+      "step": 1339
+    },
+    {
+      "epoch": 0.7334428024083196,
+      "grad_norm": 0.0017923095729202032,
+      "learning_rate": 1.653069301643918e-06,
+      "loss": 0.0001,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7339901477832512,
+      "grad_norm": 0.06065789982676506,
+      "learning_rate": 1.6466869095879079e-06,
+      "loss": 0.003,
+      "step": 1341
+    },
+    {
+      "epoch": 0.7345374931581828,
+      "grad_norm": 0.004060553852468729,
+      "learning_rate": 1.6403144326216085e-06,
+      "loss": 0.0001,
+      "step": 1342
+    },
+    {
+      "epoch": 0.7350848385331143,
+      "grad_norm": 1.5886331796646118,
+      "learning_rate": 1.6339518895871853e-06,
+      "loss": 0.2001,
+      "step": 1343
+    },
+    {
+      "epoch": 0.735632183908046,
+      "grad_norm": 0.005949839483946562,
+      "learning_rate": 1.627599299297431e-06,
+      "loss": 0.0003,
+      "step": 1344
+    },
+    {
+      "epoch": 0.7361795292829776,
+      "grad_norm": 0.004496109671890736,
+      "learning_rate": 1.6212566805357094e-06,
+      "loss": 0.0002,
+      "step": 1345
+    },
+    {
+      "epoch": 0.7367268746579091,
+      "grad_norm": 0.006199996452778578,
+      "learning_rate": 1.6149240520559023e-06,
+      "loss": 0.0002,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7372742200328407,
+      "grad_norm": 0.002357217948883772,
+      "learning_rate": 1.6086014325823485e-06,
+      "loss": 0.0001,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7378215654077723,
+      "grad_norm": 0.015918180346488953,
+      "learning_rate": 1.6022888408097991e-06,
+      "loss": 0.0008,
+      "step": 1348
+    },
+    {
+      "epoch": 0.7383689107827038,
+      "grad_norm": 0.0009474227554164827,
+      "learning_rate": 1.5959862954033495e-06,
+      "loss": 0.0001,
+      "step": 1349
+    },
+    {
+      "epoch": 0.7389162561576355,
+      "grad_norm": 0.08809763938188553,
+      "learning_rate": 1.589693814998391e-06,
+      "loss": 0.0051,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7394636015325671,
+      "grad_norm": 2.6322948932647705,
+      "learning_rate": 1.5834114182005544e-06,
+      "loss": 0.346,
+      "step": 1351
+    },
+    {
+      "epoch": 0.7400109469074986,
+      "grad_norm": 0.17066499590873718,
+      "learning_rate": 1.577139123585657e-06,
+      "loss": 0.0112,
+      "step": 1352
+    },
+    {
+      "epoch": 0.7405582922824302,
+      "grad_norm": 0.0018952427199110389,
+      "learning_rate": 1.5708769496996445e-06,
+      "loss": 0.0001,
+      "step": 1353
+    },
+    {
+      "epoch": 0.7411056376573618,
+      "grad_norm": 0.005605690646916628,
+      "learning_rate": 1.5646249150585368e-06,
+      "loss": 0.0003,
+      "step": 1354
+    },
+    {
+      "epoch": 0.7416529830322933,
+      "grad_norm": 0.07938912510871887,
+      "learning_rate": 1.5583830381483789e-06,
+      "loss": 0.0039,
+      "step": 1355
+    },
+    {
+      "epoch": 0.7422003284072249,
+      "grad_norm": 2.0162599086761475,
+      "learning_rate": 1.552151337425173e-06,
+      "loss": 0.4149,
+      "step": 1356
+    },
+    {
+      "epoch": 0.7427476737821566,
+      "grad_norm": 0.0025566776748746634,
+      "learning_rate": 1.5459298313148402e-06,
+      "loss": 0.0002,
+      "step": 1357
+    },
+    {
+      "epoch": 0.7432950191570882,
+      "grad_norm": 0.301851361989975,
+      "learning_rate": 1.5397185382131524e-06,
+      "loss": 0.0229,
+      "step": 1358
+    },
+    {
+      "epoch": 0.7438423645320197,
+      "grad_norm": 0.005030965898185968,
+      "learning_rate": 1.533517476485691e-06,
+      "loss": 0.0002,
+      "step": 1359
+    },
+    {
+      "epoch": 0.7443897099069513,
+      "grad_norm": 0.05721491575241089,
+      "learning_rate": 1.5273266644677737e-06,
+      "loss": 0.0034,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7449370552818829,
+      "grad_norm": 0.005043413024395704,
+      "learning_rate": 1.521146120464424e-06,
+      "loss": 0.0002,
+      "step": 1361
+    },
+    {
+      "epoch": 0.7454844006568144,
+      "grad_norm": 0.010818198323249817,
+      "learning_rate": 1.514975862750297e-06,
+      "loss": 0.0006,
+      "step": 1362
+    },
+    {
+      "epoch": 0.746031746031746,
+      "grad_norm": 0.010304316878318787,
+      "learning_rate": 1.5088159095696365e-06,
+      "loss": 0.0005,
+      "step": 1363
+    },
+    {
+      "epoch": 0.7465790914066777,
+      "grad_norm": 0.08233784884214401,
+      "learning_rate": 1.5026662791362145e-06,
+      "loss": 0.0043,
+      "step": 1364
+    },
+    {
+      "epoch": 0.7471264367816092,
+      "grad_norm": 0.00964348390698433,
+      "learning_rate": 1.4965269896332884e-06,
+      "loss": 0.0005,
+      "step": 1365
+    },
+    {
+      "epoch": 0.7476737821565408,
+      "grad_norm": 0.005296952556818724,
+      "learning_rate": 1.4903980592135281e-06,
+      "loss": 0.0003,
+      "step": 1366
+    },
+    {
+      "epoch": 0.7482211275314724,
+      "grad_norm": 0.00572380842640996,
+      "learning_rate": 1.4842795059989845e-06,
+      "loss": 0.0003,
+      "step": 1367
+    },
+    {
+      "epoch": 0.7487684729064039,
+      "grad_norm": 0.0025136778131127357,
+      "learning_rate": 1.4781713480810184e-06,
+      "loss": 0.0001,
+      "step": 1368
+    },
+    {
+      "epoch": 0.7493158182813355,
+      "grad_norm": 0.002252168720588088,
+      "learning_rate": 1.472073603520256e-06,
+      "loss": 0.0001,
+      "step": 1369
+    },
+    {
+      "epoch": 0.7498631636562672,
+      "grad_norm": 0.12772350013256073,
+      "learning_rate": 1.4659862903465322e-06,
+      "loss": 0.0059,
+      "step": 1370
+    },
+    {
+      "epoch": 0.7504105090311987,
+      "grad_norm": 0.006171985529363155,
+      "learning_rate": 1.4599094265588432e-06,
+      "loss": 0.0004,
+      "step": 1371
+    },
+    {
+      "epoch": 0.7509578544061303,
+      "grad_norm": 0.2971583902835846,
+      "learning_rate": 1.4538430301252783e-06,
+      "loss": 0.0203,
+      "step": 1372
+    },
+    {
+      "epoch": 0.7515051997810619,
+      "grad_norm": 0.01382134947925806,
+      "learning_rate": 1.4477871189829872e-06,
+      "loss": 0.0008,
+      "step": 1373
+    },
+    {
+      "epoch": 0.7520525451559934,
+      "grad_norm": 2.908761978149414,
+      "learning_rate": 1.4417417110381126e-06,
+      "loss": 0.6791,
+      "step": 1374
+    },
+    {
+      "epoch": 0.752599890530925,
+      "grad_norm": 0.004314839839935303,
+      "learning_rate": 1.4357068241657396e-06,
+      "loss": 0.0002,
+      "step": 1375
+    },
+    {
+      "epoch": 0.7531472359058566,
+      "grad_norm": 1.0610297918319702,
+      "learning_rate": 1.4296824762098465e-06,
+      "loss": 0.0323,
+      "step": 1376
+    },
+    {
+      "epoch": 0.7536945812807881,
+      "grad_norm": 0.031116381287574768,
+      "learning_rate": 1.4236686849832497e-06,
+      "loss": 0.0017,
+      "step": 1377
+    },
+    {
+      "epoch": 0.7542419266557198,
+      "grad_norm": 2.27239990234375,
+      "learning_rate": 1.4176654682675518e-06,
+      "loss": 0.4945,
+      "step": 1378
+    },
+    {
+      "epoch": 0.7547892720306514,
+      "grad_norm": 0.006444338243454695,
+      "learning_rate": 1.411672843813086e-06,
+      "loss": 0.0003,
+      "step": 1379
+    },
+    {
+      "epoch": 0.7553366174055829,
+      "grad_norm": 0.016596568748354912,
+      "learning_rate": 1.405690829338872e-06,
+      "loss": 0.0006,
+      "step": 1380
+    },
+    {
+      "epoch": 0.7558839627805145,
+      "grad_norm": 0.0022818318102508783,
+      "learning_rate": 1.3997194425325533e-06,
+      "loss": 0.0001,
+      "step": 1381
+    },
+    {
+      "epoch": 0.7564313081554461,
+      "grad_norm": 0.25383079051971436,
+      "learning_rate": 1.39375870105035e-06,
+      "loss": 0.0177,
+      "step": 1382
+    },
+    {
+      "epoch": 0.7569786535303776,
+      "grad_norm": 0.0013886289671063423,
+      "learning_rate": 1.3878086225170067e-06,
+      "loss": 0.0001,
+      "step": 1383
+    },
+    {
+      "epoch": 0.7575259989053093,
+      "grad_norm": 0.12911343574523926,
+      "learning_rate": 1.3818692245257398e-06,
+      "loss": 0.0084,
+      "step": 1384
+    },
+    {
+      "epoch": 0.7580733442802409,
+      "grad_norm": 0.09496220201253891,
+      "learning_rate": 1.3759405246381841e-06,
+      "loss": 0.0055,
+      "step": 1385
+    },
+    {
+      "epoch": 0.7586206896551724,
+      "grad_norm": 0.11710565537214279,
+      "learning_rate": 1.370022540384347e-06,
+      "loss": 0.0087,
+      "step": 1386
+    },
+    {
+      "epoch": 0.759168035030104,
+      "grad_norm": 0.001593872788362205,
+      "learning_rate": 1.364115289262543e-06,
+      "loss": 0.0001,
+      "step": 1387
+    },
+    {
+      "epoch": 0.7597153804050356,
+      "grad_norm": 0.006262289825826883,
+      "learning_rate": 1.358218788739361e-06,
+      "loss": 0.0003,
+      "step": 1388
+    },
+    {
+      "epoch": 0.7602627257799671,
+      "grad_norm": 0.004391077905893326,
+      "learning_rate": 1.352333056249595e-06,
+      "loss": 0.0003,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7608100711548987,
+      "grad_norm": 0.012612469494342804,
+      "learning_rate": 1.3464581091962037e-06,
+      "loss": 0.0006,
+      "step": 1390
+    },
+    {
+      "epoch": 0.7613574165298304,
+      "grad_norm": 0.24891842901706696,
+      "learning_rate": 1.340593964950252e-06,
+      "loss": 0.0156,
+      "step": 1391
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 0.0030374748166650534,
+      "learning_rate": 1.3347406408508695e-06,
+      "loss": 0.0002,
+      "step": 1392
+    },
+    {
+      "epoch": 0.7624521072796935,
+      "grad_norm": 0.028773490339517593,
+      "learning_rate": 1.3288981542051844e-06,
+      "loss": 0.0016,
+      "step": 1393
+    },
+    {
+      "epoch": 0.7629994526546251,
+      "grad_norm": 0.051260218024253845,
+      "learning_rate": 1.3230665222882872e-06,
+      "loss": 0.0023,
+      "step": 1394
+    },
+    {
+      "epoch": 0.7635467980295566,
+      "grad_norm": 0.10326068848371506,
+      "learning_rate": 1.3172457623431706e-06,
+      "loss": 0.0064,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7640941434044882,
+      "grad_norm": 0.09019647538661957,
+      "learning_rate": 1.3114358915806808e-06,
+      "loss": 0.0056,
+      "step": 1396
+    },
+    {
+      "epoch": 0.7646414887794198,
+      "grad_norm": 0.06393663585186005,
+      "learning_rate": 1.3056369271794656e-06,
+      "loss": 0.0038,
+      "step": 1397
+    },
+    {
+      "epoch": 0.7651888341543513,
+      "grad_norm": 0.0010087719419971108,
+      "learning_rate": 1.2998488862859305e-06,
+      "loss": 0.0001,
+      "step": 1398
+    },
+    {
+      "epoch": 0.765736179529283,
+      "grad_norm": 0.24810169637203217,
+      "learning_rate": 1.2940717860141734e-06,
+      "loss": 0.0203,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7662835249042146,
+      "grad_norm": 0.0025587116833776236,
+      "learning_rate": 1.2883056434459506e-06,
+      "loss": 0.0001,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7668308702791461,
+      "grad_norm": 0.17504432797431946,
+      "learning_rate": 1.2825504756306156e-06,
+      "loss": 0.0092,
+      "step": 1401
+    },
+    {
+      "epoch": 0.7673782156540777,
+      "grad_norm": 0.003361350391060114,
+      "learning_rate": 1.2768062995850716e-06,
+      "loss": 0.0002,
+      "step": 1402
+    },
+    {
+      "epoch": 0.7679255610290093,
+      "grad_norm": 0.004303762689232826,
+      "learning_rate": 1.2710731322937198e-06,
+      "loss": 0.0002,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7684729064039408,
+      "grad_norm": 0.0592142678797245,
+      "learning_rate": 1.2653509907084171e-06,
+      "loss": 0.0027,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7690202517788725,
+      "grad_norm": 0.002017846331000328,
+      "learning_rate": 1.2596398917484088e-06,
+      "loss": 0.0001,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7695675971538041,
+      "grad_norm": 0.06095453351736069,
+      "learning_rate": 1.2539398523003e-06,
+      "loss": 0.0035,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7701149425287356,
+      "grad_norm": 0.05319780856370926,
+      "learning_rate": 1.2482508892179884e-06,
+      "loss": 0.0021,
+      "step": 1407
+    },
+    {
+      "epoch": 0.7706622879036672,
+      "grad_norm": 0.10127472877502441,
+      "learning_rate": 1.2425730193226237e-06,
+      "loss": 0.0067,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7712096332785988,
+      "grad_norm": 0.00393831729888916,
+      "learning_rate": 1.2369062594025549e-06,
+      "loss": 0.0002,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7717569786535303,
+      "grad_norm": 0.132782444357872,
+      "learning_rate": 1.2312506262132795e-06,
+      "loss": 0.0068,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7723043240284619,
+      "grad_norm": 0.0037302267737686634,
+      "learning_rate": 1.2256061364773958e-06,
+      "loss": 0.0002,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7728516694033936,
+      "grad_norm": 0.09026055783033371,
+      "learning_rate": 1.2199728068845574e-06,
+      "loss": 0.0045,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7733990147783252,
+      "grad_norm": 0.028559090569615364,
+      "learning_rate": 1.214350654091413e-06,
+      "loss": 0.0015,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7739463601532567,
+      "grad_norm": 0.10447684675455093,
+      "learning_rate": 1.2087396947215678e-06,
+      "loss": 0.0068,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7744937055281883,
+      "grad_norm": 0.004746158141642809,
+      "learning_rate": 1.2031399453655296e-06,
+      "loss": 0.0002,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7750410509031199,
+      "grad_norm": 0.0731428861618042,
+      "learning_rate": 1.1975514225806573e-06,
+      "loss": 0.004,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7755883962780514,
+      "grad_norm": 0.010456275194883347,
+      "learning_rate": 1.191974142891123e-06,
+      "loss": 0.0005,
+      "step": 1417
+    },
+    {
+      "epoch": 0.776135741652983,
+      "grad_norm": 0.11829491704702377,
+      "learning_rate": 1.1864081227878438e-06,
+      "loss": 0.0091,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7766830870279147,
+      "grad_norm": 2.1715807914733887,
+      "learning_rate": 1.1808533787284543e-06,
+      "loss": 0.1799,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7772304324028462,
+      "grad_norm": 0.0016666334122419357,
+      "learning_rate": 1.1753099271372432e-06,
+      "loss": 0.0001,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 0.0010274969972670078,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.0001,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7783251231527094,
+      "grad_norm": 0.001577872666530311,
+      "learning_rate": 1.1642569668895171e-06,
+      "loss": 0.0001,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7788724685276409,
+      "grad_norm": 0.0030261666979640722,
+      "learning_rate": 1.1587474909144419e-06,
+      "loss": 0.0002,
+      "step": 1423
+    },
+    {
+      "epoch": 0.7794198139025725,
+      "grad_norm": 0.0762784481048584,
+      "learning_rate": 1.1532493727703214e-06,
+      "loss": 0.0037,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7799671592775042,
+      "grad_norm": 2.675801992416382,
+      "learning_rate": 1.1477626287140164e-06,
+      "loss": 0.1384,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7805145046524357,
+      "grad_norm": 0.04527585953474045,
+      "learning_rate": 1.1422872749687542e-06,
+      "loss": 0.0028,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7810618500273673,
+      "grad_norm": 0.005558805540204048,
+      "learning_rate": 1.136823327724081e-06,
+      "loss": 0.0003,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7816091954022989,
+      "grad_norm": 0.27077722549438477,
+      "learning_rate": 1.1313708031358183e-06,
+      "loss": 0.0131,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7821565407772304,
+      "grad_norm": 0.025989117100834846,
+      "learning_rate": 1.1259297173260158e-06,
+      "loss": 0.0013,
+      "step": 1429
+    },
+    {
+      "epoch": 0.782703886152162,
+      "grad_norm": 0.026661472395062447,
+      "learning_rate": 1.1205000863828936e-06,
+      "loss": 0.0014,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7832512315270936,
+      "grad_norm": 0.001215687021613121,
+      "learning_rate": 1.1150819263608098e-06,
+      "loss": 0.0001,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7837985769020251,
+      "grad_norm": 0.0007609071908518672,
+      "learning_rate": 1.1096752532802007e-06,
+      "loss": 0.0001,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7843459222769568,
+      "grad_norm": 0.014400236308574677,
+      "learning_rate": 1.104280083127539e-06,
+      "loss": 0.0008,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7848932676518884,
+      "grad_norm": 2.9239320755004883,
+      "learning_rate": 1.0988964318552848e-06,
+      "loss": 0.1112,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7854406130268199,
+      "grad_norm": 0.001832630136050284,
+      "learning_rate": 1.0935243153818437e-06,
+      "loss": 0.0001,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7859879584017515,
+      "grad_norm": 1.0065650939941406,
+      "learning_rate": 1.0881637495915055e-06,
+      "loss": 0.0454,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7865353037766831,
+      "grad_norm": 0.00475548068061471,
+      "learning_rate": 1.0828147503344177e-06,
+      "loss": 0.0002,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7870826491516146,
+      "grad_norm": 0.0034952079877257347,
+      "learning_rate": 1.077477333426521e-06,
+      "loss": 0.0002,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7876299945265463,
+      "grad_norm": 0.47412243485450745,
+      "learning_rate": 1.072151514649512e-06,
+      "loss": 0.0393,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7881773399014779,
+      "grad_norm": 0.0020336441230028868,
+      "learning_rate": 1.0668373097507922e-06,
+      "loss": 0.0001,
+      "step": 1440
+    },
+    {
+      "epoch": 0.7887246852764094,
+      "grad_norm": 0.40038520097732544,
+      "learning_rate": 1.061534734443425e-06,
+      "loss": 0.0264,
+      "step": 1441
+    },
+    {
+      "epoch": 0.789272030651341,
+      "grad_norm": 1.4300894737243652,
+      "learning_rate": 1.0562438044060846e-06,
+      "loss": 0.0774,
+      "step": 1442
+    },
+    {
+      "epoch": 0.7898193760262726,
+      "grad_norm": 0.001692896126769483,
+      "learning_rate": 1.0509645352830178e-06,
+      "loss": 0.0001,
+      "step": 1443
+    },
+    {
+      "epoch": 0.7903667214012041,
+      "grad_norm": 0.05487797036767006,
+      "learning_rate": 1.0456969426839869e-06,
+      "loss": 0.0028,
+      "step": 1444
+    },
+    {
+      "epoch": 0.7909140667761357,
+      "grad_norm": 2.1209144592285156,
+      "learning_rate": 1.040441042184231e-06,
+      "loss": 0.5426,
+      "step": 1445
+    },
+    {
+      "epoch": 0.7914614121510674,
+      "grad_norm": 0.001598983071744442,
+      "learning_rate": 1.035196849324418e-06,
+      "loss": 0.0001,
+      "step": 1446
+    },
+    {
+      "epoch": 0.7920087575259989,
+      "grad_norm": 0.7857943773269653,
+      "learning_rate": 1.0299643796105985e-06,
+      "loss": 0.0651,
+      "step": 1447
+    },
+    {
+      "epoch": 0.7925561029009305,
+      "grad_norm": 0.0008221364114433527,
+      "learning_rate": 1.0247436485141605e-06,
+      "loss": 0.0001,
+      "step": 1448
+    },
+    {
+      "epoch": 0.7931034482758621,
+      "grad_norm": 2.5932040214538574,
+      "learning_rate": 1.0195346714717813e-06,
+      "loss": 0.1756,
+      "step": 1449
+    },
+    {
+      "epoch": 0.7936507936507936,
+      "grad_norm": 2.633694648742676,
+      "learning_rate": 1.0143374638853892e-06,
+      "loss": 0.4555,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7941981390257252,
+      "grad_norm": 0.0076095834374427795,
+      "learning_rate": 1.0091520411221028e-06,
+      "loss": 0.0004,
+      "step": 1451
+    },
+    {
+      "epoch": 0.7947454844006568,
+      "grad_norm": 0.0011043306440114975,
+      "learning_rate": 1.0039784185142065e-06,
+      "loss": 0.0001,
+      "step": 1452
+    },
+    {
+      "epoch": 0.7952928297755884,
+      "grad_norm": 0.0019398522563278675,
+      "learning_rate": 9.988166113590857e-07,
+      "loss": 0.0001,
+      "step": 1453
+    },
+    {
+      "epoch": 0.79584017515052,
+      "grad_norm": 3.7462940216064453,
+      "learning_rate": 9.936666349191936e-07,
+      "loss": 0.2701,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7963875205254516,
+      "grad_norm": 0.921699047088623,
+      "learning_rate": 9.88528504422e-07,
+      "loss": 0.0569,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7969348659003831,
+      "grad_norm": 0.0006251604063436389,
+      "learning_rate": 9.834022350599538e-07,
+      "loss": 0.0,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7974822112753147,
+      "grad_norm": 0.0008446245919913054,
+      "learning_rate": 9.78287841990423e-07,
+      "loss": 0.0001,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7980295566502463,
+      "grad_norm": 0.10622086375951767,
+      "learning_rate": 9.731853403356705e-07,
+      "loss": 0.0066,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7985769020251778,
+      "grad_norm": 0.0006699857767671347,
+      "learning_rate": 9.68094745182792e-07,
+      "loss": 0.0001,
+      "step": 1459
+    },
+    {
+      "epoch": 0.7991242474001095,
+      "grad_norm": 1.367194652557373,
+      "learning_rate": 9.630160715836805e-07,
+      "loss": 0.2876,
+      "step": 1460
+    },
+    {
+      "epoch": 0.7996715927750411,
+      "grad_norm": 0.030433079227805138,
+      "learning_rate": 9.579493345549772e-07,
+      "loss": 0.0015,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8002189381499726,
+      "grad_norm": 0.0018854098161682487,
+      "learning_rate": 9.528945490780339e-07,
+      "loss": 0.0001,
+      "step": 1462
+    },
+    {
+      "epoch": 0.8007662835249042,
+      "grad_norm": 0.006027139723300934,
+      "learning_rate": 9.47851730098856e-07,
+      "loss": 0.0003,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8013136288998358,
+      "grad_norm": 0.7355122566223145,
+      "learning_rate": 9.428208925280746e-07,
+      "loss": 0.0696,
+      "step": 1464
+    },
+    {
+      "epoch": 0.8018609742747673,
+      "grad_norm": 2.0000030994415283,
+      "learning_rate": 9.378020512408903e-07,
+      "loss": 0.0859,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8024083196496989,
+      "grad_norm": 2.8742313385009766,
+      "learning_rate": 9.327952210770319e-07,
+      "loss": 0.5494,
+      "step": 1466
+    },
+    {
+      "epoch": 0.8029556650246306,
+      "grad_norm": 0.16476808488368988,
+      "learning_rate": 9.278004168407151e-07,
+      "loss": 0.0098,
+      "step": 1467
+    },
+    {
+      "epoch": 0.8035030103995622,
+      "grad_norm": 0.003957406617701054,
+      "learning_rate": 9.228176533005984e-07,
+      "loss": 0.0002,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8040503557744937,
+      "grad_norm": 0.00669764494523406,
+      "learning_rate": 9.178469451897376e-07,
+      "loss": 0.0003,
+      "step": 1469
+    },
+    {
+      "epoch": 0.8045977011494253,
+      "grad_norm": 0.005619878880679607,
+      "learning_rate": 9.128883072055411e-07,
+      "loss": 0.0003,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8051450465243569,
+      "grad_norm": 3.6218132972717285,
+      "learning_rate": 9.079417540097307e-07,
+      "loss": 0.7148,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8056923918992884,
+      "grad_norm": 0.005683012772351503,
+      "learning_rate": 9.030073002282941e-07,
+      "loss": 0.0002,
+      "step": 1472
+    },
+    {
+      "epoch": 0.80623973727422,
+      "grad_norm": 0.005043988116085529,
+      "learning_rate": 8.980849604514453e-07,
+      "loss": 0.0003,
+      "step": 1473
+    },
+    {
+      "epoch": 0.8067870826491517,
+      "grad_norm": 0.5382112860679626,
+      "learning_rate": 8.931747492335758e-07,
+      "loss": 0.0256,
+      "step": 1474
+    },
+    {
+      "epoch": 0.8073344280240832,
+      "grad_norm": 0.46263471245765686,
+      "learning_rate": 8.882766810932214e-07,
+      "loss": 0.0309,
+      "step": 1475
+    },
+    {
+      "epoch": 0.8078817733990148,
+      "grad_norm": 0.0011781550711020827,
+      "learning_rate": 8.833907705130091e-07,
+      "loss": 0.0001,
+      "step": 1476
+    },
+    {
+      "epoch": 0.8084291187739464,
+      "grad_norm": 0.006095638498663902,
+      "learning_rate": 8.785170319396174e-07,
+      "loss": 0.0003,
+      "step": 1477
+    },
+    {
+      "epoch": 0.8089764641488779,
+      "grad_norm": 0.03239554166793823,
+      "learning_rate": 8.736554797837376e-07,
+      "loss": 0.0009,
+      "step": 1478
+    },
+    {
+      "epoch": 0.8095238095238095,
+      "grad_norm": 0.000690631044562906,
+      "learning_rate": 8.688061284200266e-07,
+      "loss": 0.0001,
+      "step": 1479
+    },
+    {
+      "epoch": 0.8100711548987412,
+      "grad_norm": 1.924971580505371,
+      "learning_rate": 8.639689921870642e-07,
+      "loss": 0.0994,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8106185002736727,
+      "grad_norm": 0.0030510155484080315,
+      "learning_rate": 8.591440853873184e-07,
+      "loss": 0.0002,
+      "step": 1481
+    },
+    {
+      "epoch": 0.8111658456486043,
+      "grad_norm": 0.010571463964879513,
+      "learning_rate": 8.543314222870891e-07,
+      "loss": 0.0005,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8117131910235359,
+      "grad_norm": 0.7898827791213989,
+      "learning_rate": 8.495310171164805e-07,
+      "loss": 0.1773,
+      "step": 1483
+    },
+    {
+      "epoch": 0.8122605363984674,
+      "grad_norm": 1.2340046167373657,
+      "learning_rate": 8.447428840693489e-07,
+      "loss": 0.184,
+      "step": 1484
+    },
+    {
+      "epoch": 0.812807881773399,
+      "grad_norm": 0.004051442723721266,
+      "learning_rate": 8.399670373032665e-07,
+      "loss": 0.0002,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8133552271483306,
+      "grad_norm": 0.0019139654468744993,
+      "learning_rate": 8.35203490939474e-07,
+      "loss": 0.0001,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8139025725232621,
+      "grad_norm": 0.00680742971599102,
+      "learning_rate": 8.304522590628489e-07,
+      "loss": 0.0003,
+      "step": 1487
+    },
+    {
+      "epoch": 0.8144499178981938,
+      "grad_norm": 0.1572553664445877,
+      "learning_rate": 8.257133557218471e-07,
+      "loss": 0.0091,
+      "step": 1488
+    },
+    {
+      "epoch": 0.8149972632731254,
+      "grad_norm": 0.00778138916939497,
+      "learning_rate": 8.209867949284822e-07,
+      "loss": 0.0005,
+      "step": 1489
+    },
+    {
+      "epoch": 0.8155446086480569,
+      "grad_norm": 1.3799161911010742,
+      "learning_rate": 8.162725906582658e-07,
+      "loss": 0.3446,
+      "step": 1490
+    },
+    {
+      "epoch": 0.8160919540229885,
+      "grad_norm": 0.007187338080257177,
+      "learning_rate": 8.115707568501768e-07,
+      "loss": 0.0003,
+      "step": 1491
+    },
+    {
+      "epoch": 0.8166392993979201,
+      "grad_norm": 1.3800119161605835,
+      "learning_rate": 8.068813074066151e-07,
+      "loss": 0.1066,
+      "step": 1492
+    },
+    {
+      "epoch": 0.8171866447728516,
+      "grad_norm": 0.10583969950675964,
+      "learning_rate": 8.022042561933674e-07,
+      "loss": 0.0055,
+      "step": 1493
+    },
+    {
+      "epoch": 0.8177339901477833,
+      "grad_norm": 0.013661573641002178,
+      "learning_rate": 7.975396170395522e-07,
+      "loss": 0.0007,
+      "step": 1494
+    },
+    {
+      "epoch": 0.8182813355227149,
+      "grad_norm": 0.1210581511259079,
+      "learning_rate": 7.928874037375983e-07,
+      "loss": 0.0075,
+      "step": 1495
+    },
+    {
+      "epoch": 0.8188286808976464,
+      "grad_norm": 0.005312152206897736,
+      "learning_rate": 7.882476300431868e-07,
+      "loss": 0.0002,
+      "step": 1496
+    },
+    {
+      "epoch": 0.819376026272578,
+      "grad_norm": 0.026137027889490128,
+      "learning_rate": 7.836203096752193e-07,
+      "loss": 0.0012,
+      "step": 1497
+    },
+    {
+      "epoch": 0.8199233716475096,
+      "grad_norm": 1.857779622077942,
+      "learning_rate": 7.790054563157745e-07,
+      "loss": 0.3924,
+      "step": 1498
+    },
+    {
+      "epoch": 0.8204707170224411,
+      "grad_norm": 0.026395171880722046,
+      "learning_rate": 7.744030836100724e-07,
+      "loss": 0.0014,
+      "step": 1499
+    },
+    {
+      "epoch": 0.8210180623973727,
+      "grad_norm": 0.014956077560782433,
+      "learning_rate": 7.698132051664236e-07,
+      "loss": 0.0007,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8215654077723044,
+      "grad_norm": 1.9142413139343262,
+      "learning_rate": 7.652358345562016e-07,
+      "loss": 0.2772,
+      "step": 1501
+    },
+    {
+      "epoch": 0.8221127531472359,
+      "grad_norm": 0.0007567325956188142,
+      "learning_rate": 7.606709853137939e-07,
+      "loss": 0.0001,
+      "step": 1502
+    },
+    {
+      "epoch": 0.8226600985221675,
+      "grad_norm": 0.0009599222685210407,
+      "learning_rate": 7.561186709365653e-07,
+      "loss": 0.0001,
+      "step": 1503
+    },
+    {
+      "epoch": 0.8232074438970991,
+      "grad_norm": 0.13788969814777374,
+      "learning_rate": 7.515789048848171e-07,
+      "loss": 0.0076,
+      "step": 1504
+    },
+    {
+      "epoch": 0.8237547892720306,
+      "grad_norm": 0.005967889446765184,
+      "learning_rate": 7.470517005817473e-07,
+      "loss": 0.0002,
+      "step": 1505
+    },
+    {
+      "epoch": 0.8243021346469622,
+      "grad_norm": 0.001601370982825756,
+      "learning_rate": 7.425370714134122e-07,
+      "loss": 0.0001,
+      "step": 1506
+    },
+    {
+      "epoch": 0.8248494800218938,
+      "grad_norm": 0.0028195439372211695,
+      "learning_rate": 7.380350307286865e-07,
+      "loss": 0.0001,
+      "step": 1507
+    },
+    {
+      "epoch": 0.8253968253968254,
+      "grad_norm": 0.016938157379627228,
+      "learning_rate": 7.33545591839222e-07,
+      "loss": 0.0009,
+      "step": 1508
+    },
+    {
+      "epoch": 0.825944170771757,
+      "grad_norm": 0.003891898086294532,
+      "learning_rate": 7.290687680194092e-07,
+      "loss": 0.0002,
+      "step": 1509
+    },
+    {
+      "epoch": 0.8264915161466886,
+      "grad_norm": 2.533268928527832,
+      "learning_rate": 7.246045725063394e-07,
+      "loss": 0.2119,
+      "step": 1510
+    },
+    {
+      "epoch": 0.8270388615216201,
+      "grad_norm": 0.0018059737049043179,
+      "learning_rate": 7.201530184997635e-07,
+      "loss": 0.0001,
+      "step": 1511
+    },
+    {
+      "epoch": 0.8275862068965517,
+      "grad_norm": 0.017416877672076225,
+      "learning_rate": 7.157141191620548e-07,
+      "loss": 0.0008,
+      "step": 1512
+    },
+    {
+      "epoch": 0.8281335522714833,
+      "grad_norm": 0.04424600675702095,
+      "learning_rate": 7.112878876181673e-07,
+      "loss": 0.0027,
+      "step": 1513
+    },
+    {
+      "epoch": 0.8286808976464148,
+      "grad_norm": 0.04984992370009422,
+      "learning_rate": 7.068743369556042e-07,
+      "loss": 0.0025,
+      "step": 1514
+    },
+    {
+      "epoch": 0.8292282430213465,
+      "grad_norm": 0.01709463633596897,
+      "learning_rate": 7.024734802243649e-07,
+      "loss": 0.0007,
+      "step": 1515
+    },
+    {
+      "epoch": 0.8297755883962781,
+      "grad_norm": 1.7468068599700928,
+      "learning_rate": 6.980853304369239e-07,
+      "loss": 0.2574,
+      "step": 1516
+    },
+    {
+      "epoch": 0.8303229337712096,
+      "grad_norm": 0.002679870929569006,
+      "learning_rate": 6.937099005681792e-07,
+      "loss": 0.0001,
+      "step": 1517
+    },
+    {
+      "epoch": 0.8308702791461412,
+      "grad_norm": 0.001303765457123518,
+      "learning_rate": 6.8934720355542e-07,
+      "loss": 0.0001,
+      "step": 1518
+    },
+    {
+      "epoch": 0.8314176245210728,
+      "grad_norm": 1.4065778255462646,
+      "learning_rate": 6.849972522982845e-07,
+      "loss": 0.2669,
+      "step": 1519
+    },
+    {
+      "epoch": 0.8319649698960043,
+      "grad_norm": 0.0015437363181263208,
+      "learning_rate": 6.806600596587299e-07,
+      "loss": 0.0001,
+      "step": 1520
+    },
+    {
+      "epoch": 0.8325123152709359,
+      "grad_norm": 0.0023011069279164076,
+      "learning_rate": 6.763356384609809e-07,
+      "loss": 0.0001,
+      "step": 1521
+    },
+    {
+      "epoch": 0.8330596606458676,
+      "grad_norm": 0.0042620436288416386,
+      "learning_rate": 6.720240014915063e-07,
+      "loss": 0.0002,
+      "step": 1522
+    },
+    {
+      "epoch": 0.8336070060207992,
+      "grad_norm": 0.1685761958360672,
+      "learning_rate": 6.677251614989699e-07,
+      "loss": 0.0096,
+      "step": 1523
+    },
+    {
+      "epoch": 0.8341543513957307,
+      "grad_norm": 0.22431378066539764,
+      "learning_rate": 6.634391311942024e-07,
+      "loss": 0.0173,
+      "step": 1524
+    },
+    {
+      "epoch": 0.8347016967706623,
+      "grad_norm": 2.5347378253936768,
+      "learning_rate": 6.591659232501507e-07,
+      "loss": 0.3766,
+      "step": 1525
+    },
+    {
+      "epoch": 0.8352490421455939,
+      "grad_norm": 0.06536459177732468,
+      "learning_rate": 6.549055503018575e-07,
+      "loss": 0.0037,
+      "step": 1526
+    },
+    {
+      "epoch": 0.8357963875205254,
+      "grad_norm": 0.044048044830560684,
+      "learning_rate": 6.506580249464089e-07,
+      "loss": 0.0023,
+      "step": 1527
+    },
+    {
+      "epoch": 0.836343732895457,
+      "grad_norm": 3.255218029022217,
+      "learning_rate": 6.464233597429054e-07,
+      "loss": 0.3087,
+      "step": 1528
+    },
+    {
+      "epoch": 0.8368910782703887,
+      "grad_norm": 0.05638457089662552,
+      "learning_rate": 6.42201567212421e-07,
+      "loss": 0.0029,
+      "step": 1529
+    },
+    {
+      "epoch": 0.8374384236453202,
+      "grad_norm": 0.021221015602350235,
+      "learning_rate": 6.379926598379727e-07,
+      "loss": 0.0012,
+      "step": 1530
+    },
+    {
+      "epoch": 0.8379857690202518,
+      "grad_norm": 0.0007102249655872583,
+      "learning_rate": 6.337966500644699e-07,
+      "loss": 0.0001,
+      "step": 1531
+    },
+    {
+      "epoch": 0.8385331143951834,
+      "grad_norm": 2.1397488117218018,
+      "learning_rate": 6.296135502986944e-07,
+      "loss": 0.11,
+      "step": 1532
+    },
+    {
+      "epoch": 0.8390804597701149,
+      "grad_norm": 0.005235095042735338,
+      "learning_rate": 6.254433729092518e-07,
+      "loss": 0.0002,
+      "step": 1533
+    },
+    {
+      "epoch": 0.8396278051450465,
+      "grad_norm": 0.006238035392016172,
+      "learning_rate": 6.212861302265393e-07,
+      "loss": 0.0003,
+      "step": 1534
+    },
+    {
+      "epoch": 0.8401751505199782,
+      "grad_norm": 0.030654199421405792,
+      "learning_rate": 6.171418345427088e-07,
+      "loss": 0.0019,
+      "step": 1535
+    },
+    {
+      "epoch": 0.8407224958949097,
+      "grad_norm": 1.587546706199646,
+      "learning_rate": 6.130104981116314e-07,
+      "loss": 0.2527,
+      "step": 1536
+    },
+    {
+      "epoch": 0.8412698412698413,
+      "grad_norm": 0.003837469732388854,
+      "learning_rate": 6.088921331488568e-07,
+      "loss": 0.0002,
+      "step": 1537
+    },
+    {
+      "epoch": 0.8418171866447729,
+      "grad_norm": 0.10386902838945389,
+      "learning_rate": 6.04786751831587e-07,
+      "loss": 0.006,
+      "step": 1538
+    },
+    {
+      "epoch": 0.8423645320197044,
+      "grad_norm": 0.0031864827033132315,
+      "learning_rate": 6.006943662986275e-07,
+      "loss": 0.0002,
+      "step": 1539
+    },
+    {
+      "epoch": 0.842911877394636,
+      "grad_norm": 1.3453859090805054,
+      "learning_rate": 5.966149886503614e-07,
+      "loss": 0.2179,
+      "step": 1540
+    },
+    {
+      "epoch": 0.8434592227695676,
+      "grad_norm": 0.002893652068451047,
+      "learning_rate": 5.925486309487083e-07,
+      "loss": 0.0002,
+      "step": 1541
+    },
+    {
+      "epoch": 0.8440065681444991,
+      "grad_norm": 0.003957205452024937,
+      "learning_rate": 5.884953052170917e-07,
+      "loss": 0.0002,
+      "step": 1542
+    },
+    {
+      "epoch": 0.8445539135194308,
+      "grad_norm": 0.011327473446726799,
+      "learning_rate": 5.844550234404012e-07,
+      "loss": 0.0006,
+      "step": 1543
+    },
+    {
+      "epoch": 0.8451012588943624,
+      "grad_norm": 0.0016348304925486445,
+      "learning_rate": 5.804277975649574e-07,
+      "loss": 0.0001,
+      "step": 1544
+    },
+    {
+      "epoch": 0.8456486042692939,
+      "grad_norm": 0.018026404082775116,
+      "learning_rate": 5.764136394984809e-07,
+      "loss": 0.0006,
+      "step": 1545
+    },
+    {
+      "epoch": 0.8461959496442255,
+      "grad_norm": 0.06777101755142212,
+      "learning_rate": 5.724125611100467e-07,
+      "loss": 0.0028,
+      "step": 1546
+    },
+    {
+      "epoch": 0.8467432950191571,
+      "grad_norm": 0.005540602374821901,
+      "learning_rate": 5.684245742300625e-07,
+      "loss": 0.0002,
+      "step": 1547
+    },
+    {
+      "epoch": 0.8472906403940886,
+      "grad_norm": 0.04115450754761696,
+      "learning_rate": 5.644496906502233e-07,
+      "loss": 0.0019,
+      "step": 1548
+    },
+    {
+      "epoch": 0.8478379857690203,
+      "grad_norm": 2.1989943981170654,
+      "learning_rate": 5.60487922123481e-07,
+      "loss": 0.2341,
+      "step": 1549
+    },
+    {
+      "epoch": 0.8483853311439519,
+      "grad_norm": 0.006643370725214481,
+      "learning_rate": 5.565392803640069e-07,
+      "loss": 0.0003,
+      "step": 1550
+    },
+    {
+      "epoch": 0.8489326765188834,
+      "grad_norm": 0.14747342467308044,
+      "learning_rate": 5.526037770471649e-07,
+      "loss": 0.0081,
+      "step": 1551
+    },
+    {
+      "epoch": 0.849480021893815,
+      "grad_norm": 0.05744752287864685,
+      "learning_rate": 5.486814238094629e-07,
+      "loss": 0.0032,
+      "step": 1552
+    },
+    {
+      "epoch": 0.8500273672687466,
+      "grad_norm": 0.02027791179716587,
+      "learning_rate": 5.447722322485333e-07,
+      "loss": 0.001,
+      "step": 1553
+    },
+    {
+      "epoch": 0.8505747126436781,
+      "grad_norm": 0.0011862594401463866,
+      "learning_rate": 5.408762139230889e-07,
+      "loss": 0.0001,
+      "step": 1554
+    },
+    {
+      "epoch": 0.8511220580186097,
+      "grad_norm": 0.052090659737586975,
+      "learning_rate": 5.369933803528926e-07,
+      "loss": 0.0035,
+      "step": 1555
+    },
+    {
+      "epoch": 0.8516694033935414,
+      "grad_norm": 2.1469361782073975,
+      "learning_rate": 5.331237430187214e-07,
+      "loss": 0.1273,
+      "step": 1556
+    },
+    {
+      "epoch": 0.8522167487684729,
+      "grad_norm": 0.02452508918941021,
+      "learning_rate": 5.292673133623372e-07,
+      "loss": 0.001,
+      "step": 1557
+    },
+    {
+      "epoch": 0.8527640941434045,
+      "grad_norm": 0.0034144010860472918,
+      "learning_rate": 5.254241027864432e-07,
+      "loss": 0.0002,
+      "step": 1558
+    },
+    {
+      "epoch": 0.8533114395183361,
+      "grad_norm": 0.14860154688358307,
+      "learning_rate": 5.215941226546628e-07,
+      "loss": 0.0095,
+      "step": 1559
+    },
+    {
+      "epoch": 0.8538587848932676,
+      "grad_norm": 0.0034301080740988255,
+      "learning_rate": 5.177773842914963e-07,
+      "loss": 0.0002,
+      "step": 1560
+    },
+    {
+      "epoch": 0.8544061302681992,
+      "grad_norm": 0.001376735046505928,
+      "learning_rate": 5.139738989822901e-07,
+      "loss": 0.0001,
+      "step": 1561
+    },
+    {
+      "epoch": 0.8549534756431308,
+      "grad_norm": 0.06300047785043716,
+      "learning_rate": 5.101836779732044e-07,
+      "loss": 0.003,
+      "step": 1562
+    },
+    {
+      "epoch": 0.8555008210180624,
+      "grad_norm": 0.10772773623466492,
+      "learning_rate": 5.064067324711836e-07,
+      "loss": 0.0074,
+      "step": 1563
+    },
+    {
+      "epoch": 0.856048166392994,
+      "grad_norm": 3.098621368408203,
+      "learning_rate": 5.026430736439102e-07,
+      "loss": 0.8447,
+      "step": 1564
+    },
+    {
+      "epoch": 0.8565955117679256,
+      "grad_norm": 0.020347200334072113,
+      "learning_rate": 4.988927126197901e-07,
+      "loss": 0.0012,
+      "step": 1565
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.004135144874453545,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.0002,
+      "step": 1566
+    },
+    {
+      "epoch": 0.8576902025177887,
+      "grad_norm": 0.13048239052295685,
+      "learning_rate": 4.91431928297984e-07,
+      "loss": 0.0085,
+      "step": 1567
+    },
+    {
+      "epoch": 0.8582375478927203,
+      "grad_norm": 0.0032235216349363327,
+      "learning_rate": 4.877215270603752e-07,
+      "loss": 0.0002,
+      "step": 1568
+    },
+    {
+      "epoch": 0.8587848932676518,
+      "grad_norm": 0.7628779411315918,
+      "learning_rate": 4.840244677460076e-07,
+      "loss": 0.0259,
+      "step": 1569
+    },
+    {
+      "epoch": 0.8593322386425835,
+      "grad_norm": 0.03926008939743042,
+      "learning_rate": 4.803407612863603e-07,
+      "loss": 0.0022,
+      "step": 1570
+    },
+    {
+      "epoch": 0.8598795840175151,
+      "grad_norm": 0.003697637002915144,
+      "learning_rate": 4.7667041857343276e-07,
+      "loss": 0.0002,
+      "step": 1571
+    },
+    {
+      "epoch": 0.8604269293924466,
+      "grad_norm": 0.007138053886592388,
+      "learning_rate": 4.730134504597084e-07,
+      "loss": 0.0003,
+      "step": 1572
+    },
+    {
+      "epoch": 0.8609742747673782,
+      "grad_norm": 0.024003850296139717,
+      "learning_rate": 4.69369867758126e-07,
+      "loss": 0.0011,
+      "step": 1573
+    },
+    {
+      "epoch": 0.8615216201423098,
+      "grad_norm": 0.013022515922784805,
+      "learning_rate": 4.6573968124204506e-07,
+      "loss": 0.0006,
+      "step": 1574
+    },
+    {
+      "epoch": 0.8620689655172413,
+      "grad_norm": 0.025911865755915642,
+      "learning_rate": 4.6212290164521554e-07,
+      "loss": 0.0011,
+      "step": 1575
+    },
+    {
+      "epoch": 0.8626163108921729,
+      "grad_norm": 1.6997803449630737,
+      "learning_rate": 4.585195396617464e-07,
+      "loss": 0.0852,
+      "step": 1576
+    },
+    {
+      "epoch": 0.8631636562671046,
+      "grad_norm": 0.0011990342754870653,
+      "learning_rate": 4.549296059460717e-07,
+      "loss": 0.0001,
+      "step": 1577
+    },
+    {
+      "epoch": 0.8637110016420362,
+      "grad_norm": 2.725243330001831,
+      "learning_rate": 4.5135311111292435e-07,
+      "loss": 0.4184,
+      "step": 1578
+    },
+    {
+      "epoch": 0.8642583470169677,
+      "grad_norm": 0.001103404094465077,
+      "learning_rate": 4.477900657372969e-07,
+      "loss": 0.0001,
+      "step": 1579
+    },
+    {
+      "epoch": 0.8648056923918993,
+      "grad_norm": 0.18278087675571442,
+      "learning_rate": 4.442404803544176e-07,
+      "loss": 0.0114,
+      "step": 1580
+    },
+    {
+      "epoch": 0.8653530377668309,
+      "grad_norm": 0.0011701801558956504,
+      "learning_rate": 4.407043654597126e-07,
+      "loss": 0.0001,
+      "step": 1581
+    },
+    {
+      "epoch": 0.8659003831417624,
+      "grad_norm": 0.0017612545052543283,
+      "learning_rate": 4.371817315087845e-07,
+      "loss": 0.0001,
+      "step": 1582
+    },
+    {
+      "epoch": 0.866447728516694,
+      "grad_norm": 2.4826674461364746,
+      "learning_rate": 4.336725889173676e-07,
+      "loss": 0.5371,
+      "step": 1583
+    },
+    {
+      "epoch": 0.8669950738916257,
+      "grad_norm": 0.010181195102632046,
+      "learning_rate": 4.3017694806131163e-07,
+      "loss": 0.0004,
+      "step": 1584
+    },
+    {
+      "epoch": 0.8675424192665572,
+      "grad_norm": 0.003270001383498311,
+      "learning_rate": 4.266948192765402e-07,
+      "loss": 0.0002,
+      "step": 1585
+    },
+    {
+      "epoch": 0.8680897646414888,
+      "grad_norm": 0.3261997699737549,
+      "learning_rate": 4.2322621285902697e-07,
+      "loss": 0.0192,
+      "step": 1586
+    },
+    {
+      "epoch": 0.8686371100164204,
+      "grad_norm": 0.004310634918510914,
+      "learning_rate": 4.1977113906475965e-07,
+      "loss": 0.0002,
+      "step": 1587
+    },
+    {
+      "epoch": 0.8691844553913519,
+      "grad_norm": 0.009598495438694954,
+      "learning_rate": 4.163296081097168e-07,
+      "loss": 0.0005,
+      "step": 1588
+    },
+    {
+      "epoch": 0.8697318007662835,
+      "grad_norm": 0.004435474518686533,
+      "learning_rate": 4.1290163016982855e-07,
+      "loss": 0.0002,
+      "step": 1589
+    },
+    {
+      "epoch": 0.8702791461412152,
+      "grad_norm": 0.001486207009293139,
+      "learning_rate": 4.0948721538095593e-07,
+      "loss": 0.0001,
+      "step": 1590
+    },
+    {
+      "epoch": 0.8708264915161467,
+      "grad_norm": 0.12514939904212952,
+      "learning_rate": 4.060863738388532e-07,
+      "loss": 0.0074,
+      "step": 1591
+    },
+    {
+      "epoch": 0.8713738368910783,
+      "grad_norm": 0.006256112828850746,
+      "learning_rate": 4.026991155991433e-07,
+      "loss": 0.0003,
+      "step": 1592
+    },
+    {
+      "epoch": 0.8719211822660099,
+      "grad_norm": 0.023347727954387665,
+      "learning_rate": 3.9932545067728366e-07,
+      "loss": 0.0012,
+      "step": 1593
+    },
+    {
+      "epoch": 0.8724685276409414,
+      "grad_norm": 0.06584785878658295,
+      "learning_rate": 3.9596538904854263e-07,
+      "loss": 0.0032,
+      "step": 1594
+    },
+    {
+      "epoch": 0.873015873015873,
+      "grad_norm": 0.08080962300300598,
+      "learning_rate": 3.9261894064796136e-07,
+      "loss": 0.0051,
+      "step": 1595
+    },
+    {
+      "epoch": 0.8735632183908046,
+      "grad_norm": 0.002836831146851182,
+      "learning_rate": 3.8928611537033424e-07,
+      "loss": 0.0002,
+      "step": 1596
+    },
+    {
+      "epoch": 0.8741105637657361,
+      "grad_norm": 0.002772917505353689,
+      "learning_rate": 3.859669230701718e-07,
+      "loss": 0.0001,
+      "step": 1597
+    },
+    {
+      "epoch": 0.8746579091406678,
+      "grad_norm": 2.930410385131836,
+      "learning_rate": 3.8266137356167466e-07,
+      "loss": 0.1096,
+      "step": 1598
+    },
+    {
+      "epoch": 0.8752052545155994,
+      "grad_norm": 0.0008128105546347797,
+      "learning_rate": 3.7936947661870616e-07,
+      "loss": 0.0001,
+      "step": 1599
+    },
+    {
+      "epoch": 0.8757525998905309,
+      "grad_norm": 0.005456219427287579,
+      "learning_rate": 3.760912419747592e-07,
+      "loss": 0.0003,
+      "step": 1600
+    },
+    {
+      "epoch": 0.8762999452654625,
+      "grad_norm": 0.014633177779614925,
+      "learning_rate": 3.728266793229307e-07,
+      "loss": 0.0007,
+      "step": 1601
+    },
+    {
+      "epoch": 0.8768472906403941,
+      "grad_norm": 2.828840732574463,
+      "learning_rate": 3.695757983158954e-07,
+      "loss": 0.5272,
+      "step": 1602
+    },
+    {
+      "epoch": 0.8773946360153256,
+      "grad_norm": 0.0014424819964915514,
+      "learning_rate": 3.663386085658693e-07,
+      "loss": 0.0001,
+      "step": 1603
+    },
+    {
+      "epoch": 0.8779419813902573,
+      "grad_norm": 0.10093405842781067,
+      "learning_rate": 3.631151196445887e-07,
+      "loss": 0.0051,
+      "step": 1604
+    },
+    {
+      "epoch": 0.8784893267651889,
+      "grad_norm": 0.0005839603254571557,
+      "learning_rate": 3.5990534108327926e-07,
+      "loss": 0.0,
+      "step": 1605
+    },
+    {
+      "epoch": 0.8790366721401204,
+      "grad_norm": 0.022238705307245255,
+      "learning_rate": 3.567092823726259e-07,
+      "loss": 0.0011,
+      "step": 1606
+    },
+    {
+      "epoch": 0.879584017515052,
+      "grad_norm": 0.14515052735805511,
+      "learning_rate": 3.5352695296274884e-07,
+      "loss": 0.0091,
+      "step": 1607
+    },
+    {
+      "epoch": 0.8801313628899836,
+      "grad_norm": 0.0035123212728649378,
+      "learning_rate": 3.5035836226317177e-07,
+      "loss": 0.0002,
+      "step": 1608
+    },
+    {
+      "epoch": 0.8806787082649151,
+      "grad_norm": 1.9213814735412598,
+      "learning_rate": 3.4720351964279863e-07,
+      "loss": 0.3315,
+      "step": 1609
+    },
+    {
+      "epoch": 0.8812260536398467,
+      "grad_norm": 0.0053222388960421085,
+      "learning_rate": 3.4406243442987765e-07,
+      "loss": 0.0003,
+      "step": 1610
+    },
+    {
+      "epoch": 0.8817733990147784,
+      "grad_norm": 0.0038527492433786392,
+      "learning_rate": 3.409351159119845e-07,
+      "loss": 0.0002,
+      "step": 1611
+    },
+    {
+      "epoch": 0.8823207443897099,
+      "grad_norm": 0.0030256537720561028,
+      "learning_rate": 3.3782157333598687e-07,
+      "loss": 0.0002,
+      "step": 1612
+    },
+    {
+      "epoch": 0.8828680897646415,
+      "grad_norm": 0.01982075721025467,
+      "learning_rate": 3.347218159080201e-07,
+      "loss": 0.001,
+      "step": 1613
+    },
+    {
+      "epoch": 0.8834154351395731,
+      "grad_norm": 0.6527382135391235,
+      "learning_rate": 3.3163585279345823e-07,
+      "loss": 0.0446,
+      "step": 1614
+    },
+    {
+      "epoch": 0.8839627805145046,
+      "grad_norm": 0.09229915589094162,
+      "learning_rate": 3.2856369311689174e-07,
+      "loss": 0.0065,
+      "step": 1615
+    },
+    {
+      "epoch": 0.8845101258894362,
+      "grad_norm": 0.045364126563072205,
+      "learning_rate": 3.2550534596209217e-07,
+      "loss": 0.002,
+      "step": 1616
+    },
+    {
+      "epoch": 0.8850574712643678,
+      "grad_norm": 4.158183574676514,
+      "learning_rate": 3.224608203719953e-07,
+      "loss": 0.1492,
+      "step": 1617
+    },
+    {
+      "epoch": 0.8856048166392994,
+      "grad_norm": 1.665842056274414,
+      "learning_rate": 3.1943012534866536e-07,
+      "loss": 0.2008,
+      "step": 1618
+    },
+    {
+      "epoch": 0.886152162014231,
+      "grad_norm": 2.252728223800659,
+      "learning_rate": 3.164132698532735e-07,
+      "loss": 0.3352,
+      "step": 1619
+    },
+    {
+      "epoch": 0.8866995073891626,
+      "grad_norm": 0.2482650727033615,
+      "learning_rate": 3.134102628060698e-07,
+      "loss": 0.0132,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8872468527640941,
+      "grad_norm": 0.015394785441458225,
+      "learning_rate": 3.1042111308636047e-07,
+      "loss": 0.0006,
+      "step": 1621
+    },
+    {
+      "epoch": 0.8877941981390257,
+      "grad_norm": 0.17973975837230682,
+      "learning_rate": 3.074458295324717e-07,
+      "loss": 0.011,
+      "step": 1622
+    },
+    {
+      "epoch": 0.8883415435139573,
+      "grad_norm": 0.006913541350513697,
+      "learning_rate": 3.0448442094173634e-07,
+      "loss": 0.0004,
+      "step": 1623
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 1.447867751121521,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.1321,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8894362342638205,
+      "grad_norm": 0.08996415138244629,
+      "learning_rate": 2.98603263633892e-07,
+      "loss": 0.0063,
+      "step": 1625
+    },
+    {
+      "epoch": 0.8899835796387521,
+      "grad_norm": 0.0666528046131134,
+      "learning_rate": 2.9568353230621185e-07,
+      "loss": 0.0035,
+      "step": 1626
+    },
+    {
+      "epoch": 0.8905309250136836,
+      "grad_norm": 0.01801203191280365,
+      "learning_rate": 2.9277771072049433e-07,
+      "loss": 0.0008,
+      "step": 1627
+    },
+    {
+      "epoch": 0.8910782703886152,
+      "grad_norm": 0.004558671731501818,
+      "learning_rate": 2.898858074686806e-07,
+      "loss": 0.0002,
+      "step": 1628
+    },
+    {
+      "epoch": 0.8916256157635468,
+      "grad_norm": 0.13645006716251373,
+      "learning_rate": 2.8700783110156507e-07,
+      "loss": 0.0044,
+      "step": 1629
+    },
+    {
+      "epoch": 0.8921729611384783,
+      "grad_norm": 2.16013503074646,
+      "learning_rate": 2.841437901287586e-07,
+      "loss": 0.4084,
+      "step": 1630
+    },
+    {
+      "epoch": 0.89272030651341,
+      "grad_norm": 0.23281130194664001,
+      "learning_rate": 2.812936930186688e-07,
+      "loss": 0.0129,
+      "step": 1631
+    },
+    {
+      "epoch": 0.8932676518883416,
+      "grad_norm": 0.05298805981874466,
+      "learning_rate": 2.784575481984747e-07,
+      "loss": 0.0028,
+      "step": 1632
+    },
+    {
+      "epoch": 0.8938149972632731,
+      "grad_norm": 1.8807384967803955,
+      "learning_rate": 2.756353640541021e-07,
+      "loss": 0.3986,
+      "step": 1633
+    },
+    {
+      "epoch": 0.8943623426382047,
+      "grad_norm": 0.4551723003387451,
+      "learning_rate": 2.728271489301937e-07,
+      "loss": 0.0514,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8949096880131363,
+      "grad_norm": 1.692553162574768,
+      "learning_rate": 2.700329111300937e-07,
+      "loss": 0.1173,
+      "step": 1635
+    },
+    {
+      "epoch": 0.8954570333880679,
+      "grad_norm": 0.0010673926444724202,
+      "learning_rate": 2.672526589158153e-07,
+      "loss": 0.0001,
+      "step": 1636
+    },
+    {
+      "epoch": 0.8960043787629994,
+      "grad_norm": 0.026626106351614,
+      "learning_rate": 2.644864005080183e-07,
+      "loss": 0.0013,
+      "step": 1637
+    },
+    {
+      "epoch": 0.896551724137931,
+      "grad_norm": 0.0034642897080630064,
+      "learning_rate": 2.617341440859883e-07,
+      "loss": 0.0002,
+      "step": 1638
+    },
+    {
+      "epoch": 0.8970990695128627,
+      "grad_norm": 4.151165962219238,
+      "learning_rate": 2.5899589778760614e-07,
+      "loss": 0.3504,
+      "step": 1639
+    },
+    {
+      "epoch": 0.8976464148877942,
+      "grad_norm": 1.1645855903625488,
+      "learning_rate": 2.5627166970933257e-07,
+      "loss": 0.0355,
+      "step": 1640
+    },
+    {
+      "epoch": 0.8981937602627258,
+      "grad_norm": 0.024651098996400833,
+      "learning_rate": 2.535614679061732e-07,
+      "loss": 0.0012,
+      "step": 1641
+    },
+    {
+      "epoch": 0.8987411056376574,
+      "grad_norm": 2.9519765377044678,
+      "learning_rate": 2.5086530039166615e-07,
+      "loss": 0.6242,
+      "step": 1642
+    },
+    {
+      "epoch": 0.8992884510125889,
+      "grad_norm": 0.0605325847864151,
+      "learning_rate": 2.4818317513784886e-07,
+      "loss": 0.0029,
+      "step": 1643
+    },
+    {
+      "epoch": 0.8998357963875205,
+      "grad_norm": 0.002007205504924059,
+      "learning_rate": 2.4551510007524035e-07,
+      "loss": 0.0001,
+      "step": 1644
+    },
+    {
+      "epoch": 0.9003831417624522,
+      "grad_norm": 4.2551140785217285,
+      "learning_rate": 2.428610830928152e-07,
+      "loss": 0.5177,
+      "step": 1645
+    },
+    {
+      "epoch": 0.9009304871373837,
+      "grad_norm": 0.015494787134230137,
+      "learning_rate": 2.402211320379838e-07,
+      "loss": 0.0008,
+      "step": 1646
+    },
+    {
+      "epoch": 0.9014778325123153,
+      "grad_norm": 0.02462865225970745,
+      "learning_rate": 2.3759525471656163e-07,
+      "loss": 0.0013,
+      "step": 1647
+    },
+    {
+      "epoch": 0.9020251778872469,
+      "grad_norm": 0.11775901913642883,
+      "learning_rate": 2.3498345889275465e-07,
+      "loss": 0.0072,
+      "step": 1648
+    },
+    {
+      "epoch": 0.9025725232621784,
+      "grad_norm": 0.009577536955475807,
+      "learning_rate": 2.3238575228913152e-07,
+      "loss": 0.0005,
+      "step": 1649
+    },
+    {
+      "epoch": 0.90311986863711,
+      "grad_norm": 0.25204992294311523,
+      "learning_rate": 2.2980214258660038e-07,
+      "loss": 0.0117,
+      "step": 1650
+    },
+    {
+      "epoch": 0.9036672140120416,
+      "grad_norm": 0.04809905216097832,
+      "learning_rate": 2.2723263742438938e-07,
+      "loss": 0.0021,
+      "step": 1651
+    },
+    {
+      "epoch": 0.9042145593869731,
+      "grad_norm": 0.001318107359111309,
+      "learning_rate": 2.2467724440002336e-07,
+      "loss": 0.0001,
+      "step": 1652
+    },
+    {
+      "epoch": 0.9047619047619048,
+      "grad_norm": 0.005981622263789177,
+      "learning_rate": 2.2213597106929608e-07,
+      "loss": 0.0003,
+      "step": 1653
+    },
+    {
+      "epoch": 0.9053092501368364,
+      "grad_norm": 2.8996164798736572,
+      "learning_rate": 2.1960882494625692e-07,
+      "loss": 0.7142,
+      "step": 1654
+    },
+    {
+      "epoch": 0.9058565955117679,
+      "grad_norm": 0.01610439084470272,
+      "learning_rate": 2.1709581350318089e-07,
+      "loss": 0.0008,
+      "step": 1655
+    },
+    {
+      "epoch": 0.9064039408866995,
+      "grad_norm": 0.02391280047595501,
+      "learning_rate": 2.1459694417055033e-07,
+      "loss": 0.0013,
+      "step": 1656
+    },
+    {
+      "epoch": 0.9069512862616311,
+      "grad_norm": 3.049898624420166,
+      "learning_rate": 2.1211222433703217e-07,
+      "loss": 0.1284,
+      "step": 1657
+    },
+    {
+      "epoch": 0.9074986316365626,
+      "grad_norm": 0.0016047470271587372,
+      "learning_rate": 2.0964166134945674e-07,
+      "loss": 0.0001,
+      "step": 1658
+    },
+    {
+      "epoch": 0.9080459770114943,
+      "grad_norm": 0.00969348568469286,
+      "learning_rate": 2.0718526251279346e-07,
+      "loss": 0.0005,
+      "step": 1659
+    },
+    {
+      "epoch": 0.9085933223864259,
+      "grad_norm": 0.007032374385744333,
+      "learning_rate": 2.0474303509013361e-07,
+      "loss": 0.0004,
+      "step": 1660
+    },
+    {
+      "epoch": 0.9091406677613574,
+      "grad_norm": 0.07196672260761261,
+      "learning_rate": 2.0231498630266467e-07,
+      "loss": 0.0036,
+      "step": 1661
+    },
+    {
+      "epoch": 0.909688013136289,
+      "grad_norm": 0.009290819987654686,
+      "learning_rate": 1.999011233296505e-07,
+      "loss": 0.0004,
+      "step": 1662
+    },
+    {
+      "epoch": 0.9102353585112206,
+      "grad_norm": 0.0019148091087117791,
+      "learning_rate": 1.9750145330841186e-07,
+      "loss": 0.0001,
+      "step": 1663
+    },
+    {
+      "epoch": 0.9107827038861521,
+      "grad_norm": 0.00228285463526845,
+      "learning_rate": 1.9511598333430194e-07,
+      "loss": 0.0001,
+      "step": 1664
+    },
+    {
+      "epoch": 0.9113300492610837,
+      "grad_norm": 0.0029851458966732025,
+      "learning_rate": 1.9274472046068805e-07,
+      "loss": 0.0002,
+      "step": 1665
+    },
+    {
+      "epoch": 0.9118773946360154,
+      "grad_norm": 0.14858005940914154,
+      "learning_rate": 1.9038767169893058e-07,
+      "loss": 0.0082,
+      "step": 1666
+    },
+    {
+      "epoch": 0.9124247400109469,
+      "grad_norm": 0.0019396482966840267,
+      "learning_rate": 1.8804484401836077e-07,
+      "loss": 0.0001,
+      "step": 1667
+    },
+    {
+      "epoch": 0.9129720853858785,
+      "grad_norm": 0.0032133522909134626,
+      "learning_rate": 1.857162443462601e-07,
+      "loss": 0.0002,
+      "step": 1668
+    },
+    {
+      "epoch": 0.9135194307608101,
+      "grad_norm": 0.012534908019006252,
+      "learning_rate": 1.834018795678427e-07,
+      "loss": 0.0006,
+      "step": 1669
+    },
+    {
+      "epoch": 0.9140667761357416,
+      "grad_norm": 0.0011830313596874475,
+      "learning_rate": 1.8110175652623075e-07,
+      "loss": 0.0001,
+      "step": 1670
+    },
+    {
+      "epoch": 0.9146141215106732,
+      "grad_norm": 1.939831256866455,
+      "learning_rate": 1.7881588202243782e-07,
+      "loss": 0.4046,
+      "step": 1671
+    },
+    {
+      "epoch": 0.9151614668856048,
+      "grad_norm": 0.04055565223097801,
+      "learning_rate": 1.7654426281534576e-07,
+      "loss": 0.002,
+      "step": 1672
+    },
+    {
+      "epoch": 0.9157088122605364,
+      "grad_norm": 0.02508675493299961,
+      "learning_rate": 1.7428690562169003e-07,
+      "loss": 0.0015,
+      "step": 1673
+    },
+    {
+      "epoch": 0.916256157635468,
+      "grad_norm": 0.008100957609713078,
+      "learning_rate": 1.7204381711603046e-07,
+      "loss": 0.0004,
+      "step": 1674
+    },
+    {
+      "epoch": 0.9168035030103996,
+      "grad_norm": 0.0611899308860302,
+      "learning_rate": 1.698150039307428e-07,
+      "loss": 0.0033,
+      "step": 1675
+    },
+    {
+      "epoch": 0.9173508483853311,
+      "grad_norm": 0.14495496451854706,
+      "learning_rate": 1.6760047265598933e-07,
+      "loss": 0.0075,
+      "step": 1676
+    },
+    {
+      "epoch": 0.9178981937602627,
+      "grad_norm": 0.16413192451000214,
+      "learning_rate": 1.6540022983970505e-07,
+      "loss": 0.0083,
+      "step": 1677
+    },
+    {
+      "epoch": 0.9184455391351943,
+      "grad_norm": 0.006359891500324011,
+      "learning_rate": 1.632142819875776e-07,
+      "loss": 0.0003,
+      "step": 1678
+    },
+    {
+      "epoch": 0.9189928845101258,
+      "grad_norm": 0.27122026681900024,
+      "learning_rate": 1.610426355630268e-07,
+      "loss": 0.0159,
+      "step": 1679
+    },
+    {
+      "epoch": 0.9195402298850575,
+      "grad_norm": 0.011510318145155907,
+      "learning_rate": 1.5888529698718347e-07,
+      "loss": 0.0005,
+      "step": 1680
+    },
+    {
+      "epoch": 0.9200875752599891,
+      "grad_norm": 0.7412816286087036,
+      "learning_rate": 1.5674227263887732e-07,
+      "loss": 0.0462,
+      "step": 1681
+    },
+    {
+      "epoch": 0.9206349206349206,
+      "grad_norm": 2.5012526512145996,
+      "learning_rate": 1.5461356885461077e-07,
+      "loss": 0.2889,
+      "step": 1682
+    },
+    {
+      "epoch": 0.9211822660098522,
+      "grad_norm": 2.3169915676116943,
+      "learning_rate": 1.524991919285429e-07,
+      "loss": 0.2376,
+      "step": 1683
+    },
+    {
+      "epoch": 0.9217296113847838,
+      "grad_norm": 0.22899430990219116,
+      "learning_rate": 1.503991481124728e-07,
+      "loss": 0.0108,
+      "step": 1684
+    },
+    {
+      "epoch": 0.9222769567597153,
+      "grad_norm": 0.017632320523262024,
+      "learning_rate": 1.48313443615819e-07,
+      "loss": 0.0008,
+      "step": 1685
+    },
+    {
+      "epoch": 0.922824302134647,
+      "grad_norm": 0.76988285779953,
+      "learning_rate": 1.4624208460559897e-07,
+      "loss": 0.0616,
+      "step": 1686
+    },
+    {
+      "epoch": 0.9233716475095786,
+      "grad_norm": 0.002816552296280861,
+      "learning_rate": 1.4418507720641794e-07,
+      "loss": 0.0001,
+      "step": 1687
+    },
+    {
+      "epoch": 0.9239189928845101,
+      "grad_norm": 0.04683569818735123,
+      "learning_rate": 1.4214242750044238e-07,
+      "loss": 0.0021,
+      "step": 1688
+    },
+    {
+      "epoch": 0.9244663382594417,
+      "grad_norm": 0.009630477987229824,
+      "learning_rate": 1.401141415273871e-07,
+      "loss": 0.0005,
+      "step": 1689
+    },
+    {
+      "epoch": 0.9250136836343733,
+      "grad_norm": 0.005564996041357517,
+      "learning_rate": 1.3810022528449597e-07,
+      "loss": 0.0003,
+      "step": 1690
+    },
+    {
+      "epoch": 0.9255610290093049,
+      "grad_norm": 0.035956043750047684,
+      "learning_rate": 1.3610068472652615e-07,
+      "loss": 0.0018,
+      "step": 1691
+    },
+    {
+      "epoch": 0.9261083743842364,
+      "grad_norm": 2.302034854888916,
+      "learning_rate": 1.3411552576572562e-07,
+      "loss": 0.3058,
+      "step": 1692
+    },
+    {
+      "epoch": 0.926655719759168,
+      "grad_norm": 0.0033738012425601482,
+      "learning_rate": 1.3214475427182182e-07,
+      "loss": 0.0002,
+      "step": 1693
+    },
+    {
+      "epoch": 0.9272030651340997,
+      "grad_norm": 2.1513004302978516,
+      "learning_rate": 1.3018837607199909e-07,
+      "loss": 0.3423,
+      "step": 1694
+    },
+    {
+      "epoch": 0.9277504105090312,
+      "grad_norm": 0.06418784707784653,
+      "learning_rate": 1.2824639695088403e-07,
+      "loss": 0.0028,
+      "step": 1695
+    },
+    {
+      "epoch": 0.9282977558839628,
+      "grad_norm": 0.0017216246342286468,
+      "learning_rate": 1.2631882265052908e-07,
+      "loss": 0.0001,
+      "step": 1696
+    },
+    {
+      "epoch": 0.9288451012588944,
+      "grad_norm": 0.025960400700569153,
+      "learning_rate": 1.2440565887039347e-07,
+      "loss": 0.0011,
+      "step": 1697
+    },
+    {
+      "epoch": 0.9293924466338259,
+      "grad_norm": 0.027499064803123474,
+      "learning_rate": 1.2250691126732772e-07,
+      "loss": 0.0012,
+      "step": 1698
+    },
+    {
+      "epoch": 0.9299397920087575,
+      "grad_norm": 0.011645481921732426,
+      "learning_rate": 1.2062258545555649e-07,
+      "loss": 0.0007,
+      "step": 1699
+    },
+    {
+      "epoch": 0.9304871373836892,
+      "grad_norm": 2.105513334274292,
+      "learning_rate": 1.1875268700666187e-07,
+      "loss": 0.1763,
+      "step": 1700
+    },
+    {
+      "epoch": 0.9310344827586207,
+      "grad_norm": 0.10536600649356842,
+      "learning_rate": 1.1689722144956672e-07,
+      "loss": 0.0074,
+      "step": 1701
+    },
+    {
+      "epoch": 0.9315818281335523,
+      "grad_norm": 0.02680140547454357,
+      "learning_rate": 1.1505619427051973e-07,
+      "loss": 0.0012,
+      "step": 1702
+    },
+    {
+      "epoch": 0.9321291735084839,
+      "grad_norm": 0.010312930680811405,
+      "learning_rate": 1.1322961091307705e-07,
+      "loss": 0.0006,
+      "step": 1703
+    },
+    {
+      "epoch": 0.9326765188834154,
+      "grad_norm": 0.03216090425848961,
+      "learning_rate": 1.1141747677808845e-07,
+      "loss": 0.0015,
+      "step": 1704
+    },
+    {
+      "epoch": 0.933223864258347,
+      "grad_norm": 0.0016830548411235213,
+      "learning_rate": 1.0961979722367789e-07,
+      "loss": 0.0001,
+      "step": 1705
+    },
+    {
+      "epoch": 0.9337712096332786,
+      "grad_norm": 0.003885602578520775,
+      "learning_rate": 1.0783657756523347e-07,
+      "loss": 0.0002,
+      "step": 1706
+    },
+    {
+      "epoch": 0.9343185550082101,
+      "grad_norm": 0.004555482417345047,
+      "learning_rate": 1.0606782307538532e-07,
+      "loss": 0.0002,
+      "step": 1707
+    },
+    {
+      "epoch": 0.9348659003831418,
+      "grad_norm": 0.0035665256436914206,
+      "learning_rate": 1.0431353898399388e-07,
+      "loss": 0.0002,
+      "step": 1708
+    },
+    {
+      "epoch": 0.9354132457580734,
+      "grad_norm": 0.0027156653814017773,
+      "learning_rate": 1.0257373047813324e-07,
+      "loss": 0.0001,
+      "step": 1709
+    },
+    {
+      "epoch": 0.9359605911330049,
+      "grad_norm": 0.007819607853889465,
+      "learning_rate": 1.008484027020773e-07,
+      "loss": 0.0004,
+      "step": 1710
+    },
+    {
+      "epoch": 0.9365079365079365,
+      "grad_norm": 0.10022959858179092,
+      "learning_rate": 9.913756075728088e-08,
+      "loss": 0.0063,
+      "step": 1711
+    },
+    {
+      "epoch": 0.9370552818828681,
+      "grad_norm": 0.0077483695931732655,
+      "learning_rate": 9.744120970236914e-08,
+      "loss": 0.0003,
+      "step": 1712
+    },
+    {
+      "epoch": 0.9376026272577996,
+      "grad_norm": 1.458129644393921,
+      "learning_rate": 9.575935455311935e-08,
+      "loss": 0.1238,
+      "step": 1713
+    },
+    {
+      "epoch": 0.9381499726327313,
+      "grad_norm": 0.000479287700727582,
+      "learning_rate": 9.409200028244803e-08,
+      "loss": 0.0001,
+      "step": 1714
+    },
+    {
+      "epoch": 0.9386973180076629,
+      "grad_norm": 0.0025392011739313602,
+      "learning_rate": 9.243915182039431e-08,
+      "loss": 0.0001,
+      "step": 1715
+    },
+    {
+      "epoch": 0.9392446633825944,
+      "grad_norm": 1.5274864435195923,
+      "learning_rate": 9.08008140541089e-08,
+      "loss": 0.1416,
+      "step": 1716
+    },
+    {
+      "epoch": 0.939792008757526,
+      "grad_norm": 2.8007185459136963,
+      "learning_rate": 8.917699182783346e-08,
+      "loss": 0.5872,
+      "step": 1717
+    },
+    {
+      "epoch": 0.9403393541324576,
+      "grad_norm": 0.030282529070973396,
+      "learning_rate": 8.756768994289289e-08,
+      "loss": 0.0015,
+      "step": 1718
+    },
+    {
+      "epoch": 0.9408866995073891,
+      "grad_norm": 0.013260996900498867,
+      "learning_rate": 8.597291315767808e-08,
+      "loss": 0.0007,
+      "step": 1719
+    },
+    {
+      "epoch": 0.9414340448823207,
+      "grad_norm": 0.27540886402130127,
+      "learning_rate": 8.439266618763098e-08,
+      "loss": 0.0146,
+      "step": 1720
+    },
+    {
+      "epoch": 0.9419813902572524,
+      "grad_norm": 0.20886638760566711,
+      "learning_rate": 8.282695370523175e-08,
+      "loss": 0.0109,
+      "step": 1721
+    },
+    {
+      "epoch": 0.9425287356321839,
+      "grad_norm": 0.006277718581259251,
+      "learning_rate": 8.127578033998663e-08,
+      "loss": 0.0003,
+      "step": 1722
+    },
+    {
+      "epoch": 0.9430760810071155,
+      "grad_norm": 0.31388920545578003,
+      "learning_rate": 7.973915067840954e-08,
+      "loss": 0.0189,
+      "step": 1723
+    },
+    {
+      "epoch": 0.9436234263820471,
+      "grad_norm": 0.028467781841754913,
+      "learning_rate": 7.821706926401496e-08,
+      "loss": 0.0014,
+      "step": 1724
+    },
+    {
+      "epoch": 0.9441707717569786,
+      "grad_norm": 0.10319659858942032,
+      "learning_rate": 7.670954059729896e-08,
+      "loss": 0.0057,
+      "step": 1725
+    },
+    {
+      "epoch": 0.9447181171319102,
+      "grad_norm": 1.4109545946121216,
+      "learning_rate": 7.521656913572817e-08,
+      "loss": 0.242,
+      "step": 1726
+    },
+    {
+      "epoch": 0.9452654625068418,
+      "grad_norm": 2.2406203746795654,
+      "learning_rate": 7.373815929372586e-08,
+      "loss": 0.2529,
+      "step": 1727
+    },
+    {
+      "epoch": 0.9458128078817734,
+      "grad_norm": 3.631880760192871,
+      "learning_rate": 7.227431544266194e-08,
+      "loss": 0.139,
+      "step": 1728
+    },
+    {
+      "epoch": 0.946360153256705,
+      "grad_norm": 0.0025556655600667,
+      "learning_rate": 7.082504191083417e-08,
+      "loss": 0.0002,
+      "step": 1729
+    },
+    {
+      "epoch": 0.9469074986316366,
+      "grad_norm": 0.007169991731643677,
+      "learning_rate": 6.939034298346192e-08,
+      "loss": 0.0003,
+      "step": 1730
+    },
+    {
+      "epoch": 0.9474548440065681,
+      "grad_norm": 0.005089812446385622,
+      "learning_rate": 6.797022290266741e-08,
+      "loss": 0.0002,
+      "step": 1731
+    },
+    {
+      "epoch": 0.9480021893814997,
+      "grad_norm": 0.04695936292409897,
+      "learning_rate": 6.656468586746789e-08,
+      "loss": 0.002,
+      "step": 1732
+    },
+    {
+      "epoch": 0.9485495347564313,
+      "grad_norm": 0.23610833287239075,
+      "learning_rate": 6.517373603376176e-08,
+      "loss": 0.0122,
+      "step": 1733
+    },
+    {
+      "epoch": 0.9490968801313628,
+      "grad_norm": 0.13638249039649963,
+      "learning_rate": 6.379737751431415e-08,
+      "loss": 0.0083,
+      "step": 1734
+    },
+    {
+      "epoch": 0.9496442255062945,
+      "grad_norm": 0.9663567543029785,
+      "learning_rate": 6.243561437874745e-08,
+      "loss": 0.0553,
+      "step": 1735
+    },
+    {
+      "epoch": 0.9501915708812261,
+      "grad_norm": 0.03125873580574989,
+      "learning_rate": 6.108845065352864e-08,
+      "loss": 0.0015,
+      "step": 1736
+    },
+    {
+      "epoch": 0.9507389162561576,
+      "grad_norm": 1.2737540006637573,
+      "learning_rate": 5.97558903219575e-08,
+      "loss": 0.0837,
+      "step": 1737
+    },
+    {
+      "epoch": 0.9512862616310892,
+      "grad_norm": 0.003754893783479929,
+      "learning_rate": 5.843793732415282e-08,
+      "loss": 0.0002,
+      "step": 1738
+    },
+    {
+      "epoch": 0.9518336070060208,
+      "grad_norm": 0.014718581922352314,
+      "learning_rate": 5.713459555704404e-08,
+      "loss": 0.0007,
+      "step": 1739
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 2.769636869430542,
+      "learning_rate": 5.584586887435739e-08,
+      "loss": 0.2708,
+      "step": 1740
+    },
+    {
+      "epoch": 0.952928297755884,
+      "grad_norm": 2.3508386611938477,
+      "learning_rate": 5.457176108660478e-08,
+      "loss": 0.5224,
+      "step": 1741
+    },
+    {
+      "epoch": 0.9534756431308156,
+      "grad_norm": 0.01864919811487198,
+      "learning_rate": 5.331227596107325e-08,
+      "loss": 0.0009,
+      "step": 1742
+    },
+    {
+      "epoch": 0.9540229885057471,
+      "grad_norm": 0.022848375141620636,
+      "learning_rate": 5.206741722181385e-08,
+      "loss": 0.0012,
+      "step": 1743
+    },
+    {
+      "epoch": 0.9545703338806787,
+      "grad_norm": 0.8749512434005737,
+      "learning_rate": 5.0837188549628934e-08,
+      "loss": 0.036,
+      "step": 1744
+    },
+    {
+      "epoch": 0.9551176792556103,
+      "grad_norm": 0.026627959683537483,
+      "learning_rate": 4.9621593582065416e-08,
+      "loss": 0.0015,
+      "step": 1745
+    },
+    {
+      "epoch": 0.9556650246305419,
+      "grad_norm": 1.6593846082687378,
+      "learning_rate": 4.842063591339763e-08,
+      "loss": 0.3074,
+      "step": 1746
+    },
+    {
+      "epoch": 0.9562123700054734,
+      "grad_norm": 1.2600449323654175,
+      "learning_rate": 4.723431909462339e-08,
+      "loss": 0.059,
+      "step": 1747
+    },
+    {
+      "epoch": 0.956759715380405,
+      "grad_norm": 0.00087725929915905,
+      "learning_rate": 4.606264663344851e-08,
+      "loss": 0.0001,
+      "step": 1748
+    },
+    {
+      "epoch": 0.9573070607553367,
+      "grad_norm": 0.08448724448680878,
+      "learning_rate": 4.490562199427839e-08,
+      "loss": 0.0044,
+      "step": 1749
+    },
+    {
+      "epoch": 0.9578544061302682,
+      "grad_norm": 2.0202221870422363,
+      "learning_rate": 4.376324859820924e-08,
+      "loss": 0.2994,
+      "step": 1750
+    },
+    {
+      "epoch": 0.9584017515051998,
+      "grad_norm": 2.787095785140991,
+      "learning_rate": 4.2635529823014664e-08,
+      "loss": 0.5762,
+      "step": 1751
+    },
+    {
+      "epoch": 0.9589490968801314,
+      "grad_norm": 0.17930848896503448,
+      "learning_rate": 4.1522469003137946e-08,
+      "loss": 0.0087,
+      "step": 1752
+    },
+    {
+      "epoch": 0.9594964422550629,
+      "grad_norm": 1.5180494785308838,
+      "learning_rate": 4.0424069429682024e-08,
+      "loss": 0.3157,
+      "step": 1753
+    },
+    {
+      "epoch": 0.9600437876299945,
+      "grad_norm": 0.0052284374833106995,
+      "learning_rate": 3.9340334350399525e-08,
+      "loss": 0.0002,
+      "step": 1754
+    },
+    {
+      "epoch": 0.9605911330049262,
+      "grad_norm": 0.0017134948866441846,
+      "learning_rate": 3.82712669696822e-08,
+      "loss": 0.0001,
+      "step": 1755
+    },
+    {
+      "epoch": 0.9611384783798577,
+      "grad_norm": 2.9423959255218506,
+      "learning_rate": 3.721687044855315e-08,
+      "loss": 0.1727,
+      "step": 1756
+    },
+    {
+      "epoch": 0.9616858237547893,
+      "grad_norm": 0.25030413269996643,
+      "learning_rate": 3.617714790465576e-08,
+      "loss": 0.0137,
+      "step": 1757
+    },
+    {
+      "epoch": 0.9622331691297209,
+      "grad_norm": 2.802866220474243,
+      "learning_rate": 3.515210241224698e-08,
+      "loss": 0.4985,
+      "step": 1758
+    },
+    {
+      "epoch": 0.9627805145046524,
+      "grad_norm": 0.20648542046546936,
+      "learning_rate": 3.4141737002184036e-08,
+      "loss": 0.011,
+      "step": 1759
+    },
+    {
+      "epoch": 0.963327859879584,
+      "grad_norm": 0.24056246876716614,
+      "learning_rate": 3.3146054661920556e-08,
+      "loss": 0.0174,
+      "step": 1760
+    },
+    {
+      "epoch": 0.9638752052545156,
+      "grad_norm": 0.2007961869239807,
+      "learning_rate": 3.216505833549377e-08,
+      "loss": 0.0112,
+      "step": 1761
+    },
+    {
+      "epoch": 0.9644225506294472,
+      "grad_norm": 0.24780268967151642,
+      "learning_rate": 3.1198750923517316e-08,
+      "loss": 0.0167,
+      "step": 1762
+    },
+    {
+      "epoch": 0.9649698960043788,
+      "grad_norm": 0.002492027822881937,
+      "learning_rate": 3.0247135283172914e-08,
+      "loss": 0.0001,
+      "step": 1763
+    },
+    {
+      "epoch": 0.9655172413793104,
+      "grad_norm": 0.00669819675385952,
+      "learning_rate": 2.9310214228202016e-08,
+      "loss": 0.0003,
+      "step": 1764
+    },
+    {
+      "epoch": 0.9660645867542419,
+      "grad_norm": 0.03622516244649887,
+      "learning_rate": 2.8387990528896404e-08,
+      "loss": 0.0023,
+      "step": 1765
+    },
+    {
+      "epoch": 0.9666119321291735,
+      "grad_norm": 0.022351013496518135,
+      "learning_rate": 2.7480466912090386e-08,
+      "loss": 0.0012,
+      "step": 1766
+    },
+    {
+      "epoch": 0.9671592775041051,
+      "grad_norm": 0.011430719867348671,
+      "learning_rate": 2.6587646061153604e-08,
+      "loss": 0.0005,
+      "step": 1767
+    },
+    {
+      "epoch": 0.9677066228790366,
+      "grad_norm": 0.03099585883319378,
+      "learning_rate": 2.5709530615983246e-08,
+      "loss": 0.0014,
+      "step": 1768
+    },
+    {
+      "epoch": 0.9682539682539683,
+      "grad_norm": 0.044385433197021484,
+      "learning_rate": 2.4846123172992953e-08,
+      "loss": 0.0021,
+      "step": 1769
+    },
+    {
+      "epoch": 0.9688013136288999,
+      "grad_norm": 0.0769612118601799,
+      "learning_rate": 2.3997426285110592e-08,
+      "loss": 0.0036,
+      "step": 1770
+    },
+    {
+      "epoch": 0.9693486590038314,
+      "grad_norm": 0.007691814098507166,
+      "learning_rate": 2.3163442461766604e-08,
+      "loss": 0.0004,
+      "step": 1771
+    },
+    {
+      "epoch": 0.969896004378763,
+      "grad_norm": 0.03668086975812912,
+      "learning_rate": 2.2344174168887346e-08,
+      "loss": 0.0018,
+      "step": 1772
+    },
+    {
+      "epoch": 0.9704433497536946,
+      "grad_norm": 0.03234627842903137,
+      "learning_rate": 2.153962382888841e-08,
+      "loss": 0.0016,
+      "step": 1773
+    },
+    {
+      "epoch": 0.9709906951286261,
+      "grad_norm": 1.0921375751495361,
+      "learning_rate": 2.0749793820667995e-08,
+      "loss": 0.2131,
+      "step": 1774
+    },
+    {
+      "epoch": 0.9715380405035577,
+      "grad_norm": 0.0049126651138067245,
+      "learning_rate": 1.9974686479597993e-08,
+      "loss": 0.0003,
+      "step": 1775
+    },
+    {
+      "epoch": 0.9720853858784894,
+      "grad_norm": 0.030328121036291122,
+      "learning_rate": 1.921430409752012e-08,
+      "loss": 0.0015,
+      "step": 1776
+    },
+    {
+      "epoch": 0.9726327312534209,
+      "grad_norm": 0.010860969312489033,
+      "learning_rate": 1.846864892273481e-08,
+      "loss": 0.0005,
+      "step": 1777
+    },
+    {
+      "epoch": 0.9731800766283525,
+      "grad_norm": 0.06575039029121399,
+      "learning_rate": 1.7737723159999e-08,
+      "loss": 0.0038,
+      "step": 1778
+    },
+    {
+      "epoch": 0.9737274220032841,
+      "grad_norm": 0.08788035809993744,
+      "learning_rate": 1.702152897051612e-08,
+      "loss": 0.0051,
+      "step": 1779
+    },
+    {
+      "epoch": 0.9742747673782156,
+      "grad_norm": 0.0011737667955458164,
+      "learning_rate": 1.632006847193335e-08,
+      "loss": 0.0001,
+      "step": 1780
+    },
+    {
+      "epoch": 0.9748221127531472,
+      "grad_norm": 0.009627328254282475,
+      "learning_rate": 1.563334373833103e-08,
+      "loss": 0.0004,
+      "step": 1781
+    },
+    {
+      "epoch": 0.9753694581280788,
+      "grad_norm": 0.17805936932563782,
+      "learning_rate": 1.496135680021993e-08,
+      "loss": 0.0106,
+      "step": 1782
+    },
+    {
+      "epoch": 0.9759168035030104,
+      "grad_norm": 0.009691783227026463,
+      "learning_rate": 1.4304109644533438e-08,
+      "loss": 0.0004,
+      "step": 1783
+    },
+    {
+      "epoch": 0.976464148877942,
+      "grad_norm": 0.001948626129887998,
+      "learning_rate": 1.3661604214623147e-08,
+      "loss": 0.0001,
+      "step": 1784
+    },
+    {
+      "epoch": 0.9770114942528736,
+      "grad_norm": 0.0033026135060936213,
+      "learning_rate": 1.3033842410251074e-08,
+      "loss": 0.0002,
+      "step": 1785
+    },
+    {
+      "epoch": 0.9775588396278051,
+      "grad_norm": 1.8588638305664062,
+      "learning_rate": 1.2420826087586324e-08,
+      "loss": 0.4028,
+      "step": 1786
+    },
+    {
+      "epoch": 0.9781061850027367,
+      "grad_norm": 0.002413011621683836,
+      "learning_rate": 1.182255705919788e-08,
+      "loss": 0.0001,
+      "step": 1787
+    },
+    {
+      "epoch": 0.9786535303776683,
+      "grad_norm": 0.0020728744566440582,
+      "learning_rate": 1.123903709404961e-08,
+      "loss": 0.0001,
+      "step": 1788
+    },
+    {
+      "epoch": 0.9792008757525998,
+      "grad_norm": 1.5994459390640259,
+      "learning_rate": 1.0670267917496923e-08,
+      "loss": 0.3538,
+      "step": 1789
+    },
+    {
+      "epoch": 0.9797482211275315,
+      "grad_norm": 0.06763313710689545,
+      "learning_rate": 1.011625121127735e-08,
+      "loss": 0.004,
+      "step": 1790
+    },
+    {
+      "epoch": 0.9802955665024631,
+      "grad_norm": 0.07340343296527863,
+      "learning_rate": 9.576988613511084e-09,
+      "loss": 0.0041,
+      "step": 1791
+    },
+    {
+      "epoch": 0.9808429118773946,
+      "grad_norm": 0.004961518570780754,
+      "learning_rate": 9.052481718690998e-09,
+      "loss": 0.0002,
+      "step": 1792
+    },
+    {
+      "epoch": 0.9813902572523262,
+      "grad_norm": 0.3102180063724518,
+      "learning_rate": 8.542732077680971e-09,
+      "loss": 0.0071,
+      "step": 1793
+    },
+    {
+      "epoch": 0.9819376026272578,
+      "grad_norm": 0.002447548322379589,
+      "learning_rate": 8.04774119771201e-09,
+      "loss": 0.0001,
+      "step": 1794
+    },
+    {
+      "epoch": 0.9824849480021893,
+      "grad_norm": 0.003043276723474264,
+      "learning_rate": 7.567510542373923e-09,
+      "loss": 0.0002,
+      "step": 1795
+    },
+    {
+      "epoch": 0.983032293377121,
+      "grad_norm": 0.10866506397724152,
+      "learning_rate": 7.102041531615867e-09,
+      "loss": 0.0067,
+      "step": 1796
+    },
+    {
+      "epoch": 0.9835796387520526,
+      "grad_norm": 0.11161591857671738,
+      "learning_rate": 6.65133554173747e-09,
+      "loss": 0.0061,
+      "step": 1797
+    },
+    {
+      "epoch": 0.9841269841269841,
+      "grad_norm": 0.057195790112018585,
+      "learning_rate": 6.215393905388278e-09,
+      "loss": 0.0026,
+      "step": 1798
+    },
+    {
+      "epoch": 0.9846743295019157,
+      "grad_norm": 0.009329413995146751,
+      "learning_rate": 5.794217911562205e-09,
+      "loss": 0.0005,
+      "step": 1799
+    },
+    {
+      "epoch": 0.9852216748768473,
+      "grad_norm": 0.030702760443091393,
+      "learning_rate": 5.387808805594752e-09,
+      "loss": 0.0012,
+      "step": 1800
+    },
+    {
+      "epoch": 0.9857690202517789,
+      "grad_norm": 1.5946861505508423,
+      "learning_rate": 4.996167789157457e-09,
+      "loss": 0.0419,
+      "step": 1801
+    },
+    {
+      "epoch": 0.9863163656267104,
+      "grad_norm": 2.047494649887085,
+      "learning_rate": 4.619296020256236e-09,
+      "loss": 0.4583,
+      "step": 1802
+    },
+    {
+      "epoch": 0.986863711001642,
+      "grad_norm": 0.001278059440664947,
+      "learning_rate": 4.257194613226379e-09,
+      "loss": 0.0001,
+      "step": 1803
+    },
+    {
+      "epoch": 0.9874110563765737,
+      "grad_norm": 0.005214928183704615,
+      "learning_rate": 3.9098646387319974e-09,
+      "loss": 0.0002,
+      "step": 1804
+    },
+    {
+      "epoch": 0.9879584017515052,
+      "grad_norm": 0.08367697894573212,
+      "learning_rate": 3.577307123759366e-09,
+      "loss": 0.005,
+      "step": 1805
+    },
+    {
+      "epoch": 0.9885057471264368,
+      "grad_norm": 0.03230416402220726,
+      "learning_rate": 3.2595230516152543e-09,
+      "loss": 0.0014,
+      "step": 1806
+    },
+    {
+      "epoch": 0.9890530925013684,
+      "grad_norm": 2.5367720127105713,
+      "learning_rate": 2.956513361925262e-09,
+      "loss": 0.2684,
+      "step": 1807
+    },
+    {
+      "epoch": 0.9896004378762999,
+      "grad_norm": 0.4515298008918762,
+      "learning_rate": 2.6682789506299322e-09,
+      "loss": 0.024,
+      "step": 1808
+    },
+    {
+      "epoch": 0.9901477832512315,
+      "grad_norm": 0.0051987855695188046,
+      "learning_rate": 2.3948206699819787e-09,
+      "loss": 0.0002,
+      "step": 1809
+    },
+    {
+      "epoch": 0.9906951286261632,
+      "grad_norm": 0.011141418479382992,
+      "learning_rate": 2.136139328543507e-09,
+      "loss": 0.0005,
+      "step": 1810
+    },
+    {
+      "epoch": 0.9912424740010947,
+      "grad_norm": 0.003159766783937812,
+      "learning_rate": 1.892235691184907e-09,
+      "loss": 0.0002,
+      "step": 1811
+    },
+    {
+      "epoch": 0.9917898193760263,
+      "grad_norm": 0.0010058737825602293,
+      "learning_rate": 1.6631104790809648e-09,
+      "loss": 0.0001,
+      "step": 1812
+    },
+    {
+      "epoch": 0.9923371647509579,
+      "grad_norm": 0.14787787199020386,
+      "learning_rate": 1.4487643697103092e-09,
+      "loss": 0.0097,
+      "step": 1813
+    },
+    {
+      "epoch": 0.9928845101258894,
+      "grad_norm": 0.05396431311964989,
+      "learning_rate": 1.2491979968526358e-09,
+      "loss": 0.0028,
+      "step": 1814
+    },
+    {
+      "epoch": 0.993431855500821,
+      "grad_norm": 0.008193233981728554,
+      "learning_rate": 1.0644119505864858e-09,
+      "loss": 0.0004,
+      "step": 1815
+    },
+    {
+      "epoch": 0.9939792008757526,
+      "grad_norm": 0.001475210883654654,
+      "learning_rate": 8.944067772881371e-10,
+      "loss": 0.0001,
+      "step": 1816
+    },
+    {
+      "epoch": 0.9945265462506842,
+      "grad_norm": 0.0464748851954937,
+      "learning_rate": 7.391829796288275e-10,
+      "loss": 0.0026,
+      "step": 1817
+    },
+    {
+      "epoch": 0.9950738916256158,
+      "grad_norm": 0.0010864358628168702,
+      "learning_rate": 5.987410165758656e-10,
+      "loss": 0.0001,
+      "step": 1818
+    },
+    {
+      "epoch": 0.9956212370005474,
+      "grad_norm": 0.10535825043916702,
+      "learning_rate": 4.730813033881898e-10,
+      "loss": 0.0065,
+      "step": 1819
+    },
+    {
+      "epoch": 0.9961685823754789,
+      "grad_norm": 0.10840434581041336,
+      "learning_rate": 3.6220421161692333e-10,
+      "loss": 0.0041,
+      "step": 1820
+    },
+    {
+      "epoch": 0.9967159277504105,
+      "grad_norm": 0.057124633342027664,
+      "learning_rate": 2.6611006910370884e-10,
+      "loss": 0.0028,
+      "step": 1821
+    },
+    {
+      "epoch": 0.9972632731253421,
+      "grad_norm": 0.3710714280605316,
+      "learning_rate": 1.847991599801535e-10,
+      "loss": 0.0237,
+      "step": 1822
+    },
+    {
+      "epoch": 0.9978106185002736,
+      "grad_norm": 0.09165342897176743,
+      "learning_rate": 1.1827172466727376e-10,
+      "loss": 0.0048,
+      "step": 1823
+    },
+    {
+      "epoch": 0.9983579638752053,
+      "grad_norm": 0.01384245976805687,
+      "learning_rate": 6.652795987271975e-11,
+      "loss": 0.0007,
+      "step": 1824
+    },
+    {
+      "epoch": 0.9989053092501369,
+      "grad_norm": 0.0028413371182978153,
+      "learning_rate": 2.9568018593550965e-11,
+      "loss": 0.0002,
+      "step": 1825
+    },
+    {
+      "epoch": 0.9994526546250684,
+      "grad_norm": 0.20825180411338806,
+      "learning_rate": 7.392010112350356e-12,
+      "loss": 0.0144,
+      "step": 1826
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.01608341373503208,
+      "learning_rate": 0.0,
+      "loss": 0.0008,
+      "step": 1827
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.27450980392156865,
+      "eval_loss": 0.03834722936153412,
+      "eval_runtime": 1062.6197,
+      "eval_samples_per_second": 0.192,
+      "eval_steps_per_second": 0.192,
+      "step": 1827
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1827,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.437810384269312e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}