diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7730 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9995456610631531,
+  "eval_steps": 500,
+  "global_step": 1100,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0009086778736937755,
+      "grad_norm": 4.391181014616113,
+      "learning_rate": 0.0,
+      "loss": -0.1479,
+      "step": 1
+    },
+    {
+      "epoch": 0.001817355747387551,
+      "grad_norm": 4.065568909195848,
+      "learning_rate": 3.964797263411211e-07,
+      "loss": -0.1677,
+      "step": 2
+    },
+    {
+      "epoch": 0.0027260336210813267,
+      "grad_norm": 3.727827261381386,
+      "learning_rate": 6.28405498546863e-07,
+      "loss": -0.2008,
+      "step": 3
+    },
+    {
+      "epoch": 0.003634711494775102,
+      "grad_norm": 5.186434716837906,
+      "learning_rate": 7.929594526822422e-07,
+      "loss": -0.2258,
+      "step": 4
+    },
+    {
+      "epoch": 0.004543389368468878,
+      "grad_norm": 3.4872309864709177,
+      "learning_rate": 9.205974156447019e-07,
+      "loss": -0.1794,
+      "step": 5
+    },
+    {
+      "epoch": 0.005452067242162653,
+      "grad_norm": 4.674909033884794,
+      "learning_rate": 1.024885224887984e-06,
+      "loss": -0.1939,
+      "step": 6
+    },
+    {
+      "epoch": 0.006360745115856429,
+      "grad_norm": 4.089473674830758,
+      "learning_rate": 1.113059311239798e-06,
+      "loss": -0.0352,
+      "step": 7
+    },
+    {
+      "epoch": 0.007269422989550204,
+      "grad_norm": 4.933823194325695,
+      "learning_rate": 1.1894391790233632e-06,
+      "loss": -0.2375,
+      "step": 8
+    },
+    {
+      "epoch": 0.00817810086324398,
+      "grad_norm": 3.616340818243807,
+      "learning_rate": 1.256810997093726e-06,
+      "loss": -0.2643,
+      "step": 9
+    },
+    {
+      "epoch": 0.009086778736937756,
+      "grad_norm": 3.971857251263452,
+      "learning_rate": 1.3170771419858234e-06,
+      "loss": -0.1523,
+      "step": 10
+    },
+    {
+      "epoch": 0.009995456610631531,
+      "grad_norm": 3.347238734948954,
+      "learning_rate": 1.3715945014531372e-06,
+      "loss": -0.2438,
+      "step": 11
+    },
+    {
+      "epoch": 0.010904134484325307,
+      "grad_norm": 3.3600717221280023,
+      "learning_rate": 1.4213649512291051e-06,
+      "loss": -0.1756,
+      "step": 12
+    },
+    {
+      "epoch": 0.011812812358019082,
+      "grad_norm": 3.3676310023687623,
+      "learning_rate": 1.4671493267903953e-06,
+      "loss": -0.1479,
+      "step": 13
+    },
+    {
+      "epoch": 0.012721490231712857,
+      "grad_norm": 4.336527356415436,
+      "learning_rate": 1.509539037580919e-06,
+      "loss": -0.1993,
+      "step": 14
+    },
+    {
+      "epoch": 0.013630168105406633,
+      "grad_norm": 5.037825558273214,
+      "learning_rate": 1.549002914191565e-06,
+      "loss": -0.2917,
+      "step": 15
+    },
+    {
+      "epoch": 0.014538845979100408,
+      "grad_norm": 4.239590631552123,
+      "learning_rate": 1.5859189053644844e-06,
+      "loss": -0.1143,
+      "step": 16
+    },
+    {
+      "epoch": 0.015447523852794184,
+      "grad_norm": 4.0191058201580026,
+      "learning_rate": 1.6205961487284358e-06,
+      "loss": -0.2751,
+      "step": 17
+    },
+    {
+      "epoch": 0.01635620172648796,
+      "grad_norm": 3.4615415988810088,
+      "learning_rate": 1.6532907234348468e-06,
+      "loss": -0.1722,
+      "step": 18
+    },
+    {
+      "epoch": 0.017264879600181735,
+      "grad_norm": 3.5435676110643874,
+      "learning_rate": 1.6842171380470315e-06,
+      "loss": -0.2177,
+      "step": 19
+    },
+    {
+      "epoch": 0.01817355747387551,
+      "grad_norm": 3.8179413783757052,
+      "learning_rate": 1.7135568683269441e-06,
+      "loss": -0.1256,
+      "step": 20
+    },
+    {
+      "epoch": 0.019082235347569285,
+      "grad_norm": 4.225855290446067,
+      "learning_rate": 1.7414648097866612e-06,
+      "loss": -0.2481,
+      "step": 21
+    },
+    {
+      "epoch": 0.019990913221263062,
+      "grad_norm": 4.7304215747802525,
+      "learning_rate": 1.7680742277942584e-06,
+      "loss": -0.1884,
+      "step": 22
+    },
+    {
+      "epoch": 0.020899591094956836,
+      "grad_norm": 4.087148687448322,
+      "learning_rate": 1.7935006064245909e-06,
+      "loss": -0.3307,
+      "step": 23
+    },
+    {
+      "epoch": 0.021808268968650613,
+      "grad_norm": 4.824147187601813,
+      "learning_rate": 1.8178446775702263e-06,
+      "loss": -0.1294,
+      "step": 24
+    },
+    {
+      "epoch": 0.02271694684234439,
+      "grad_norm": 6.91202079670921,
+      "learning_rate": 1.8411948312894039e-06,
+      "loss": -0.1682,
+      "step": 25
+    },
+    {
+      "epoch": 0.023625624716038164,
+      "grad_norm": 6.104099659557386,
+      "learning_rate": 1.8636290531315166e-06,
+      "loss": -0.0907,
+      "step": 26
+    },
+    {
+      "epoch": 0.02453430258973194,
+      "grad_norm": 3.4748485669409233,
+      "learning_rate": 1.8852164956405888e-06,
+      "loss": -0.226,
+      "step": 27
+    },
+    {
+      "epoch": 0.025442980463425715,
+      "grad_norm": 5.057665741977059,
+      "learning_rate": 1.9060187639220403e-06,
+      "loss": -0.1975,
+      "step": 28
+    },
+    {
+      "epoch": 0.026351658337119492,
+      "grad_norm": 3.5329005342200595,
+      "learning_rate": 1.926090975518547e-06,
+      "loss": -0.1301,
+      "step": 29
+    },
+    {
+      "epoch": 0.027260336210813266,
+      "grad_norm": 4.27802907252409,
+      "learning_rate": 1.945482640532686e-06,
+      "loss": -0.1237,
+      "step": 30
+    },
+    {
+      "epoch": 0.028169014084507043,
+      "grad_norm": 4.836873544746256,
+      "learning_rate": 1.96423839738238e-06,
+      "loss": -0.1665,
+      "step": 31
+    },
+    {
+      "epoch": 0.029077691958200817,
+      "grad_norm": 4.677275633168609,
+      "learning_rate": 1.982398631705605e-06,
+      "loss": -0.064,
+      "step": 32
+    },
+    {
+      "epoch": 0.029986369831894594,
+      "grad_norm": 3.8299668221493315,
+      "learning_rate": 2e-06,
+      "loss": -0.1263,
+      "step": 33
+    },
+    {
+      "epoch": 0.030895047705588367,
+      "grad_norm": 4.453931875912948,
+      "learning_rate": 2e-06,
+      "loss": -0.1919,
+      "step": 34
+    },
+    {
+      "epoch": 0.03180372557928214,
+      "grad_norm": 5.464880048128201,
+      "learning_rate": 1.9981255857544518e-06,
+      "loss": -0.1316,
+      "step": 35
+    },
+    {
+      "epoch": 0.03271240345297592,
+      "grad_norm": 4.457912508393936,
+      "learning_rate": 1.9962511715089036e-06,
+      "loss": -0.2277,
+      "step": 36
+    },
+    {
+      "epoch": 0.033621081326669695,
+      "grad_norm": 4.521768054343375,
+      "learning_rate": 1.994376757263355e-06,
+      "loss": -0.2221,
+      "step": 37
+    },
+    {
+      "epoch": 0.03452975920036347,
+      "grad_norm": 5.077359655366783,
+      "learning_rate": 1.992502343017807e-06,
+      "loss": -0.2109,
+      "step": 38
+    },
+    {
+      "epoch": 0.03543843707405725,
+      "grad_norm": 3.8828763591956905,
+      "learning_rate": 1.9906279287722583e-06,
+      "loss": -0.1177,
+      "step": 39
+    },
+    {
+      "epoch": 0.03634711494775102,
+      "grad_norm": 3.798323366811302,
+      "learning_rate": 1.98875351452671e-06,
+      "loss": -0.259,
+      "step": 40
+    },
+    {
+      "epoch": 0.0372557928214448,
+      "grad_norm": 3.7131545340765317,
+      "learning_rate": 1.986879100281162e-06,
+      "loss": -0.1806,
+      "step": 41
+    },
+    {
+      "epoch": 0.03816447069513857,
+      "grad_norm": 4.719071939361814,
+      "learning_rate": 1.985004686035614e-06,
+      "loss": -0.1506,
+      "step": 42
+    },
+    {
+      "epoch": 0.03907314856883235,
+      "grad_norm": 3.4905070586330558,
+      "learning_rate": 1.9831302717900657e-06,
+      "loss": -0.1734,
+      "step": 43
+    },
+    {
+      "epoch": 0.039981826442526125,
+      "grad_norm": 4.911769410762973,
+      "learning_rate": 1.981255857544517e-06,
+      "loss": -0.2468,
+      "step": 44
+    },
+    {
+      "epoch": 0.0408905043162199,
+      "grad_norm": 4.1411475661551655,
+      "learning_rate": 1.979381443298969e-06,
+      "loss": -0.0777,
+      "step": 45
+    },
+    {
+      "epoch": 0.04179918218991367,
+      "grad_norm": 4.901846884058687,
+      "learning_rate": 1.9775070290534208e-06,
+      "loss": -0.1697,
+      "step": 46
+    },
+    {
+      "epoch": 0.04270786006360745,
+      "grad_norm": 5.754655031790533,
+      "learning_rate": 1.975632614807872e-06,
+      "loss": -0.2279,
+      "step": 47
+    },
+    {
+      "epoch": 0.04361653793730123,
+      "grad_norm": 4.481142587052776,
+      "learning_rate": 1.973758200562324e-06,
+      "loss": -0.2417,
+      "step": 48
+    },
+    {
+      "epoch": 0.044525215810995,
+      "grad_norm": 4.048442286698734,
+      "learning_rate": 1.971883786316776e-06,
+      "loss": -0.2257,
+      "step": 49
+    },
+    {
+      "epoch": 0.04543389368468878,
+      "grad_norm": 5.530669618319463,
+      "learning_rate": 1.9700093720712277e-06,
+      "loss": -0.1677,
+      "step": 50
+    },
+    {
+      "epoch": 0.046342571558382555,
+      "grad_norm": 5.877365934529612,
+      "learning_rate": 1.9681349578256796e-06,
+      "loss": -0.1516,
+      "step": 51
+    },
+    {
+      "epoch": 0.04725124943207633,
+      "grad_norm": 4.2167853407814695,
+      "learning_rate": 1.966260543580131e-06,
+      "loss": -0.2476,
+      "step": 52
+    },
+    {
+      "epoch": 0.0481599273057701,
+      "grad_norm": 3.4824211903655384,
+      "learning_rate": 1.964386129334583e-06,
+      "loss": -0.2049,
+      "step": 53
+    },
+    {
+      "epoch": 0.04906860517946388,
+      "grad_norm": 5.205244486084406,
+      "learning_rate": 1.9625117150890347e-06,
+      "loss": -0.1593,
+      "step": 54
+    },
+    {
+      "epoch": 0.049977283053157656,
+      "grad_norm": 4.983858566077597,
+      "learning_rate": 1.960637300843486e-06,
+      "loss": -0.2446,
+      "step": 55
+    },
+    {
+      "epoch": 0.05088596092685143,
+      "grad_norm": 5.024932306697241,
+      "learning_rate": 1.958762886597938e-06,
+      "loss": -0.1608,
+      "step": 56
+    },
+    {
+      "epoch": 0.051794638800545204,
+      "grad_norm": 4.556961041764262,
+      "learning_rate": 1.95688847235239e-06,
+      "loss": -0.1724,
+      "step": 57
+    },
+    {
+      "epoch": 0.052703316674238984,
+      "grad_norm": 3.760474643982415,
+      "learning_rate": 1.9550140581068416e-06,
+      "loss": -0.1767,
+      "step": 58
+    },
+    {
+      "epoch": 0.05361199454793276,
+      "grad_norm": 4.500337588045154,
+      "learning_rate": 1.953139643861293e-06,
+      "loss": -0.1803,
+      "step": 59
+    },
+    {
+      "epoch": 0.05452067242162653,
+      "grad_norm": 4.771738602263339,
+      "learning_rate": 1.951265229615745e-06,
+      "loss": -0.1913,
+      "step": 60
+    },
+    {
+      "epoch": 0.05542935029532031,
+      "grad_norm": 4.989048981640783,
+      "learning_rate": 1.9493908153701968e-06,
+      "loss": -0.2773,
+      "step": 61
+    },
+    {
+      "epoch": 0.056338028169014086,
+      "grad_norm": 5.733095613622837,
+      "learning_rate": 1.9475164011246486e-06,
+      "loss": -0.1921,
+      "step": 62
+    },
+    {
+      "epoch": 0.05724670604270786,
+      "grad_norm": 3.16445822419275,
+      "learning_rate": 1.9456419868791e-06,
+      "loss": -0.2395,
+      "step": 63
+    },
+    {
+      "epoch": 0.05815538391640163,
+      "grad_norm": 4.412805969177259,
+      "learning_rate": 1.943767572633552e-06,
+      "loss": -0.2266,
+      "step": 64
+    },
+    {
+      "epoch": 0.059064061790095414,
+      "grad_norm": 4.379926251352317,
+      "learning_rate": 1.9418931583880037e-06,
+      "loss": -0.133,
+      "step": 65
+    },
+    {
+      "epoch": 0.05997273966378919,
+      "grad_norm": 4.2932104401938025,
+      "learning_rate": 1.9400187441424556e-06,
+      "loss": -0.1427,
+      "step": 66
+    },
+    {
+      "epoch": 0.06088141753748296,
+      "grad_norm": 3.308170728426876,
+      "learning_rate": 1.938144329896907e-06,
+      "loss": -0.3091,
+      "step": 67
+    },
+    {
+      "epoch": 0.061790095411176735,
+      "grad_norm": 5.405383090153589,
+      "learning_rate": 1.936269915651359e-06,
+      "loss": -0.1587,
+      "step": 68
+    },
+    {
+      "epoch": 0.06269877328487052,
+      "grad_norm": 5.150825930540036,
+      "learning_rate": 1.9343955014058107e-06,
+      "loss": -0.1797,
+      "step": 69
+    },
+    {
+      "epoch": 0.06360745115856428,
+      "grad_norm": 5.635319208887778,
+      "learning_rate": 1.9325210871602625e-06,
+      "loss": -0.1216,
+      "step": 70
+    },
+    {
+      "epoch": 0.06451612903225806,
+      "grad_norm": 5.269436416168523,
+      "learning_rate": 1.9306466729147144e-06,
+      "loss": -0.2444,
+      "step": 71
+    },
+    {
+      "epoch": 0.06542480690595184,
+      "grad_norm": 6.856346208980578,
+      "learning_rate": 1.9287722586691658e-06,
+      "loss": -0.1544,
+      "step": 72
+    },
+    {
+      "epoch": 0.06633348477964561,
+      "grad_norm": 3.330806028400352,
+      "learning_rate": 1.9268978444236176e-06,
+      "loss": -0.1875,
+      "step": 73
+    },
+    {
+      "epoch": 0.06724216265333939,
+      "grad_norm": 3.3511054876714867,
+      "learning_rate": 1.925023430178069e-06,
+      "loss": -0.2371,
+      "step": 74
+    },
+    {
+      "epoch": 0.06815084052703317,
+      "grad_norm": 4.757599824308139,
+      "learning_rate": 1.923149015932521e-06,
+      "loss": -0.2061,
+      "step": 75
+    },
+    {
+      "epoch": 0.06905951840072694,
+      "grad_norm": 3.7595653609408144,
+      "learning_rate": 1.9212746016869727e-06,
+      "loss": -0.2426,
+      "step": 76
+    },
+    {
+      "epoch": 0.06996819627442072,
+      "grad_norm": 5.5564164744991835,
+      "learning_rate": 1.9194001874414246e-06,
+      "loss": -0.24,
+      "step": 77
+    },
+    {
+      "epoch": 0.0708768741481145,
+      "grad_norm": 4.155434664726136,
+      "learning_rate": 1.9175257731958764e-06,
+      "loss": -0.2378,
+      "step": 78
+    },
+    {
+      "epoch": 0.07178555202180827,
+      "grad_norm": 3.8134124164339345,
+      "learning_rate": 1.915651358950328e-06,
+      "loss": -0.1358,
+      "step": 79
+    },
+    {
+      "epoch": 0.07269422989550205,
+      "grad_norm": 5.073539659078385,
+      "learning_rate": 1.9137769447047797e-06,
+      "loss": -0.2952,
+      "step": 80
+    },
+    {
+      "epoch": 0.07360290776919581,
+      "grad_norm": 4.841031597637149,
+      "learning_rate": 1.911902530459231e-06,
+      "loss": -0.1026,
+      "step": 81
+    },
+    {
+      "epoch": 0.0745115856428896,
+      "grad_norm": 5.221331112272461,
+      "learning_rate": 1.910028116213683e-06,
+      "loss": -0.375,
+      "step": 82
+    },
+    {
+      "epoch": 0.07542026351658337,
+      "grad_norm": 3.662232419807882,
+      "learning_rate": 1.908153701968135e-06,
+      "loss": -0.1293,
+      "step": 83
+    },
+    {
+      "epoch": 0.07632894139027714,
+      "grad_norm": 4.342771131696319,
+      "learning_rate": 1.9062792877225867e-06,
+      "loss": -0.1471,
+      "step": 84
+    },
+    {
+      "epoch": 0.07723761926397092,
+      "grad_norm": 4.312415080685647,
+      "learning_rate": 1.9044048734770383e-06,
+      "loss": -0.1627,
+      "step": 85
+    },
+    {
+      "epoch": 0.0781462971376647,
+      "grad_norm": 3.342877568309456,
+      "learning_rate": 1.9025304592314901e-06,
+      "loss": -0.1638,
+      "step": 86
+    },
+    {
+      "epoch": 0.07905497501135847,
+      "grad_norm": 4.4233484408118935,
+      "learning_rate": 1.900656044985942e-06,
+      "loss": -0.263,
+      "step": 87
+    },
+    {
+      "epoch": 0.07996365288505225,
+      "grad_norm": 3.1082036097834074,
+      "learning_rate": 1.8987816307403934e-06,
+      "loss": -0.2073,
+      "step": 88
+    },
+    {
+      "epoch": 0.08087233075874603,
+      "grad_norm": 4.686860511392148,
+      "learning_rate": 1.8969072164948452e-06,
+      "loss": -0.0681,
+      "step": 89
+    },
+    {
+      "epoch": 0.0817810086324398,
+      "grad_norm": 5.011308660699104,
+      "learning_rate": 1.8950328022492969e-06,
+      "loss": -0.2656,
+      "step": 90
+    },
+    {
+      "epoch": 0.08268968650613358,
+      "grad_norm": 4.304785199449947,
+      "learning_rate": 1.8931583880037487e-06,
+      "loss": -0.1953,
+      "step": 91
+    },
+    {
+      "epoch": 0.08359836437982734,
+      "grad_norm": 3.8940677960121555,
+      "learning_rate": 1.8912839737582006e-06,
+      "loss": -0.1494,
+      "step": 92
+    },
+    {
+      "epoch": 0.08450704225352113,
+      "grad_norm": 4.5124984529867715,
+      "learning_rate": 1.8894095595126522e-06,
+      "loss": -0.0636,
+      "step": 93
+    },
+    {
+      "epoch": 0.0854157201272149,
+      "grad_norm": 4.95671841691654,
+      "learning_rate": 1.887535145267104e-06,
+      "loss": -0.1829,
+      "step": 94
+    },
+    {
+      "epoch": 0.08632439800090867,
+      "grad_norm": 4.375928507777039,
+      "learning_rate": 1.8856607310215557e-06,
+      "loss": -0.0305,
+      "step": 95
+    },
+    {
+      "epoch": 0.08723307587460245,
+      "grad_norm": 6.041048145422489,
+      "learning_rate": 1.8837863167760075e-06,
+      "loss": -0.0909,
+      "step": 96
+    },
+    {
+      "epoch": 0.08814175374829623,
+      "grad_norm": 3.8791825886586397,
+      "learning_rate": 1.8819119025304592e-06,
+      "loss": -0.1961,
+      "step": 97
+    },
+    {
+      "epoch": 0.08905043162199,
+      "grad_norm": 5.041660319342703,
+      "learning_rate": 1.8800374882849108e-06,
+      "loss": -0.2236,
+      "step": 98
+    },
+    {
+      "epoch": 0.08995910949568378,
+      "grad_norm": 4.712811503752029,
+      "learning_rate": 1.8781630740393626e-06,
+      "loss": -0.2103,
+      "step": 99
+    },
+    {
+      "epoch": 0.09086778736937756,
+      "grad_norm": 4.263708110101409,
+      "learning_rate": 1.8762886597938143e-06,
+      "loss": -0.0561,
+      "step": 100
+    },
+    {
+      "epoch": 0.09177646524307133,
+      "grad_norm": 4.086170197200939,
+      "learning_rate": 1.8744142455482661e-06,
+      "loss": -0.1583,
+      "step": 101
+    },
+    {
+      "epoch": 0.09268514311676511,
+      "grad_norm": 7.249399620716177,
+      "learning_rate": 1.872539831302718e-06,
+      "loss": -0.1699,
+      "step": 102
+    },
+    {
+      "epoch": 0.09359382099045888,
+      "grad_norm": 3.366899022182538,
+      "learning_rate": 1.8706654170571696e-06,
+      "loss": -0.1975,
+      "step": 103
+    },
+    {
+      "epoch": 0.09450249886415266,
+      "grad_norm": 5.490878083372899,
+      "learning_rate": 1.8687910028116214e-06,
+      "loss": -0.1296,
+      "step": 104
+    },
+    {
+      "epoch": 0.09541117673784644,
+      "grad_norm": 3.9332129740215227,
+      "learning_rate": 1.8669165885660729e-06,
+      "loss": -0.141,
+      "step": 105
+    },
+    {
+      "epoch": 0.0963198546115402,
+      "grad_norm": 3.9096381790285872,
+      "learning_rate": 1.8650421743205247e-06,
+      "loss": -0.2522,
+      "step": 106
+    },
+    {
+      "epoch": 0.09722853248523398,
+      "grad_norm": 4.547483003740808,
+      "learning_rate": 1.8631677600749766e-06,
+      "loss": -0.1741,
+      "step": 107
+    },
+    {
+      "epoch": 0.09813721035892777,
+      "grad_norm": 4.323418988683492,
+      "learning_rate": 1.8612933458294282e-06,
+      "loss": -0.0643,
+      "step": 108
+    },
+    {
+      "epoch": 0.09904588823262153,
+      "grad_norm": 4.4852232446469475,
+      "learning_rate": 1.85941893158388e-06,
+      "loss": -0.2309,
+      "step": 109
+    },
+    {
+      "epoch": 0.09995456610631531,
+      "grad_norm": 3.942204284027452,
+      "learning_rate": 1.8575445173383317e-06,
+      "loss": -0.1376,
+      "step": 110
+    },
+    {
+      "epoch": 0.10086324398000909,
+      "grad_norm": 5.267802255266036,
+      "learning_rate": 1.8556701030927835e-06,
+      "loss": -0.2095,
+      "step": 111
+    },
+    {
+      "epoch": 0.10177192185370286,
+      "grad_norm": 4.8580966261468586,
+      "learning_rate": 1.8537956888472354e-06,
+      "loss": -0.1527,
+      "step": 112
+    },
+    {
+      "epoch": 0.10268059972739664,
+      "grad_norm": 6.524133336765768,
+      "learning_rate": 1.8519212746016868e-06,
+      "loss": -0.2725,
+      "step": 113
+    },
+    {
+      "epoch": 0.10358927760109041,
+      "grad_norm": 5.808451565488262,
+      "learning_rate": 1.8500468603561386e-06,
+      "loss": -0.322,
+      "step": 114
+    },
+    {
+      "epoch": 0.10449795547478419,
+      "grad_norm": 4.037196541829292,
+      "learning_rate": 1.8481724461105903e-06,
+      "loss": -0.2171,
+      "step": 115
+    },
+    {
+      "epoch": 0.10540663334847797,
+      "grad_norm": 3.585111901147578,
+      "learning_rate": 1.846298031865042e-06,
+      "loss": -0.1782,
+      "step": 116
+    },
+    {
+      "epoch": 0.10631531122217174,
+      "grad_norm": 4.596404469964939,
+      "learning_rate": 1.8444236176194937e-06,
+      "loss": -0.3113,
+      "step": 117
+    },
+    {
+      "epoch": 0.10722398909586552,
+      "grad_norm": 5.1264733778517995,
+      "learning_rate": 1.8425492033739456e-06,
+      "loss": -0.1256,
+      "step": 118
+    },
+    {
+      "epoch": 0.1081326669695593,
+      "grad_norm": 3.514532843594382,
+      "learning_rate": 1.8406747891283974e-06,
+      "loss": -0.2381,
+      "step": 119
+    },
+    {
+      "epoch": 0.10904134484325306,
+      "grad_norm": 3.694003819040783,
+      "learning_rate": 1.838800374882849e-06,
+      "loss": -0.1902,
+      "step": 120
+    },
+    {
+      "epoch": 0.10995002271694684,
+      "grad_norm": 6.100816590730332,
+      "learning_rate": 1.8369259606373007e-06,
+      "loss": -0.1483,
+      "step": 121
+    },
+    {
+      "epoch": 0.11085870059064062,
+      "grad_norm": 4.043964768238335,
+      "learning_rate": 1.8350515463917523e-06,
+      "loss": -0.1452,
+      "step": 122
+    },
+    {
+      "epoch": 0.11176737846433439,
+      "grad_norm": 4.870105731920806,
+      "learning_rate": 1.8331771321462042e-06,
+      "loss": -0.1754,
+      "step": 123
+    },
+    {
+      "epoch": 0.11267605633802817,
+      "grad_norm": 4.748219005582576,
+      "learning_rate": 1.831302717900656e-06,
+      "loss": -0.2018,
+      "step": 124
+    },
+    {
+      "epoch": 0.11358473421172194,
+      "grad_norm": 5.687045236369042,
+      "learning_rate": 1.8294283036551076e-06,
+      "loss": -0.1945,
+      "step": 125
+    },
+    {
+      "epoch": 0.11449341208541572,
+      "grad_norm": 6.189698108219693,
+      "learning_rate": 1.8275538894095595e-06,
+      "loss": -0.2811,
+      "step": 126
+    },
+    {
+      "epoch": 0.1154020899591095,
+      "grad_norm": 6.279845730775857,
+      "learning_rate": 1.8256794751640111e-06,
+      "loss": -0.1218,
+      "step": 127
+    },
+    {
+      "epoch": 0.11631076783280327,
+      "grad_norm": 5.884669685784497,
+      "learning_rate": 1.823805060918463e-06,
+      "loss": -0.2064,
+      "step": 128
+    },
+    {
+      "epoch": 0.11721944570649705,
+      "grad_norm": 3.9615538228586837,
+      "learning_rate": 1.8219306466729148e-06,
+      "loss": -0.2156,
+      "step": 129
+    },
+    {
+      "epoch": 0.11812812358019083,
+      "grad_norm": 5.313031697312787,
+      "learning_rate": 1.8200562324273662e-06,
+      "loss": -0.1922,
+      "step": 130
+    },
+    {
+      "epoch": 0.1190368014538846,
+      "grad_norm": 6.594190537619879,
+      "learning_rate": 1.818181818181818e-06,
+      "loss": -0.175,
+      "step": 131
+    },
+    {
+      "epoch": 0.11994547932757837,
+      "grad_norm": 3.924716239439436,
+      "learning_rate": 1.8163074039362697e-06,
+      "loss": -0.2409,
+      "step": 132
+    },
+    {
+      "epoch": 0.12085415720127216,
+      "grad_norm": 4.264924044184273,
+      "learning_rate": 1.8144329896907216e-06,
+      "loss": -0.184,
+      "step": 133
+    },
+    {
+      "epoch": 0.12176283507496592,
+      "grad_norm": 4.293485212976499,
+      "learning_rate": 1.8125585754451734e-06,
+      "loss": -0.2318,
+      "step": 134
+    },
+    {
+      "epoch": 0.1226715129486597,
+      "grad_norm": 3.8274070845076764,
+      "learning_rate": 1.810684161199625e-06,
+      "loss": -0.227,
+      "step": 135
+    },
+    {
+      "epoch": 0.12358019082235347,
+      "grad_norm": 3.5756874277192505,
+      "learning_rate": 1.8088097469540769e-06,
+      "loss": -0.1972,
+      "step": 136
+    },
+    {
+      "epoch": 0.12448886869604725,
+      "grad_norm": 5.30460875597391,
+      "learning_rate": 1.8069353327085285e-06,
+      "loss": -0.1793,
+      "step": 137
+    },
+    {
+      "epoch": 0.12539754656974103,
+      "grad_norm": 5.929899149096638,
+      "learning_rate": 1.8050609184629801e-06,
+      "loss": -0.1971,
+      "step": 138
+    },
+    {
+      "epoch": 0.1263062244434348,
+      "grad_norm": 4.774254362197915,
+      "learning_rate": 1.803186504217432e-06,
+      "loss": -0.26,
+      "step": 139
+    },
+    {
+      "epoch": 0.12721490231712856,
+      "grad_norm": 5.0055871265096465,
+      "learning_rate": 1.8013120899718836e-06,
+      "loss": -0.1757,
+      "step": 140
+    },
+    {
+      "epoch": 0.12812358019082234,
+      "grad_norm": 3.9182579108772746,
+      "learning_rate": 1.7994376757263355e-06,
+      "loss": -0.3008,
+      "step": 141
+    },
+    {
+      "epoch": 0.12903225806451613,
+      "grad_norm": 5.794770138833297,
+      "learning_rate": 1.797563261480787e-06,
+      "loss": -0.2026,
+      "step": 142
+    },
+    {
+      "epoch": 0.1299409359382099,
+      "grad_norm": 3.7056697037358894,
+      "learning_rate": 1.795688847235239e-06,
+      "loss": -0.1745,
+      "step": 143
+    },
+    {
+      "epoch": 0.1308496138119037,
+      "grad_norm": 5.750551611937131,
+      "learning_rate": 1.7938144329896908e-06,
+      "loss": -0.1468,
+      "step": 144
+    },
+    {
+      "epoch": 0.13175829168559747,
+      "grad_norm": 6.603536710166722,
+      "learning_rate": 1.7919400187441424e-06,
+      "loss": -0.1709,
+      "step": 145
+    },
+    {
+      "epoch": 0.13266696955929122,
+      "grad_norm": 5.701994483412553,
+      "learning_rate": 1.790065604498594e-06,
+      "loss": -0.1089,
+      "step": 146
+    },
+    {
+      "epoch": 0.133575647432985,
+      "grad_norm": 3.945486778715715,
+      "learning_rate": 1.7881911902530457e-06,
+      "loss": -0.1258,
+      "step": 147
+    },
+    {
+      "epoch": 0.13448432530667878,
+      "grad_norm": 4.389461050818243,
+      "learning_rate": 1.7863167760074975e-06,
+      "loss": -0.1579,
+      "step": 148
+    },
+    {
+      "epoch": 0.13539300318037256,
+      "grad_norm": 8.987642493464682,
+      "learning_rate": 1.7844423617619494e-06,
+      "loss": -0.2373,
+      "step": 149
+    },
+    {
+      "epoch": 0.13630168105406634,
+      "grad_norm": 4.304136708325501,
+      "learning_rate": 1.782567947516401e-06,
+      "loss": -0.263,
+      "step": 150
+    },
+    {
+      "epoch": 0.1372103589277601,
+      "grad_norm": 4.2679772231862625,
+      "learning_rate": 1.7806935332708529e-06,
+      "loss": -0.1582,
+      "step": 151
+    },
+    {
+      "epoch": 0.13811903680145388,
+      "grad_norm": 4.560525336204491,
+      "learning_rate": 1.7788191190253045e-06,
+      "loss": -0.2722,
+      "step": 152
+    },
+    {
+      "epoch": 0.13902771467514766,
+      "grad_norm": 6.509811039539305,
+      "learning_rate": 1.7769447047797563e-06,
+      "loss": -0.1166,
+      "step": 153
+    },
+    {
+      "epoch": 0.13993639254884144,
+      "grad_norm": 9.903796352406555,
+      "learning_rate": 1.775070290534208e-06,
+      "loss": -0.1835,
+      "step": 154
+    },
+    {
+      "epoch": 0.14084507042253522,
+      "grad_norm": 4.51832684531333,
+      "learning_rate": 1.7731958762886596e-06,
+      "loss": -0.0601,
+      "step": 155
+    },
+    {
+      "epoch": 0.141753748296229,
+      "grad_norm": 4.691529843292891,
+      "learning_rate": 1.7713214620431115e-06,
+      "loss": -0.0937,
+      "step": 156
+    },
+    {
+      "epoch": 0.14266242616992275,
+      "grad_norm": 5.343781409812659,
+      "learning_rate": 1.769447047797563e-06,
+      "loss": -0.1619,
+      "step": 157
+    },
+    {
+      "epoch": 0.14357110404361653,
+      "grad_norm": 7.187280535457158,
+      "learning_rate": 1.767572633552015e-06,
+      "loss": -0.1912,
+      "step": 158
+    },
+    {
+      "epoch": 0.1444797819173103,
+      "grad_norm": 4.7984580742293375,
+      "learning_rate": 1.7656982193064668e-06,
+      "loss": -0.1931,
+      "step": 159
+    },
+    {
+      "epoch": 0.1453884597910041,
+      "grad_norm": 3.501311601384926,
+      "learning_rate": 1.7638238050609184e-06,
+      "loss": -0.2596,
+      "step": 160
+    },
+    {
+      "epoch": 0.14629713766469787,
+      "grad_norm": 4.884811309637428,
+      "learning_rate": 1.7619493908153703e-06,
+      "loss": -0.1312,
+      "step": 161
+    },
+    {
+      "epoch": 0.14720581553839163,
+      "grad_norm": 4.543380810458526,
+      "learning_rate": 1.7600749765698219e-06,
+      "loss": -0.1469,
+      "step": 162
+    },
+    {
+      "epoch": 0.1481144934120854,
+      "grad_norm": 3.3950214624639443,
+      "learning_rate": 1.7582005623242735e-06,
+      "loss": -0.1728,
+      "step": 163
+    },
+    {
+      "epoch": 0.1490231712857792,
+      "grad_norm": 72.58024586513709,
+      "learning_rate": 1.7563261480787252e-06,
+      "loss": -0.163,
+      "step": 164
+    },
+    {
+      "epoch": 0.14993184915947297,
+      "grad_norm": 3.932097225139742,
+      "learning_rate": 1.754451733833177e-06,
+      "loss": -0.1423,
+      "step": 165
+    },
+    {
+      "epoch": 0.15084052703316675,
+      "grad_norm": 8.725837697910352,
+      "learning_rate": 1.7525773195876288e-06,
+      "loss": -0.1997,
+      "step": 166
+    },
+    {
+      "epoch": 0.15174920490686053,
+      "grad_norm": 3.6178915109052365,
+      "learning_rate": 1.7507029053420805e-06,
+      "loss": -0.2216,
+      "step": 167
+    },
+    {
+      "epoch": 0.15265788278055428,
+      "grad_norm": 4.0221141824092825,
+      "learning_rate": 1.7488284910965323e-06,
+      "loss": -0.1252,
+      "step": 168
+    },
+    {
+      "epoch": 0.15356656065424806,
+      "grad_norm": 4.483665489929633,
+      "learning_rate": 1.746954076850984e-06,
+      "loss": -0.22,
+      "step": 169
+    },
+    {
+      "epoch": 0.15447523852794184,
+      "grad_norm": 4.822009512804506,
+      "learning_rate": 1.7450796626054358e-06,
+      "loss": -0.3071,
+      "step": 170
+    },
+    {
+      "epoch": 0.15538391640163562,
+      "grad_norm": 3.2816648675136437,
+      "learning_rate": 1.7432052483598874e-06,
+      "loss": -0.1151,
+      "step": 171
+    },
+    {
+      "epoch": 0.1562925942753294,
+      "grad_norm": 5.384566731285535,
+      "learning_rate": 1.741330834114339e-06,
+      "loss": -0.1446,
+      "step": 172
+    },
+    {
+      "epoch": 0.15720127214902316,
+      "grad_norm": 4.039698065626413,
+      "learning_rate": 1.739456419868791e-06,
+      "loss": -0.2863,
+      "step": 173
+    },
+    {
+      "epoch": 0.15810995002271694,
+      "grad_norm": 6.048751500463172,
+      "learning_rate": 1.7375820056232425e-06,
+      "loss": -0.2781,
+      "step": 174
+    },
+    {
+      "epoch": 0.15901862789641072,
+      "grad_norm": 4.097092523335956,
+      "learning_rate": 1.7357075913776944e-06,
+      "loss": -0.1484,
+      "step": 175
+    },
+    {
+      "epoch": 0.1599273057701045,
+      "grad_norm": 5.650475520825676,
+      "learning_rate": 1.7338331771321462e-06,
+      "loss": -0.155,
+      "step": 176
+    },
+    {
+      "epoch": 0.16083598364379828,
+      "grad_norm": 4.330244626365804,
+      "learning_rate": 1.7319587628865979e-06,
+      "loss": -0.1997,
+      "step": 177
+    },
+    {
+      "epoch": 0.16174466151749206,
+      "grad_norm": 5.92176291326971,
+      "learning_rate": 1.7300843486410497e-06,
+      "loss": -0.2391,
+      "step": 178
+    },
+    {
+      "epoch": 0.1626533393911858,
+      "grad_norm": 4.863694212511953,
+      "learning_rate": 1.7282099343955011e-06,
+      "loss": -0.117,
+      "step": 179
+    },
+    {
+      "epoch": 0.1635620172648796,
+      "grad_norm": 3.844912965580707,
+      "learning_rate": 1.726335520149953e-06,
+      "loss": -0.2181,
+      "step": 180
+    },
+    {
+      "epoch": 0.16447069513857338,
+      "grad_norm": 6.016921025162859,
+      "learning_rate": 1.7244611059044048e-06,
+      "loss": -0.2675,
+      "step": 181
+    },
+    {
+      "epoch": 0.16537937301226716,
+      "grad_norm": 4.479626630483299,
+      "learning_rate": 1.7225866916588565e-06,
+      "loss": -0.1808,
+      "step": 182
+    },
+    {
+      "epoch": 0.16628805088596094,
+      "grad_norm": 4.034597973087698,
+      "learning_rate": 1.7207122774133083e-06,
+      "loss": -0.2191,
+      "step": 183
+    },
+    {
+      "epoch": 0.1671967287596547,
+      "grad_norm": 6.612035770381811,
+      "learning_rate": 1.71883786316776e-06,
+      "loss": -0.1623,
+      "step": 184
+    },
+    {
+      "epoch": 0.16810540663334847,
+      "grad_norm": 4.542105254014905,
+      "learning_rate": 1.7169634489222118e-06,
+      "loss": -0.1731,
+      "step": 185
+    },
+    {
+      "epoch": 0.16901408450704225,
+      "grad_norm": 6.970830085177519,
+      "learning_rate": 1.7150890346766636e-06,
+      "loss": -0.1504,
+      "step": 186
+    },
+    {
+      "epoch": 0.16992276238073603,
+      "grad_norm": 5.053148840268244,
+      "learning_rate": 1.7132146204311153e-06,
+      "loss": -0.1579,
+      "step": 187
+    },
+    {
+      "epoch": 0.1708314402544298,
+      "grad_norm": 4.340730564489818,
+      "learning_rate": 1.711340206185567e-06,
+      "loss": -0.1321,
+      "step": 188
+    },
+    {
+      "epoch": 0.1717401181281236,
+      "grad_norm": 5.267824703975675,
+      "learning_rate": 1.7094657919400185e-06,
+      "loss": -0.2367,
+      "step": 189
+    },
+    {
+      "epoch": 0.17264879600181735,
+      "grad_norm": 5.164628498494913,
+      "learning_rate": 1.7075913776944704e-06,
+      "loss": -0.1951,
+      "step": 190
+    },
+    {
+      "epoch": 0.17355747387551113,
+      "grad_norm": 4.785435658296668,
+      "learning_rate": 1.7057169634489222e-06,
+      "loss": -0.2579,
+      "step": 191
+    },
+    {
+      "epoch": 0.1744661517492049,
+      "grad_norm": 4.797465235279545,
+      "learning_rate": 1.7038425492033739e-06,
+      "loss": -0.1116,
+      "step": 192
+    },
+    {
+      "epoch": 0.1753748296228987,
+      "grad_norm": 4.385823186424985,
+      "learning_rate": 1.7019681349578257e-06,
+      "loss": -0.1276,
+      "step": 193
+    },
+    {
+      "epoch": 0.17628350749659247,
+      "grad_norm": 4.032823595566027,
+      "learning_rate": 1.7000937207122773e-06,
+      "loss": -0.0895,
+      "step": 194
+    },
+    {
+      "epoch": 0.17719218537028622,
+      "grad_norm": 16.741859393658498,
+      "learning_rate": 1.6982193064667292e-06,
+      "loss": -0.3349,
+      "step": 195
+    },
+    {
+      "epoch": 0.17810086324398,
+      "grad_norm": 6.252384798924977,
+      "learning_rate": 1.6963448922211808e-06,
+      "loss": -0.1732,
+      "step": 196
+    },
+    {
+      "epoch": 0.17900954111767378,
+      "grad_norm": 8.902951349109427,
+      "learning_rate": 1.6944704779756324e-06,
+      "loss": -0.1492,
+      "step": 197
+    },
+    {
+      "epoch": 0.17991821899136756,
+      "grad_norm": 5.481126306719264,
+      "learning_rate": 1.6925960637300843e-06,
+      "loss": -0.2905,
+      "step": 198
+    },
+    {
+      "epoch": 0.18082689686506134,
+      "grad_norm": 6.971445426205846,
+      "learning_rate": 1.690721649484536e-06,
+      "loss": -0.1708,
+      "step": 199
+    },
+    {
+      "epoch": 0.18173557473875512,
+      "grad_norm": 4.806748359151746,
+      "learning_rate": 1.6888472352389878e-06,
+      "loss": -0.2564,
+      "step": 200
+    },
+    {
+      "epoch": 0.18264425261244888,
+      "grad_norm": 6.543923962456778,
+      "learning_rate": 1.6869728209934396e-06,
+      "loss": -0.2293,
+      "step": 201
+    },
+    {
+      "epoch": 0.18355293048614266,
+      "grad_norm": 4.847910023700975,
+      "learning_rate": 1.6850984067478912e-06,
+      "loss": -0.1178,
+      "step": 202
+    },
+    {
+      "epoch": 0.18446160835983644,
+      "grad_norm": 4.81636189659221,
+      "learning_rate": 1.683223992502343e-06,
+      "loss": -0.223,
+      "step": 203
+    },
+    {
+      "epoch": 0.18537028623353022,
+      "grad_norm": 6.230629773585354,
+      "learning_rate": 1.6813495782567945e-06,
+      "loss": -0.225,
+      "step": 204
+    },
+    {
+      "epoch": 0.186278964107224,
+      "grad_norm": 4.640285903409493,
+      "learning_rate": 1.6794751640112464e-06,
+      "loss": -0.1866,
+      "step": 205
+    },
+    {
+      "epoch": 0.18718764198091775,
+      "grad_norm": 3.9999279969410564,
+      "learning_rate": 1.6776007497656982e-06,
+      "loss": -0.2697,
+      "step": 206
+    },
+    {
+      "epoch": 0.18809631985461153,
+      "grad_norm": 4.82707938813021,
+      "learning_rate": 1.6757263355201498e-06,
+      "loss": -0.2037,
+      "step": 207
+    },
+    {
+      "epoch": 0.1890049977283053,
+      "grad_norm": 5.049309205501164,
+      "learning_rate": 1.6738519212746017e-06,
+      "loss": -0.2202,
+      "step": 208
+    },
+    {
+      "epoch": 0.1899136756019991,
+      "grad_norm": 13.900346578764346,
+      "learning_rate": 1.6719775070290533e-06,
+      "loss": -0.2232,
+      "step": 209
+    },
+    {
+      "epoch": 0.19082235347569287,
+      "grad_norm": 5.749723013135781,
+      "learning_rate": 1.6701030927835052e-06,
+      "loss": -0.2292,
+      "step": 210
+    },
+    {
+      "epoch": 0.19173103134938665,
+      "grad_norm": 5.760779849348198,
+      "learning_rate": 1.668228678537957e-06,
+      "loss": -0.268,
+      "step": 211
+    },
+    {
+      "epoch": 0.1926397092230804,
+      "grad_norm": 4.757833947444988,
+      "learning_rate": 1.6663542642924084e-06,
+      "loss": -0.2611,
+      "step": 212
+    },
+    {
+      "epoch": 0.1935483870967742,
+      "grad_norm": 3.7303566481438573,
+      "learning_rate": 1.6644798500468603e-06,
+      "loss": -0.1823,
+      "step": 213
+    },
+    {
+      "epoch": 0.19445706497046797,
+      "grad_norm": 4.160431787383456,
+      "learning_rate": 1.662605435801312e-06,
+      "loss": -0.2508,
+      "step": 214
+    },
+    {
+      "epoch": 0.19536574284416175,
+      "grad_norm": 4.2685494371607495,
+      "learning_rate": 1.6607310215557637e-06,
+      "loss": -0.2681,
+      "step": 215
+    },
+    {
+      "epoch": 0.19627442071785553,
+      "grad_norm": 4.752016542938927,
+      "learning_rate": 1.6588566073102154e-06,
+      "loss": -0.2493,
+      "step": 216
+    },
+    {
+      "epoch": 0.19718309859154928,
+      "grad_norm": 4.634741756154299,
+      "learning_rate": 1.6569821930646672e-06,
+      "loss": -0.0705,
+      "step": 217
+    },
+    {
+      "epoch": 0.19809177646524306,
+      "grad_norm": 5.129669109477676,
+      "learning_rate": 1.655107778819119e-06,
+      "loss": -0.2276,
+      "step": 218
+    },
+    {
+      "epoch": 0.19900045433893684,
+      "grad_norm": 3.7418866444743104,
+      "learning_rate": 1.6532333645735707e-06,
+      "loss": -0.194,
+      "step": 219
+    },
+    {
+      "epoch": 0.19990913221263062,
+      "grad_norm": 4.410207283528905,
+      "learning_rate": 1.6513589503280225e-06,
+      "loss": -0.2639,
+      "step": 220
+    },
+    {
+      "epoch": 0.2008178100863244,
+      "grad_norm": 7.633397965247823,
+      "learning_rate": 1.649484536082474e-06,
+      "loss": -0.2059,
+      "step": 221
+    },
+    {
+      "epoch": 0.20172648796001819,
+      "grad_norm": 4.98001338298179,
+      "learning_rate": 1.6476101218369258e-06,
+      "loss": -0.1242,
+      "step": 222
+    },
+    {
+      "epoch": 0.20263516583371194,
+      "grad_norm": 5.559403832946767,
+      "learning_rate": 1.6457357075913777e-06,
+      "loss": -0.1631,
+      "step": 223
+    },
+    {
+      "epoch": 0.20354384370740572,
+      "grad_norm": 9.76836289745058,
+      "learning_rate": 1.6438612933458293e-06,
+      "loss": -0.1708,
+      "step": 224
+    },
+    {
+      "epoch": 0.2044525215810995,
+      "grad_norm": 4.5942496722162565,
+      "learning_rate": 1.6419868791002811e-06,
+      "loss": -0.2033,
+      "step": 225
+    },
+    {
+      "epoch": 0.20536119945479328,
+      "grad_norm": 4.752049054338954,
+      "learning_rate": 1.6401124648547328e-06,
+      "loss": -0.2656,
+      "step": 226
+    },
+    {
+      "epoch": 0.20626987732848706,
+      "grad_norm": 3.773314922229436,
+      "learning_rate": 1.6382380506091846e-06,
+      "loss": -0.1851,
+      "step": 227
+    },
+    {
+      "epoch": 0.20717855520218081,
+      "grad_norm": 9.106420544956265,
+      "learning_rate": 1.6363636363636365e-06,
+      "loss": -0.1096,
+      "step": 228
+    },
+    {
+      "epoch": 0.2080872330758746,
+      "grad_norm": 4.829813541580622,
+      "learning_rate": 1.6344892221180879e-06,
+      "loss": -0.1406,
+      "step": 229
+    },
+    {
+      "epoch": 0.20899591094956838,
+      "grad_norm": 4.71033338529089,
+      "learning_rate": 1.6326148078725397e-06,
+      "loss": -0.1784,
+      "step": 230
+    },
+    {
+      "epoch": 0.20990458882326216,
+      "grad_norm": 5.003460830768457,
+      "learning_rate": 1.6307403936269914e-06,
+      "loss": -0.1612,
+      "step": 231
+    },
+    {
+      "epoch": 0.21081326669695594,
+      "grad_norm": 4.253905353248299,
+      "learning_rate": 1.6288659793814432e-06,
+      "loss": -0.267,
+      "step": 232
+    },
+    {
+      "epoch": 0.21172194457064972,
+      "grad_norm": 5.400002458359901,
+      "learning_rate": 1.626991565135895e-06,
+      "loss": -0.2044,
+      "step": 233
+    },
+    {
+      "epoch": 0.21263062244434347,
+      "grad_norm": 5.6811910997617145,
+      "learning_rate": 1.6251171508903467e-06,
+      "loss": -0.185,
+      "step": 234
+    },
+    {
+      "epoch": 0.21353930031803725,
+      "grad_norm": 9.537558747156881,
+      "learning_rate": 1.6232427366447985e-06,
+      "loss": -0.2198,
+      "step": 235
+    },
+    {
+      "epoch": 0.21444797819173103,
+      "grad_norm": 8.605246660065381,
+      "learning_rate": 1.6213683223992502e-06,
+      "loss": -0.1635,
+      "step": 236
+    },
+    {
+      "epoch": 0.2153566560654248,
+      "grad_norm": 4.788561846154162,
+      "learning_rate": 1.6194939081537018e-06,
+      "loss": -0.2256,
+      "step": 237
+    },
+    {
+      "epoch": 0.2162653339391186,
+      "grad_norm": 4.068096590093825,
+      "learning_rate": 1.6176194939081536e-06,
+      "loss": -0.1502,
+      "step": 238
+    },
+    {
+      "epoch": 0.21717401181281235,
+      "grad_norm": 9.017080417619248,
+      "learning_rate": 1.6157450796626053e-06,
+      "loss": -0.1792,
+      "step": 239
+    },
+    {
+      "epoch": 0.21808268968650613,
+      "grad_norm": 7.695583404574833,
+      "learning_rate": 1.6138706654170571e-06,
+      "loss": -0.2208,
+      "step": 240
+    },
+    {
+      "epoch": 0.2189913675601999,
+      "grad_norm": 4.627921547754764,
+      "learning_rate": 1.6119962511715088e-06,
+      "loss": -0.1925,
+      "step": 241
+    },
+    {
+      "epoch": 0.2199000454338937,
+      "grad_norm": 5.086632746978763,
+      "learning_rate": 1.6101218369259606e-06,
+      "loss": -0.2913,
+      "step": 242
+    },
+    {
+      "epoch": 0.22080872330758747,
+      "grad_norm": 3.8045053928039727,
+      "learning_rate": 1.6082474226804124e-06,
+      "loss": -0.2023,
+      "step": 243
+    },
+    {
+      "epoch": 0.22171740118128125,
+      "grad_norm": 4.559248510309541,
+      "learning_rate": 1.606373008434864e-06,
+      "loss": -0.1941,
+      "step": 244
+    },
+    {
+      "epoch": 0.222626079054975,
+      "grad_norm": 5.581952147193629,
+      "learning_rate": 1.6044985941893157e-06,
+      "loss": -0.1746,
+      "step": 245
+    },
+    {
+      "epoch": 0.22353475692866878,
+      "grad_norm": 5.331283572807903,
+      "learning_rate": 1.6026241799437673e-06,
+      "loss": -0.1638,
+      "step": 246
+    },
+    {
+      "epoch": 0.22444343480236256,
+      "grad_norm": 3.8788988741432893,
+      "learning_rate": 1.6007497656982192e-06,
+      "loss": -0.305,
+      "step": 247
+    },
+    {
+      "epoch": 0.22535211267605634,
+      "grad_norm": 7.156194561739413,
+      "learning_rate": 1.598875351452671e-06,
+      "loss": -0.1809,
+      "step": 248
+    },
+    {
+      "epoch": 0.22626079054975012,
+      "grad_norm": 4.134327371082181,
+      "learning_rate": 1.5970009372071227e-06,
+      "loss": -0.1183,
+      "step": 249
+    },
+    {
+      "epoch": 0.22716946842344388,
+      "grad_norm": 4.928218374776762,
+      "learning_rate": 1.5951265229615745e-06,
+      "loss": -0.1349,
+      "step": 250
+    },
+    {
+      "epoch": 0.22807814629713766,
+      "grad_norm": 4.098560331702163,
+      "learning_rate": 1.5932521087160261e-06,
+      "loss": -0.1052,
+      "step": 251
+    },
+    {
+      "epoch": 0.22898682417083144,
+      "grad_norm": 5.627077524332124,
+      "learning_rate": 1.591377694470478e-06,
+      "loss": -0.2399,
+      "step": 252
+    },
+    {
+      "epoch": 0.22989550204452522,
+      "grad_norm": 6.5538735503825745,
+      "learning_rate": 1.5895032802249298e-06,
+      "loss": -0.1436,
+      "step": 253
+    },
+    {
+      "epoch": 0.230804179918219,
+      "grad_norm": 3.519742732833534,
+      "learning_rate": 1.5876288659793813e-06,
+      "loss": -0.2724,
+      "step": 254
+    },
+    {
+      "epoch": 0.23171285779191278,
+      "grad_norm": 3.4391626671879587,
+      "learning_rate": 1.585754451733833e-06,
+      "loss": -0.1909,
+      "step": 255
+    },
+    {
+      "epoch": 0.23262153566560653,
+      "grad_norm": 5.0095010133577205,
+      "learning_rate": 1.5838800374882847e-06,
+      "loss": -0.2237,
+      "step": 256
+    },
+    {
+      "epoch": 0.2335302135393003,
+      "grad_norm": 5.772705031661358,
+      "learning_rate": 1.5820056232427366e-06,
+      "loss": -0.0852,
+      "step": 257
+    },
+    {
+      "epoch": 0.2344388914129941,
+      "grad_norm": 4.886559499803229,
+      "learning_rate": 1.5801312089971884e-06,
+      "loss": -0.1729,
+      "step": 258
+    },
+    {
+      "epoch": 0.23534756928668787,
+      "grad_norm": 5.983557104886809,
+      "learning_rate": 1.57825679475164e-06,
+      "loss": -0.3057,
+      "step": 259
+    },
+    {
+      "epoch": 0.23625624716038165,
+      "grad_norm": 4.884102368693348,
+      "learning_rate": 1.576382380506092e-06,
+      "loss": -0.133,
+      "step": 260
+    },
+    {
+      "epoch": 0.2371649250340754,
+      "grad_norm": 8.287916041269295,
+      "learning_rate": 1.5745079662605435e-06,
+      "loss": -0.0757,
+      "step": 261
+    },
+    {
+      "epoch": 0.2380736029077692,
+      "grad_norm": 6.325053206510246,
+      "learning_rate": 1.5726335520149952e-06,
+      "loss": -0.1852,
+      "step": 262
+    },
+    {
+      "epoch": 0.23898228078146297,
+      "grad_norm": 4.459260481691903,
+      "learning_rate": 1.5707591377694468e-06,
+      "loss": -0.2391,
+      "step": 263
+    },
+    {
+      "epoch": 0.23989095865515675,
+      "grad_norm": 5.011735495392913,
+      "learning_rate": 1.5688847235238986e-06,
+      "loss": -0.295,
+      "step": 264
+    },
+    {
+      "epoch": 0.24079963652885053,
+      "grad_norm": 15.188642294040903,
+      "learning_rate": 1.5670103092783505e-06,
+      "loss": -0.2031,
+      "step": 265
+    },
+    {
+      "epoch": 0.2417083144025443,
+      "grad_norm": 4.381398127469313,
+      "learning_rate": 1.5651358950328021e-06,
+      "loss": -0.2088,
+      "step": 266
+    },
+    {
+      "epoch": 0.24261699227623806,
+      "grad_norm": 4.565777372288936,
+      "learning_rate": 1.563261480787254e-06,
+      "loss": -0.2197,
+      "step": 267
+    },
+    {
+      "epoch": 0.24352567014993184,
+      "grad_norm": 6.161956156474695,
+      "learning_rate": 1.5613870665417056e-06,
+      "loss": -0.1399,
+      "step": 268
+    },
+    {
+      "epoch": 0.24443434802362562,
+      "grad_norm": 5.70969829343467,
+      "learning_rate": 1.5595126522961575e-06,
+      "loss": -0.1889,
+      "step": 269
+    },
+    {
+      "epoch": 0.2453430258973194,
+      "grad_norm": 5.450033191264864,
+      "learning_rate": 1.557638238050609e-06,
+      "loss": -0.2661,
+      "step": 270
+    },
+    {
+      "epoch": 0.2462517037710132,
+      "grad_norm": 5.349195951132132,
+      "learning_rate": 1.5557638238050607e-06,
+      "loss": -0.0542,
+      "step": 271
+    },
+    {
+      "epoch": 0.24716038164470694,
+      "grad_norm": 4.106675570424654,
+      "learning_rate": 1.5538894095595126e-06,
+      "loss": -0.2924,
+      "step": 272
+    },
+    {
+      "epoch": 0.24806905951840072,
+      "grad_norm": 4.31801506888661,
+      "learning_rate": 1.5520149953139642e-06,
+      "loss": -0.2105,
+      "step": 273
+    },
+    {
+      "epoch": 0.2489777373920945,
+      "grad_norm": 4.778779432488999,
+      "learning_rate": 1.550140581068416e-06,
+      "loss": -0.1074,
+      "step": 274
+    },
+    {
+      "epoch": 0.24988641526578828,
+      "grad_norm": 8.674476576962864,
+      "learning_rate": 1.5482661668228679e-06,
+      "loss": -0.3648,
+      "step": 275
+    },
+    {
+      "epoch": 0.25079509313948206,
+      "grad_norm": 7.079915946418405,
+      "learning_rate": 1.5463917525773195e-06,
+      "loss": -0.2242,
+      "step": 276
+    },
+    {
+      "epoch": 0.2517037710131758,
+      "grad_norm": 3.784447445164216,
+      "learning_rate": 1.5445173383317714e-06,
+      "loss": -0.1672,
+      "step": 277
+    },
+    {
+      "epoch": 0.2526124488868696,
+      "grad_norm": 3.5989990379287695,
+      "learning_rate": 1.542642924086223e-06,
+      "loss": -0.144,
+      "step": 278
+    },
+    {
+      "epoch": 0.2535211267605634,
+      "grad_norm": 4.3454298983629815,
+      "learning_rate": 1.5407685098406746e-06,
+      "loss": -0.0438,
+      "step": 279
+    },
+    {
+      "epoch": 0.25442980463425713,
+      "grad_norm": 4.649405550132672,
+      "learning_rate": 1.5388940955951265e-06,
+      "loss": -0.2014,
+      "step": 280
+    },
+    {
+      "epoch": 0.25533848250795094,
+      "grad_norm": 3.8724429399831437,
+      "learning_rate": 1.5370196813495781e-06,
+      "loss": -0.1143,
+      "step": 281
+    },
+    {
+      "epoch": 0.2562471603816447,
+      "grad_norm": 5.170395377294075,
+      "learning_rate": 1.53514526710403e-06,
+      "loss": -0.1806,
+      "step": 282
+    },
+    {
+      "epoch": 0.2571558382553385,
+      "grad_norm": 4.163679552630364,
+      "learning_rate": 1.5332708528584816e-06,
+      "loss": -0.2692,
+      "step": 283
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 5.327021599229004,
+      "learning_rate": 1.5313964386129334e-06,
+      "loss": -0.1822,
+      "step": 284
+    },
+    {
+      "epoch": 0.25897319400272606,
+      "grad_norm": 4.2138523636399166,
+      "learning_rate": 1.5295220243673853e-06,
+      "loss": -0.0823,
+      "step": 285
+    },
+    {
+      "epoch": 0.2598818718764198,
+      "grad_norm": 5.172555008637686,
+      "learning_rate": 1.527647610121837e-06,
+      "loss": -0.0804,
+      "step": 286
+    },
+    {
+      "epoch": 0.26079054975011356,
+      "grad_norm": 4.064900789258413,
+      "learning_rate": 1.5257731958762885e-06,
+      "loss": -0.1827,
+      "step": 287
+    },
+    {
+      "epoch": 0.2616992276238074,
+      "grad_norm": 4.619024179293337,
+      "learning_rate": 1.5238987816307402e-06,
+      "loss": -0.1035,
+      "step": 288
+    },
+    {
+      "epoch": 0.2626079054975011,
+      "grad_norm": 4.758946476240377,
+      "learning_rate": 1.522024367385192e-06,
+      "loss": -0.2036,
+      "step": 289
+    },
+    {
+      "epoch": 0.26351658337119493,
+      "grad_norm": 6.483092399507111,
+      "learning_rate": 1.5201499531396439e-06,
+      "loss": -0.1383,
+      "step": 290
+    },
+    {
+      "epoch": 0.2644252612448887,
+      "grad_norm": 4.437222109407841,
+      "learning_rate": 1.5182755388940955e-06,
+      "loss": -0.3327,
+      "step": 291
+    },
+    {
+      "epoch": 0.26533393911858244,
+      "grad_norm": 5.249574189856229,
+      "learning_rate": 1.5164011246485473e-06,
+      "loss": -0.2909,
+      "step": 292
+    },
+    {
+      "epoch": 0.26624261699227625,
+      "grad_norm": 3.933278560795867,
+      "learning_rate": 1.514526710402999e-06,
+      "loss": -0.1498,
+      "step": 293
+    },
+    {
+      "epoch": 0.26715129486597,
+      "grad_norm": 6.961728514522245,
+      "learning_rate": 1.5126522961574508e-06,
+      "loss": -0.1467,
+      "step": 294
+    },
+    {
+      "epoch": 0.2680599727396638,
+      "grad_norm": 3.7284941395399813,
+      "learning_rate": 1.5107778819119025e-06,
+      "loss": -0.1051,
+      "step": 295
+    },
+    {
+      "epoch": 0.26896865061335756,
+      "grad_norm": 5.247162733430195,
+      "learning_rate": 1.508903467666354e-06,
+      "loss": -0.2269,
+      "step": 296
+    },
+    {
+      "epoch": 0.2698773284870513,
+      "grad_norm": 4.803217469158592,
+      "learning_rate": 1.507029053420806e-06,
+      "loss": -0.0861,
+      "step": 297
+    },
+    {
+      "epoch": 0.2707860063607451,
+      "grad_norm": 4.210784399756234,
+      "learning_rate": 1.5051546391752576e-06,
+      "loss": -0.0973,
+      "step": 298
+    },
+    {
+      "epoch": 0.2716946842344389,
+      "grad_norm": 4.994698956861939,
+      "learning_rate": 1.5032802249297094e-06,
+      "loss": -0.1295,
+      "step": 299
+    },
+    {
+      "epoch": 0.2726033621081327,
+      "grad_norm": 4.75556318860813,
+      "learning_rate": 1.5014058106841613e-06,
+      "loss": -0.183,
+      "step": 300
+    },
+    {
+      "epoch": 0.27351203998182644,
+      "grad_norm": 3.9227623410627936,
+      "learning_rate": 1.4995313964386129e-06,
+      "loss": -0.1386,
+      "step": 301
+    },
+    {
+      "epoch": 0.2744207178555202,
+      "grad_norm": 3.4783498905545147,
+      "learning_rate": 1.4976569821930647e-06,
+      "loss": -0.2092,
+      "step": 302
+    },
+    {
+      "epoch": 0.275329395729214,
+      "grad_norm": 5.411332051551037,
+      "learning_rate": 1.4957825679475162e-06,
+      "loss": -0.2603,
+      "step": 303
+    },
+    {
+      "epoch": 0.27623807360290775,
+      "grad_norm": 3.815903192199877,
+      "learning_rate": 1.493908153701968e-06,
+      "loss": -0.1264,
+      "step": 304
+    },
+    {
+      "epoch": 0.27714675147660156,
+      "grad_norm": 3.70034479648235,
+      "learning_rate": 1.4920337394564199e-06,
+      "loss": -0.2,
+      "step": 305
+    },
+    {
+      "epoch": 0.2780554293502953,
+      "grad_norm": 4.260197242862788,
+      "learning_rate": 1.4901593252108715e-06,
+      "loss": -0.1566,
+      "step": 306
+    },
+    {
+      "epoch": 0.2789641072239891,
+      "grad_norm": 6.028847329471058,
+      "learning_rate": 1.4882849109653233e-06,
+      "loss": -0.1446,
+      "step": 307
+    },
+    {
+      "epoch": 0.2798727850976829,
+      "grad_norm": 3.766460658378537,
+      "learning_rate": 1.486410496719775e-06,
+      "loss": -0.2499,
+      "step": 308
+    },
+    {
+      "epoch": 0.2807814629713766,
+      "grad_norm": 5.290464950637221,
+      "learning_rate": 1.4845360824742268e-06,
+      "loss": -0.0989,
+      "step": 309
+    },
+    {
+      "epoch": 0.28169014084507044,
+      "grad_norm": 4.653753853262741,
+      "learning_rate": 1.4826616682286787e-06,
+      "loss": -0.2276,
+      "step": 310
+    },
+    {
+      "epoch": 0.2825988187187642,
+      "grad_norm": 4.930907665210537,
+      "learning_rate": 1.4807872539831303e-06,
+      "loss": -0.2037,
+      "step": 311
+    },
+    {
+      "epoch": 0.283507496592458,
+      "grad_norm": 5.222745907455462,
+      "learning_rate": 1.478912839737582e-06,
+      "loss": -0.2654,
+      "step": 312
+    },
+    {
+      "epoch": 0.28441617446615175,
+      "grad_norm": 6.697021835931794,
+      "learning_rate": 1.4770384254920336e-06,
+      "loss": -0.1482,
+      "step": 313
+    },
+    {
+      "epoch": 0.2853248523398455,
+      "grad_norm": 4.536201875651594,
+      "learning_rate": 1.4751640112464854e-06,
+      "loss": -0.1631,
+      "step": 314
+    },
+    {
+      "epoch": 0.2862335302135393,
+      "grad_norm": 4.483214331161934,
+      "learning_rate": 1.473289597000937e-06,
+      "loss": -0.1153,
+      "step": 315
+    },
+    {
+      "epoch": 0.28714220808723306,
+      "grad_norm": 4.643132723500063,
+      "learning_rate": 1.4714151827553889e-06,
+      "loss": -0.1322,
+      "step": 316
+    },
+    {
+      "epoch": 0.2880508859609269,
+      "grad_norm": 4.265675974111639,
+      "learning_rate": 1.4695407685098407e-06,
+      "loss": -0.1103,
+      "step": 317
+    },
+    {
+      "epoch": 0.2889595638346206,
+      "grad_norm": 7.2385510776744235,
+      "learning_rate": 1.4676663542642924e-06,
+      "loss": -0.133,
+      "step": 318
+    },
+    {
+      "epoch": 0.2898682417083144,
+      "grad_norm": 3.879766424561571,
+      "learning_rate": 1.4657919400187442e-06,
+      "loss": -0.2204,
+      "step": 319
+    },
+    {
+      "epoch": 0.2907769195820082,
+      "grad_norm": 3.9738731658461903,
+      "learning_rate": 1.4639175257731956e-06,
+      "loss": -0.1872,
+      "step": 320
+    },
+    {
+      "epoch": 0.29168559745570194,
+      "grad_norm": 4.545431111015253,
+      "learning_rate": 1.4620431115276475e-06,
+      "loss": -0.1742,
+      "step": 321
+    },
+    {
+      "epoch": 0.29259427532939575,
+      "grad_norm": 3.77102919152452,
+      "learning_rate": 1.4601686972820993e-06,
+      "loss": -0.2573,
+      "step": 322
+    },
+    {
+      "epoch": 0.2935029532030895,
+      "grad_norm": 5.832586876475211,
+      "learning_rate": 1.458294283036551e-06,
+      "loss": -0.1758,
+      "step": 323
+    },
+    {
+      "epoch": 0.29441163107678325,
+      "grad_norm": 4.528645575949113,
+      "learning_rate": 1.4564198687910028e-06,
+      "loss": -0.2488,
+      "step": 324
+    },
+    {
+      "epoch": 0.29532030895047706,
+      "grad_norm": 5.12728402949993,
+      "learning_rate": 1.4545454545454544e-06,
+      "loss": -0.1507,
+      "step": 325
+    },
+    {
+      "epoch": 0.2962289868241708,
+      "grad_norm": 5.7958506375714265,
+      "learning_rate": 1.4526710402999063e-06,
+      "loss": -0.2222,
+      "step": 326
+    },
+    {
+      "epoch": 0.2971376646978646,
+      "grad_norm": 3.9327065491073196,
+      "learning_rate": 1.4507966260543581e-06,
+      "loss": -0.1882,
+      "step": 327
+    },
+    {
+      "epoch": 0.2980463425715584,
+      "grad_norm": 5.232081397563729,
+      "learning_rate": 1.4489222118088095e-06,
+      "loss": -0.2592,
+      "step": 328
+    },
+    {
+      "epoch": 0.2989550204452522,
+      "grad_norm": 5.276903700252939,
+      "learning_rate": 1.4470477975632614e-06,
+      "loss": -0.1086,
+      "step": 329
+    },
+    {
+      "epoch": 0.29986369831894594,
+      "grad_norm": 6.1825952067065035,
+      "learning_rate": 1.445173383317713e-06,
+      "loss": -0.2041,
+      "step": 330
+    },
+    {
+      "epoch": 0.3007723761926397,
+      "grad_norm": 9.78885167557793,
+      "learning_rate": 1.4432989690721649e-06,
+      "loss": -0.17,
+      "step": 331
+    },
+    {
+      "epoch": 0.3016810540663335,
+      "grad_norm": 4.342962179849771,
+      "learning_rate": 1.4414245548266167e-06,
+      "loss": -0.198,
+      "step": 332
+    },
+    {
+      "epoch": 0.30258973194002725,
+      "grad_norm": 3.7356128636137975,
+      "learning_rate": 1.4395501405810683e-06,
+      "loss": -0.1572,
+      "step": 333
+    },
+    {
+      "epoch": 0.30349840981372106,
+      "grad_norm": 4.839666404892442,
+      "learning_rate": 1.4376757263355202e-06,
+      "loss": -0.2409,
+      "step": 334
+    },
+    {
+      "epoch": 0.3044070876874148,
+      "grad_norm": 4.973548347659235,
+      "learning_rate": 1.4358013120899718e-06,
+      "loss": -0.1374,
+      "step": 335
+    },
+    {
+      "epoch": 0.30531576556110857,
+      "grad_norm": 5.688005026727915,
+      "learning_rate": 1.4339268978444234e-06,
+      "loss": -0.204,
+      "step": 336
+    },
+    {
+      "epoch": 0.3062244434348024,
+      "grad_norm": 5.632302355052114,
+      "learning_rate": 1.4320524835988753e-06,
+      "loss": -0.2057,
+      "step": 337
+    },
+    {
+      "epoch": 0.3071331213084961,
+      "grad_norm": 4.943189602230883,
+      "learning_rate": 1.430178069353327e-06,
+      "loss": -0.2562,
+      "step": 338
+    },
+    {
+      "epoch": 0.30804179918218993,
+      "grad_norm": 7.537469978335994,
+      "learning_rate": 1.4283036551077788e-06,
+      "loss": -0.1345,
+      "step": 339
+    },
+    {
+      "epoch": 0.3089504770558837,
+      "grad_norm": 9.000143261934866,
+      "learning_rate": 1.4264292408622304e-06,
+      "loss": -0.1996,
+      "step": 340
+    },
+    {
+      "epoch": 0.30985915492957744,
+      "grad_norm": 3.492962847154135,
+      "learning_rate": 1.4245548266166822e-06,
+      "loss": -0.2311,
+      "step": 341
+    },
+    {
+      "epoch": 0.31076783280327125,
+      "grad_norm": 5.332799388065755,
+      "learning_rate": 1.422680412371134e-06,
+      "loss": -0.1632,
+      "step": 342
+    },
+    {
+      "epoch": 0.311676510676965,
+      "grad_norm": 4.742480298827358,
+      "learning_rate": 1.4208059981255857e-06,
+      "loss": -0.221,
+      "step": 343
+    },
+    {
+      "epoch": 0.3125851885506588,
+      "grad_norm": 4.639938355069254,
+      "learning_rate": 1.4189315838800376e-06,
+      "loss": -0.127,
+      "step": 344
+    },
+    {
+      "epoch": 0.31349386642435256,
+      "grad_norm": 9.109826837395111,
+      "learning_rate": 1.417057169634489e-06,
+      "loss": -0.1594,
+      "step": 345
+    },
+    {
+      "epoch": 0.3144025442980463,
+      "grad_norm": 4.254132225402842,
+      "learning_rate": 1.4151827553889408e-06,
+      "loss": -0.2692,
+      "step": 346
+    },
+    {
+      "epoch": 0.3153112221717401,
+      "grad_norm": 6.5561643989436815,
+      "learning_rate": 1.4133083411433927e-06,
+      "loss": -0.1985,
+      "step": 347
+    },
+    {
+      "epoch": 0.3162199000454339,
+      "grad_norm": 4.046735253001578,
+      "learning_rate": 1.4114339268978443e-06,
+      "loss": -0.1805,
+      "step": 348
+    },
+    {
+      "epoch": 0.3171285779191277,
+      "grad_norm": 4.75837711685979,
+      "learning_rate": 1.4095595126522962e-06,
+      "loss": -0.1146,
+      "step": 349
+    },
+    {
+      "epoch": 0.31803725579282144,
+      "grad_norm": 4.602100282600152,
+      "learning_rate": 1.4076850984067478e-06,
+      "loss": -0.1624,
+      "step": 350
+    },
+    {
+      "epoch": 0.31894593366651525,
+      "grad_norm": 6.743739014794256,
+      "learning_rate": 1.4058106841611996e-06,
+      "loss": -0.1908,
+      "step": 351
+    },
+    {
+      "epoch": 0.319854611540209,
+      "grad_norm": 4.66055672164694,
+      "learning_rate": 1.4039362699156515e-06,
+      "loss": -0.1344,
+      "step": 352
+    },
+    {
+      "epoch": 0.32076328941390275,
+      "grad_norm": 4.92050704111131,
+      "learning_rate": 1.402061855670103e-06,
+      "loss": -0.2282,
+      "step": 353
+    },
+    {
+      "epoch": 0.32167196728759656,
+      "grad_norm": 5.961061008918696,
+      "learning_rate": 1.4001874414245548e-06,
+      "loss": -0.1413,
+      "step": 354
+    },
+    {
+      "epoch": 0.3225806451612903,
+      "grad_norm": 6.344661853819054,
+      "learning_rate": 1.3983130271790064e-06,
+      "loss": -0.1712,
+      "step": 355
+    },
+    {
+      "epoch": 0.3234893230349841,
+      "grad_norm": 4.705413363031157,
+      "learning_rate": 1.3964386129334582e-06,
+      "loss": -0.2427,
+      "step": 356
+    },
+    {
+      "epoch": 0.3243980009086779,
+      "grad_norm": 4.195968264115199,
+      "learning_rate": 1.39456419868791e-06,
+      "loss": -0.2012,
+      "step": 357
+    },
+    {
+      "epoch": 0.3253066787823716,
+      "grad_norm": 3.874068240461245,
+      "learning_rate": 1.3926897844423617e-06,
+      "loss": -0.2102,
+      "step": 358
+    },
+    {
+      "epoch": 0.32621535665606544,
+      "grad_norm": 5.201982069067464,
+      "learning_rate": 1.3908153701968136e-06,
+      "loss": -0.2469,
+      "step": 359
+    },
+    {
+      "epoch": 0.3271240345297592,
+      "grad_norm": 5.094624075755403,
+      "learning_rate": 1.3889409559512652e-06,
+      "loss": -0.2124,
+      "step": 360
+    },
+    {
+      "epoch": 0.328032712403453,
+      "grad_norm": 3.505907114708039,
+      "learning_rate": 1.3870665417057168e-06,
+      "loss": -0.1893,
+      "step": 361
+    },
+    {
+      "epoch": 0.32894139027714675,
+      "grad_norm": 4.6281770671050975,
+      "learning_rate": 1.3851921274601685e-06,
+      "loss": -0.3312,
+      "step": 362
+    },
+    {
+      "epoch": 0.3298500681508405,
+      "grad_norm": 4.573419541824475,
+      "learning_rate": 1.3833177132146203e-06,
+      "loss": -0.2159,
+      "step": 363
+    },
+    {
+      "epoch": 0.3307587460245343,
+      "grad_norm": 4.577887935815037,
+      "learning_rate": 1.3814432989690721e-06,
+      "loss": -0.2138,
+      "step": 364
+    },
+    {
+      "epoch": 0.33166742389822806,
+      "grad_norm": 4.563179488076846,
+      "learning_rate": 1.3795688847235238e-06,
+      "loss": -0.1574,
+      "step": 365
+    },
+    {
+      "epoch": 0.3325761017719219,
+      "grad_norm": 6.231211995245187,
+      "learning_rate": 1.3776944704779756e-06,
+      "loss": -0.2428,
+      "step": 366
+    },
+    {
+      "epoch": 0.3334847796456156,
+      "grad_norm": 4.627766580760203,
+      "learning_rate": 1.3758200562324275e-06,
+      "loss": -0.1749,
+      "step": 367
+    },
+    {
+      "epoch": 0.3343934575193094,
+      "grad_norm": 5.510457933508148,
+      "learning_rate": 1.373945641986879e-06,
+      "loss": -0.2243,
+      "step": 368
+    },
+    {
+      "epoch": 0.3353021353930032,
+      "grad_norm": 4.133883995373725,
+      "learning_rate": 1.3720712277413307e-06,
+      "loss": -0.1958,
+      "step": 369
+    },
+    {
+      "epoch": 0.33621081326669694,
+      "grad_norm": 4.958484722145072,
+      "learning_rate": 1.3701968134957824e-06,
+      "loss": -0.1847,
+      "step": 370
+    },
+    {
+      "epoch": 0.33711949114039075,
+      "grad_norm": 4.8581580695878515,
+      "learning_rate": 1.3683223992502342e-06,
+      "loss": -0.1029,
+      "step": 371
+    },
+    {
+      "epoch": 0.3380281690140845,
+      "grad_norm": 4.78108923149338,
+      "learning_rate": 1.3664479850046858e-06,
+      "loss": -0.2093,
+      "step": 372
+    },
+    {
+      "epoch": 0.3389368468877783,
+      "grad_norm": 4.640194034794449,
+      "learning_rate": 1.3645735707591377e-06,
+      "loss": -0.2417,
+      "step": 373
+    },
+    {
+      "epoch": 0.33984552476147206,
+      "grad_norm": 4.222006889897828,
+      "learning_rate": 1.3626991565135895e-06,
+      "loss": -0.1458,
+      "step": 374
+    },
+    {
+      "epoch": 0.3407542026351658,
+      "grad_norm": 5.081899701350368,
+      "learning_rate": 1.3608247422680412e-06,
+      "loss": -0.0925,
+      "step": 375
+    },
+    {
+      "epoch": 0.3416628805088596,
+      "grad_norm": 5.036644927877743,
+      "learning_rate": 1.358950328022493e-06,
+      "loss": -0.1406,
+      "step": 376
+    },
+    {
+      "epoch": 0.3425715583825534,
+      "grad_norm": 6.126317408046278,
+      "learning_rate": 1.3570759137769446e-06,
+      "loss": -0.2134,
+      "step": 377
+    },
+    {
+      "epoch": 0.3434802362562472,
+      "grad_norm": 4.076894757265192,
+      "learning_rate": 1.3552014995313963e-06,
+      "loss": -0.1784,
+      "step": 378
+    },
+    {
+      "epoch": 0.34438891412994094,
+      "grad_norm": 3.924685986829372,
+      "learning_rate": 1.3533270852858481e-06,
+      "loss": -0.2526,
+      "step": 379
+    },
+    {
+      "epoch": 0.3452975920036347,
+      "grad_norm": 3.798126014607645,
+      "learning_rate": 1.3514526710402998e-06,
+      "loss": -0.229,
+      "step": 380
+    },
+    {
+      "epoch": 0.3462062698773285,
+      "grad_norm": 4.6426077052591515,
+      "learning_rate": 1.3495782567947516e-06,
+      "loss": -0.1043,
+      "step": 381
+    },
+    {
+      "epoch": 0.34711494775102225,
+      "grad_norm": 11.147589562401274,
+      "learning_rate": 1.3477038425492032e-06,
+      "loss": -0.1573,
+      "step": 382
+    },
+    {
+      "epoch": 0.34802362562471606,
+      "grad_norm": 5.579529828461307,
+      "learning_rate": 1.345829428303655e-06,
+      "loss": -0.2248,
+      "step": 383
+    },
+    {
+      "epoch": 0.3489323034984098,
+      "grad_norm": 3.8773939059885656,
+      "learning_rate": 1.343955014058107e-06,
+      "loss": -0.29,
+      "step": 384
+    },
+    {
+      "epoch": 0.34984098137210357,
+      "grad_norm": 10.664922770360501,
+      "learning_rate": 1.3420805998125586e-06,
+      "loss": -0.2079,
+      "step": 385
+    },
+    {
+      "epoch": 0.3507496592457974,
+      "grad_norm": 4.050570534915445,
+      "learning_rate": 1.3402061855670102e-06,
+      "loss": -0.1832,
+      "step": 386
+    },
+    {
+      "epoch": 0.3516583371194911,
+      "grad_norm": 4.676431226688182,
+      "learning_rate": 1.3383317713214618e-06,
+      "loss": -0.2125,
+      "step": 387
+    },
+    {
+      "epoch": 0.35256701499318494,
+      "grad_norm": 5.069606361532291,
+      "learning_rate": 1.3364573570759137e-06,
+      "loss": -0.2082,
+      "step": 388
+    },
+    {
+      "epoch": 0.3534756928668787,
+      "grad_norm": 3.371948416744109,
+      "learning_rate": 1.3345829428303655e-06,
+      "loss": -0.2975,
+      "step": 389
+    },
+    {
+      "epoch": 0.35438437074057244,
+      "grad_norm": 4.059722420359685,
+      "learning_rate": 1.3327085285848172e-06,
+      "loss": -0.2188,
+      "step": 390
+    },
+    {
+      "epoch": 0.35529304861426625,
+      "grad_norm": 7.1519685289514054,
+      "learning_rate": 1.330834114339269e-06,
+      "loss": -0.1885,
+      "step": 391
+    },
+    {
+      "epoch": 0.35620172648796,
+      "grad_norm": 4.786519456652311,
+      "learning_rate": 1.3289597000937206e-06,
+      "loss": -0.2334,
+      "step": 392
+    },
+    {
+      "epoch": 0.3571104043616538,
+      "grad_norm": 8.182019445082432,
+      "learning_rate": 1.3270852858481725e-06,
+      "loss": -0.1618,
+      "step": 393
+    },
+    {
+      "epoch": 0.35801908223534756,
+      "grad_norm": 3.4964296659064042,
+      "learning_rate": 1.3252108716026241e-06,
+      "loss": -0.1436,
+      "step": 394
+    },
+    {
+      "epoch": 0.35892776010904137,
+      "grad_norm": 4.190879255371773,
+      "learning_rate": 1.3233364573570757e-06,
+      "loss": -0.1521,
+      "step": 395
+    },
+    {
+      "epoch": 0.3598364379827351,
+      "grad_norm": 5.603550003706844,
+      "learning_rate": 1.3214620431115276e-06,
+      "loss": -0.2002,
+      "step": 396
+    },
+    {
+      "epoch": 0.3607451158564289,
+      "grad_norm": 4.961113393473815,
+      "learning_rate": 1.3195876288659792e-06,
+      "loss": -0.1989,
+      "step": 397
+    },
+    {
+      "epoch": 0.3616537937301227,
+      "grad_norm": 3.8811125996666864,
+      "learning_rate": 1.317713214620431e-06,
+      "loss": -0.0482,
+      "step": 398
+    },
+    {
+      "epoch": 0.36256247160381644,
+      "grad_norm": 3.6207721653824616,
+      "learning_rate": 1.315838800374883e-06,
+      "loss": -0.2161,
+      "step": 399
+    },
+    {
+      "epoch": 0.36347114947751025,
+      "grad_norm": 4.472223385849532,
+      "learning_rate": 1.3139643861293345e-06,
+      "loss": -0.2479,
+      "step": 400
+    },
+    {
+      "epoch": 0.364379827351204,
+      "grad_norm": 3.419060930412622,
+      "learning_rate": 1.3120899718837864e-06,
+      "loss": -0.0728,
+      "step": 401
+    },
+    {
+      "epoch": 0.36528850522489775,
+      "grad_norm": 4.082554308219969,
+      "learning_rate": 1.310215557638238e-06,
+      "loss": -0.1431,
+      "step": 402
+    },
+    {
+      "epoch": 0.36619718309859156,
+      "grad_norm": 4.147713750719806,
+      "learning_rate": 1.3083411433926897e-06,
+      "loss": -0.1491,
+      "step": 403
+    },
+    {
+      "epoch": 0.3671058609722853,
+      "grad_norm": 5.530371644705666,
+      "learning_rate": 1.3064667291471415e-06,
+      "loss": -0.2055,
+      "step": 404
+    },
+    {
+      "epoch": 0.3680145388459791,
+      "grad_norm": 5.4613761671095205,
+      "learning_rate": 1.3045923149015931e-06,
+      "loss": -0.2091,
+      "step": 405
+    },
+    {
+      "epoch": 0.3689232167196729,
+      "grad_norm": 4.82018567974125,
+      "learning_rate": 1.302717900656045e-06,
+      "loss": -0.1908,
+      "step": 406
+    },
+    {
+      "epoch": 0.36983189459336663,
+      "grad_norm": 3.590612883252714,
+      "learning_rate": 1.3008434864104966e-06,
+      "loss": -0.1217,
+      "step": 407
+    },
+    {
+      "epoch": 0.37074057246706044,
+      "grad_norm": 4.0510239707747955,
+      "learning_rate": 1.2989690721649485e-06,
+      "loss": -0.1283,
+      "step": 408
+    },
+    {
+      "epoch": 0.3716492503407542,
+      "grad_norm": 4.463516692273821,
+      "learning_rate": 1.2970946579194003e-06,
+      "loss": -0.1458,
+      "step": 409
+    },
+    {
+      "epoch": 0.372557928214448,
+      "grad_norm": 4.569678316376453,
+      "learning_rate": 1.295220243673852e-06,
+      "loss": -0.3171,
+      "step": 410
+    },
+    {
+      "epoch": 0.37346660608814175,
+      "grad_norm": 4.083151339672692,
+      "learning_rate": 1.2933458294283036e-06,
+      "loss": -0.2515,
+      "step": 411
+    },
+    {
+      "epoch": 0.3743752839618355,
+      "grad_norm": 6.7073579951896685,
+      "learning_rate": 1.2914714151827552e-06,
+      "loss": -0.1769,
+      "step": 412
+    },
+    {
+      "epoch": 0.3752839618355293,
+      "grad_norm": 3.406767079714257,
+      "learning_rate": 1.289597000937207e-06,
+      "loss": -0.2123,
+      "step": 413
+    },
+    {
+      "epoch": 0.37619263970922306,
+      "grad_norm": 4.3239049065338,
+      "learning_rate": 1.2877225866916589e-06,
+      "loss": -0.1612,
+      "step": 414
+    },
+    {
+      "epoch": 0.3771013175829169,
+      "grad_norm": 6.62022112381992,
+      "learning_rate": 1.2858481724461105e-06,
+      "loss": -0.1965,
+      "step": 415
+    },
+    {
+      "epoch": 0.3780099954566106,
+      "grad_norm": 4.239404148123641,
+      "learning_rate": 1.2839737582005624e-06,
+      "loss": -0.1351,
+      "step": 416
+    },
+    {
+      "epoch": 0.37891867333030443,
+      "grad_norm": 5.421347116942239,
+      "learning_rate": 1.282099343955014e-06,
+      "loss": -0.2542,
+      "step": 417
+    },
+    {
+      "epoch": 0.3798273512039982,
+      "grad_norm": 4.969870003241644,
+      "learning_rate": 1.2802249297094658e-06,
+      "loss": -0.1208,
+      "step": 418
+    },
+    {
+      "epoch": 0.38073602907769194,
+      "grad_norm": 3.8531011947304856,
+      "learning_rate": 1.2783505154639173e-06,
+      "loss": -0.2276,
+      "step": 419
+    },
+    {
+      "epoch": 0.38164470695138575,
+      "grad_norm": 5.617035886286214,
+      "learning_rate": 1.2764761012183691e-06,
+      "loss": -0.1872,
+      "step": 420
+    },
+    {
+      "epoch": 0.3825533848250795,
+      "grad_norm": 5.161391220913907,
+      "learning_rate": 1.274601686972821e-06,
+      "loss": -0.2059,
+      "step": 421
+    },
+    {
+      "epoch": 0.3834620626987733,
+      "grad_norm": 3.5573568864077934,
+      "learning_rate": 1.2727272727272726e-06,
+      "loss": -0.1921,
+      "step": 422
+    },
+    {
+      "epoch": 0.38437074057246706,
+      "grad_norm": 4.844248296886695,
+      "learning_rate": 1.2708528584817244e-06,
+      "loss": -0.174,
+      "step": 423
+    },
+    {
+      "epoch": 0.3852794184461608,
+      "grad_norm": 4.306898598070033,
+      "learning_rate": 1.268978444236176e-06,
+      "loss": -0.2192,
+      "step": 424
+    },
+    {
+      "epoch": 0.3861880963198546,
+      "grad_norm": 3.7653413404572302,
+      "learning_rate": 1.267104029990628e-06,
+      "loss": -0.1619,
+      "step": 425
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 4.045999911041938,
+      "learning_rate": 1.2652296157450798e-06,
+      "loss": -0.2766,
+      "step": 426
+    },
+    {
+      "epoch": 0.3880054520672422,
+      "grad_norm": 6.963095248132853,
+      "learning_rate": 1.2633552014995312e-06,
+      "loss": -0.2158,
+      "step": 427
+    },
+    {
+      "epoch": 0.38891412994093594,
+      "grad_norm": 4.334081193995168,
+      "learning_rate": 1.261480787253983e-06,
+      "loss": -0.2669,
+      "step": 428
+    },
+    {
+      "epoch": 0.3898228078146297,
+      "grad_norm": 3.8167121007177927,
+      "learning_rate": 1.2596063730084347e-06,
+      "loss": -0.1872,
+      "step": 429
+    },
+    {
+      "epoch": 0.3907314856883235,
+      "grad_norm": 4.132337575352469,
+      "learning_rate": 1.2577319587628865e-06,
+      "loss": -0.0656,
+      "step": 430
+    },
+    {
+      "epoch": 0.39164016356201725,
+      "grad_norm": 3.663156490922787,
+      "learning_rate": 1.2558575445173384e-06,
+      "loss": -0.2011,
+      "step": 431
+    },
+    {
+      "epoch": 0.39254884143571106,
+      "grad_norm": 5.413868381823933,
+      "learning_rate": 1.25398313027179e-06,
+      "loss": -0.0988,
+      "step": 432
+    },
+    {
+      "epoch": 0.3934575193094048,
+      "grad_norm": 4.741710255472237,
+      "learning_rate": 1.2521087160262418e-06,
+      "loss": -0.0857,
+      "step": 433
+    },
+    {
+      "epoch": 0.39436619718309857,
+      "grad_norm": 3.2042107951511527,
+      "learning_rate": 1.2502343017806935e-06,
+      "loss": -0.0789,
+      "step": 434
+    },
+    {
+      "epoch": 0.3952748750567924,
+      "grad_norm": 3.9855127243587356,
+      "learning_rate": 1.2483598875351453e-06,
+      "loss": -0.1967,
+      "step": 435
+    },
+    {
+      "epoch": 0.3961835529304861,
+      "grad_norm": 4.577034986492689,
+      "learning_rate": 1.246485473289597e-06,
+      "loss": -0.1788,
+      "step": 436
+    },
+    {
+      "epoch": 0.39709223080417994,
+      "grad_norm": 5.850716584487917,
+      "learning_rate": 1.2446110590440486e-06,
+      "loss": -0.1658,
+      "step": 437
+    },
+    {
+      "epoch": 0.3980009086778737,
+      "grad_norm": 4.5117931541483935,
+      "learning_rate": 1.2427366447985004e-06,
+      "loss": -0.2723,
+      "step": 438
+    },
+    {
+      "epoch": 0.3989095865515675,
+      "grad_norm": 4.713307456701343,
+      "learning_rate": 1.240862230552952e-06,
+      "loss": -0.1559,
+      "step": 439
+    },
+    {
+      "epoch": 0.39981826442526125,
+      "grad_norm": 7.781201956114232,
+      "learning_rate": 1.238987816307404e-06,
+      "loss": -0.1496,
+      "step": 440
+    },
+    {
+      "epoch": 0.400726942298955,
+      "grad_norm": 4.128578252404872,
+      "learning_rate": 1.2371134020618557e-06,
+      "loss": -0.1851,
+      "step": 441
+    },
+    {
+      "epoch": 0.4016356201726488,
+      "grad_norm": 4.225440026116409,
+      "learning_rate": 1.2352389878163074e-06,
+      "loss": -0.2081,
+      "step": 442
+    },
+    {
+      "epoch": 0.40254429804634256,
+      "grad_norm": 3.652218283063869,
+      "learning_rate": 1.2333645735707592e-06,
+      "loss": -0.1863,
+      "step": 443
+    },
+    {
+      "epoch": 0.40345297592003637,
+      "grad_norm": 5.3013004506918735,
+      "learning_rate": 1.2314901593252106e-06,
+      "loss": -0.2019,
+      "step": 444
+    },
+    {
+      "epoch": 0.4043616537937301,
+      "grad_norm": 6.0559650869837105,
+      "learning_rate": 1.2296157450796625e-06,
+      "loss": -0.2064,
+      "step": 445
+    },
+    {
+      "epoch": 0.4052703316674239,
+      "grad_norm": 11.442799825509475,
+      "learning_rate": 1.2277413308341143e-06,
+      "loss": -0.1324,
+      "step": 446
+    },
+    {
+      "epoch": 0.4061790095411177,
+      "grad_norm": 4.7853904199717725,
+      "learning_rate": 1.225866916588566e-06,
+      "loss": -0.2481,
+      "step": 447
+    },
+    {
+      "epoch": 0.40708768741481144,
+      "grad_norm": 5.846815967516663,
+      "learning_rate": 1.2239925023430178e-06,
+      "loss": -0.1997,
+      "step": 448
+    },
+    {
+      "epoch": 0.40799636528850525,
+      "grad_norm": 3.963056309875824,
+      "learning_rate": 1.2221180880974694e-06,
+      "loss": -0.1628,
+      "step": 449
+    },
+    {
+      "epoch": 0.408905043162199,
+      "grad_norm": 4.841343435680537,
+      "learning_rate": 1.2202436738519213e-06,
+      "loss": -0.1545,
+      "step": 450
+    },
+    {
+      "epoch": 0.40981372103589275,
+      "grad_norm": 6.001546978204186,
+      "learning_rate": 1.2183692596063731e-06,
+      "loss": -0.0715,
+      "step": 451
+    },
+    {
+      "epoch": 0.41072239890958656,
+      "grad_norm": 5.578380076669674,
+      "learning_rate": 1.2164948453608246e-06,
+      "loss": -0.2178,
+      "step": 452
+    },
+    {
+      "epoch": 0.4116310767832803,
+      "grad_norm": 4.1049709145262465,
+      "learning_rate": 1.2146204311152764e-06,
+      "loss": -0.2159,
+      "step": 453
+    },
+    {
+      "epoch": 0.4125397546569741,
+      "grad_norm": 5.970437016170801,
+      "learning_rate": 1.212746016869728e-06,
+      "loss": -0.2216,
+      "step": 454
+    },
+    {
+      "epoch": 0.4134484325306679,
+      "grad_norm": 5.2133326023293165,
+      "learning_rate": 1.2108716026241799e-06,
+      "loss": -0.1866,
+      "step": 455
+    },
+    {
+      "epoch": 0.41435711040436163,
+      "grad_norm": 4.947817296151278,
+      "learning_rate": 1.2089971883786317e-06,
+      "loss": -0.1818,
+      "step": 456
+    },
+    {
+      "epoch": 0.41526578827805544,
+      "grad_norm": 5.087266036518865,
+      "learning_rate": 1.2071227741330834e-06,
+      "loss": -0.0788,
+      "step": 457
+    },
+    {
+      "epoch": 0.4161744661517492,
+      "grad_norm": 4.999213729071357,
+      "learning_rate": 1.2052483598875352e-06,
+      "loss": -0.2006,
+      "step": 458
+    },
+    {
+      "epoch": 0.417083144025443,
+      "grad_norm": 3.4540979732098966,
+      "learning_rate": 1.2033739456419868e-06,
+      "loss": -0.2898,
+      "step": 459
+    },
+    {
+      "epoch": 0.41799182189913675,
+      "grad_norm": 3.7851815710018006,
+      "learning_rate": 1.2014995313964385e-06,
+      "loss": -0.1236,
+      "step": 460
+    },
+    {
+      "epoch": 0.41890049977283056,
+      "grad_norm": 3.8917963800044224,
+      "learning_rate": 1.1996251171508903e-06,
+      "loss": -0.0274,
+      "step": 461
+    },
+    {
+      "epoch": 0.4198091776465243,
+      "grad_norm": 5.20247941694537,
+      "learning_rate": 1.197750702905342e-06,
+      "loss": -0.252,
+      "step": 462
+    },
+    {
+      "epoch": 0.42071785552021806,
+      "grad_norm": 5.937866842080356,
+      "learning_rate": 1.1958762886597938e-06,
+      "loss": -0.2282,
+      "step": 463
+    },
+    {
+      "epoch": 0.4216265333939119,
+      "grad_norm": 4.464159548475297,
+      "learning_rate": 1.1940018744142454e-06,
+      "loss": -0.1432,
+      "step": 464
+    },
+    {
+      "epoch": 0.4225352112676056,
+      "grad_norm": 11.670850675643317,
+      "learning_rate": 1.1921274601686973e-06,
+      "loss": -0.0756,
+      "step": 465
+    },
+    {
+      "epoch": 0.42344388914129943,
+      "grad_norm": 4.784631269033721,
+      "learning_rate": 1.1902530459231491e-06,
+      "loss": -0.1903,
+      "step": 466
+    },
+    {
+      "epoch": 0.4243525670149932,
+      "grad_norm": 4.106995564538471,
+      "learning_rate": 1.1883786316776007e-06,
+      "loss": -0.126,
+      "step": 467
+    },
+    {
+      "epoch": 0.42526124488868694,
+      "grad_norm": 3.811457381969888,
+      "learning_rate": 1.1865042174320526e-06,
+      "loss": -0.1858,
+      "step": 468
+    },
+    {
+      "epoch": 0.42616992276238075,
+      "grad_norm": 4.85507590928926,
+      "learning_rate": 1.184629803186504e-06,
+      "loss": -0.0895,
+      "step": 469
+    },
+    {
+      "epoch": 0.4270786006360745,
+      "grad_norm": 4.751604110500454,
+      "learning_rate": 1.1827553889409559e-06,
+      "loss": -0.172,
+      "step": 470
+    },
+    {
+      "epoch": 0.4279872785097683,
+      "grad_norm": 3.4826885983456624,
+      "learning_rate": 1.1808809746954075e-06,
+      "loss": -0.1564,
+      "step": 471
+    },
+    {
+      "epoch": 0.42889595638346206,
+      "grad_norm": 4.4248319518512815,
+      "learning_rate": 1.1790065604498593e-06,
+      "loss": -0.1087,
+      "step": 472
+    },
+    {
+      "epoch": 0.4298046342571558,
+      "grad_norm": 4.06216992724606,
+      "learning_rate": 1.1771321462043112e-06,
+      "loss": -0.2302,
+      "step": 473
+    },
+    {
+      "epoch": 0.4307133121308496,
+      "grad_norm": 4.91791348354322,
+      "learning_rate": 1.1752577319587628e-06,
+      "loss": -0.2286,
+      "step": 474
+    },
+    {
+      "epoch": 0.4316219900045434,
+      "grad_norm": 5.814520043912864,
+      "learning_rate": 1.1733833177132147e-06,
+      "loss": -0.0977,
+      "step": 475
+    },
+    {
+      "epoch": 0.4325306678782372,
+      "grad_norm": 3.437724851710342,
+      "learning_rate": 1.1715089034676663e-06,
+      "loss": -0.2244,
+      "step": 476
+    },
+    {
+      "epoch": 0.43343934575193094,
+      "grad_norm": 5.88345345580726,
+      "learning_rate": 1.169634489222118e-06,
+      "loss": -0.2032,
+      "step": 477
+    },
+    {
+      "epoch": 0.4343480236256247,
+      "grad_norm": 3.9753346524229736,
+      "learning_rate": 1.1677600749765698e-06,
+      "loss": -0.1939,
+      "step": 478
+    },
+    {
+      "epoch": 0.4352567014993185,
+      "grad_norm": 6.388427356506096,
+      "learning_rate": 1.1658856607310214e-06,
+      "loss": -0.1359,
+      "step": 479
+    },
+    {
+      "epoch": 0.43616537937301225,
+      "grad_norm": 4.639731375712341,
+      "learning_rate": 1.1640112464854733e-06,
+      "loss": -0.1335,
+      "step": 480
+    },
+    {
+      "epoch": 0.43707405724670606,
+      "grad_norm": 3.8538878193926314,
+      "learning_rate": 1.1621368322399249e-06,
+      "loss": -0.2333,
+      "step": 481
+    },
+    {
+      "epoch": 0.4379827351203998,
+      "grad_norm": 3.5538742308852456,
+      "learning_rate": 1.1602624179943767e-06,
+      "loss": -0.0918,
+      "step": 482
+    },
+    {
+      "epoch": 0.4388914129940936,
+      "grad_norm": 5.872253750071144,
+      "learning_rate": 1.1583880037488286e-06,
+      "loss": -0.0509,
+      "step": 483
+    },
+    {
+      "epoch": 0.4398000908677874,
+      "grad_norm": 7.165461978646711,
+      "learning_rate": 1.1565135895032802e-06,
+      "loss": -0.1956,
+      "step": 484
+    },
+    {
+      "epoch": 0.4407087687414811,
+      "grad_norm": 7.312005963713303,
+      "learning_rate": 1.1546391752577318e-06,
+      "loss": -0.0751,
+      "step": 485
+    },
+    {
+      "epoch": 0.44161744661517494,
+      "grad_norm": 4.681862862046344,
+      "learning_rate": 1.1527647610121835e-06,
+      "loss": -0.0558,
+      "step": 486
+    },
+    {
+      "epoch": 0.4425261244888687,
+      "grad_norm": 5.479032952401898,
+      "learning_rate": 1.1508903467666353e-06,
+      "loss": -0.2781,
+      "step": 487
+    },
+    {
+      "epoch": 0.4434348023625625,
+      "grad_norm": 6.642946802287188,
+      "learning_rate": 1.1490159325210872e-06,
+      "loss": -0.2191,
+      "step": 488
+    },
+    {
+      "epoch": 0.44434348023625625,
+      "grad_norm": 4.788368261900109,
+      "learning_rate": 1.1471415182755388e-06,
+      "loss": -0.2368,
+      "step": 489
+    },
+    {
+      "epoch": 0.44525215810995,
+      "grad_norm": 12.068778346071385,
+      "learning_rate": 1.1452671040299906e-06,
+      "loss": -0.1796,
+      "step": 490
+    },
+    {
+      "epoch": 0.4461608359836438,
+      "grad_norm": 3.9536312216405514,
+      "learning_rate": 1.1433926897844423e-06,
+      "loss": -0.1197,
+      "step": 491
+    },
+    {
+      "epoch": 0.44706951385733756,
+      "grad_norm": 6.614987274411418,
+      "learning_rate": 1.1415182755388941e-06,
+      "loss": -0.248,
+      "step": 492
+    },
+    {
+      "epoch": 0.44797819173103137,
+      "grad_norm": 3.7943923431537994,
+      "learning_rate": 1.139643861293346e-06,
+      "loss": -0.1646,
+      "step": 493
+    },
+    {
+      "epoch": 0.4488868696047251,
+      "grad_norm": 5.855016608210593,
+      "learning_rate": 1.1377694470477974e-06,
+      "loss": -0.2134,
+      "step": 494
+    },
+    {
+      "epoch": 0.4497955474784189,
+      "grad_norm": 4.708656885063392,
+      "learning_rate": 1.1358950328022492e-06,
+      "loss": -0.2292,
+      "step": 495
+    },
+    {
+      "epoch": 0.4507042253521127,
+      "grad_norm": 4.501034299877835,
+      "learning_rate": 1.1340206185567009e-06,
+      "loss": -0.2709,
+      "step": 496
+    },
+    {
+      "epoch": 0.45161290322580644,
+      "grad_norm": 4.712901147418838,
+      "learning_rate": 1.1321462043111527e-06,
+      "loss": -0.3342,
+      "step": 497
+    },
+    {
+      "epoch": 0.45252158109950025,
+      "grad_norm": 4.683578478459777,
+      "learning_rate": 1.1302717900656046e-06,
+      "loss": -0.1525,
+      "step": 498
+    },
+    {
+      "epoch": 0.453430258973194,
+      "grad_norm": 5.03241161892293,
+      "learning_rate": 1.1283973758200562e-06,
+      "loss": -0.2474,
+      "step": 499
+    },
+    {
+      "epoch": 0.45433893684688775,
+      "grad_norm": 5.504151424797876,
+      "learning_rate": 1.126522961574508e-06,
+      "loss": -0.3001,
+      "step": 500
+    },
+    {
+      "epoch": 0.45524761472058156,
+      "grad_norm": 4.309639465759184,
+      "learning_rate": 1.1246485473289597e-06,
+      "loss": -0.2877,
+      "step": 501
+    },
+    {
+      "epoch": 0.4561562925942753,
+      "grad_norm": 5.4149995243428055,
+      "learning_rate": 1.1227741330834113e-06,
+      "loss": -0.2794,
+      "step": 502
+    },
+    {
+      "epoch": 0.4570649704679691,
+      "grad_norm": 4.059552576000164,
+      "learning_rate": 1.1208997188378631e-06,
+      "loss": -0.1102,
+      "step": 503
+    },
+    {
+      "epoch": 0.4579736483416629,
+      "grad_norm": 3.869571944819325,
+      "learning_rate": 1.1190253045923148e-06,
+      "loss": -0.2288,
+      "step": 504
+    },
+    {
+      "epoch": 0.45888232621535663,
+      "grad_norm": 4.17252736223779,
+      "learning_rate": 1.1171508903467666e-06,
+      "loss": -0.09,
+      "step": 505
+    },
+    {
+      "epoch": 0.45979100408905044,
+      "grad_norm": 6.0016804567090505,
+      "learning_rate": 1.1152764761012183e-06,
+      "loss": -0.3503,
+      "step": 506
+    },
+    {
+      "epoch": 0.4606996819627442,
+      "grad_norm": 4.3717861905461035,
+      "learning_rate": 1.11340206185567e-06,
+      "loss": -0.2743,
+      "step": 507
+    },
+    {
+      "epoch": 0.461608359836438,
+      "grad_norm": 5.097581791242609,
+      "learning_rate": 1.111527647610122e-06,
+      "loss": -0.1792,
+      "step": 508
+    },
+    {
+      "epoch": 0.46251703771013175,
+      "grad_norm": 4.615262064162602,
+      "learning_rate": 1.1096532333645736e-06,
+      "loss": -0.256,
+      "step": 509
+    },
+    {
+      "epoch": 0.46342571558382556,
+      "grad_norm": 5.078595230632341,
+      "learning_rate": 1.1077788191190252e-06,
+      "loss": -0.1023,
+      "step": 510
+    },
+    {
+      "epoch": 0.4643343934575193,
+      "grad_norm": 3.9871220711533675,
+      "learning_rate": 1.1059044048734769e-06,
+      "loss": -0.2304,
+      "step": 511
+    },
+    {
+      "epoch": 0.46524307133121307,
+      "grad_norm": 5.028613614084974,
+      "learning_rate": 1.1040299906279287e-06,
+      "loss": -0.2209,
+      "step": 512
+    },
+    {
+      "epoch": 0.4661517492049069,
+      "grad_norm": 4.1773472018465725,
+      "learning_rate": 1.1021555763823805e-06,
+      "loss": -0.1468,
+      "step": 513
+    },
+    {
+      "epoch": 0.4670604270786006,
+      "grad_norm": 5.228789763020145,
+      "learning_rate": 1.1002811621368322e-06,
+      "loss": -0.2038,
+      "step": 514
+    },
+    {
+      "epoch": 0.46796910495229443,
+      "grad_norm": 3.7515154001885107,
+      "learning_rate": 1.098406747891284e-06,
+      "loss": -0.25,
+      "step": 515
+    },
+    {
+      "epoch": 0.4688777828259882,
+      "grad_norm": 4.396684585276414,
+      "learning_rate": 1.0965323336457357e-06,
+      "loss": -0.2023,
+      "step": 516
+    },
+    {
+      "epoch": 0.46978646069968194,
+      "grad_norm": 5.3191666006241345,
+      "learning_rate": 1.0946579194001875e-06,
+      "loss": -0.1258,
+      "step": 517
+    },
+    {
+      "epoch": 0.47069513857337575,
+      "grad_norm": 4.270346240047485,
+      "learning_rate": 1.092783505154639e-06,
+      "loss": -0.1124,
+      "step": 518
+    },
+    {
+      "epoch": 0.4716038164470695,
+      "grad_norm": 4.907395344578748,
+      "learning_rate": 1.0909090909090908e-06,
+      "loss": -0.2615,
+      "step": 519
+    },
+    {
+      "epoch": 0.4725124943207633,
+      "grad_norm": 5.466243321998926,
+      "learning_rate": 1.0890346766635426e-06,
+      "loss": -0.2162,
+      "step": 520
+    },
+    {
+      "epoch": 0.47342117219445706,
+      "grad_norm": 4.276872164714029,
+      "learning_rate": 1.0871602624179942e-06,
+      "loss": -0.2768,
+      "step": 521
+    },
+    {
+      "epoch": 0.4743298500681508,
+      "grad_norm": 4.098910275298337,
+      "learning_rate": 1.085285848172446e-06,
+      "loss": -0.1488,
+      "step": 522
+    },
+    {
+      "epoch": 0.4752385279418446,
+      "grad_norm": 4.069513224142383,
+      "learning_rate": 1.0834114339268977e-06,
+      "loss": -0.2399,
+      "step": 523
+    },
+    {
+      "epoch": 0.4761472058155384,
+      "grad_norm": 4.960475918179512,
+      "learning_rate": 1.0815370196813496e-06,
+      "loss": -0.2082,
+      "step": 524
+    },
+    {
+      "epoch": 0.4770558836892322,
+      "grad_norm": 4.028222181569904,
+      "learning_rate": 1.0796626054358014e-06,
+      "loss": -0.2219,
+      "step": 525
+    },
+    {
+      "epoch": 0.47796456156292594,
+      "grad_norm": 4.095040853107945,
+      "learning_rate": 1.077788191190253e-06,
+      "loss": -0.129,
+      "step": 526
+    },
+    {
+      "epoch": 0.4788732394366197,
+      "grad_norm": 4.200161367677005,
+      "learning_rate": 1.0759137769447047e-06,
+      "loss": -0.1787,
+      "step": 527
+    },
+    {
+      "epoch": 0.4797819173103135,
+      "grad_norm": 5.33086246633786,
+      "learning_rate": 1.0740393626991563e-06,
+      "loss": -0.1801,
+      "step": 528
+    },
+    {
+      "epoch": 0.48069059518400725,
+      "grad_norm": 4.724223619082885,
+      "learning_rate": 1.0721649484536082e-06,
+      "loss": -0.2577,
+      "step": 529
+    },
+    {
+      "epoch": 0.48159927305770106,
+      "grad_norm": 5.390221091355742,
+      "learning_rate": 1.07029053420806e-06,
+      "loss": -0.1355,
+      "step": 530
+    },
+    {
+      "epoch": 0.4825079509313948,
+      "grad_norm": 4.936929211036962,
+      "learning_rate": 1.0684161199625116e-06,
+      "loss": -0.1106,
+      "step": 531
+    },
+    {
+      "epoch": 0.4834166288050886,
+      "grad_norm": 4.532409730023912,
+      "learning_rate": 1.0665417057169635e-06,
+      "loss": -0.22,
+      "step": 532
+    },
+    {
+      "epoch": 0.4843253066787824,
+      "grad_norm": 5.3280482132823765,
+      "learning_rate": 1.0646672914714151e-06,
+      "loss": -0.2152,
+      "step": 533
+    },
+    {
+      "epoch": 0.4852339845524761,
+      "grad_norm": 5.469767971661902,
+      "learning_rate": 1.062792877225867e-06,
+      "loss": -0.2924,
+      "step": 534
+    },
+    {
+      "epoch": 0.48614266242616994,
+      "grad_norm": 4.970232088745073,
+      "learning_rate": 1.0609184629803186e-06,
+      "loss": -0.2166,
+      "step": 535
+    },
+    {
+      "epoch": 0.4870513402998637,
+      "grad_norm": 4.926586596535583,
+      "learning_rate": 1.0590440487347702e-06,
+      "loss": -0.1833,
+      "step": 536
+    },
+    {
+      "epoch": 0.4879600181735575,
+      "grad_norm": 5.729211462597125,
+      "learning_rate": 1.057169634489222e-06,
+      "loss": -0.1348,
+      "step": 537
+    },
+    {
+      "epoch": 0.48886869604725125,
+      "grad_norm": 5.289882669571879,
+      "learning_rate": 1.0552952202436737e-06,
+      "loss": -0.1282,
+      "step": 538
+    },
+    {
+      "epoch": 0.489777373920945,
+      "grad_norm": 3.7550531038676684,
+      "learning_rate": 1.0534208059981255e-06,
+      "loss": -0.1749,
+      "step": 539
+    },
+    {
+      "epoch": 0.4906860517946388,
+      "grad_norm": 3.6339154281726658,
+      "learning_rate": 1.0515463917525774e-06,
+      "loss": -0.1992,
+      "step": 540
+    },
+    {
+      "epoch": 0.49159472966833256,
+      "grad_norm": 4.4321912790742415,
+      "learning_rate": 1.049671977507029e-06,
+      "loss": -0.247,
+      "step": 541
+    },
+    {
+      "epoch": 0.4925034075420264,
+      "grad_norm": 5.877518783532831,
+      "learning_rate": 1.0477975632614809e-06,
+      "loss": -0.1868,
+      "step": 542
+    },
+    {
+      "epoch": 0.4934120854157201,
+      "grad_norm": 4.72496986985061,
+      "learning_rate": 1.0459231490159323e-06,
+      "loss": -0.0836,
+      "step": 543
+    },
+    {
+      "epoch": 0.4943207632894139,
+      "grad_norm": 5.120428139548038,
+      "learning_rate": 1.0440487347703841e-06,
+      "loss": -0.1283,
+      "step": 544
+    },
+    {
+      "epoch": 0.4952294411631077,
+      "grad_norm": 4.818257838347981,
+      "learning_rate": 1.042174320524836e-06,
+      "loss": -0.2365,
+      "step": 545
+    },
+    {
+      "epoch": 0.49613811903680144,
+      "grad_norm": 4.0017281613807665,
+      "learning_rate": 1.0402999062792876e-06,
+      "loss": -0.1042,
+      "step": 546
+    },
+    {
+      "epoch": 0.49704679691049525,
+      "grad_norm": 4.516154221934723,
+      "learning_rate": 1.0384254920337395e-06,
+      "loss": -0.1229,
+      "step": 547
+    },
+    {
+      "epoch": 0.497955474784189,
+      "grad_norm": 4.4349456742495645,
+      "learning_rate": 1.036551077788191e-06,
+      "loss": -0.1578,
+      "step": 548
+    },
+    {
+      "epoch": 0.49886415265788275,
+      "grad_norm": 4.268493582088082,
+      "learning_rate": 1.034676663542643e-06,
+      "loss": -0.2186,
+      "step": 549
+    },
+    {
+      "epoch": 0.49977283053157656,
+      "grad_norm": 4.146709994818535,
+      "learning_rate": 1.0328022492970948e-06,
+      "loss": -0.1732,
+      "step": 550
+    },
+    {
+      "epoch": 0.5006815084052704,
+      "grad_norm": 5.550814580850844,
+      "learning_rate": 1.0309278350515462e-06,
+      "loss": -0.2143,
+      "step": 551
+    },
+    {
+      "epoch": 0.5015901862789641,
+      "grad_norm": 6.215217238639879,
+      "learning_rate": 1.029053420805998e-06,
+      "loss": -0.2036,
+      "step": 552
+    },
+    {
+      "epoch": 0.5024988641526579,
+      "grad_norm": 4.771900685059496,
+      "learning_rate": 1.0271790065604497e-06,
+      "loss": -0.2799,
+      "step": 553
+    },
+    {
+      "epoch": 0.5034075420263516,
+      "grad_norm": 3.939001130258127,
+      "learning_rate": 1.0253045923149015e-06,
+      "loss": -0.1153,
+      "step": 554
+    },
+    {
+      "epoch": 0.5043162199000454,
+      "grad_norm": 5.29392666752735,
+      "learning_rate": 1.0234301780693534e-06,
+      "loss": -0.2497,
+      "step": 555
+    },
+    {
+      "epoch": 0.5052248977737392,
+      "grad_norm": 3.4061028684984627,
+      "learning_rate": 1.021555763823805e-06,
+      "loss": -0.1171,
+      "step": 556
+    },
+    {
+      "epoch": 0.506133575647433,
+      "grad_norm": 4.697810113322999,
+      "learning_rate": 1.0196813495782569e-06,
+      "loss": -0.1885,
+      "step": 557
+    },
+    {
+      "epoch": 0.5070422535211268,
+      "grad_norm": 5.886307244450821,
+      "learning_rate": 1.0178069353327085e-06,
+      "loss": -0.237,
+      "step": 558
+    },
+    {
+      "epoch": 0.5079509313948205,
+      "grad_norm": 3.822197260023513,
+      "learning_rate": 1.0159325210871603e-06,
+      "loss": -0.1806,
+      "step": 559
+    },
+    {
+      "epoch": 0.5088596092685143,
+      "grad_norm": 6.76464836701043,
+      "learning_rate": 1.014058106841612e-06,
+      "loss": -0.1412,
+      "step": 560
+    },
+    {
+      "epoch": 0.5097682871422081,
+      "grad_norm": 6.8697567452754695,
+      "learning_rate": 1.0121836925960636e-06,
+      "loss": -0.1427,
+      "step": 561
+    },
+    {
+      "epoch": 0.5106769650159019,
+      "grad_norm": 4.555903434393186,
+      "learning_rate": 1.0103092783505154e-06,
+      "loss": -0.2413,
+      "step": 562
+    },
+    {
+      "epoch": 0.5115856428895956,
+      "grad_norm": 3.6788342047389184,
+      "learning_rate": 1.008434864104967e-06,
+      "loss": -0.1408,
+      "step": 563
+    },
+    {
+      "epoch": 0.5124943207632894,
+      "grad_norm": 8.168112594381409,
+      "learning_rate": 1.006560449859419e-06,
+      "loss": -0.1763,
+      "step": 564
+    },
+    {
+      "epoch": 0.5134029986369832,
+      "grad_norm": 4.856309911488444,
+      "learning_rate": 1.0046860356138708e-06,
+      "loss": -0.112,
+      "step": 565
+    },
+    {
+      "epoch": 0.514311676510677,
+      "grad_norm": 4.443125637739637,
+      "learning_rate": 1.0028116213683224e-06,
+      "loss": -0.1384,
+      "step": 566
+    },
+    {
+      "epoch": 0.5152203543843707,
+      "grad_norm": 3.66386936874402,
+      "learning_rate": 1.0009372071227742e-06,
+      "loss": -0.2491,
+      "step": 567
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 5.1612029360479825,
+      "learning_rate": 9.990627928772259e-07,
+      "loss": -0.2171,
+      "step": 568
+    },
+    {
+      "epoch": 0.5170377101317583,
+      "grad_norm": 4.510341206096292,
+      "learning_rate": 9.971883786316775e-07,
+      "loss": -0.3559,
+      "step": 569
+    },
+    {
+      "epoch": 0.5179463880054521,
+      "grad_norm": 3.873329417762375,
+      "learning_rate": 9.953139643861291e-07,
+      "loss": -0.1712,
+      "step": 570
+    },
+    {
+      "epoch": 0.5188550658791459,
+      "grad_norm": 4.4512290800267715,
+      "learning_rate": 9.93439550140581e-07,
+      "loss": -0.2435,
+      "step": 571
+    },
+    {
+      "epoch": 0.5197637437528396,
+      "grad_norm": 5.0176087256389765,
+      "learning_rate": 9.915651358950328e-07,
+      "loss": -0.1425,
+      "step": 572
+    },
+    {
+      "epoch": 0.5206724216265334,
+      "grad_norm": 4.073162464942973,
+      "learning_rate": 9.896907216494845e-07,
+      "loss": -0.1998,
+      "step": 573
+    },
+    {
+      "epoch": 0.5215810995002271,
+      "grad_norm": 4.716207059072772,
+      "learning_rate": 9.87816307403936e-07,
+      "loss": -0.2262,
+      "step": 574
+    },
+    {
+      "epoch": 0.522489777373921,
+      "grad_norm": 3.9997863712484696,
+      "learning_rate": 9.85941893158388e-07,
+      "loss": -0.3002,
+      "step": 575
+    },
+    {
+      "epoch": 0.5233984552476147,
+      "grad_norm": 4.1829141586655405,
+      "learning_rate": 9.840674789128398e-07,
+      "loss": -0.1131,
+      "step": 576
+    },
+    {
+      "epoch": 0.5243071331213085,
+      "grad_norm": 3.831099495101847,
+      "learning_rate": 9.821930646672914e-07,
+      "loss": -0.1784,
+      "step": 577
+    },
+    {
+      "epoch": 0.5252158109950023,
+      "grad_norm": 5.184012838065181,
+      "learning_rate": 9.80318650421743e-07,
+      "loss": -0.2398,
+      "step": 578
+    },
+    {
+      "epoch": 0.526124488868696,
+      "grad_norm": 6.0904331668337,
+      "learning_rate": 9.78444236176195e-07,
+      "loss": -0.0988,
+      "step": 579
+    },
+    {
+      "epoch": 0.5270331667423899,
+      "grad_norm": 3.8564720201183316,
+      "learning_rate": 9.765698219306465e-07,
+      "loss": -0.1478,
+      "step": 580
+    },
+    {
+      "epoch": 0.5279418446160836,
+      "grad_norm": 4.278960792370793,
+      "learning_rate": 9.746954076850984e-07,
+      "loss": -0.1976,
+      "step": 581
+    },
+    {
+      "epoch": 0.5288505224897774,
+      "grad_norm": 5.334598490542456,
+      "learning_rate": 9.7282099343955e-07,
+      "loss": -0.1793,
+      "step": 582
+    },
+    {
+      "epoch": 0.5297592003634711,
+      "grad_norm": 3.987478446304282,
+      "learning_rate": 9.709465791940019e-07,
+      "loss": -0.1429,
+      "step": 583
+    },
+    {
+      "epoch": 0.5306678782371649,
+      "grad_norm": 4.641496672790511,
+      "learning_rate": 9.690721649484535e-07,
+      "loss": -0.1218,
+      "step": 584
+    },
+    {
+      "epoch": 0.5315765561108587,
+      "grad_norm": 5.731720770051507,
+      "learning_rate": 9.671977507029053e-07,
+      "loss": -0.1258,
+      "step": 585
+    },
+    {
+      "epoch": 0.5324852339845525,
+      "grad_norm": 11.624391601393127,
+      "learning_rate": 9.653233364573572e-07,
+      "loss": -0.0567,
+      "step": 586
+    },
+    {
+      "epoch": 0.5333939118582463,
+      "grad_norm": 5.955677036134534,
+      "learning_rate": 9.634489222118088e-07,
+      "loss": -0.165,
+      "step": 587
+    },
+    {
+      "epoch": 0.53430258973194,
+      "grad_norm": 6.902435215774139,
+      "learning_rate": 9.615745079662604e-07,
+      "loss": -0.1064,
+      "step": 588
+    },
+    {
+      "epoch": 0.5352112676056338,
+      "grad_norm": 4.242301925324311,
+      "learning_rate": 9.597000937207123e-07,
+      "loss": -0.1578,
+      "step": 589
+    },
+    {
+      "epoch": 0.5361199454793276,
+      "grad_norm": 4.273440847447843,
+      "learning_rate": 9.57825679475164e-07,
+      "loss": -0.1642,
+      "step": 590
+    },
+    {
+      "epoch": 0.5370286233530214,
+      "grad_norm": 4.870140588129313,
+      "learning_rate": 9.559512652296156e-07,
+      "loss": -0.2159,
+      "step": 591
+    },
+    {
+      "epoch": 0.5379373012267151,
+      "grad_norm": 3.7449161400974393,
+      "learning_rate": 9.540768509840674e-07,
+      "loss": -0.1467,
+      "step": 592
+    },
+    {
+      "epoch": 0.5388459791004089,
+      "grad_norm": 8.05987174791097,
+      "learning_rate": 9.522024367385191e-07,
+      "loss": -0.1108,
+      "step": 593
+    },
+    {
+      "epoch": 0.5397546569741026,
+      "grad_norm": 6.293399429028084,
+      "learning_rate": 9.50328022492971e-07,
+      "loss": -0.0882,
+      "step": 594
+    },
+    {
+      "epoch": 0.5406633348477965,
+      "grad_norm": 3.460711262974522,
+      "learning_rate": 9.484536082474226e-07,
+      "loss": -0.2517,
+      "step": 595
+    },
+    {
+      "epoch": 0.5415720127214902,
+      "grad_norm": 3.636467651483462,
+      "learning_rate": 9.465791940018744e-07,
+      "loss": -0.1595,
+      "step": 596
+    },
+    {
+      "epoch": 0.542480690595184,
+      "grad_norm": 5.734445514297377,
+      "learning_rate": 9.447047797563261e-07,
+      "loss": -0.1867,
+      "step": 597
+    },
+    {
+      "epoch": 0.5433893684688778,
+      "grad_norm": 4.484079809807432,
+      "learning_rate": 9.428303655107778e-07,
+      "loss": -0.162,
+      "step": 598
+    },
+    {
+      "epoch": 0.5442980463425715,
+      "grad_norm": 4.467073352765685,
+      "learning_rate": 9.409559512652296e-07,
+      "loss": -0.1856,
+      "step": 599
+    },
+    {
+      "epoch": 0.5452067242162654,
+      "grad_norm": 4.633329777521832,
+      "learning_rate": 9.390815370196813e-07,
+      "loss": -0.1866,
+      "step": 600
+    },
+    {
+      "epoch": 0.5461154020899591,
+      "grad_norm": 5.561064288416153,
+      "learning_rate": 9.372071227741331e-07,
+      "loss": -0.1792,
+      "step": 601
+    },
+    {
+      "epoch": 0.5470240799636529,
+      "grad_norm": 3.8631525124293957,
+      "learning_rate": 9.353327085285848e-07,
+      "loss": -0.1398,
+      "step": 602
+    },
+    {
+      "epoch": 0.5479327578373466,
+      "grad_norm": 4.647747084691338,
+      "learning_rate": 9.334582942830364e-07,
+      "loss": -0.1798,
+      "step": 603
+    },
+    {
+      "epoch": 0.5488414357110404,
+      "grad_norm": 3.9624514122520176,
+      "learning_rate": 9.315838800374883e-07,
+      "loss": -0.1884,
+      "step": 604
+    },
+    {
+      "epoch": 0.5497501135847342,
+      "grad_norm": 3.4242751530939626,
+      "learning_rate": 9.2970946579194e-07,
+      "loss": -0.2381,
+      "step": 605
+    },
+    {
+      "epoch": 0.550658791458428,
+      "grad_norm": 4.371186692627831,
+      "learning_rate": 9.278350515463918e-07,
+      "loss": -0.3577,
+      "step": 606
+    },
+    {
+      "epoch": 0.5515674693321218,
+      "grad_norm": 5.5484886723540505,
+      "learning_rate": 9.259606373008434e-07,
+      "loss": -0.1239,
+      "step": 607
+    },
+    {
+      "epoch": 0.5524761472058155,
+      "grad_norm": 4.670799310147966,
+      "learning_rate": 9.240862230552951e-07,
+      "loss": -0.1944,
+      "step": 608
+    },
+    {
+      "epoch": 0.5533848250795094,
+      "grad_norm": 3.812132145379466,
+      "learning_rate": 9.222118088097469e-07,
+      "loss": -0.2146,
+      "step": 609
+    },
+    {
+      "epoch": 0.5542935029532031,
+      "grad_norm": 6.121334614058881,
+      "learning_rate": 9.203373945641987e-07,
+      "loss": -0.1604,
+      "step": 610
+    },
+    {
+      "epoch": 0.5552021808268969,
+      "grad_norm": 5.4376916961203,
+      "learning_rate": 9.184629803186503e-07,
+      "loss": -0.2165,
+      "step": 611
+    },
+    {
+      "epoch": 0.5561108587005906,
+      "grad_norm": 4.350257006539868,
+      "learning_rate": 9.165885660731021e-07,
+      "loss": -0.1754,
+      "step": 612
+    },
+    {
+      "epoch": 0.5570195365742844,
+      "grad_norm": 4.299463664473049,
+      "learning_rate": 9.147141518275538e-07,
+      "loss": -0.1626,
+      "step": 613
+    },
+    {
+      "epoch": 0.5579282144479782,
+      "grad_norm": 4.004089887175981,
+      "learning_rate": 9.128397375820056e-07,
+      "loss": -0.2304,
+      "step": 614
+    },
+    {
+      "epoch": 0.558836892321672,
+      "grad_norm": 3.5103731388521378,
+      "learning_rate": 9.109653233364574e-07,
+      "loss": -0.2567,
+      "step": 615
+    },
+    {
+      "epoch": 0.5597455701953657,
+      "grad_norm": 5.3093837798586545,
+      "learning_rate": 9.09090909090909e-07,
+      "loss": -0.1589,
+      "step": 616
+    },
+    {
+      "epoch": 0.5606542480690595,
+      "grad_norm": 5.407253282538505,
+      "learning_rate": 9.072164948453608e-07,
+      "loss": -0.2232,
+      "step": 617
+    },
+    {
+      "epoch": 0.5615629259427533,
+      "grad_norm": 5.24488072349705,
+      "learning_rate": 9.053420805998125e-07,
+      "loss": -0.1575,
+      "step": 618
+    },
+    {
+      "epoch": 0.5624716038164471,
+      "grad_norm": 4.111342970578898,
+      "learning_rate": 9.034676663542643e-07,
+      "loss": -0.2456,
+      "step": 619
+    },
+    {
+      "epoch": 0.5633802816901409,
+      "grad_norm": 3.8995345106851427,
+      "learning_rate": 9.01593252108716e-07,
+      "loss": -0.2413,
+      "step": 620
+    },
+    {
+      "epoch": 0.5642889595638346,
+      "grad_norm": 7.207016429524057,
+      "learning_rate": 8.997188378631677e-07,
+      "loss": -0.2513,
+      "step": 621
+    },
+    {
+      "epoch": 0.5651976374375284,
+      "grad_norm": 5.315847441758262,
+      "learning_rate": 8.978444236176195e-07,
+      "loss": -0.2215,
+      "step": 622
+    },
+    {
+      "epoch": 0.5661063153112221,
+      "grad_norm": 4.025379726907858,
+      "learning_rate": 8.959700093720712e-07,
+      "loss": -0.1783,
+      "step": 623
+    },
+    {
+      "epoch": 0.567014993184916,
+      "grad_norm": 5.741293576783411,
+      "learning_rate": 8.940955951265228e-07,
+      "loss": -0.1972,
+      "step": 624
+    },
+    {
+      "epoch": 0.5679236710586097,
+      "grad_norm": 4.753061211613801,
+      "learning_rate": 8.922211808809747e-07,
+      "loss": -0.1872,
+      "step": 625
+    },
+    {
+      "epoch": 0.5688323489323035,
+      "grad_norm": 4.336842045640238,
+      "learning_rate": 8.903467666354264e-07,
+      "loss": -0.341,
+      "step": 626
+    },
+    {
+      "epoch": 0.5697410268059973,
+      "grad_norm": 4.782319591288078,
+      "learning_rate": 8.884723523898782e-07,
+      "loss": -0.0698,
+      "step": 627
+    },
+    {
+      "epoch": 0.570649704679691,
+      "grad_norm": 4.7092847059841905,
+      "learning_rate": 8.865979381443298e-07,
+      "loss": -0.3309,
+      "step": 628
+    },
+    {
+      "epoch": 0.5715583825533849,
+      "grad_norm": 4.741338763885763,
+      "learning_rate": 8.847235238987815e-07,
+      "loss": -0.2014,
+      "step": 629
+    },
+    {
+      "epoch": 0.5724670604270786,
+      "grad_norm": 4.4393448554219495,
+      "learning_rate": 8.828491096532334e-07,
+      "loss": -0.2238,
+      "step": 630
+    },
+    {
+      "epoch": 0.5733757383007724,
+      "grad_norm": 5.116418306036815,
+      "learning_rate": 8.809746954076851e-07,
+      "loss": -0.2524,
+      "step": 631
+    },
+    {
+      "epoch": 0.5742844161744661,
+      "grad_norm": 6.708241464052935,
+      "learning_rate": 8.791002811621368e-07,
+      "loss": -0.0821,
+      "step": 632
+    },
+    {
+      "epoch": 0.5751930940481599,
+      "grad_norm": 4.156001033353264,
+      "learning_rate": 8.772258669165885e-07,
+      "loss": -0.1872,
+      "step": 633
+    },
+    {
+      "epoch": 0.5761017719218537,
+      "grad_norm": 4.224638044213286,
+      "learning_rate": 8.753514526710402e-07,
+      "loss": -0.1332,
+      "step": 634
+    },
+    {
+      "epoch": 0.5770104497955475,
+      "grad_norm": 3.3778650344347585,
+      "learning_rate": 8.73477038425492e-07,
+      "loss": -0.1928,
+      "step": 635
+    },
+    {
+      "epoch": 0.5779191276692413,
+      "grad_norm": 4.775655827247423,
+      "learning_rate": 8.716026241799437e-07,
+      "loss": -0.162,
+      "step": 636
+    },
+    {
+      "epoch": 0.578827805542935,
+      "grad_norm": 4.30751279965191,
+      "learning_rate": 8.697282099343955e-07,
+      "loss": -0.0161,
+      "step": 637
+    },
+    {
+      "epoch": 0.5797364834166288,
+      "grad_norm": 5.037579458665458,
+      "learning_rate": 8.678537956888472e-07,
+      "loss": -0.2092,
+      "step": 638
+    },
+    {
+      "epoch": 0.5806451612903226,
+      "grad_norm": 5.445299539865316,
+      "learning_rate": 8.659793814432989e-07,
+      "loss": -0.2664,
+      "step": 639
+    },
+    {
+      "epoch": 0.5815538391640164,
+      "grad_norm": 4.353134227559427,
+      "learning_rate": 8.641049671977506e-07,
+      "loss": -0.159,
+      "step": 640
+    },
+    {
+      "epoch": 0.5824625170377101,
+      "grad_norm": 5.187830902943837,
+      "learning_rate": 8.622305529522024e-07,
+      "loss": -0.1676,
+      "step": 641
+    },
+    {
+      "epoch": 0.5833711949114039,
+      "grad_norm": 3.7386813692038707,
+      "learning_rate": 8.603561387066542e-07,
+      "loss": -0.3193,
+      "step": 642
+    },
+    {
+      "epoch": 0.5842798727850976,
+      "grad_norm": 7.951747812520577,
+      "learning_rate": 8.584817244611059e-07,
+      "loss": -0.1379,
+      "step": 643
+    },
+    {
+      "epoch": 0.5851885506587915,
+      "grad_norm": 6.415640744850991,
+      "learning_rate": 8.566073102155576e-07,
+      "loss": -0.172,
+      "step": 644
+    },
+    {
+      "epoch": 0.5860972285324852,
+      "grad_norm": 5.845885187609229,
+      "learning_rate": 8.547328959700093e-07,
+      "loss": -0.1652,
+      "step": 645
+    },
+    {
+      "epoch": 0.587005906406179,
+      "grad_norm": 4.819169612283129,
+      "learning_rate": 8.528584817244611e-07,
+      "loss": -0.1732,
+      "step": 646
+    },
+    {
+      "epoch": 0.5879145842798728,
+      "grad_norm": 3.6616192380333508,
+      "learning_rate": 8.509840674789128e-07,
+      "loss": -0.2016,
+      "step": 647
+    },
+    {
+      "epoch": 0.5888232621535665,
+      "grad_norm": 4.213639165868573,
+      "learning_rate": 8.491096532333646e-07,
+      "loss": -0.2572,
+      "step": 648
+    },
+    {
+      "epoch": 0.5897319400272604,
+      "grad_norm": 4.015802876729755,
+      "learning_rate": 8.472352389878162e-07,
+      "loss": -0.1288,
+      "step": 649
+    },
+    {
+      "epoch": 0.5906406179009541,
+      "grad_norm": 5.319478556382797,
+      "learning_rate": 8.45360824742268e-07,
+      "loss": -0.1614,
+      "step": 650
+    },
+    {
+      "epoch": 0.5915492957746479,
+      "grad_norm": 4.4138397430656475,
+      "learning_rate": 8.434864104967198e-07,
+      "loss": -0.2832,
+      "step": 651
+    },
+    {
+      "epoch": 0.5924579736483416,
+      "grad_norm": 4.230611219323102,
+      "learning_rate": 8.416119962511715e-07,
+      "loss": -0.2148,
+      "step": 652
+    },
+    {
+      "epoch": 0.5933666515220355,
+      "grad_norm": 4.2425116596050465,
+      "learning_rate": 8.397375820056232e-07,
+      "loss": -0.1719,
+      "step": 653
+    },
+    {
+      "epoch": 0.5942753293957292,
+      "grad_norm": 4.568490889647156,
+      "learning_rate": 8.378631677600749e-07,
+      "loss": -0.0674,
+      "step": 654
+    },
+    {
+      "epoch": 0.595184007269423,
+      "grad_norm": 3.8982234429740608,
+      "learning_rate": 8.359887535145267e-07,
+      "loss": -0.2537,
+      "step": 655
+    },
+    {
+      "epoch": 0.5960926851431168,
+      "grad_norm": 3.845633564751491,
+      "learning_rate": 8.341143392689785e-07,
+      "loss": -0.0815,
+      "step": 656
+    },
+    {
+      "epoch": 0.5970013630168105,
+      "grad_norm": 4.219887813143565,
+      "learning_rate": 8.322399250234301e-07,
+      "loss": -0.2272,
+      "step": 657
+    },
+    {
+      "epoch": 0.5979100408905044,
+      "grad_norm": 3.794109828921785,
+      "learning_rate": 8.303655107778819e-07,
+      "loss": -0.2465,
+      "step": 658
+    },
+    {
+      "epoch": 0.5988187187641981,
+      "grad_norm": 5.445383079651538,
+      "learning_rate": 8.284910965323336e-07,
+      "loss": -0.1715,
+      "step": 659
+    },
+    {
+      "epoch": 0.5997273966378919,
+      "grad_norm": 5.608097536629216,
+      "learning_rate": 8.266166822867854e-07,
+      "loss": -0.2313,
+      "step": 660
+    },
+    {
+      "epoch": 0.6006360745115856,
+      "grad_norm": 6.940809680417364,
+      "learning_rate": 8.24742268041237e-07,
+      "loss": 0.0072,
+      "step": 661
+    },
+    {
+      "epoch": 0.6015447523852794,
+      "grad_norm": 5.072062938167316,
+      "learning_rate": 8.228678537956888e-07,
+      "loss": -0.1812,
+      "step": 662
+    },
+    {
+      "epoch": 0.6024534302589732,
+      "grad_norm": 5.743823839445634,
+      "learning_rate": 8.209934395501406e-07,
+      "loss": -0.1304,
+      "step": 663
+    },
+    {
+      "epoch": 0.603362108132667,
+      "grad_norm": 5.345745048874242,
+      "learning_rate": 8.191190253045923e-07,
+      "loss": -0.1887,
+      "step": 664
+    },
+    {
+      "epoch": 0.6042707860063607,
+      "grad_norm": 4.238961638935553,
+      "learning_rate": 8.172446110590439e-07,
+      "loss": -0.2295,
+      "step": 665
+    },
+    {
+      "epoch": 0.6051794638800545,
+      "grad_norm": 4.670105053457272,
+      "learning_rate": 8.153701968134957e-07,
+      "loss": -0.1319,
+      "step": 666
+    },
+    {
+      "epoch": 0.6060881417537483,
+      "grad_norm": 4.981464361271411,
+      "learning_rate": 8.134957825679475e-07,
+      "loss": -0.1353,
+      "step": 667
+    },
+    {
+      "epoch": 0.6069968196274421,
+      "grad_norm": 4.4975944023127346,
+      "learning_rate": 8.116213683223993e-07,
+      "loss": -0.0629,
+      "step": 668
+    },
+    {
+      "epoch": 0.6079054975011359,
+      "grad_norm": 5.154255429210025,
+      "learning_rate": 8.097469540768509e-07,
+      "loss": -0.1329,
+      "step": 669
+    },
+    {
+      "epoch": 0.6088141753748296,
+      "grad_norm": 5.896113017247736,
+      "learning_rate": 8.078725398313026e-07,
+      "loss": -0.1933,
+      "step": 670
+    },
+    {
+      "epoch": 0.6097228532485234,
+      "grad_norm": 4.1149891842297865,
+      "learning_rate": 8.059981255857544e-07,
+      "loss": -0.1621,
+      "step": 671
+    },
+    {
+      "epoch": 0.6106315311222171,
+      "grad_norm": 4.052389622008176,
+      "learning_rate": 8.041237113402062e-07,
+      "loss": -0.1774,
+      "step": 672
+    },
+    {
+      "epoch": 0.611540208995911,
+      "grad_norm": 4.212263531938618,
+      "learning_rate": 8.022492970946579e-07,
+      "loss": -0.1126,
+      "step": 673
+    },
+    {
+      "epoch": 0.6124488868696047,
+      "grad_norm": 5.4128275642209625,
+      "learning_rate": 8.003748828491096e-07,
+      "loss": -0.1689,
+      "step": 674
+    },
+    {
+      "epoch": 0.6133575647432985,
+      "grad_norm": 5.622386918819888,
+      "learning_rate": 7.985004686035613e-07,
+      "loss": -0.1813,
+      "step": 675
+    },
+    {
+      "epoch": 0.6142662426169923,
+      "grad_norm": 5.312954512955883,
+      "learning_rate": 7.966260543580131e-07,
+      "loss": -0.1472,
+      "step": 676
+    },
+    {
+      "epoch": 0.615174920490686,
+      "grad_norm": 3.82649613584181,
+      "learning_rate": 7.947516401124649e-07,
+      "loss": -0.2111,
+      "step": 677
+    },
+    {
+      "epoch": 0.6160835983643799,
+      "grad_norm": 5.299165213718126,
+      "learning_rate": 7.928772258669166e-07,
+      "loss": -0.2532,
+      "step": 678
+    },
+    {
+      "epoch": 0.6169922762380736,
+      "grad_norm": 5.186543939571839,
+      "learning_rate": 7.910028116213683e-07,
+      "loss": -0.1865,
+      "step": 679
+    },
+    {
+      "epoch": 0.6179009541117674,
+      "grad_norm": 5.91700016568306,
+      "learning_rate": 7.8912839737582e-07,
+      "loss": -0.1086,
+      "step": 680
+    },
+    {
+      "epoch": 0.6188096319854611,
+      "grad_norm": 3.5406789561753724,
+      "learning_rate": 7.872539831302718e-07,
+      "loss": -0.2013,
+      "step": 681
+    },
+    {
+      "epoch": 0.6197183098591549,
+      "grad_norm": 11.996120461546294,
+      "learning_rate": 7.853795688847234e-07,
+      "loss": -0.0952,
+      "step": 682
+    },
+    {
+      "epoch": 0.6206269877328487,
+      "grad_norm": 8.870006607208028,
+      "learning_rate": 7.835051546391752e-07,
+      "loss": -0.1189,
+      "step": 683
+    },
+    {
+      "epoch": 0.6215356656065425,
+      "grad_norm": 4.208271038740024,
+      "learning_rate": 7.81630740393627e-07,
+      "loss": -0.1102,
+      "step": 684
+    },
+    {
+      "epoch": 0.6224443434802363,
+      "grad_norm": 5.164274687454701,
+      "learning_rate": 7.797563261480787e-07,
+      "loss": -0.143,
+      "step": 685
+    },
+    {
+      "epoch": 0.62335302135393,
+      "grad_norm": 5.030543876047875,
+      "learning_rate": 7.778819119025304e-07,
+      "loss": -0.1806,
+      "step": 686
+    },
+    {
+      "epoch": 0.6242616992276238,
+      "grad_norm": 4.3151732189644445,
+      "learning_rate": 7.760074976569821e-07,
+      "loss": -0.1554,
+      "step": 687
+    },
+    {
+      "epoch": 0.6251703771013176,
+      "grad_norm": 6.643407907228501,
+      "learning_rate": 7.741330834114339e-07,
+      "loss": -0.2601,
+      "step": 688
+    },
+    {
+      "epoch": 0.6260790549750114,
+      "grad_norm": 4.853875191032129,
+      "learning_rate": 7.722586691658857e-07,
+      "loss": -0.1626,
+      "step": 689
+    },
+    {
+      "epoch": 0.6269877328487051,
+      "grad_norm": 7.27515755708897,
+      "learning_rate": 7.703842549203373e-07,
+      "loss": -0.1848,
+      "step": 690
+    },
+    {
+      "epoch": 0.6278964107223989,
+      "grad_norm": 6.11590523757522,
+      "learning_rate": 7.685098406747891e-07,
+      "loss": -0.2326,
+      "step": 691
+    },
+    {
+      "epoch": 0.6288050885960926,
+      "grad_norm": 6.2239131484260275,
+      "learning_rate": 7.666354264292408e-07,
+      "loss": -0.2246,
+      "step": 692
+    },
+    {
+      "epoch": 0.6297137664697865,
+      "grad_norm": 3.796069903631496,
+      "learning_rate": 7.647610121836926e-07,
+      "loss": -0.2737,
+      "step": 693
+    },
+    {
+      "epoch": 0.6306224443434802,
+      "grad_norm": 4.966897871195926,
+      "learning_rate": 7.628865979381443e-07,
+      "loss": -0.135,
+      "step": 694
+    },
+    {
+      "epoch": 0.631531122217174,
+      "grad_norm": 4.8994260335026665,
+      "learning_rate": 7.61012183692596e-07,
+      "loss": -0.1161,
+      "step": 695
+    },
+    {
+      "epoch": 0.6324398000908678,
+      "grad_norm": 4.869091463663273,
+      "learning_rate": 7.591377694470478e-07,
+      "loss": -0.0589,
+      "step": 696
+    },
+    {
+      "epoch": 0.6333484779645616,
+      "grad_norm": 5.318999678851106,
+      "learning_rate": 7.572633552014995e-07,
+      "loss": -0.138,
+      "step": 697
+    },
+    {
+      "epoch": 0.6342571558382554,
+      "grad_norm": 3.542764441204673,
+      "learning_rate": 7.553889409559512e-07,
+      "loss": -0.1443,
+      "step": 698
+    },
+    {
+      "epoch": 0.6351658337119491,
+      "grad_norm": 4.664245704107069,
+      "learning_rate": 7.53514526710403e-07,
+      "loss": -0.3273,
+      "step": 699
+    },
+    {
+      "epoch": 0.6360745115856429,
+      "grad_norm": 6.057151716282373,
+      "learning_rate": 7.516401124648547e-07,
+      "loss": -0.2828,
+      "step": 700
+    },
+    {
+      "epoch": 0.6369831894593366,
+      "grad_norm": 3.680197257269913,
+      "learning_rate": 7.497656982193064e-07,
+      "loss": -0.263,
+      "step": 701
+    },
+    {
+      "epoch": 0.6378918673330305,
+      "grad_norm": 3.882182571001447,
+      "learning_rate": 7.478912839737581e-07,
+      "loss": -0.21,
+      "step": 702
+    },
+    {
+      "epoch": 0.6388005452067242,
+      "grad_norm": 5.5626611579094805,
+      "learning_rate": 7.460168697282099e-07,
+      "loss": -0.2584,
+      "step": 703
+    },
+    {
+      "epoch": 0.639709223080418,
+      "grad_norm": 4.360876556265868,
+      "learning_rate": 7.441424554826617e-07,
+      "loss": -0.0565,
+      "step": 704
+    },
+    {
+      "epoch": 0.6406179009541118,
+      "grad_norm": 4.3875876132661915,
+      "learning_rate": 7.422680412371134e-07,
+      "loss": -0.2379,
+      "step": 705
+    },
+    {
+      "epoch": 0.6415265788278055,
+      "grad_norm": 5.244049514746247,
+      "learning_rate": 7.403936269915651e-07,
+      "loss": -0.1259,
+      "step": 706
+    },
+    {
+      "epoch": 0.6424352567014994,
+      "grad_norm": 4.161872903594466,
+      "learning_rate": 7.385192127460168e-07,
+      "loss": -0.1395,
+      "step": 707
+    },
+    {
+      "epoch": 0.6433439345751931,
+      "grad_norm": 4.555693683752843,
+      "learning_rate": 7.366447985004685e-07,
+      "loss": -0.106,
+      "step": 708
+    },
+    {
+      "epoch": 0.6442526124488869,
+      "grad_norm": 5.073797171731263,
+      "learning_rate": 7.347703842549204e-07,
+      "loss": -0.1288,
+      "step": 709
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 4.328383455237219,
+      "learning_rate": 7.328959700093721e-07,
+      "loss": -0.1757,
+      "step": 710
+    },
+    {
+      "epoch": 0.6460699681962744,
+      "grad_norm": 5.2344914209884195,
+      "learning_rate": 7.310215557638237e-07,
+      "loss": -0.0138,
+      "step": 711
+    },
+    {
+      "epoch": 0.6469786460699682,
+      "grad_norm": 4.270567995838502,
+      "learning_rate": 7.291471415182755e-07,
+      "loss": -0.3114,
+      "step": 712
+    },
+    {
+      "epoch": 0.647887323943662,
+      "grad_norm": 8.178193138926112,
+      "learning_rate": 7.272727272727272e-07,
+      "loss": -0.2516,
+      "step": 713
+    },
+    {
+      "epoch": 0.6487960018173557,
+      "grad_norm": 4.439756008008158,
+      "learning_rate": 7.253983130271791e-07,
+      "loss": -0.2647,
+      "step": 714
+    },
+    {
+      "epoch": 0.6497046796910495,
+      "grad_norm": 4.850485370377594,
+      "learning_rate": 7.235238987816307e-07,
+      "loss": -0.2376,
+      "step": 715
+    },
+    {
+      "epoch": 0.6506133575647433,
+      "grad_norm": 3.7991854698171585,
+      "learning_rate": 7.216494845360824e-07,
+      "loss": -0.1406,
+      "step": 716
+    },
+    {
+      "epoch": 0.6515220354384371,
+      "grad_norm": 4.813139439496646,
+      "learning_rate": 7.197750702905342e-07,
+      "loss": -0.1435,
+      "step": 717
+    },
+    {
+      "epoch": 0.6524307133121309,
+      "grad_norm": 5.562204760313512,
+      "learning_rate": 7.179006560449859e-07,
+      "loss": -0.1414,
+      "step": 718
+    },
+    {
+      "epoch": 0.6533393911858246,
+      "grad_norm": 5.087563906556447,
+      "learning_rate": 7.160262417994376e-07,
+      "loss": -0.2874,
+      "step": 719
+    },
+    {
+      "epoch": 0.6542480690595184,
+      "grad_norm": 4.748446963552802,
+      "learning_rate": 7.141518275538894e-07,
+      "loss": -0.2348,
+      "step": 720
+    },
+    {
+      "epoch": 0.6551567469332121,
+      "grad_norm": 5.75179992368133,
+      "learning_rate": 7.122774133083411e-07,
+      "loss": -0.213,
+      "step": 721
+    },
+    {
+      "epoch": 0.656065424806906,
+      "grad_norm": 5.886802343674883,
+      "learning_rate": 7.104029990627929e-07,
+      "loss": -0.2127,
+      "step": 722
+    },
+    {
+      "epoch": 0.6569741026805997,
+      "grad_norm": 4.232388526633732,
+      "learning_rate": 7.085285848172445e-07,
+      "loss": -0.1753,
+      "step": 723
+    },
+    {
+      "epoch": 0.6578827805542935,
+      "grad_norm": 4.036929602235258,
+      "learning_rate": 7.066541705716963e-07,
+      "loss": -0.1835,
+      "step": 724
+    },
+    {
+      "epoch": 0.6587914584279873,
+      "grad_norm": 4.7764721104109045,
+      "learning_rate": 7.047797563261481e-07,
+      "loss": -0.2007,
+      "step": 725
+    },
+    {
+      "epoch": 0.659700136301681,
+      "grad_norm": 4.136506187734053,
+      "learning_rate": 7.029053420805998e-07,
+      "loss": -0.1015,
+      "step": 726
+    },
+    {
+      "epoch": 0.6606088141753749,
+      "grad_norm": 3.7784208953318412,
+      "learning_rate": 7.010309278350515e-07,
+      "loss": -0.1487,
+      "step": 727
+    },
+    {
+      "epoch": 0.6615174920490686,
+      "grad_norm": 5.064315282061638,
+      "learning_rate": 6.991565135895032e-07,
+      "loss": -0.2022,
+      "step": 728
+    },
+    {
+      "epoch": 0.6624261699227624,
+      "grad_norm": 5.953620309324206,
+      "learning_rate": 6.97282099343955e-07,
+      "loss": -0.1728,
+      "step": 729
+    },
+    {
+      "epoch": 0.6633348477964561,
+      "grad_norm": 3.969787341939385,
+      "learning_rate": 6.954076850984068e-07,
+      "loss": -0.0717,
+      "step": 730
+    },
+    {
+      "epoch": 0.6642435256701499,
+      "grad_norm": 4.2043781893964605,
+      "learning_rate": 6.935332708528584e-07,
+      "loss": -0.1888,
+      "step": 731
+    },
+    {
+      "epoch": 0.6651522035438437,
+      "grad_norm": 4.211561620318321,
+      "learning_rate": 6.916588566073101e-07,
+      "loss": -0.1699,
+      "step": 732
+    },
+    {
+      "epoch": 0.6660608814175375,
+      "grad_norm": 8.738720844860222,
+      "learning_rate": 6.897844423617619e-07,
+      "loss": -0.2468,
+      "step": 733
+    },
+    {
+      "epoch": 0.6669695592912313,
+      "grad_norm": 4.802237926761294,
+      "learning_rate": 6.879100281162137e-07,
+      "loss": -0.1894,
+      "step": 734
+    },
+    {
+      "epoch": 0.667878237164925,
+      "grad_norm": 9.395410724166414,
+      "learning_rate": 6.860356138706654e-07,
+      "loss": -0.1652,
+      "step": 735
+    },
+    {
+      "epoch": 0.6687869150386188,
+      "grad_norm": 4.020560352500097,
+      "learning_rate": 6.841611996251171e-07,
+      "loss": -0.0603,
+      "step": 736
+    },
+    {
+      "epoch": 0.6696955929123126,
+      "grad_norm": 4.657330483644758,
+      "learning_rate": 6.822867853795688e-07,
+      "loss": -0.1442,
+      "step": 737
+    },
+    {
+      "epoch": 0.6706042707860064,
+      "grad_norm": 5.090926622385331,
+      "learning_rate": 6.804123711340206e-07,
+      "loss": -0.1991,
+      "step": 738
+    },
+    {
+      "epoch": 0.6715129486597001,
+      "grad_norm": 4.2626520213652235,
+      "learning_rate": 6.785379568884723e-07,
+      "loss": -0.1406,
+      "step": 739
+    },
+    {
+      "epoch": 0.6724216265333939,
+      "grad_norm": 5.230399099260861,
+      "learning_rate": 6.766635426429241e-07,
+      "loss": -0.1873,
+      "step": 740
+    },
+    {
+      "epoch": 0.6733303044070876,
+      "grad_norm": 4.746757002588818,
+      "learning_rate": 6.747891283973758e-07,
+      "loss": -0.2428,
+      "step": 741
+    },
+    {
+      "epoch": 0.6742389822807815,
+      "grad_norm": 4.512651883683853,
+      "learning_rate": 6.729147141518275e-07,
+      "loss": -0.132,
+      "step": 742
+    },
+    {
+      "epoch": 0.6751476601544752,
+      "grad_norm": 4.150743091651256,
+      "learning_rate": 6.710402999062793e-07,
+      "loss": -0.265,
+      "step": 743
+    },
+    {
+      "epoch": 0.676056338028169,
+      "grad_norm": 4.0510606954449235,
+      "learning_rate": 6.691658856607309e-07,
+      "loss": -0.1626,
+      "step": 744
+    },
+    {
+      "epoch": 0.6769650159018628,
+      "grad_norm": 5.135150997983869,
+      "learning_rate": 6.672914714151828e-07,
+      "loss": -0.171,
+      "step": 745
+    },
+    {
+      "epoch": 0.6778736937755566,
+      "grad_norm": 3.524217066831024,
+      "learning_rate": 6.654170571696345e-07,
+      "loss": -0.2425,
+      "step": 746
+    },
+    {
+      "epoch": 0.6787823716492504,
+      "grad_norm": 4.541984943588343,
+      "learning_rate": 6.635426429240862e-07,
+      "loss": -0.1773,
+      "step": 747
+    },
+    {
+      "epoch": 0.6796910495229441,
+      "grad_norm": 5.116702550374965,
+      "learning_rate": 6.616682286785379e-07,
+      "loss": -0.1476,
+      "step": 748
+    },
+    {
+      "epoch": 0.6805997273966379,
+      "grad_norm": 4.912952873962101,
+      "learning_rate": 6.597938144329896e-07,
+      "loss": -0.0343,
+      "step": 749
+    },
+    {
+      "epoch": 0.6815084052703316,
+      "grad_norm": 4.161196182008382,
+      "learning_rate": 6.579194001874415e-07,
+      "loss": -0.2363,
+      "step": 750
+    },
+    {
+      "epoch": 0.6824170831440255,
+      "grad_norm": 4.058419158675565,
+      "learning_rate": 6.560449859418932e-07,
+      "loss": -0.1669,
+      "step": 751
+    },
+    {
+      "epoch": 0.6833257610177192,
+      "grad_norm": 4.048218257478302,
+      "learning_rate": 6.541705716963448e-07,
+      "loss": -0.1141,
+      "step": 752
+    },
+    {
+      "epoch": 0.684234438891413,
+      "grad_norm": 3.402260500111671,
+      "learning_rate": 6.522961574507966e-07,
+      "loss": -0.1721,
+      "step": 753
+    },
+    {
+      "epoch": 0.6851431167651068,
+      "grad_norm": 4.958602812467101,
+      "learning_rate": 6.504217432052483e-07,
+      "loss": -0.2631,
+      "step": 754
+    },
+    {
+      "epoch": 0.6860517946388005,
+      "grad_norm": 4.152820520312162,
+      "learning_rate": 6.485473289597002e-07,
+      "loss": -0.1606,
+      "step": 755
+    },
+    {
+      "epoch": 0.6869604725124944,
+      "grad_norm": 4.551687728253896,
+      "learning_rate": 6.466729147141518e-07,
+      "loss": -0.0881,
+      "step": 756
+    },
+    {
+      "epoch": 0.6878691503861881,
+      "grad_norm": 3.596084102109655,
+      "learning_rate": 6.447985004686035e-07,
+      "loss": -0.1848,
+      "step": 757
+    },
+    {
+      "epoch": 0.6887778282598819,
+      "grad_norm": 4.101759900383676,
+      "learning_rate": 6.429240862230553e-07,
+      "loss": -0.1266,
+      "step": 758
+    },
+    {
+      "epoch": 0.6896865061335756,
+      "grad_norm": 5.295338635733303,
+      "learning_rate": 6.41049671977507e-07,
+      "loss": -0.1543,
+      "step": 759
+    },
+    {
+      "epoch": 0.6905951840072694,
+      "grad_norm": 4.57579611433421,
+      "learning_rate": 6.391752577319586e-07,
+      "loss": -0.0422,
+      "step": 760
+    },
+    {
+      "epoch": 0.6915038618809632,
+      "grad_norm": 5.241149890615925,
+      "learning_rate": 6.373008434864105e-07,
+      "loss": -0.0995,
+      "step": 761
+    },
+    {
+      "epoch": 0.692412539754657,
+      "grad_norm": 5.68693221056215,
+      "learning_rate": 6.354264292408622e-07,
+      "loss": -0.2656,
+      "step": 762
+    },
+    {
+      "epoch": 0.6933212176283507,
+      "grad_norm": 5.33526937789619,
+      "learning_rate": 6.33552014995314e-07,
+      "loss": -0.1486,
+      "step": 763
+    },
+    {
+      "epoch": 0.6942298955020445,
+      "grad_norm": 3.709601681860402,
+      "learning_rate": 6.316776007497656e-07,
+      "loss": -0.1987,
+      "step": 764
+    },
+    {
+      "epoch": 0.6951385733757383,
+      "grad_norm": 3.4434568636936835,
+      "learning_rate": 6.298031865042173e-07,
+      "loss": -0.2286,
+      "step": 765
+    },
+    {
+      "epoch": 0.6960472512494321,
+      "grad_norm": 5.728898179670097,
+      "learning_rate": 6.279287722586692e-07,
+      "loss": -0.2089,
+      "step": 766
+    },
+    {
+      "epoch": 0.6969559291231259,
+      "grad_norm": 5.590598982382078,
+      "learning_rate": 6.260543580131209e-07,
+      "loss": -0.1553,
+      "step": 767
+    },
+    {
+      "epoch": 0.6978646069968196,
+      "grad_norm": 4.182833220438032,
+      "learning_rate": 6.241799437675727e-07,
+      "loss": -0.0979,
+      "step": 768
+    },
+    {
+      "epoch": 0.6987732848705134,
+      "grad_norm": 4.465831310292876,
+      "learning_rate": 6.223055295220243e-07,
+      "loss": -0.2693,
+      "step": 769
+    },
+    {
+      "epoch": 0.6996819627442071,
+      "grad_norm": 5.269815922224037,
+      "learning_rate": 6.20431115276476e-07,
+      "loss": -0.2432,
+      "step": 770
+    },
+    {
+      "epoch": 0.700590640617901,
+      "grad_norm": 8.286021344362943,
+      "learning_rate": 6.185567010309279e-07,
+      "loss": -0.2065,
+      "step": 771
+    },
+    {
+      "epoch": 0.7014993184915947,
+      "grad_norm": 5.001768562340795,
+      "learning_rate": 6.166822867853796e-07,
+      "loss": -0.2061,
+      "step": 772
+    },
+    {
+      "epoch": 0.7024079963652885,
+      "grad_norm": 4.414030956054325,
+      "learning_rate": 6.148078725398312e-07,
+      "loss": -0.17,
+      "step": 773
+    },
+    {
+      "epoch": 0.7033166742389823,
+      "grad_norm": 4.963341798657709,
+      "learning_rate": 6.12933458294283e-07,
+      "loss": -0.1922,
+      "step": 774
+    },
+    {
+      "epoch": 0.704225352112676,
+      "grad_norm": 4.291202087934926,
+      "learning_rate": 6.110590440487347e-07,
+      "loss": -0.115,
+      "step": 775
+    },
+    {
+      "epoch": 0.7051340299863699,
+      "grad_norm": 4.635498091879065,
+      "learning_rate": 6.091846298031866e-07,
+      "loss": -0.1882,
+      "step": 776
+    },
+    {
+      "epoch": 0.7060427078600636,
+      "grad_norm": 3.350954938830899,
+      "learning_rate": 6.073102155576382e-07,
+      "loss": -0.2007,
+      "step": 777
+    },
+    {
+      "epoch": 0.7069513857337574,
+      "grad_norm": 5.090567126802804,
+      "learning_rate": 6.054358013120899e-07,
+      "loss": -0.137,
+      "step": 778
+    },
+    {
+      "epoch": 0.7078600636074511,
+      "grad_norm": 4.673086156108414,
+      "learning_rate": 6.035613870665417e-07,
+      "loss": -0.2859,
+      "step": 779
+    },
+    {
+      "epoch": 0.7087687414811449,
+      "grad_norm": 5.1142196784966005,
+      "learning_rate": 6.016869728209934e-07,
+      "loss": -0.1273,
+      "step": 780
+    },
+    {
+      "epoch": 0.7096774193548387,
+      "grad_norm": 3.8046076646767006,
+      "learning_rate": 5.998125585754452e-07,
+      "loss": -0.1988,
+      "step": 781
+    },
+    {
+      "epoch": 0.7105860972285325,
+      "grad_norm": 4.623577620899155,
+      "learning_rate": 5.979381443298969e-07,
+      "loss": -0.1504,
+      "step": 782
+    },
+    {
+      "epoch": 0.7114947751022263,
+      "grad_norm": 3.9078669749005104,
+      "learning_rate": 5.960637300843486e-07,
+      "loss": -0.2073,
+      "step": 783
+    },
+    {
+      "epoch": 0.71240345297592,
+      "grad_norm": 4.675933809906734,
+      "learning_rate": 5.941893158388004e-07,
+      "loss": -0.2384,
+      "step": 784
+    },
+    {
+      "epoch": 0.7133121308496138,
+      "grad_norm": 6.275300353867534,
+      "learning_rate": 5.92314901593252e-07,
+      "loss": -0.1592,
+      "step": 785
+    },
+    {
+      "epoch": 0.7142208087233076,
+      "grad_norm": 4.727416914961722,
+      "learning_rate": 5.904404873477037e-07,
+      "loss": -0.1907,
+      "step": 786
+    },
+    {
+      "epoch": 0.7151294865970014,
+      "grad_norm": 5.480545489317065,
+      "learning_rate": 5.885660731021556e-07,
+      "loss": -0.1336,
+      "step": 787
+    },
+    {
+      "epoch": 0.7160381644706951,
+      "grad_norm": 5.6065010210228685,
+      "learning_rate": 5.866916588566073e-07,
+      "loss": -0.0857,
+      "step": 788
+    },
+    {
+      "epoch": 0.7169468423443889,
+      "grad_norm": 3.8634450355116265,
+      "learning_rate": 5.84817244611059e-07,
+      "loss": -0.2584,
+      "step": 789
+    },
+    {
+      "epoch": 0.7178555202180827,
+      "grad_norm": 4.278234380378446,
+      "learning_rate": 5.829428303655107e-07,
+      "loss": -0.2173,
+      "step": 790
+    },
+    {
+      "epoch": 0.7187641980917765,
+      "grad_norm": 3.7860289080483955,
+      "learning_rate": 5.810684161199624e-07,
+      "loss": -0.115,
+      "step": 791
+    },
+    {
+      "epoch": 0.7196728759654702,
+      "grad_norm": 4.683711440999416,
+      "learning_rate": 5.791940018744143e-07,
+      "loss": -0.1912,
+      "step": 792
+    },
+    {
+      "epoch": 0.720581553839164,
+      "grad_norm": 3.9812002660977575,
+      "learning_rate": 5.773195876288659e-07,
+      "loss": -0.1376,
+      "step": 793
+    },
+    {
+      "epoch": 0.7214902317128578,
+      "grad_norm": 6.3106441980269965,
+      "learning_rate": 5.754451733833177e-07,
+      "loss": -0.2311,
+      "step": 794
+    },
+    {
+      "epoch": 0.7223989095865516,
+      "grad_norm": 6.2843061764902135,
+      "learning_rate": 5.735707591377694e-07,
+      "loss": -0.0512,
+      "step": 795
+    },
+    {
+      "epoch": 0.7233075874602454,
+      "grad_norm": 4.420414444147995,
+      "learning_rate": 5.716963448922211e-07,
+      "loss": -0.2191,
+      "step": 796
+    },
+    {
+      "epoch": 0.7242162653339391,
+      "grad_norm": 4.040328102385592,
+      "learning_rate": 5.69821930646673e-07,
+      "loss": -0.2314,
+      "step": 797
+    },
+    {
+      "epoch": 0.7251249432076329,
+      "grad_norm": 7.638516062876497,
+      "learning_rate": 5.679475164011246e-07,
+      "loss": -0.1768,
+      "step": 798
+    },
+    {
+      "epoch": 0.7260336210813266,
+      "grad_norm": 5.001018611147918,
+      "learning_rate": 5.660731021555764e-07,
+      "loss": -0.2176,
+      "step": 799
+    },
+    {
+      "epoch": 0.7269422989550205,
+      "grad_norm": 4.368885017458845,
+      "learning_rate": 5.641986879100281e-07,
+      "loss": -0.1503,
+      "step": 800
+    },
+    {
+      "epoch": 0.7278509768287142,
+      "grad_norm": 4.6666927109854415,
+      "learning_rate": 5.623242736644798e-07,
+      "loss": -0.2479,
+      "step": 801
+    },
+    {
+      "epoch": 0.728759654702408,
+      "grad_norm": 4.467476830147558,
+      "learning_rate": 5.604498594189316e-07,
+      "loss": -0.1877,
+      "step": 802
+    },
+    {
+      "epoch": 0.7296683325761018,
+      "grad_norm": 4.910510114908944,
+      "learning_rate": 5.585754451733833e-07,
+      "loss": -0.2513,
+      "step": 803
+    },
+    {
+      "epoch": 0.7305770104497955,
+      "grad_norm": 4.323978793026055,
+      "learning_rate": 5.56701030927835e-07,
+      "loss": -0.1836,
+      "step": 804
+    },
+    {
+      "epoch": 0.7314856883234894,
+      "grad_norm": 4.968778598151289,
+      "learning_rate": 5.548266166822868e-07,
+      "loss": -0.192,
+      "step": 805
+    },
+    {
+      "epoch": 0.7323943661971831,
+      "grad_norm": 5.428965041582925,
+      "learning_rate": 5.529522024367384e-07,
+      "loss": -0.2909,
+      "step": 806
+    },
+    {
+      "epoch": 0.7333030440708769,
+      "grad_norm": 3.1583574265607006,
+      "learning_rate": 5.510777881911903e-07,
+      "loss": -0.2073,
+      "step": 807
+    },
+    {
+      "epoch": 0.7342117219445706,
+      "grad_norm": 6.184096816535079,
+      "learning_rate": 5.49203373945642e-07,
+      "loss": -0.1523,
+      "step": 808
+    },
+    {
+      "epoch": 0.7351203998182644,
+      "grad_norm": 5.289572213656905,
+      "learning_rate": 5.473289597000937e-07,
+      "loss": -0.1148,
+      "step": 809
+    },
+    {
+      "epoch": 0.7360290776919582,
+      "grad_norm": 4.788437371580405,
+      "learning_rate": 5.454545454545454e-07,
+      "loss": -0.1839,
+      "step": 810
+    },
+    {
+      "epoch": 0.736937755565652,
+      "grad_norm": 4.09461604851014,
+      "learning_rate": 5.435801312089971e-07,
+      "loss": -0.1175,
+      "step": 811
+    },
+    {
+      "epoch": 0.7378464334393458,
+      "grad_norm": 4.346728723781199,
+      "learning_rate": 5.417057169634489e-07,
+      "loss": -0.153,
+      "step": 812
+    },
+    {
+      "epoch": 0.7387551113130395,
+      "grad_norm": 4.03177775395666,
+      "learning_rate": 5.398313027179007e-07,
+      "loss": -0.2904,
+      "step": 813
+    },
+    {
+      "epoch": 0.7396637891867333,
+      "grad_norm": 4.571084503440913,
+      "learning_rate": 5.379568884723523e-07,
+      "loss": -0.1895,
+      "step": 814
+    },
+    {
+      "epoch": 0.7405724670604271,
+      "grad_norm": 4.892564285471943,
+      "learning_rate": 5.360824742268041e-07,
+      "loss": -0.2635,
+      "step": 815
+    },
+    {
+      "epoch": 0.7414811449341209,
+      "grad_norm": 5.054787117449594,
+      "learning_rate": 5.342080599812558e-07,
+      "loss": -0.2093,
+      "step": 816
+    },
+    {
+      "epoch": 0.7423898228078146,
+      "grad_norm": 3.9686622910306975,
+      "learning_rate": 5.323336457357076e-07,
+      "loss": -0.1207,
+      "step": 817
+    },
+    {
+      "epoch": 0.7432985006815084,
+      "grad_norm": 7.144336095014385,
+      "learning_rate": 5.304592314901593e-07,
+      "loss": -0.2041,
+      "step": 818
+    },
+    {
+      "epoch": 0.7442071785552021,
+      "grad_norm": 4.547614492995674,
+      "learning_rate": 5.28584817244611e-07,
+      "loss": -0.2582,
+      "step": 819
+    },
+    {
+      "epoch": 0.745115856428896,
+      "grad_norm": 4.580614057796659,
+      "learning_rate": 5.267104029990628e-07,
+      "loss": -0.199,
+      "step": 820
+    },
+    {
+      "epoch": 0.7460245343025897,
+      "grad_norm": 3.597914960839081,
+      "learning_rate": 5.248359887535145e-07,
+      "loss": -0.1993,
+      "step": 821
+    },
+    {
+      "epoch": 0.7469332121762835,
+      "grad_norm": 7.989021875049507,
+      "learning_rate": 5.229615745079661e-07,
+      "loss": -0.1165,
+      "step": 822
+    },
+    {
+      "epoch": 0.7478418900499773,
+      "grad_norm": 5.7161976272449415,
+      "learning_rate": 5.21087160262418e-07,
+      "loss": -0.2158,
+      "step": 823
+    },
+    {
+      "epoch": 0.748750567923671,
+      "grad_norm": 12.826420957732555,
+      "learning_rate": 5.192127460168697e-07,
+      "loss": -0.1076,
+      "step": 824
+    },
+    {
+      "epoch": 0.7496592457973649,
+      "grad_norm": 4.505167007991764,
+      "learning_rate": 5.173383317713215e-07,
+      "loss": -0.1702,
+      "step": 825
+    },
+    {
+      "epoch": 0.7505679236710586,
+      "grad_norm": 4.011135099720279,
+      "learning_rate": 5.154639175257731e-07,
+      "loss": -0.0292,
+      "step": 826
+    },
+    {
+      "epoch": 0.7514766015447524,
+      "grad_norm": 4.269314801452496,
+      "learning_rate": 5.135895032802248e-07,
+      "loss": -0.1825,
+      "step": 827
+    },
+    {
+      "epoch": 0.7523852794184461,
+      "grad_norm": 4.415261650929307,
+      "learning_rate": 5.117150890346767e-07,
+      "loss": -0.264,
+      "step": 828
+    },
+    {
+      "epoch": 0.7532939572921399,
+      "grad_norm": 4.531051736637742,
+      "learning_rate": 5.098406747891284e-07,
+      "loss": -0.2159,
+      "step": 829
+    },
+    {
+      "epoch": 0.7542026351658337,
+      "grad_norm": 4.84775067897554,
+      "learning_rate": 5.079662605435802e-07,
+      "loss": -0.1994,
+      "step": 830
+    },
+    {
+      "epoch": 0.7551113130395275,
+      "grad_norm": 4.394413843917777,
+      "learning_rate": 5.060918462980318e-07,
+      "loss": -0.2004,
+      "step": 831
+    },
+    {
+      "epoch": 0.7560199909132213,
+      "grad_norm": 5.891535438180435,
+      "learning_rate": 5.042174320524835e-07,
+      "loss": -0.2658,
+      "step": 832
+    },
+    {
+      "epoch": 0.756928668786915,
+      "grad_norm": 4.048230743134364,
+      "learning_rate": 5.023430178069354e-07,
+      "loss": -0.1691,
+      "step": 833
+    },
+    {
+      "epoch": 0.7578373466606089,
+      "grad_norm": 5.2363829448011625,
+      "learning_rate": 5.004686035613871e-07,
+      "loss": -0.3204,
+      "step": 834
+    },
+    {
+      "epoch": 0.7587460245343026,
+      "grad_norm": 4.226595443061311,
+      "learning_rate": 4.985941893158388e-07,
+      "loss": -0.1536,
+      "step": 835
+    },
+    {
+      "epoch": 0.7596547024079964,
+      "grad_norm": 5.1354375481584,
+      "learning_rate": 4.967197750702905e-07,
+      "loss": -0.1829,
+      "step": 836
+    },
+    {
+      "epoch": 0.7605633802816901,
+      "grad_norm": 8.15301800413096,
+      "learning_rate": 4.948453608247422e-07,
+      "loss": -0.1271,
+      "step": 837
+    },
+    {
+      "epoch": 0.7614720581553839,
+      "grad_norm": 5.28735738791913,
+      "learning_rate": 4.92970946579194e-07,
+      "loss": -0.1065,
+      "step": 838
+    },
+    {
+      "epoch": 0.7623807360290777,
+      "grad_norm": 4.967374791264953,
+      "learning_rate": 4.910965323336457e-07,
+      "loss": -0.1342,
+      "step": 839
+    },
+    {
+      "epoch": 0.7632894139027715,
+      "grad_norm": 6.0248356840275115,
+      "learning_rate": 4.892221180880975e-07,
+      "loss": -0.1427,
+      "step": 840
+    },
+    {
+      "epoch": 0.7641980917764652,
+      "grad_norm": 4.099103848283035,
+      "learning_rate": 4.873477038425492e-07,
+      "loss": -0.1147,
+      "step": 841
+    },
+    {
+      "epoch": 0.765106769650159,
+      "grad_norm": 12.536355212994213,
+      "learning_rate": 4.854732895970009e-07,
+      "loss": -0.0922,
+      "step": 842
+    },
+    {
+      "epoch": 0.7660154475238528,
+      "grad_norm": 5.704591327442726,
+      "learning_rate": 4.835988753514527e-07,
+      "loss": -0.2026,
+      "step": 843
+    },
+    {
+      "epoch": 0.7669241253975466,
+      "grad_norm": 3.746621071910253,
+      "learning_rate": 4.817244611059044e-07,
+      "loss": -0.1517,
+      "step": 844
+    },
+    {
+      "epoch": 0.7678328032712404,
+      "grad_norm": 4.171267118707498,
+      "learning_rate": 4.798500468603561e-07,
+      "loss": -0.1717,
+      "step": 845
+    },
+    {
+      "epoch": 0.7687414811449341,
+      "grad_norm": 4.498835942919968,
+      "learning_rate": 4.779756326148078e-07,
+      "loss": -0.1507,
+      "step": 846
+    },
+    {
+      "epoch": 0.7696501590186279,
+      "grad_norm": 5.530786527192584,
+      "learning_rate": 4.7610121836925957e-07,
+      "loss": -0.1426,
+      "step": 847
+    },
+    {
+      "epoch": 0.7705588368923216,
+      "grad_norm": 5.370620185088995,
+      "learning_rate": 4.742268041237113e-07,
+      "loss": -0.1238,
+      "step": 848
+    },
+    {
+      "epoch": 0.7714675147660155,
+      "grad_norm": 4.6222548714935545,
+      "learning_rate": 4.7235238987816305e-07,
+      "loss": -0.2007,
+      "step": 849
+    },
+    {
+      "epoch": 0.7723761926397092,
+      "grad_norm": 5.361721581646309,
+      "learning_rate": 4.704779756326148e-07,
+      "loss": -0.1071,
+      "step": 850
+    },
+    {
+      "epoch": 0.773284870513403,
+      "grad_norm": 5.3821121003757195,
+      "learning_rate": 4.6860356138706653e-07,
+      "loss": -0.264,
+      "step": 851
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 4.721464677272347,
+      "learning_rate": 4.667291471415182e-07,
+      "loss": -0.2299,
+      "step": 852
+    },
+    {
+      "epoch": 0.7751022262607905,
+      "grad_norm": 5.1027645546765985,
+      "learning_rate": 4.6485473289597e-07,
+      "loss": -0.2356,
+      "step": 853
+    },
+    {
+      "epoch": 0.7760109041344844,
+      "grad_norm": 4.17843489371448,
+      "learning_rate": 4.629803186504217e-07,
+      "loss": -0.2622,
+      "step": 854
+    },
+    {
+      "epoch": 0.7769195820081781,
+      "grad_norm": 3.636621197335537,
+      "learning_rate": 4.6110590440487343e-07,
+      "loss": -0.2195,
+      "step": 855
+    },
+    {
+      "epoch": 0.7778282598818719,
+      "grad_norm": 4.829566913087877,
+      "learning_rate": 4.5923149015932517e-07,
+      "loss": -0.1972,
+      "step": 856
+    },
+    {
+      "epoch": 0.7787369377555656,
+      "grad_norm": 4.206724532908753,
+      "learning_rate": 4.573570759137769e-07,
+      "loss": -0.2214,
+      "step": 857
+    },
+    {
+      "epoch": 0.7796456156292594,
+      "grad_norm": 5.8584556570436614,
+      "learning_rate": 4.554826616682287e-07,
+      "loss": -0.1782,
+      "step": 858
+    },
+    {
+      "epoch": 0.7805542935029532,
+      "grad_norm": 5.422581167708395,
+      "learning_rate": 4.536082474226804e-07,
+      "loss": -0.2344,
+      "step": 859
+    },
+    {
+      "epoch": 0.781462971376647,
+      "grad_norm": 4.962775711356546,
+      "learning_rate": 4.5173383317713213e-07,
+      "loss": -0.147,
+      "step": 860
+    },
+    {
+      "epoch": 0.7823716492503408,
+      "grad_norm": 4.68680475801176,
+      "learning_rate": 4.4985941893158387e-07,
+      "loss": -0.2765,
+      "step": 861
+    },
+    {
+      "epoch": 0.7832803271240345,
+      "grad_norm": 4.113032928535318,
+      "learning_rate": 4.479850046860356e-07,
+      "loss": -0.251,
+      "step": 862
+    },
+    {
+      "epoch": 0.7841890049977283,
+      "grad_norm": 3.6400706642753473,
+      "learning_rate": 4.4611059044048735e-07,
+      "loss": -0.1793,
+      "step": 863
+    },
+    {
+      "epoch": 0.7850976828714221,
+      "grad_norm": 6.680984888624674,
+      "learning_rate": 4.442361761949391e-07,
+      "loss": -0.1826,
+      "step": 864
+    },
+    {
+      "epoch": 0.7860063607451159,
+      "grad_norm": 6.6588035306073206,
+      "learning_rate": 4.4236176194939077e-07,
+      "loss": -0.1784,
+      "step": 865
+    },
+    {
+      "epoch": 0.7869150386188096,
+      "grad_norm": 4.442026542889997,
+      "learning_rate": 4.4048734770384256e-07,
+      "loss": -0.1066,
+      "step": 866
+    },
+    {
+      "epoch": 0.7878237164925034,
+      "grad_norm": 3.516229602959332,
+      "learning_rate": 4.3861293345829425e-07,
+      "loss": -0.1667,
+      "step": 867
+    },
+    {
+      "epoch": 0.7887323943661971,
+      "grad_norm": 3.3906417354047673,
+      "learning_rate": 4.36738519212746e-07,
+      "loss": -0.2445,
+      "step": 868
+    },
+    {
+      "epoch": 0.789641072239891,
+      "grad_norm": 4.656360599305183,
+      "learning_rate": 4.3486410496719773e-07,
+      "loss": -0.1588,
+      "step": 869
+    },
+    {
+      "epoch": 0.7905497501135847,
+      "grad_norm": 5.187707000646871,
+      "learning_rate": 4.3298969072164947e-07,
+      "loss": -0.1133,
+      "step": 870
+    },
+    {
+      "epoch": 0.7914584279872785,
+      "grad_norm": 4.070746400517259,
+      "learning_rate": 4.311152764761012e-07,
+      "loss": -0.1936,
+      "step": 871
+    },
+    {
+      "epoch": 0.7923671058609723,
+      "grad_norm": 4.284727591510021,
+      "learning_rate": 4.2924086223055295e-07,
+      "loss": -0.2703,
+      "step": 872
+    },
+    {
+      "epoch": 0.793275783734666,
+      "grad_norm": 4.402438632379245,
+      "learning_rate": 4.2736644798500463e-07,
+      "loss": -0.252,
+      "step": 873
+    },
+    {
+      "epoch": 0.7941844616083599,
+      "grad_norm": 5.213935321449134,
+      "learning_rate": 4.254920337394564e-07,
+      "loss": -0.1745,
+      "step": 874
+    },
+    {
+      "epoch": 0.7950931394820536,
+      "grad_norm": 4.666928329398072,
+      "learning_rate": 4.236176194939081e-07,
+      "loss": -0.2497,
+      "step": 875
+    },
+    {
+      "epoch": 0.7960018173557474,
+      "grad_norm": 4.416174783246982,
+      "learning_rate": 4.217432052483599e-07,
+      "loss": -0.1092,
+      "step": 876
+    },
+    {
+      "epoch": 0.7969104952294411,
+      "grad_norm": 7.4926304532119055,
+      "learning_rate": 4.198687910028116e-07,
+      "loss": -0.2762,
+      "step": 877
+    },
+    {
+      "epoch": 0.797819173103135,
+      "grad_norm": 4.964336810579241,
+      "learning_rate": 4.1799437675726333e-07,
+      "loss": -0.1835,
+      "step": 878
+    },
+    {
+      "epoch": 0.7987278509768287,
+      "grad_norm": 5.070967762695102,
+      "learning_rate": 4.1611996251171507e-07,
+      "loss": -0.2205,
+      "step": 879
+    },
+    {
+      "epoch": 0.7996365288505225,
+      "grad_norm": 3.6854960281915807,
+      "learning_rate": 4.142455482661668e-07,
+      "loss": -0.2029,
+      "step": 880
+    },
+    {
+      "epoch": 0.8005452067242163,
+      "grad_norm": 6.171734385759739,
+      "learning_rate": 4.123711340206185e-07,
+      "loss": -0.2369,
+      "step": 881
+    },
+    {
+      "epoch": 0.80145388459791,
+      "grad_norm": 5.572337163819609,
+      "learning_rate": 4.104967197750703e-07,
+      "loss": -0.1034,
+      "step": 882
+    },
+    {
+      "epoch": 0.8023625624716039,
+      "grad_norm": 3.852386200825096,
+      "learning_rate": 4.0862230552952197e-07,
+      "loss": -0.1836,
+      "step": 883
+    },
+    {
+      "epoch": 0.8032712403452976,
+      "grad_norm": 3.2868936595978266,
+      "learning_rate": 4.0674789128397376e-07,
+      "loss": -0.2453,
+      "step": 884
+    },
+    {
+      "epoch": 0.8041799182189914,
+      "grad_norm": 3.7088459103547793,
+      "learning_rate": 4.0487347703842545e-07,
+      "loss": -0.1535,
+      "step": 885
+    },
+    {
+      "epoch": 0.8050885960926851,
+      "grad_norm": 4.644558969836855,
+      "learning_rate": 4.029990627928772e-07,
+      "loss": -0.061,
+      "step": 886
+    },
+    {
+      "epoch": 0.8059972739663789,
+      "grad_norm": 3.545369370644503,
+      "learning_rate": 4.0112464854732893e-07,
+      "loss": -0.2026,
+      "step": 887
+    },
+    {
+      "epoch": 0.8069059518400727,
+      "grad_norm": 5.459939890100955,
+      "learning_rate": 3.9925023430178067e-07,
+      "loss": -0.1753,
+      "step": 888
+    },
+    {
+      "epoch": 0.8078146297137665,
+      "grad_norm": 5.2516164561810905,
+      "learning_rate": 3.9737582005623246e-07,
+      "loss": -0.1943,
+      "step": 889
+    },
+    {
+      "epoch": 0.8087233075874602,
+      "grad_norm": 5.805392606274642,
+      "learning_rate": 3.9550140581068415e-07,
+      "loss": -0.1378,
+      "step": 890
+    },
+    {
+      "epoch": 0.809631985461154,
+      "grad_norm": 4.939061557729861,
+      "learning_rate": 3.936269915651359e-07,
+      "loss": -0.2057,
+      "step": 891
+    },
+    {
+      "epoch": 0.8105406633348478,
+      "grad_norm": 4.86562541552513,
+      "learning_rate": 3.917525773195876e-07,
+      "loss": -0.0368,
+      "step": 892
+    },
+    {
+      "epoch": 0.8114493412085416,
+      "grad_norm": 4.114908764037489,
+      "learning_rate": 3.8987816307403936e-07,
+      "loss": -0.1641,
+      "step": 893
+    },
+    {
+      "epoch": 0.8123580190822354,
+      "grad_norm": 3.9656572681794566,
+      "learning_rate": 3.8800374882849105e-07,
+      "loss": -0.1747,
+      "step": 894
+    },
+    {
+      "epoch": 0.8132666969559291,
+      "grad_norm": 6.736809168910777,
+      "learning_rate": 3.8612933458294284e-07,
+      "loss": -0.2796,
+      "step": 895
+    },
+    {
+      "epoch": 0.8141753748296229,
+      "grad_norm": 5.775672325202103,
+      "learning_rate": 3.8425492033739453e-07,
+      "loss": -0.2429,
+      "step": 896
+    },
+    {
+      "epoch": 0.8150840527033166,
+      "grad_norm": 4.273179961818129,
+      "learning_rate": 3.823805060918463e-07,
+      "loss": -0.0765,
+      "step": 897
+    },
+    {
+      "epoch": 0.8159927305770105,
+      "grad_norm": 4.936812340859959,
+      "learning_rate": 3.80506091846298e-07,
+      "loss": -0.235,
+      "step": 898
+    },
+    {
+      "epoch": 0.8169014084507042,
+      "grad_norm": 5.76070204218191,
+      "learning_rate": 3.7863167760074974e-07,
+      "loss": -0.2066,
+      "step": 899
+    },
+    {
+      "epoch": 0.817810086324398,
+      "grad_norm": 4.445946190690749,
+      "learning_rate": 3.767572633552015e-07,
+      "loss": -0.2028,
+      "step": 900
+    },
+    {
+      "epoch": 0.8187187641980918,
+      "grad_norm": 4.535611703954907,
+      "learning_rate": 3.748828491096532e-07,
+      "loss": -0.1981,
+      "step": 901
+    },
+    {
+      "epoch": 0.8196274420717855,
+      "grad_norm": 4.4708294065279865,
+      "learning_rate": 3.7300843486410496e-07,
+      "loss": -0.213,
+      "step": 902
+    },
+    {
+      "epoch": 0.8205361199454794,
+      "grad_norm": 3.9397948102239386,
+      "learning_rate": 3.711340206185567e-07,
+      "loss": -0.1959,
+      "step": 903
+    },
+    {
+      "epoch": 0.8214447978191731,
+      "grad_norm": 4.651580722351568,
+      "learning_rate": 3.692596063730084e-07,
+      "loss": -0.21,
+      "step": 904
+    },
+    {
+      "epoch": 0.8223534756928669,
+      "grad_norm": 5.141290592596331,
+      "learning_rate": 3.673851921274602e-07,
+      "loss": -0.2316,
+      "step": 905
+    },
+    {
+      "epoch": 0.8232621535665606,
+      "grad_norm": 3.795012971154764,
+      "learning_rate": 3.6551077788191187e-07,
+      "loss": -0.2323,
+      "step": 906
+    },
+    {
+      "epoch": 0.8241708314402544,
+      "grad_norm": 3.175777646446295,
+      "learning_rate": 3.636363636363636e-07,
+      "loss": -0.2084,
+      "step": 907
+    },
+    {
+      "epoch": 0.8250795093139482,
+      "grad_norm": 6.210860725744045,
+      "learning_rate": 3.6176194939081534e-07,
+      "loss": -0.1732,
+      "step": 908
+    },
+    {
+      "epoch": 0.825988187187642,
+      "grad_norm": 7.027762535725358,
+      "learning_rate": 3.598875351452671e-07,
+      "loss": -0.2145,
+      "step": 909
+    },
+    {
+      "epoch": 0.8268968650613358,
+      "grad_norm": 4.415451506427766,
+      "learning_rate": 3.580131208997188e-07,
+      "loss": -0.2354,
+      "step": 910
+    },
+    {
+      "epoch": 0.8278055429350295,
+      "grad_norm": 3.848505939056306,
+      "learning_rate": 3.5613870665417056e-07,
+      "loss": -0.1206,
+      "step": 911
+    },
+    {
+      "epoch": 0.8287142208087233,
+      "grad_norm": 5.256975172460603,
+      "learning_rate": 3.5426429240862225e-07,
+      "loss": -0.3057,
+      "step": 912
+    },
+    {
+      "epoch": 0.8296228986824171,
+      "grad_norm": 4.166035082950926,
+      "learning_rate": 3.5238987816307404e-07,
+      "loss": -0.0963,
+      "step": 913
+    },
+    {
+      "epoch": 0.8305315765561109,
+      "grad_norm": 4.241957291175413,
+      "learning_rate": 3.5051546391752573e-07,
+      "loss": -0.1243,
+      "step": 914
+    },
+    {
+      "epoch": 0.8314402544298046,
+      "grad_norm": 4.313626349596769,
+      "learning_rate": 3.486410496719775e-07,
+      "loss": -0.0677,
+      "step": 915
+    },
+    {
+      "epoch": 0.8323489323034984,
+      "grad_norm": 4.4036017501612745,
+      "learning_rate": 3.467666354264292e-07,
+      "loss": -0.1575,
+      "step": 916
+    },
+    {
+      "epoch": 0.8332576101771921,
+      "grad_norm": 4.4995498432069745,
+      "learning_rate": 3.4489222118088094e-07,
+      "loss": -0.4124,
+      "step": 917
+    },
+    {
+      "epoch": 0.834166288050886,
+      "grad_norm": 3.957373586383963,
+      "learning_rate": 3.430178069353327e-07,
+      "loss": -0.0751,
+      "step": 918
+    },
+    {
+      "epoch": 0.8350749659245797,
+      "grad_norm": 3.8584479944536403,
+      "learning_rate": 3.411433926897844e-07,
+      "loss": -0.2115,
+      "step": 919
+    },
+    {
+      "epoch": 0.8359836437982735,
+      "grad_norm": 8.492139884583622,
+      "learning_rate": 3.3926897844423616e-07,
+      "loss": -0.292,
+      "step": 920
+    },
+    {
+      "epoch": 0.8368923216719673,
+      "grad_norm": 3.8307461231443347,
+      "learning_rate": 3.373945641986879e-07,
+      "loss": -0.164,
+      "step": 921
+    },
+    {
+      "epoch": 0.8378009995456611,
+      "grad_norm": 4.559619987447599,
+      "learning_rate": 3.3552014995313964e-07,
+      "loss": -0.2368,
+      "step": 922
+    },
+    {
+      "epoch": 0.8387096774193549,
+      "grad_norm": 4.691791451908134,
+      "learning_rate": 3.336457357075914e-07,
+      "loss": -0.1918,
+      "step": 923
+    },
+    {
+      "epoch": 0.8396183552930486,
+      "grad_norm": 4.235945181732361,
+      "learning_rate": 3.317713214620431e-07,
+      "loss": -0.1806,
+      "step": 924
+    },
+    {
+      "epoch": 0.8405270331667424,
+      "grad_norm": 4.430706573006611,
+      "learning_rate": 3.298969072164948e-07,
+      "loss": -0.1359,
+      "step": 925
+    },
+    {
+      "epoch": 0.8414357110404361,
+      "grad_norm": 4.293323061399631,
+      "learning_rate": 3.280224929709466e-07,
+      "loss": -0.2142,
+      "step": 926
+    },
+    {
+      "epoch": 0.84234438891413,
+      "grad_norm": 4.1227651379641435,
+      "learning_rate": 3.261480787253983e-07,
+      "loss": -0.2696,
+      "step": 927
+    },
+    {
+      "epoch": 0.8432530667878237,
+      "grad_norm": 9.87520782638309,
+      "learning_rate": 3.242736644798501e-07,
+      "loss": -0.2556,
+      "step": 928
+    },
+    {
+      "epoch": 0.8441617446615175,
+      "grad_norm": 6.107769455151787,
+      "learning_rate": 3.2239925023430176e-07,
+      "loss": -0.18,
+      "step": 929
+    },
+    {
+      "epoch": 0.8450704225352113,
+      "grad_norm": 2.89882852636308,
+      "learning_rate": 3.205248359887535e-07,
+      "loss": -0.2191,
+      "step": 930
+    },
+    {
+      "epoch": 0.845979100408905,
+      "grad_norm": 3.3080606042276,
+      "learning_rate": 3.1865042174320524e-07,
+      "loss": -0.1841,
+      "step": 931
+    },
+    {
+      "epoch": 0.8468877782825989,
+      "grad_norm": 4.2603931130999415,
+      "learning_rate": 3.16776007497657e-07,
+      "loss": -0.1081,
+      "step": 932
+    },
+    {
+      "epoch": 0.8477964561562926,
+      "grad_norm": 7.081133272668505,
+      "learning_rate": 3.1490159325210867e-07,
+      "loss": -0.2153,
+      "step": 933
+    },
+    {
+      "epoch": 0.8487051340299864,
+      "grad_norm": 5.2967509477430585,
+      "learning_rate": 3.1302717900656046e-07,
+      "loss": -0.1281,
+      "step": 934
+    },
+    {
+      "epoch": 0.8496138119036801,
+      "grad_norm": 4.96805342104043,
+      "learning_rate": 3.1115276476101214e-07,
+      "loss": -0.1544,
+      "step": 935
+    },
+    {
+      "epoch": 0.8505224897773739,
+      "grad_norm": 6.267592428018114,
+      "learning_rate": 3.0927835051546394e-07,
+      "loss": -0.231,
+      "step": 936
+    },
+    {
+      "epoch": 0.8514311676510677,
+      "grad_norm": 3.697317083759369,
+      "learning_rate": 3.074039362699156e-07,
+      "loss": -0.1947,
+      "step": 937
+    },
+    {
+      "epoch": 0.8523398455247615,
+      "grad_norm": 6.826016517366576,
+      "learning_rate": 3.0552952202436736e-07,
+      "loss": -0.0579,
+      "step": 938
+    },
+    {
+      "epoch": 0.8532485233984552,
+      "grad_norm": 4.287805591958851,
+      "learning_rate": 3.036551077788191e-07,
+      "loss": -0.361,
+      "step": 939
+    },
+    {
+      "epoch": 0.854157201272149,
+      "grad_norm": 4.47942075969731,
+      "learning_rate": 3.0178069353327084e-07,
+      "loss": -0.1699,
+      "step": 940
+    },
+    {
+      "epoch": 0.8550658791458428,
+      "grad_norm": 4.0121656427085926,
+      "learning_rate": 2.999062792877226e-07,
+      "loss": -0.1246,
+      "step": 941
+    },
+    {
+      "epoch": 0.8559745570195366,
+      "grad_norm": 4.313162269049853,
+      "learning_rate": 2.980318650421743e-07,
+      "loss": -0.1783,
+      "step": 942
+    },
+    {
+      "epoch": 0.8568832348932304,
+      "grad_norm": 3.464839758972828,
+      "learning_rate": 2.96157450796626e-07,
+      "loss": -0.1256,
+      "step": 943
+    },
+    {
+      "epoch": 0.8577919127669241,
+      "grad_norm": 5.12671707045866,
+      "learning_rate": 2.942830365510778e-07,
+      "loss": -0.0828,
+      "step": 944
+    },
+    {
+      "epoch": 0.8587005906406179,
+      "grad_norm": 4.779448126054769,
+      "learning_rate": 2.924086223055295e-07,
+      "loss": -0.1721,
+      "step": 945
+    },
+    {
+      "epoch": 0.8596092685143116,
+      "grad_norm": 6.1071673450919155,
+      "learning_rate": 2.905342080599812e-07,
+      "loss": -0.0802,
+      "step": 946
+    },
+    {
+      "epoch": 0.8605179463880055,
+      "grad_norm": 4.111634768107279,
+      "learning_rate": 2.8865979381443296e-07,
+      "loss": -0.1387,
+      "step": 947
+    },
+    {
+      "epoch": 0.8614266242616992,
+      "grad_norm": 4.055517681299156,
+      "learning_rate": 2.867853795688847e-07,
+      "loss": -0.1301,
+      "step": 948
+    },
+    {
+      "epoch": 0.862335302135393,
+      "grad_norm": 3.7171077147263865,
+      "learning_rate": 2.849109653233365e-07,
+      "loss": -0.1439,
+      "step": 949
+    },
+    {
+      "epoch": 0.8632439800090868,
+      "grad_norm": 3.9221139052374214,
+      "learning_rate": 2.830365510777882e-07,
+      "loss": -0.1468,
+      "step": 950
+    },
+    {
+      "epoch": 0.8641526578827805,
+      "grad_norm": 4.928064336022368,
+      "learning_rate": 2.811621368322399e-07,
+      "loss": -0.26,
+      "step": 951
+    },
+    {
+      "epoch": 0.8650613357564744,
+      "grad_norm": 4.267749299341662,
+      "learning_rate": 2.7928772258669166e-07,
+      "loss": -0.1711,
+      "step": 952
+    },
+    {
+      "epoch": 0.8659700136301681,
+      "grad_norm": 3.9970987050054383,
+      "learning_rate": 2.774133083411434e-07,
+      "loss": -0.2217,
+      "step": 953
+    },
+    {
+      "epoch": 0.8668786915038619,
+      "grad_norm": 5.375105391067964,
+      "learning_rate": 2.7553889409559514e-07,
+      "loss": -0.1431,
+      "step": 954
+    },
+    {
+      "epoch": 0.8677873693775556,
+      "grad_norm": 4.0346464294796744,
+      "learning_rate": 2.736644798500469e-07,
+      "loss": -0.1624,
+      "step": 955
+    },
+    {
+      "epoch": 0.8686960472512494,
+      "grad_norm": 6.9083748565027205,
+      "learning_rate": 2.7179006560449856e-07,
+      "loss": -0.2672,
+      "step": 956
+    },
+    {
+      "epoch": 0.8696047251249432,
+      "grad_norm": 4.569100817611069,
+      "learning_rate": 2.6991565135895035e-07,
+      "loss": -0.1725,
+      "step": 957
+    },
+    {
+      "epoch": 0.870513402998637,
+      "grad_norm": 4.9125281335544395,
+      "learning_rate": 2.6804123711340204e-07,
+      "loss": -0.2727,
+      "step": 958
+    },
+    {
+      "epoch": 0.8714220808723308,
+      "grad_norm": 5.616041212414323,
+      "learning_rate": 2.661668228678538e-07,
+      "loss": -0.2261,
+      "step": 959
+    },
+    {
+      "epoch": 0.8723307587460245,
+      "grad_norm": 5.7055896188886495,
+      "learning_rate": 2.642924086223055e-07,
+      "loss": -0.1945,
+      "step": 960
+    },
+    {
+      "epoch": 0.8732394366197183,
+      "grad_norm": 3.913294063828155,
+      "learning_rate": 2.6241799437675726e-07,
+      "loss": -0.165,
+      "step": 961
+    },
+    {
+      "epoch": 0.8741481144934121,
+      "grad_norm": 3.868026488259794,
+      "learning_rate": 2.60543580131209e-07,
+      "loss": -0.2315,
+      "step": 962
+    },
+    {
+      "epoch": 0.8750567923671059,
+      "grad_norm": 5.345848519156759,
+      "learning_rate": 2.5866916588566073e-07,
+      "loss": -0.1267,
+      "step": 963
+    },
+    {
+      "epoch": 0.8759654702407996,
+      "grad_norm": 4.069524707082233,
+      "learning_rate": 2.567947516401124e-07,
+      "loss": -0.1839,
+      "step": 964
+    },
+    {
+      "epoch": 0.8768741481144934,
+      "grad_norm": 4.955496331693992,
+      "learning_rate": 2.549203373945642e-07,
+      "loss": -0.2333,
+      "step": 965
+    },
+    {
+      "epoch": 0.8777828259881872,
+      "grad_norm": 4.040528730409505,
+      "learning_rate": 2.530459231490159e-07,
+      "loss": -0.1941,
+      "step": 966
+    },
+    {
+      "epoch": 0.878691503861881,
+      "grad_norm": 3.597418728587077,
+      "learning_rate": 2.511715089034677e-07,
+      "loss": -0.224,
+      "step": 967
+    },
+    {
+      "epoch": 0.8796001817355747,
+      "grad_norm": 4.100640707429241,
+      "learning_rate": 2.492970946579194e-07,
+      "loss": -0.1674,
+      "step": 968
+    },
+    {
+      "epoch": 0.8805088596092685,
+      "grad_norm": 3.2854314409857905,
+      "learning_rate": 2.474226804123711e-07,
+      "loss": -0.1007,
+      "step": 969
+    },
+    {
+      "epoch": 0.8814175374829623,
+      "grad_norm": 3.848543357245078,
+      "learning_rate": 2.4554826616682286e-07,
+      "loss": -0.1386,
+      "step": 970
+    },
+    {
+      "epoch": 0.8823262153566561,
+      "grad_norm": 3.571751544517642,
+      "learning_rate": 2.436738519212746e-07,
+      "loss": -0.3022,
+      "step": 971
+    },
+    {
+      "epoch": 0.8832348932303499,
+      "grad_norm": 4.40266088309814,
+      "learning_rate": 2.4179943767572633e-07,
+      "loss": -0.0664,
+      "step": 972
+    },
+    {
+      "epoch": 0.8841435711040436,
+      "grad_norm": 6.986681120083445,
+      "learning_rate": 2.3992502343017807e-07,
+      "loss": -0.0705,
+      "step": 973
+    },
+    {
+      "epoch": 0.8850522489777374,
+      "grad_norm": 3.5165786466988007,
+      "learning_rate": 2.3805060918462979e-07,
+      "loss": -0.2037,
+      "step": 974
+    },
+    {
+      "epoch": 0.8859609268514311,
+      "grad_norm": 3.809825146485013,
+      "learning_rate": 2.3617619493908153e-07,
+      "loss": -0.1509,
+      "step": 975
+    },
+    {
+      "epoch": 0.886869604725125,
+      "grad_norm": 5.094711868180122,
+      "learning_rate": 2.3430178069353326e-07,
+      "loss": -0.1886,
+      "step": 976
+    },
+    {
+      "epoch": 0.8877782825988187,
+      "grad_norm": 2.9924418289024826,
+      "learning_rate": 2.32427366447985e-07,
+      "loss": -0.1926,
+      "step": 977
+    },
+    {
+      "epoch": 0.8886869604725125,
+      "grad_norm": 4.412754533466805,
+      "learning_rate": 2.3055295220243672e-07,
+      "loss": -0.2539,
+      "step": 978
+    },
+    {
+      "epoch": 0.8895956383462063,
+      "grad_norm": 5.081401437001887,
+      "learning_rate": 2.2867853795688846e-07,
+      "loss": -0.1269,
+      "step": 979
+    },
+    {
+      "epoch": 0.8905043162199,
+      "grad_norm": 4.437758962032812,
+      "learning_rate": 2.268041237113402e-07,
+      "loss": -0.1801,
+      "step": 980
+    },
+    {
+      "epoch": 0.8914129940935939,
+      "grad_norm": 5.618180273448062,
+      "learning_rate": 2.2492970946579193e-07,
+      "loss": -0.1343,
+      "step": 981
+    },
+    {
+      "epoch": 0.8923216719672876,
+      "grad_norm": 6.945956188534943,
+      "learning_rate": 2.2305529522024367e-07,
+      "loss": -0.1205,
+      "step": 982
+    },
+    {
+      "epoch": 0.8932303498409814,
+      "grad_norm": 5.786521364195505,
+      "learning_rate": 2.2118088097469539e-07,
+      "loss": -0.2628,
+      "step": 983
+    },
+    {
+      "epoch": 0.8941390277146751,
+      "grad_norm": 4.760013666213476,
+      "learning_rate": 2.1930646672914713e-07,
+      "loss": -0.149,
+      "step": 984
+    },
+    {
+      "epoch": 0.8950477055883689,
+      "grad_norm": 4.086092944133271,
+      "learning_rate": 2.1743205248359886e-07,
+      "loss": -0.1062,
+      "step": 985
+    },
+    {
+      "epoch": 0.8959563834620627,
+      "grad_norm": 4.567391895336542,
+      "learning_rate": 2.155576382380506e-07,
+      "loss": -0.2168,
+      "step": 986
+    },
+    {
+      "epoch": 0.8968650613357565,
+      "grad_norm": 4.311439494334307,
+      "learning_rate": 2.1368322399250232e-07,
+      "loss": -0.1771,
+      "step": 987
+    },
+    {
+      "epoch": 0.8977737392094502,
+      "grad_norm": 4.8444461445314495,
+      "learning_rate": 2.1180880974695406e-07,
+      "loss": -0.2171,
+      "step": 988
+    },
+    {
+      "epoch": 0.898682417083144,
+      "grad_norm": 4.788743473646973,
+      "learning_rate": 2.099343955014058e-07,
+      "loss": -0.2004,
+      "step": 989
+    },
+    {
+      "epoch": 0.8995910949568378,
+      "grad_norm": 5.0009535833842795,
+      "learning_rate": 2.0805998125585753e-07,
+      "loss": -0.1576,
+      "step": 990
+    },
+    {
+      "epoch": 0.9004997728305316,
+      "grad_norm": 4.023962960605527,
+      "learning_rate": 2.0618556701030925e-07,
+      "loss": -0.1631,
+      "step": 991
+    },
+    {
+      "epoch": 0.9014084507042254,
+      "grad_norm": 6.732332632404599,
+      "learning_rate": 2.0431115276476099e-07,
+      "loss": -0.2434,
+      "step": 992
+    },
+    {
+      "epoch": 0.9023171285779191,
+      "grad_norm": 3.500478711768985,
+      "learning_rate": 2.0243673851921272e-07,
+      "loss": -0.2508,
+      "step": 993
+    },
+    {
+      "epoch": 0.9032258064516129,
+      "grad_norm": 4.39826534063053,
+      "learning_rate": 2.0056232427366446e-07,
+      "loss": -0.1704,
+      "step": 994
+    },
+    {
+      "epoch": 0.9041344843253066,
+      "grad_norm": 11.495352427890424,
+      "learning_rate": 1.9868791002811623e-07,
+      "loss": -0.2038,
+      "step": 995
+    },
+    {
+      "epoch": 0.9050431621990005,
+      "grad_norm": 5.086517816534688,
+      "learning_rate": 1.9681349578256794e-07,
+      "loss": -0.1918,
+      "step": 996
+    },
+    {
+      "epoch": 0.9059518400726942,
+      "grad_norm": 5.369097285201581,
+      "learning_rate": 1.9493908153701968e-07,
+      "loss": -0.2234,
+      "step": 997
+    },
+    {
+      "epoch": 0.906860517946388,
+      "grad_norm": 4.6196656265468645,
+      "learning_rate": 1.9306466729147142e-07,
+      "loss": -0.1889,
+      "step": 998
+    },
+    {
+      "epoch": 0.9077691958200818,
+      "grad_norm": 5.05612761133648,
+      "learning_rate": 1.9119025304592316e-07,
+      "loss": -0.2671,
+      "step": 999
+    },
+    {
+      "epoch": 0.9086778736937755,
+      "grad_norm": 4.230878336373238,
+      "learning_rate": 1.8931583880037487e-07,
+      "loss": -0.0625,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9095865515674694,
+      "grad_norm": 4.1636935244154065,
+      "learning_rate": 1.874414245548266e-07,
+      "loss": -0.2297,
+      "step": 1001
+    },
+    {
+      "epoch": 0.9104952294411631,
+      "grad_norm": 3.2084088172124665,
+      "learning_rate": 1.8556701030927835e-07,
+      "loss": -0.142,
+      "step": 1002
+    },
+    {
+      "epoch": 0.9114039073148569,
+      "grad_norm": 3.91327139955432,
+      "learning_rate": 1.836925960637301e-07,
+      "loss": -0.2354,
+      "step": 1003
+    },
+    {
+      "epoch": 0.9123125851885506,
+      "grad_norm": 4.250187813591967,
+      "learning_rate": 1.818181818181818e-07,
+      "loss": -0.1885,
+      "step": 1004
+    },
+    {
+      "epoch": 0.9132212630622444,
+      "grad_norm": 12.326190289015322,
+      "learning_rate": 1.7994376757263354e-07,
+      "loss": -0.1215,
+      "step": 1005
+    },
+    {
+      "epoch": 0.9141299409359382,
+      "grad_norm": 4.293655354485335,
+      "learning_rate": 1.7806935332708528e-07,
+      "loss": -0.0682,
+      "step": 1006
+    },
+    {
+      "epoch": 0.915038618809632,
+      "grad_norm": 8.84437672042264,
+      "learning_rate": 1.7619493908153702e-07,
+      "loss": -0.1395,
+      "step": 1007
+    },
+    {
+      "epoch": 0.9159472966833258,
+      "grad_norm": 4.063482312503442,
+      "learning_rate": 1.7432052483598876e-07,
+      "loss": -0.2062,
+      "step": 1008
+    },
+    {
+      "epoch": 0.9168559745570195,
+      "grad_norm": 4.527575456484426,
+      "learning_rate": 1.7244611059044047e-07,
+      "loss": -0.0703,
+      "step": 1009
+    },
+    {
+      "epoch": 0.9177646524307133,
+      "grad_norm": 5.346303407849594,
+      "learning_rate": 1.705716963448922e-07,
+      "loss": -0.1859,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9186733303044071,
+      "grad_norm": 3.391990000116255,
+      "learning_rate": 1.6869728209934395e-07,
+      "loss": -0.1513,
+      "step": 1011
+    },
+    {
+      "epoch": 0.9195820081781009,
+      "grad_norm": 4.865210460725049,
+      "learning_rate": 1.668228678537957e-07,
+      "loss": -0.2196,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9204906860517946,
+      "grad_norm": 5.056242478109265,
+      "learning_rate": 1.649484536082474e-07,
+      "loss": -0.1673,
+      "step": 1013
+    },
+    {
+      "epoch": 0.9213993639254884,
+      "grad_norm": 4.7897114399617635,
+      "learning_rate": 1.6307403936269914e-07,
+      "loss": -0.1745,
+      "step": 1014
+    },
+    {
+      "epoch": 0.9223080417991822,
+      "grad_norm": 5.442792925260937,
+      "learning_rate": 1.6119962511715088e-07,
+      "loss": -0.1723,
+      "step": 1015
+    },
+    {
+      "epoch": 0.923216719672876,
+      "grad_norm": 4.600801000100884,
+      "learning_rate": 1.5932521087160262e-07,
+      "loss": -0.2118,
+      "step": 1016
+    },
+    {
+      "epoch": 0.9241253975465697,
+      "grad_norm": 3.3452732502587117,
+      "learning_rate": 1.5745079662605433e-07,
+      "loss": -0.2108,
+      "step": 1017
+    },
+    {
+      "epoch": 0.9250340754202635,
+      "grad_norm": 4.829424340672527,
+      "learning_rate": 1.5557638238050607e-07,
+      "loss": -0.2805,
+      "step": 1018
+    },
+    {
+      "epoch": 0.9259427532939573,
+      "grad_norm": 4.899729678424195,
+      "learning_rate": 1.537019681349578e-07,
+      "loss": -0.2945,
+      "step": 1019
+    },
+    {
+      "epoch": 0.9268514311676511,
+      "grad_norm": 4.744371392081975,
+      "learning_rate": 1.5182755388940955e-07,
+      "loss": -0.2143,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9277601090413449,
+      "grad_norm": 3.733742760164989,
+      "learning_rate": 1.499531396438613e-07,
+      "loss": -0.1407,
+      "step": 1021
+    },
+    {
+      "epoch": 0.9286687869150386,
+      "grad_norm": 5.254107548711089,
+      "learning_rate": 1.48078725398313e-07,
+      "loss": -0.0944,
+      "step": 1022
+    },
+    {
+      "epoch": 0.9295774647887324,
+      "grad_norm": 3.4539985760021534,
+      "learning_rate": 1.4620431115276474e-07,
+      "loss": -0.2044,
+      "step": 1023
+    },
+    {
+      "epoch": 0.9304861426624261,
+      "grad_norm": 4.353099394077374,
+      "learning_rate": 1.4432989690721648e-07,
+      "loss": -0.0513,
+      "step": 1024
+    },
+    {
+      "epoch": 0.93139482053612,
+      "grad_norm": 5.350014745388924,
+      "learning_rate": 1.4245548266166825e-07,
+      "loss": -0.1869,
+      "step": 1025
+    },
+    {
+      "epoch": 0.9323034984098137,
+      "grad_norm": 4.456839528085071,
+      "learning_rate": 1.4058106841611996e-07,
+      "loss": -0.1937,
+      "step": 1026
+    },
+    {
+      "epoch": 0.9332121762835075,
+      "grad_norm": 8.385679353820873,
+      "learning_rate": 1.387066541705717e-07,
+      "loss": -0.0903,
+      "step": 1027
+    },
+    {
+      "epoch": 0.9341208541572013,
+      "grad_norm": 3.6286615755278024,
+      "learning_rate": 1.3683223992502344e-07,
+      "loss": -0.2141,
+      "step": 1028
+    },
+    {
+      "epoch": 0.935029532030895,
+      "grad_norm": 5.078466409797297,
+      "learning_rate": 1.3495782567947518e-07,
+      "loss": -0.2067,
+      "step": 1029
+    },
+    {
+      "epoch": 0.9359382099045889,
+      "grad_norm": 5.913472275906107,
+      "learning_rate": 1.330834114339269e-07,
+      "loss": -0.3082,
+      "step": 1030
+    },
+    {
+      "epoch": 0.9368468877782826,
+      "grad_norm": 4.284872040412638,
+      "learning_rate": 1.3120899718837863e-07,
+      "loss": -0.2409,
+      "step": 1031
+    },
+    {
+      "epoch": 0.9377555656519764,
+      "grad_norm": 4.509520526105431,
+      "learning_rate": 1.2933458294283037e-07,
+      "loss": -0.2473,
+      "step": 1032
+    },
+    {
+      "epoch": 0.9386642435256701,
+      "grad_norm": 5.857904112255415,
+      "learning_rate": 1.274601686972821e-07,
+      "loss": -0.198,
+      "step": 1033
+    },
+    {
+      "epoch": 0.9395729213993639,
+      "grad_norm": 6.633465234924294,
+      "learning_rate": 1.2558575445173385e-07,
+      "loss": -0.1618,
+      "step": 1034
+    },
+    {
+      "epoch": 0.9404815992730577,
+      "grad_norm": 4.733295380525259,
+      "learning_rate": 1.2371134020618556e-07,
+      "loss": -0.1541,
+      "step": 1035
+    },
+    {
+      "epoch": 0.9413902771467515,
+      "grad_norm": 4.3819102618292645,
+      "learning_rate": 1.218369259606373e-07,
+      "loss": -0.2616,
+      "step": 1036
+    },
+    {
+      "epoch": 0.9422989550204453,
+      "grad_norm": 5.727187306040947,
+      "learning_rate": 1.1996251171508904e-07,
+      "loss": -0.1543,
+      "step": 1037
+    },
+    {
+      "epoch": 0.943207632894139,
+      "grad_norm": 4.365513352732499,
+      "learning_rate": 1.1808809746954076e-07,
+      "loss": -0.2963,
+      "step": 1038
+    },
+    {
+      "epoch": 0.9441163107678328,
+      "grad_norm": 4.683752978293147,
+      "learning_rate": 1.162136832239925e-07,
+      "loss": -0.2271,
+      "step": 1039
+    },
+    {
+      "epoch": 0.9450249886415266,
+      "grad_norm": 3.247564577043522,
+      "learning_rate": 1.1433926897844423e-07,
+      "loss": -0.2741,
+      "step": 1040
+    },
+    {
+      "epoch": 0.9459336665152204,
+      "grad_norm": 4.311239307066237,
+      "learning_rate": 1.1246485473289597e-07,
+      "loss": -0.1925,
+      "step": 1041
+    },
+    {
+      "epoch": 0.9468423443889141,
+      "grad_norm": 8.99631127952916,
+      "learning_rate": 1.1059044048734769e-07,
+      "loss": -0.1659,
+      "step": 1042
+    },
+    {
+      "epoch": 0.9477510222626079,
+      "grad_norm": 4.3374484891883505,
+      "learning_rate": 1.0871602624179943e-07,
+      "loss": -0.0964,
+      "step": 1043
+    },
+    {
+      "epoch": 0.9486597001363016,
+      "grad_norm": 3.680653568778156,
+      "learning_rate": 1.0684161199625116e-07,
+      "loss": -0.1268,
+      "step": 1044
+    },
+    {
+      "epoch": 0.9495683780099955,
+      "grad_norm": 5.462848727270509,
+      "learning_rate": 1.049671977507029e-07,
+      "loss": -0.1627,
+      "step": 1045
+    },
+    {
+      "epoch": 0.9504770558836892,
+      "grad_norm": 4.321892723549161,
+      "learning_rate": 1.0309278350515462e-07,
+      "loss": -0.254,
+      "step": 1046
+    },
+    {
+      "epoch": 0.951385733757383,
+      "grad_norm": 8.954415486320709,
+      "learning_rate": 1.0121836925960636e-07,
+      "loss": -0.1093,
+      "step": 1047
+    },
+    {
+      "epoch": 0.9522944116310768,
+      "grad_norm": 4.001055339831913,
+      "learning_rate": 9.934395501405811e-08,
+      "loss": -0.1986,
+      "step": 1048
+    },
+    {
+      "epoch": 0.9532030895047705,
+      "grad_norm": 4.132296957299965,
+      "learning_rate": 9.746954076850984e-08,
+      "loss": -0.2327,
+      "step": 1049
+    },
+    {
+      "epoch": 0.9541117673784644,
+      "grad_norm": 4.745423772799144,
+      "learning_rate": 9.559512652296158e-08,
+      "loss": -0.1443,
+      "step": 1050
+    },
+    {
+      "epoch": 0.9550204452521581,
+      "grad_norm": 4.3652158049800756,
+      "learning_rate": 9.37207122774133e-08,
+      "loss": -0.1035,
+      "step": 1051
+    },
+    {
+      "epoch": 0.9559291231258519,
+      "grad_norm": 5.166644168107464,
+      "learning_rate": 9.184629803186505e-08,
+      "loss": -0.128,
+      "step": 1052
+    },
+    {
+      "epoch": 0.9568378009995456,
+      "grad_norm": 4.0876111310031416,
+      "learning_rate": 8.997188378631677e-08,
+      "loss": -0.2385,
+      "step": 1053
+    },
+    {
+      "epoch": 0.9577464788732394,
+      "grad_norm": 4.982550116563054,
+      "learning_rate": 8.809746954076851e-08,
+      "loss": -0.1261,
+      "step": 1054
+    },
+    {
+      "epoch": 0.9586551567469332,
+      "grad_norm": 4.277619319034154,
+      "learning_rate": 8.622305529522024e-08,
+      "loss": -0.2367,
+      "step": 1055
+    },
+    {
+      "epoch": 0.959563834620627,
+      "grad_norm": 4.647204323690001,
+      "learning_rate": 8.434864104967198e-08,
+      "loss": -0.1716,
+      "step": 1056
+    },
+    {
+      "epoch": 0.9604725124943208,
+      "grad_norm": 5.2942752365892085,
+      "learning_rate": 8.24742268041237e-08,
+      "loss": -0.151,
+      "step": 1057
+    },
+    {
+      "epoch": 0.9613811903680145,
+      "grad_norm": 5.023058934875483,
+      "learning_rate": 8.059981255857544e-08,
+      "loss": -0.2278,
+      "step": 1058
+    },
+    {
+      "epoch": 0.9622898682417084,
+      "grad_norm": 7.0575622581792485,
+      "learning_rate": 7.872539831302717e-08,
+      "loss": -0.1516,
+      "step": 1059
+    },
+    {
+      "epoch": 0.9631985461154021,
+      "grad_norm": 5.226725075421328,
+      "learning_rate": 7.68509840674789e-08,
+      "loss": -0.2607,
+      "step": 1060
+    },
+    {
+      "epoch": 0.9641072239890959,
+      "grad_norm": 3.9635219230945444,
+      "learning_rate": 7.497656982193064e-08,
+      "loss": -0.13,
+      "step": 1061
+    },
+    {
+      "epoch": 0.9650159018627896,
+      "grad_norm": 5.8784724585597,
+      "learning_rate": 7.310215557638237e-08,
+      "loss": -0.214,
+      "step": 1062
+    },
+    {
+      "epoch": 0.9659245797364834,
+      "grad_norm": 3.906377561393758,
+      "learning_rate": 7.122774133083412e-08,
+      "loss": -0.1074,
+      "step": 1063
+    },
+    {
+      "epoch": 0.9668332576101772,
+      "grad_norm": 3.9619446326156074,
+      "learning_rate": 6.935332708528585e-08,
+      "loss": -0.1757,
+      "step": 1064
+    },
+    {
+      "epoch": 0.967741935483871,
+      "grad_norm": 4.304220683130329,
+      "learning_rate": 6.747891283973759e-08,
+      "loss": -0.1609,
+      "step": 1065
+    },
+    {
+      "epoch": 0.9686506133575647,
+      "grad_norm": 3.8884251514837476,
+      "learning_rate": 6.560449859418931e-08,
+      "loss": -0.1807,
+      "step": 1066
+    },
+    {
+      "epoch": 0.9695592912312585,
+      "grad_norm": 5.199804764530434,
+      "learning_rate": 6.373008434864105e-08,
+      "loss": -0.1655,
+      "step": 1067
+    },
+    {
+      "epoch": 0.9704679691049523,
+      "grad_norm": 12.303808872400232,
+      "learning_rate": 6.185567010309278e-08,
+      "loss": -0.1721,
+      "step": 1068
+    },
+    {
+      "epoch": 0.9713766469786461,
+      "grad_norm": 4.15438984711727,
+      "learning_rate": 5.998125585754452e-08,
+      "loss": -0.1774,
+      "step": 1069
+    },
+    {
+      "epoch": 0.9722853248523399,
+      "grad_norm": 3.648814147197159,
+      "learning_rate": 5.810684161199625e-08,
+      "loss": -0.1019,
+      "step": 1070
+    },
+    {
+      "epoch": 0.9731940027260336,
+      "grad_norm": 8.309384328572236,
+      "learning_rate": 5.6232427366447984e-08,
+      "loss": -0.1126,
+      "step": 1071
+    },
+    {
+      "epoch": 0.9741026805997274,
+      "grad_norm": 5.088031765376805,
+      "learning_rate": 5.4358013120899716e-08,
+      "loss": -0.1604,
+      "step": 1072
+    },
+    {
+      "epoch": 0.9750113584734211,
+      "grad_norm": 4.68862352258091,
+      "learning_rate": 5.248359887535145e-08,
+      "loss": -0.1804,
+      "step": 1073
+    },
+    {
+      "epoch": 0.975920036347115,
+      "grad_norm": 5.2739319068005,
+      "learning_rate": 5.060918462980318e-08,
+      "loss": -0.2422,
+      "step": 1074
+    },
+    {
+      "epoch": 0.9768287142208087,
+      "grad_norm": 4.136167494918515,
+      "learning_rate": 4.873477038425492e-08,
+      "loss": -0.2008,
+      "step": 1075
+    },
+    {
+      "epoch": 0.9777373920945025,
+      "grad_norm": 6.200776623722604,
+      "learning_rate": 4.686035613870665e-08,
+      "loss": -0.0895,
+      "step": 1076
+    },
+    {
+      "epoch": 0.9786460699681963,
+      "grad_norm": 4.638407270413583,
+      "learning_rate": 4.4985941893158385e-08,
+      "loss": -0.2527,
+      "step": 1077
+    },
+    {
+      "epoch": 0.97955474784189,
+      "grad_norm": 5.128189954979775,
+      "learning_rate": 4.311152764761012e-08,
+      "loss": -0.2788,
+      "step": 1078
+    },
+    {
+      "epoch": 0.9804634257155839,
+      "grad_norm": 4.496823355223767,
+      "learning_rate": 4.123711340206185e-08,
+      "loss": -0.1804,
+      "step": 1079
+    },
+    {
+      "epoch": 0.9813721035892776,
+      "grad_norm": 3.858237280312584,
+      "learning_rate": 3.936269915651358e-08,
+      "loss": -0.2325,
+      "step": 1080
+    },
+    {
+      "epoch": 0.9822807814629714,
+      "grad_norm": 4.511255176062401,
+      "learning_rate": 3.748828491096532e-08,
+      "loss": -0.0738,
+      "step": 1081
+    },
+    {
+      "epoch": 0.9831894593366651,
+      "grad_norm": 4.842928909110805,
+      "learning_rate": 3.561387066541706e-08,
+      "loss": -0.2267,
+      "step": 1082
+    },
+    {
+      "epoch": 0.9840981372103589,
+      "grad_norm": 3.9052901652789402,
+      "learning_rate": 3.3739456419868794e-08,
+      "loss": -0.2157,
+      "step": 1083
+    },
+    {
+      "epoch": 0.9850068150840527,
+      "grad_norm": 4.196785495965395,
+      "learning_rate": 3.1865042174320527e-08,
+      "loss": -0.1555,
+      "step": 1084
+    },
+    {
+      "epoch": 0.9859154929577465,
+      "grad_norm": 4.977174347015666,
+      "learning_rate": 2.999062792877226e-08,
+      "loss": -0.2304,
+      "step": 1085
+    },
+    {
+      "epoch": 0.9868241708314403,
+      "grad_norm": 5.158787328475507,
+      "learning_rate": 2.8116213683223992e-08,
+      "loss": -0.2232,
+      "step": 1086
+    },
+    {
+      "epoch": 0.987732848705134,
+      "grad_norm": 3.9157030394648067,
+      "learning_rate": 2.6241799437675724e-08,
+      "loss": -0.2039,
+      "step": 1087
+    },
+    {
+      "epoch": 0.9886415265788278,
+      "grad_norm": 3.6149716274050285,
+      "learning_rate": 2.436738519212746e-08,
+      "loss": -0.2116,
+      "step": 1088
+    },
+    {
+      "epoch": 0.9895502044525216,
+      "grad_norm": 5.446126474761816,
+      "learning_rate": 2.2492970946579193e-08,
+      "loss": -0.2351,
+      "step": 1089
+    },
+    {
+      "epoch": 0.9904588823262154,
+      "grad_norm": 3.904635164261984,
+      "learning_rate": 2.0618556701030925e-08,
+      "loss": -0.2539,
+      "step": 1090
+    },
+    {
+      "epoch": 0.9913675601999091,
+      "grad_norm": 5.967222327814702,
+      "learning_rate": 1.874414245548266e-08,
+      "loss": -0.1553,
+      "step": 1091
+    },
+    {
+      "epoch": 0.9922762380736029,
+      "grad_norm": 3.626183546812453,
+      "learning_rate": 1.6869728209934397e-08,
+      "loss": -0.1108,
+      "step": 1092
+    },
+    {
+      "epoch": 0.9931849159472966,
+      "grad_norm": 3.779066984690834,
+      "learning_rate": 1.499531396438613e-08,
+      "loss": -0.1386,
+      "step": 1093
+    },
+    {
+      "epoch": 0.9940935938209905,
+      "grad_norm": 5.324745695104068,
+      "learning_rate": 1.3120899718837862e-08,
+      "loss": -0.1397,
+      "step": 1094
+    },
+    {
+      "epoch": 0.9950022716946842,
+      "grad_norm": 5.146278310627929,
+      "learning_rate": 1.1246485473289596e-08,
+      "loss": -0.1775,
+      "step": 1095
+    },
+    {
+      "epoch": 0.995910949568378,
+      "grad_norm": 6.2996901663023275,
+      "learning_rate": 9.37207122774133e-09,
+      "loss": -0.2014,
+      "step": 1096
+    },
+    {
+      "epoch": 0.9968196274420718,
+      "grad_norm": 6.0192802916538275,
+      "learning_rate": 7.497656982193065e-09,
+      "loss": -0.2224,
+      "step": 1097
+    },
+    {
+      "epoch": 0.9977283053157655,
+      "grad_norm": 4.008881959756454,
+      "learning_rate": 5.623242736644798e-09,
+      "loss": -0.104,
+      "step": 1098
+    },
+    {
+      "epoch": 0.9986369831894594,
+      "grad_norm": 5.426981958576142,
+      "learning_rate": 3.748828491096532e-09,
+      "loss": -0.1881,
+      "step": 1099
+    },
+    {
+      "epoch": 0.9995456610631531,
+      "grad_norm": 6.364116578746877,
+      "learning_rate": 1.874414245548266e-09,
+      "loss": -0.1804,
+      "step": 1100
+    },
+    {
+      "epoch": 0.9995456610631531,
+      "step": 1100,
+      "total_flos": 173376501841920.0,
+      "train_loss": -0.18491140658882532,
+      "train_runtime": 13008.6585,
+      "train_samples_per_second": 10.827,
+      "train_steps_per_second": 0.085
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1100,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 173376501841920.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}