diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8652 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998983843105376,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000812925515699624,
+      "grad_norm": 34.804351806640625,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 2.022,
+      "step": 1
+    },
+    {
+      "epoch": 0.001625851031399248,
+      "grad_norm": 35.288848876953125,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 2.1055,
+      "step": 2
+    },
+    {
+      "epoch": 0.002438776547098872,
+      "grad_norm": 37.58893585205078,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 2.0685,
+      "step": 3
+    },
+    {
+      "epoch": 0.003251702062798496,
+      "grad_norm": 28.51118278503418,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 2.0364,
+      "step": 4
+    },
+    {
+      "epoch": 0.00406462757849812,
+      "grad_norm": 24.874475479125977,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 1.9688,
+      "step": 5
+    },
+    {
+      "epoch": 0.004877553094197744,
+      "grad_norm": 12.156012535095215,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 1.8677,
+      "step": 6
+    },
+    {
+      "epoch": 0.005690478609897368,
+      "grad_norm": 7.017012119293213,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 1.7774,
+      "step": 7
+    },
+    {
+      "epoch": 0.006503404125596992,
+      "grad_norm": 11.024828910827637,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 1.8042,
+      "step": 8
+    },
+    {
+      "epoch": 0.007316329641296616,
+      "grad_norm": 6.988280296325684,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 1.7973,
+      "step": 9
+    },
+    {
+      "epoch": 0.00812925515699624,
+      "grad_norm": 8.541196823120117,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 1.7946,
+      "step": 10
+    },
+    {
+      "epoch": 0.008942180672695864,
+      "grad_norm": 7.084593772888184,
+      "learning_rate": 5.945945945945947e-06,
+      "loss": 1.8178,
+      "step": 11
+    },
+    {
+      "epoch": 0.009755106188395488,
+      "grad_norm": 5.755589962005615,
+      "learning_rate": 6.486486486486487e-06,
+      "loss": 1.7748,
+      "step": 12
+    },
+    {
+      "epoch": 0.010568031704095112,
+      "grad_norm": 4.855886459350586,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 1.6665,
+      "step": 13
+    },
+    {
+      "epoch": 0.011380957219794737,
+      "grad_norm": 5.280701160430908,
+      "learning_rate": 7.567567567567569e-06,
+      "loss": 1.7226,
+      "step": 14
+    },
+    {
+      "epoch": 0.01219388273549436,
+      "grad_norm": 4.513389587402344,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 1.7219,
+      "step": 15
+    },
+    {
+      "epoch": 0.013006808251193984,
+      "grad_norm": 4.92287015914917,
+      "learning_rate": 8.64864864864865e-06,
+      "loss": 1.697,
+      "step": 16
+    },
+    {
+      "epoch": 0.013819733766893608,
+      "grad_norm": 4.488801002502441,
+      "learning_rate": 9.189189189189191e-06,
+      "loss": 1.6584,
+      "step": 17
+    },
+    {
+      "epoch": 0.014632659282593233,
+      "grad_norm": 3.736851930618286,
+      "learning_rate": 9.729729729729732e-06,
+      "loss": 1.6752,
+      "step": 18
+    },
+    {
+      "epoch": 0.015445584798292857,
+      "grad_norm": 3.7089431285858154,
+      "learning_rate": 1.027027027027027e-05,
+      "loss": 1.602,
+      "step": 19
+    },
+    {
+      "epoch": 0.01625851031399248,
+      "grad_norm": 3.9821619987487793,
+      "learning_rate": 1.0810810810810812e-05,
+      "loss": 1.6492,
+      "step": 20
+    },
+    {
+      "epoch": 0.017071435829692106,
+      "grad_norm": 3.72698974609375,
+      "learning_rate": 1.1351351351351352e-05,
+      "loss": 1.6893,
+      "step": 21
+    },
+    {
+      "epoch": 0.017884361345391727,
+      "grad_norm": 3.0124993324279785,
+      "learning_rate": 1.1891891891891894e-05,
+      "loss": 1.5879,
+      "step": 22
+    },
+    {
+      "epoch": 0.01869728686109135,
+      "grad_norm": 9.361907005310059,
+      "learning_rate": 1.2432432432432433e-05,
+      "loss": 1.6172,
+      "step": 23
+    },
+    {
+      "epoch": 0.019510212376790976,
+      "grad_norm": 3.431147813796997,
+      "learning_rate": 1.2972972972972975e-05,
+      "loss": 1.6354,
+      "step": 24
+    },
+    {
+      "epoch": 0.0203231378924906,
+      "grad_norm": 3.3041067123413086,
+      "learning_rate": 1.3513513513513515e-05,
+      "loss": 1.5998,
+      "step": 25
+    },
+    {
+      "epoch": 0.021136063408190225,
+      "grad_norm": 3.4122121334075928,
+      "learning_rate": 1.4054054054054055e-05,
+      "loss": 1.5737,
+      "step": 26
+    },
+    {
+      "epoch": 0.02194898892388985,
+      "grad_norm": 3.538844585418701,
+      "learning_rate": 1.4594594594594596e-05,
+      "loss": 1.5737,
+      "step": 27
+    },
+    {
+      "epoch": 0.022761914439589474,
+      "grad_norm": 3.245887041091919,
+      "learning_rate": 1.5135135135135138e-05,
+      "loss": 1.5893,
+      "step": 28
+    },
+    {
+      "epoch": 0.023574839955289098,
+      "grad_norm": 3.7787671089172363,
+      "learning_rate": 1.5675675675675676e-05,
+      "loss": 1.5923,
+      "step": 29
+    },
+    {
+      "epoch": 0.02438776547098872,
+      "grad_norm": 3.557563066482544,
+      "learning_rate": 1.6216216216216218e-05,
+      "loss": 1.5906,
+      "step": 30
+    },
+    {
+      "epoch": 0.025200690986688343,
+      "grad_norm": 3.1536169052124023,
+      "learning_rate": 1.6756756756756757e-05,
+      "loss": 1.5976,
+      "step": 31
+    },
+    {
+      "epoch": 0.026013616502387968,
+      "grad_norm": 3.060678005218506,
+      "learning_rate": 1.72972972972973e-05,
+      "loss": 1.5239,
+      "step": 32
+    },
+    {
+      "epoch": 0.026826542018087592,
+      "grad_norm": 3.0163331031799316,
+      "learning_rate": 1.783783783783784e-05,
+      "loss": 1.5703,
+      "step": 33
+    },
+    {
+      "epoch": 0.027639467533787217,
+      "grad_norm": 3.0648066997528076,
+      "learning_rate": 1.8378378378378383e-05,
+      "loss": 1.5421,
+      "step": 34
+    },
+    {
+      "epoch": 0.02845239304948684,
+      "grad_norm": 2.8359413146972656,
+      "learning_rate": 1.891891891891892e-05,
+      "loss": 1.5698,
+      "step": 35
+    },
+    {
+      "epoch": 0.029265318565186466,
+      "grad_norm": 3.0256259441375732,
+      "learning_rate": 1.9459459459459463e-05,
+      "loss": 1.5258,
+      "step": 36
+    },
+    {
+      "epoch": 0.03007824408088609,
+      "grad_norm": 3.305952548980713,
+      "learning_rate": 2e-05,
+      "loss": 1.5979,
+      "step": 37
+    },
+    {
+      "epoch": 0.030891169596585714,
+      "grad_norm": 2.7958834171295166,
+      "learning_rate": 1.99999653272242e-05,
+      "loss": 1.5065,
+      "step": 38
+    },
+    {
+      "epoch": 0.031704095112285335,
+      "grad_norm": 3.515479564666748,
+      "learning_rate": 1.9999861309137232e-05,
+      "loss": 1.4837,
+      "step": 39
+    },
+    {
+      "epoch": 0.03251702062798496,
+      "grad_norm": 2.7845990657806396,
+      "learning_rate": 1.999968794646042e-05,
+      "loss": 1.5634,
+      "step": 40
+    },
+    {
+      "epoch": 0.033329946143684584,
+      "grad_norm": 3.0540645122528076,
+      "learning_rate": 1.9999445240395953e-05,
+      "loss": 1.5001,
+      "step": 41
+    },
+    {
+      "epoch": 0.03414287165938421,
+      "grad_norm": 3.059220790863037,
+      "learning_rate": 1.9999133192626893e-05,
+      "loss": 1.502,
+      "step": 42
+    },
+    {
+      "epoch": 0.03495579717508383,
+      "grad_norm": 2.594452142715454,
+      "learning_rate": 1.9998751805317152e-05,
+      "loss": 1.5245,
+      "step": 43
+    },
+    {
+      "epoch": 0.035768722690783454,
+      "grad_norm": 3.0076844692230225,
+      "learning_rate": 1.999830108111148e-05,
+      "loss": 1.5032,
+      "step": 44
+    },
+    {
+      "epoch": 0.03658164820648308,
+      "grad_norm": 2.9521396160125732,
+      "learning_rate": 1.999778102313545e-05,
+      "loss": 1.5381,
+      "step": 45
+    },
+    {
+      "epoch": 0.0373945737221827,
+      "grad_norm": 3.280303478240967,
+      "learning_rate": 1.999719163499543e-05,
+      "loss": 1.5478,
+      "step": 46
+    },
+    {
+      "epoch": 0.03820749923788233,
+      "grad_norm": 2.9089877605438232,
+      "learning_rate": 1.999653292077857e-05,
+      "loss": 1.4783,
+      "step": 47
+    },
+    {
+      "epoch": 0.03902042475358195,
+      "grad_norm": 5.1869635581970215,
+      "learning_rate": 1.999580488505276e-05,
+      "loss": 1.5067,
+      "step": 48
+    },
+    {
+      "epoch": 0.03983335026928158,
+      "grad_norm": 3.053921699523926,
+      "learning_rate": 1.9995007532866594e-05,
+      "loss": 1.503,
+      "step": 49
+    },
+    {
+      "epoch": 0.0406462757849812,
+      "grad_norm": 2.952059507369995,
+      "learning_rate": 1.9994140869749366e-05,
+      "loss": 1.4579,
+      "step": 50
+    },
+    {
+      "epoch": 0.04145920130068083,
+      "grad_norm": 2.609379291534424,
+      "learning_rate": 1.9993204901710995e-05,
+      "loss": 1.4679,
+      "step": 51
+    },
+    {
+      "epoch": 0.04227212681638045,
+      "grad_norm": 3.41717267036438,
+      "learning_rate": 1.9992199635241997e-05,
+      "loss": 1.5197,
+      "step": 52
+    },
+    {
+      "epoch": 0.04308505233208007,
+      "grad_norm": 2.8707101345062256,
+      "learning_rate": 1.999112507731346e-05,
+      "loss": 1.5074,
+      "step": 53
+    },
+    {
+      "epoch": 0.0438979778477797,
+      "grad_norm": 3.325697660446167,
+      "learning_rate": 1.9989981235376956e-05,
+      "loss": 1.427,
+      "step": 54
+    },
+    {
+      "epoch": 0.04471090336347932,
+      "grad_norm": 2.7196686267852783,
+      "learning_rate": 1.9988768117364526e-05,
+      "loss": 1.4868,
+      "step": 55
+    },
+    {
+      "epoch": 0.04552382887917895,
+      "grad_norm": 2.9488351345062256,
+      "learning_rate": 1.9987485731688595e-05,
+      "loss": 1.5011,
+      "step": 56
+    },
+    {
+      "epoch": 0.04633675439487857,
+      "grad_norm": 2.7776849269866943,
+      "learning_rate": 1.998613408724195e-05,
+      "loss": 1.4664,
+      "step": 57
+    },
+    {
+      "epoch": 0.047149679910578196,
+      "grad_norm": 2.719594717025757,
+      "learning_rate": 1.998471319339763e-05,
+      "loss": 1.4905,
+      "step": 58
+    },
+    {
+      "epoch": 0.04796260542627782,
+      "grad_norm": 2.8028323650360107,
+      "learning_rate": 1.9983223060008908e-05,
+      "loss": 1.4754,
+      "step": 59
+    },
+    {
+      "epoch": 0.04877553094197744,
+      "grad_norm": 3.1789817810058594,
+      "learning_rate": 1.9981663697409203e-05,
+      "loss": 1.4618,
+      "step": 60
+    },
+    {
+      "epoch": 0.049588456457677066,
+      "grad_norm": 3.077449321746826,
+      "learning_rate": 1.998003511641199e-05,
+      "loss": 1.453,
+      "step": 61
+    },
+    {
+      "epoch": 0.05040138197337669,
+      "grad_norm": 2.960418939590454,
+      "learning_rate": 1.997833732831076e-05,
+      "loss": 1.4564,
+      "step": 62
+    },
+    {
+      "epoch": 0.051214307489076315,
+      "grad_norm": 5.316094875335693,
+      "learning_rate": 1.9976570344878916e-05,
+      "loss": 1.4711,
+      "step": 63
+    },
+    {
+      "epoch": 0.052027233004775936,
+      "grad_norm": 3.257415771484375,
+      "learning_rate": 1.9974734178369702e-05,
+      "loss": 1.4606,
+      "step": 64
+    },
+    {
+      "epoch": 0.052840158520475564,
+      "grad_norm": 3.0437912940979004,
+      "learning_rate": 1.997282884151612e-05,
+      "loss": 1.5075,
+      "step": 65
+    },
+    {
+      "epoch": 0.053653084036175185,
+      "grad_norm": 3.3059332370758057,
+      "learning_rate": 1.9970854347530828e-05,
+      "loss": 1.484,
+      "step": 66
+    },
+    {
+      "epoch": 0.05446600955187481,
+      "grad_norm": 4.510897636413574,
+      "learning_rate": 1.9968810710106065e-05,
+      "loss": 1.5091,
+      "step": 67
+    },
+    {
+      "epoch": 0.05527893506757443,
+      "grad_norm": 3.3621528148651123,
+      "learning_rate": 1.9966697943413548e-05,
+      "loss": 1.4603,
+      "step": 68
+    },
+    {
+      "epoch": 0.056091860583274054,
+      "grad_norm": 2.878563642501831,
+      "learning_rate": 1.9964516062104377e-05,
+      "loss": 1.4438,
+      "step": 69
+    },
+    {
+      "epoch": 0.05690478609897368,
+      "grad_norm": 2.8587141036987305,
+      "learning_rate": 1.996226508130892e-05,
+      "loss": 1.441,
+      "step": 70
+    },
+    {
+      "epoch": 0.0577177116146733,
+      "grad_norm": 3.2675728797912598,
+      "learning_rate": 1.995994501663674e-05,
+      "loss": 1.4515,
+      "step": 71
+    },
+    {
+      "epoch": 0.05853063713037293,
+      "grad_norm": 3.018068790435791,
+      "learning_rate": 1.995755588417644e-05,
+      "loss": 1.4499,
+      "step": 72
+    },
+    {
+      "epoch": 0.05934356264607255,
+      "grad_norm": 3.715628147125244,
+      "learning_rate": 1.99550977004956e-05,
+      "loss": 1.4624,
+      "step": 73
+    },
+    {
+      "epoch": 0.06015648816177218,
+      "grad_norm": 2.7632699012756348,
+      "learning_rate": 1.9952570482640628e-05,
+      "loss": 1.4437,
+      "step": 74
+    },
+    {
+      "epoch": 0.0609694136774718,
+      "grad_norm": 3.3581650257110596,
+      "learning_rate": 1.9949974248136655e-05,
+      "loss": 1.4865,
+      "step": 75
+    },
+    {
+      "epoch": 0.06178233919317143,
+      "grad_norm": 3.090432643890381,
+      "learning_rate": 1.9947309014987414e-05,
+      "loss": 1.4416,
+      "step": 76
+    },
+    {
+      "epoch": 0.06259526470887104,
+      "grad_norm": 3.3709418773651123,
+      "learning_rate": 1.9944574801675106e-05,
+      "loss": 1.4184,
+      "step": 77
+    },
+    {
+      "epoch": 0.06340819022457067,
+      "grad_norm": 3.6959853172302246,
+      "learning_rate": 1.9941771627160287e-05,
+      "loss": 1.4694,
+      "step": 78
+    },
+    {
+      "epoch": 0.0642211157402703,
+      "grad_norm": 3.2907724380493164,
+      "learning_rate": 1.9938899510881732e-05,
+      "loss": 1.4121,
+      "step": 79
+    },
+    {
+      "epoch": 0.06503404125596993,
+      "grad_norm": 2.7885124683380127,
+      "learning_rate": 1.9935958472756283e-05,
+      "loss": 1.4033,
+      "step": 80
+    },
+    {
+      "epoch": 0.06584696677166954,
+      "grad_norm": 2.8771262168884277,
+      "learning_rate": 1.993294853317873e-05,
+      "loss": 1.4466,
+      "step": 81
+    },
+    {
+      "epoch": 0.06665989228736917,
+      "grad_norm": 3.572303056716919,
+      "learning_rate": 1.9929869713021668e-05,
+      "loss": 1.3854,
+      "step": 82
+    },
+    {
+      "epoch": 0.0674728178030688,
+      "grad_norm": 2.636934757232666,
+      "learning_rate": 1.9926722033635343e-05,
+      "loss": 1.4186,
+      "step": 83
+    },
+    {
+      "epoch": 0.06828574331876842,
+      "grad_norm": 3.1140427589416504,
+      "learning_rate": 1.9923505516847514e-05,
+      "loss": 1.424,
+      "step": 84
+    },
+    {
+      "epoch": 0.06909866883446804,
+      "grad_norm": 2.808480739593506,
+      "learning_rate": 1.9920220184963296e-05,
+      "loss": 1.4744,
+      "step": 85
+    },
+    {
+      "epoch": 0.06991159435016767,
+      "grad_norm": 3.118234872817993,
+      "learning_rate": 1.9916866060764994e-05,
+      "loss": 1.4277,
+      "step": 86
+    },
+    {
+      "epoch": 0.0707245198658673,
+      "grad_norm": 4.0702033042907715,
+      "learning_rate": 1.991344316751198e-05,
+      "loss": 1.4236,
+      "step": 87
+    },
+    {
+      "epoch": 0.07153744538156691,
+      "grad_norm": 2.938345193862915,
+      "learning_rate": 1.9909951528940485e-05,
+      "loss": 1.4119,
+      "step": 88
+    },
+    {
+      "epoch": 0.07235037089726654,
+      "grad_norm": 2.960853338241577,
+      "learning_rate": 1.990639116926348e-05,
+      "loss": 1.471,
+      "step": 89
+    },
+    {
+      "epoch": 0.07316329641296616,
+      "grad_norm": 3.146742343902588,
+      "learning_rate": 1.9902762113170467e-05,
+      "loss": 1.4751,
+      "step": 90
+    },
+    {
+      "epoch": 0.07397622192866579,
+      "grad_norm": 3.3954169750213623,
+      "learning_rate": 1.989906438582734e-05,
+      "loss": 1.467,
+      "step": 91
+    },
+    {
+      "epoch": 0.0747891474443654,
+      "grad_norm": 2.9790520668029785,
+      "learning_rate": 1.9895298012876192e-05,
+      "loss": 1.507,
+      "step": 92
+    },
+    {
+      "epoch": 0.07560207296006503,
+      "grad_norm": 2.577925682067871,
+      "learning_rate": 1.9891463020435144e-05,
+      "loss": 1.4728,
+      "step": 93
+    },
+    {
+      "epoch": 0.07641499847576466,
+      "grad_norm": 3.437133550643921,
+      "learning_rate": 1.9887559435098162e-05,
+      "loss": 1.4472,
+      "step": 94
+    },
+    {
+      "epoch": 0.07722792399146428,
+      "grad_norm": 2.806886911392212,
+      "learning_rate": 1.9883587283934875e-05,
+      "loss": 1.4497,
+      "step": 95
+    },
+    {
+      "epoch": 0.0780408495071639,
+      "grad_norm": 2.703793525695801,
+      "learning_rate": 1.9879546594490383e-05,
+      "loss": 1.4643,
+      "step": 96
+    },
+    {
+      "epoch": 0.07885377502286353,
+      "grad_norm": 3.2830615043640137,
+      "learning_rate": 1.987543739478507e-05,
+      "loss": 1.4162,
+      "step": 97
+    },
+    {
+      "epoch": 0.07966670053856316,
+      "grad_norm": 2.5376830101013184,
+      "learning_rate": 1.987125971331441e-05,
+      "loss": 1.494,
+      "step": 98
+    },
+    {
+      "epoch": 0.08047962605426277,
+      "grad_norm": 2.532893180847168,
+      "learning_rate": 1.9867013579048765e-05,
+      "loss": 1.4575,
+      "step": 99
+    },
+    {
+      "epoch": 0.0812925515699624,
+      "grad_norm": 2.838155508041382,
+      "learning_rate": 1.9862699021433186e-05,
+      "loss": 1.4007,
+      "step": 100
+    },
+    {
+      "epoch": 0.08210547708566203,
+      "grad_norm": 2.5777368545532227,
+      "learning_rate": 1.9858316070387208e-05,
+      "loss": 1.4213,
+      "step": 101
+    },
+    {
+      "epoch": 0.08291840260136166,
+      "grad_norm": 2.634209394454956,
+      "learning_rate": 1.9853864756304654e-05,
+      "loss": 1.4544,
+      "step": 102
+    },
+    {
+      "epoch": 0.08373132811706127,
+      "grad_norm": 2.9893202781677246,
+      "learning_rate": 1.9849345110053405e-05,
+      "loss": 1.4361,
+      "step": 103
+    },
+    {
+      "epoch": 0.0845442536327609,
+      "grad_norm": 2.668808698654175,
+      "learning_rate": 1.984475716297519e-05,
+      "loss": 1.4267,
+      "step": 104
+    },
+    {
+      "epoch": 0.08535717914846053,
+      "grad_norm": 3.2199463844299316,
+      "learning_rate": 1.984010094688539e-05,
+      "loss": 1.4731,
+      "step": 105
+    },
+    {
+      "epoch": 0.08617010466416014,
+      "grad_norm": 2.746006965637207,
+      "learning_rate": 1.9835376494072788e-05,
+      "loss": 1.385,
+      "step": 106
+    },
+    {
+      "epoch": 0.08698303017985977,
+      "grad_norm": 2.955232620239258,
+      "learning_rate": 1.9830583837299363e-05,
+      "loss": 1.3984,
+      "step": 107
+    },
+    {
+      "epoch": 0.0877959556955594,
+      "grad_norm": 5.357511520385742,
+      "learning_rate": 1.9825723009800058e-05,
+      "loss": 1.4562,
+      "step": 108
+    },
+    {
+      "epoch": 0.08860888121125902,
+      "grad_norm": 2.5583655834198,
+      "learning_rate": 1.9820794045282553e-05,
+      "loss": 1.4222,
+      "step": 109
+    },
+    {
+      "epoch": 0.08942180672695864,
+      "grad_norm": 2.6951992511749268,
+      "learning_rate": 1.9815796977927015e-05,
+      "loss": 1.4697,
+      "step": 110
+    },
+    {
+      "epoch": 0.09023473224265827,
+      "grad_norm": 2.714019775390625,
+      "learning_rate": 1.9810731842385892e-05,
+      "loss": 1.4696,
+      "step": 111
+    },
+    {
+      "epoch": 0.0910476577583579,
+      "grad_norm": 3.2998311519622803,
+      "learning_rate": 1.9805598673783644e-05,
+      "loss": 1.4034,
+      "step": 112
+    },
+    {
+      "epoch": 0.09186058327405751,
+      "grad_norm": 7.7324652671813965,
+      "learning_rate": 1.980039750771651e-05,
+      "loss": 1.4697,
+      "step": 113
+    },
+    {
+      "epoch": 0.09267350878975714,
+      "grad_norm": 2.6200242042541504,
+      "learning_rate": 1.9795128380252263e-05,
+      "loss": 1.451,
+      "step": 114
+    },
+    {
+      "epoch": 0.09348643430545676,
+      "grad_norm": 2.937061071395874,
+      "learning_rate": 1.978979132792996e-05,
+      "loss": 1.4348,
+      "step": 115
+    },
+    {
+      "epoch": 0.09429935982115639,
+      "grad_norm": 3.087085247039795,
+      "learning_rate": 1.9784386387759684e-05,
+      "loss": 1.4271,
+      "step": 116
+    },
+    {
+      "epoch": 0.095112285336856,
+      "grad_norm": 2.6796271800994873,
+      "learning_rate": 1.977891359722229e-05,
+      "loss": 1.4933,
+      "step": 117
+    },
+    {
+      "epoch": 0.09592521085255563,
+      "grad_norm": 3.4506633281707764,
+      "learning_rate": 1.9773372994269147e-05,
+      "loss": 1.427,
+      "step": 118
+    },
+    {
+      "epoch": 0.09673813636825526,
+      "grad_norm": 2.532562732696533,
+      "learning_rate": 1.976776461732187e-05,
+      "loss": 1.436,
+      "step": 119
+    },
+    {
+      "epoch": 0.09755106188395488,
+      "grad_norm": 3.619605541229248,
+      "learning_rate": 1.976208850527206e-05,
+      "loss": 1.4384,
+      "step": 120
+    },
+    {
+      "epoch": 0.0983639873996545,
+      "grad_norm": 3.5245602130889893,
+      "learning_rate": 1.9756344697481027e-05,
+      "loss": 1.4303,
+      "step": 121
+    },
+    {
+      "epoch": 0.09917691291535413,
+      "grad_norm": 2.649686336517334,
+      "learning_rate": 1.975053323377952e-05,
+      "loss": 1.4692,
+      "step": 122
+    },
+    {
+      "epoch": 0.09998983843105376,
+      "grad_norm": 2.516016721725464,
+      "learning_rate": 1.9744654154467468e-05,
+      "loss": 1.4154,
+      "step": 123
+    },
+    {
+      "epoch": 0.10080276394675337,
+      "grad_norm": 2.591168165206909,
+      "learning_rate": 1.9738707500313655e-05,
+      "loss": 1.403,
+      "step": 124
+    },
+    {
+      "epoch": 0.101615689462453,
+      "grad_norm": 2.522486686706543,
+      "learning_rate": 1.9732693312555492e-05,
+      "loss": 1.4575,
+      "step": 125
+    },
+    {
+      "epoch": 0.10242861497815263,
+      "grad_norm": 2.8282413482666016,
+      "learning_rate": 1.9726611632898693e-05,
+      "loss": 1.377,
+      "step": 126
+    },
+    {
+      "epoch": 0.10324154049385226,
+      "grad_norm": 2.598076820373535,
+      "learning_rate": 1.9720462503517e-05,
+      "loss": 1.4382,
+      "step": 127
+    },
+    {
+      "epoch": 0.10405446600955187,
+      "grad_norm": 2.688178777694702,
+      "learning_rate": 1.971424596705189e-05,
+      "loss": 1.4132,
+      "step": 128
+    },
+    {
+      "epoch": 0.1048673915252515,
+      "grad_norm": 2.7512471675872803,
+      "learning_rate": 1.9707962066612278e-05,
+      "loss": 1.4193,
+      "step": 129
+    },
+    {
+      "epoch": 0.10568031704095113,
+      "grad_norm": 2.8025805950164795,
+      "learning_rate": 1.970161084577422e-05,
+      "loss": 1.3829,
+      "step": 130
+    },
+    {
+      "epoch": 0.10649324255665074,
+      "grad_norm": 2.6514623165130615,
+      "learning_rate": 1.9695192348580606e-05,
+      "loss": 1.4362,
+      "step": 131
+    },
+    {
+      "epoch": 0.10730616807235037,
+      "grad_norm": 2.4559547901153564,
+      "learning_rate": 1.9688706619540863e-05,
+      "loss": 1.4357,
+      "step": 132
+    },
+    {
+      "epoch": 0.10811909358805,
+      "grad_norm": 2.8258724212646484,
+      "learning_rate": 1.968215370363063e-05,
+      "loss": 1.4501,
+      "step": 133
+    },
+    {
+      "epoch": 0.10893201910374962,
+      "grad_norm": 2.8553593158721924,
+      "learning_rate": 1.9675533646291463e-05,
+      "loss": 1.4841,
+      "step": 134
+    },
+    {
+      "epoch": 0.10974494461944924,
+      "grad_norm": 3.625079870223999,
+      "learning_rate": 1.9668846493430522e-05,
+      "loss": 1.47,
+      "step": 135
+    },
+    {
+      "epoch": 0.11055787013514887,
+      "grad_norm": 2.9793193340301514,
+      "learning_rate": 1.9662092291420233e-05,
+      "loss": 1.3969,
+      "step": 136
+    },
+    {
+      "epoch": 0.1113707956508485,
+      "grad_norm": 2.5699939727783203,
+      "learning_rate": 1.965527108709798e-05,
+      "loss": 1.4258,
+      "step": 137
+    },
+    {
+      "epoch": 0.11218372116654811,
+      "grad_norm": 2.7961106300354004,
+      "learning_rate": 1.964838292776579e-05,
+      "loss": 1.4637,
+      "step": 138
+    },
+    {
+      "epoch": 0.11299664668224774,
+      "grad_norm": 3.3331451416015625,
+      "learning_rate": 1.9641427861189973e-05,
+      "loss": 1.3976,
+      "step": 139
+    },
+    {
+      "epoch": 0.11380957219794736,
+      "grad_norm": 2.5645205974578857,
+      "learning_rate": 1.963440593560083e-05,
+      "loss": 1.409,
+      "step": 140
+    },
+    {
+      "epoch": 0.11462249771364699,
+      "grad_norm": 2.5996487140655518,
+      "learning_rate": 1.9627317199692287e-05,
+      "loss": 1.4834,
+      "step": 141
+    },
+    {
+      "epoch": 0.1154354232293466,
+      "grad_norm": 2.9811034202575684,
+      "learning_rate": 1.962016170262157e-05,
+      "loss": 1.4508,
+      "step": 142
+    },
+    {
+      "epoch": 0.11624834874504623,
+      "grad_norm": 2.4133377075195312,
+      "learning_rate": 1.961293949400888e-05,
+      "loss": 1.4077,
+      "step": 143
+    },
+    {
+      "epoch": 0.11706127426074586,
+      "grad_norm": 2.622091770172119,
+      "learning_rate": 1.960565062393701e-05,
+      "loss": 1.4046,
+      "step": 144
+    },
+    {
+      "epoch": 0.11787419977644548,
+      "grad_norm": 2.857346534729004,
+      "learning_rate": 1.9598295142951035e-05,
+      "loss": 1.4217,
+      "step": 145
+    },
+    {
+      "epoch": 0.1186871252921451,
+      "grad_norm": 2.600935220718384,
+      "learning_rate": 1.9590873102057948e-05,
+      "loss": 1.403,
+      "step": 146
+    },
+    {
+      "epoch": 0.11950005080784473,
+      "grad_norm": 2.820359945297241,
+      "learning_rate": 1.9583384552726294e-05,
+      "loss": 1.4358,
+      "step": 147
+    },
+    {
+      "epoch": 0.12031297632354436,
+      "grad_norm": 2.6272051334381104,
+      "learning_rate": 1.957582954688584e-05,
+      "loss": 1.4505,
+      "step": 148
+    },
+    {
+      "epoch": 0.12112590183924397,
+      "grad_norm": 2.8003182411193848,
+      "learning_rate": 1.9568208136927177e-05,
+      "loss": 1.3977,
+      "step": 149
+    },
+    {
+      "epoch": 0.1219388273549436,
+      "grad_norm": 3.560518264770508,
+      "learning_rate": 1.9560520375701408e-05,
+      "loss": 1.3992,
+      "step": 150
+    },
+    {
+      "epoch": 0.12275175287064323,
+      "grad_norm": 2.6377906799316406,
+      "learning_rate": 1.9552766316519726e-05,
+      "loss": 1.4022,
+      "step": 151
+    },
+    {
+      "epoch": 0.12356467838634286,
+      "grad_norm": 2.7982730865478516,
+      "learning_rate": 1.9544946013153093e-05,
+      "loss": 1.409,
+      "step": 152
+    },
+    {
+      "epoch": 0.12437760390204247,
+      "grad_norm": 2.7179160118103027,
+      "learning_rate": 1.9537059519831822e-05,
+      "loss": 1.415,
+      "step": 153
+    },
+    {
+      "epoch": 0.12519052941774209,
+      "grad_norm": 2.959397554397583,
+      "learning_rate": 1.9529106891245244e-05,
+      "loss": 1.4296,
+      "step": 154
+    },
+    {
+      "epoch": 0.12600345493344173,
+      "grad_norm": 3.3228979110717773,
+      "learning_rate": 1.9521088182541298e-05,
+      "loss": 1.4282,
+      "step": 155
+    },
+    {
+      "epoch": 0.12681638044914134,
+      "grad_norm": 2.763151168823242,
+      "learning_rate": 1.951300344932616e-05,
+      "loss": 1.3686,
+      "step": 156
+    },
+    {
+      "epoch": 0.12762930596484098,
+      "grad_norm": 2.863190174102783,
+      "learning_rate": 1.9504852747663862e-05,
+      "loss": 1.4227,
+      "step": 157
+    },
+    {
+      "epoch": 0.1284422314805406,
+      "grad_norm": 2.889604330062866,
+      "learning_rate": 1.9496636134075894e-05,
+      "loss": 1.4658,
+      "step": 158
+    },
+    {
+      "epoch": 0.1292551569962402,
+      "grad_norm": 3.024122476577759,
+      "learning_rate": 1.9488353665540813e-05,
+      "loss": 1.4081,
+      "step": 159
+    },
+    {
+      "epoch": 0.13006808251193985,
+      "grad_norm": 2.4810218811035156,
+      "learning_rate": 1.9480005399493857e-05,
+      "loss": 1.4296,
+      "step": 160
+    },
+    {
+      "epoch": 0.13088100802763947,
+      "grad_norm": 2.643673896789551,
+      "learning_rate": 1.9471591393826536e-05,
+      "loss": 1.3652,
+      "step": 161
+    },
+    {
+      "epoch": 0.13169393354333908,
+      "grad_norm": 2.888829231262207,
+      "learning_rate": 1.9463111706886234e-05,
+      "loss": 1.4003,
+      "step": 162
+    },
+    {
+      "epoch": 0.13250685905903872,
+      "grad_norm": 2.7480149269104004,
+      "learning_rate": 1.9454566397475813e-05,
+      "loss": 1.4195,
+      "step": 163
+    },
+    {
+      "epoch": 0.13331978457473834,
+      "grad_norm": 2.68972110748291,
+      "learning_rate": 1.944595552485319e-05,
+      "loss": 1.3848,
+      "step": 164
+    },
+    {
+      "epoch": 0.13413271009043795,
+      "grad_norm": 2.8888440132141113,
+      "learning_rate": 1.943727914873094e-05,
+      "loss": 1.481,
+      "step": 165
+    },
+    {
+      "epoch": 0.1349456356061376,
+      "grad_norm": 2.8409390449523926,
+      "learning_rate": 1.9428537329275862e-05,
+      "loss": 1.4176,
+      "step": 166
+    },
+    {
+      "epoch": 0.1357585611218372,
+      "grad_norm": 2.4992098808288574,
+      "learning_rate": 1.941973012710859e-05,
+      "loss": 1.395,
+      "step": 167
+    },
+    {
+      "epoch": 0.13657148663753685,
+      "grad_norm": 4.587447166442871,
+      "learning_rate": 1.941085760330316e-05,
+      "loss": 1.3905,
+      "step": 168
+    },
+    {
+      "epoch": 0.13738441215323646,
+      "grad_norm": 2.4778833389282227,
+      "learning_rate": 1.940191981938657e-05,
+      "loss": 1.3707,
+      "step": 169
+    },
+    {
+      "epoch": 0.13819733766893608,
+      "grad_norm": 2.7843387126922607,
+      "learning_rate": 1.9392916837338376e-05,
+      "loss": 1.3698,
+      "step": 170
+    },
+    {
+      "epoch": 0.13901026318463572,
+      "grad_norm": 2.731437921524048,
+      "learning_rate": 1.9383848719590257e-05,
+      "loss": 1.4358,
+      "step": 171
+    },
+    {
+      "epoch": 0.13982318870033533,
+      "grad_norm": 3.079371213912964,
+      "learning_rate": 1.9374715529025575e-05,
+      "loss": 1.4027,
+      "step": 172
+    },
+    {
+      "epoch": 0.14063611421603495,
+      "grad_norm": 3.6557998657226562,
+      "learning_rate": 1.9365517328978943e-05,
+      "loss": 1.428,
+      "step": 173
+    },
+    {
+      "epoch": 0.1414490397317346,
+      "grad_norm": 2.9291248321533203,
+      "learning_rate": 1.9356254183235785e-05,
+      "loss": 1.4039,
+      "step": 174
+    },
+    {
+      "epoch": 0.1422619652474342,
+      "grad_norm": 2.498507499694824,
+      "learning_rate": 1.93469261560319e-05,
+      "loss": 1.3731,
+      "step": 175
+    },
+    {
+      "epoch": 0.14307489076313382,
+      "grad_norm": 3.6117923259735107,
+      "learning_rate": 1.9337533312053002e-05,
+      "loss": 1.4263,
+      "step": 176
+    },
+    {
+      "epoch": 0.14388781627883346,
+      "grad_norm": 2.490755319595337,
+      "learning_rate": 1.9328075716434287e-05,
+      "loss": 1.4215,
+      "step": 177
+    },
+    {
+      "epoch": 0.14470074179453307,
+      "grad_norm": 2.9008986949920654,
+      "learning_rate": 1.931855343475998e-05,
+      "loss": 1.3968,
+      "step": 178
+    },
+    {
+      "epoch": 0.14551366731023269,
+      "grad_norm": 5.284730911254883,
+      "learning_rate": 1.930896653306286e-05,
+      "loss": 1.418,
+      "step": 179
+    },
+    {
+      "epoch": 0.14632659282593233,
+      "grad_norm": 2.826756000518799,
+      "learning_rate": 1.929931507782383e-05,
+      "loss": 1.3996,
+      "step": 180
+    },
+    {
+      "epoch": 0.14713951834163194,
+      "grad_norm": 2.8084652423858643,
+      "learning_rate": 1.9289599135971437e-05,
+      "loss": 1.374,
+      "step": 181
+    },
+    {
+      "epoch": 0.14795244385733158,
+      "grad_norm": 2.736046075820923,
+      "learning_rate": 1.9279818774881418e-05,
+      "loss": 1.3687,
+      "step": 182
+    },
+    {
+      "epoch": 0.1487653693730312,
+      "grad_norm": 2.7098567485809326,
+      "learning_rate": 1.9269974062376224e-05,
+      "loss": 1.4059,
+      "step": 183
+    },
+    {
+      "epoch": 0.1495782948887308,
+      "grad_norm": 2.7764878273010254,
+      "learning_rate": 1.926006506672456e-05,
+      "loss": 1.42,
+      "step": 184
+    },
+    {
+      "epoch": 0.15039122040443045,
+      "grad_norm": 2.7715649604797363,
+      "learning_rate": 1.9250091856640895e-05,
+      "loss": 1.4549,
+      "step": 185
+    },
+    {
+      "epoch": 0.15120414592013007,
+      "grad_norm": 2.4104158878326416,
+      "learning_rate": 1.9240054501285015e-05,
+      "loss": 1.4129,
+      "step": 186
+    },
+    {
+      "epoch": 0.15201707143582968,
+      "grad_norm": 2.75614595413208,
+      "learning_rate": 1.922995307026151e-05,
+      "loss": 1.3959,
+      "step": 187
+    },
+    {
+      "epoch": 0.15282999695152932,
+      "grad_norm": 2.813262939453125,
+      "learning_rate": 1.921978763361931e-05,
+      "loss": 1.4139,
+      "step": 188
+    },
+    {
+      "epoch": 0.15364292246722894,
+      "grad_norm": 2.5106594562530518,
+      "learning_rate": 1.9209558261851194e-05,
+      "loss": 1.3683,
+      "step": 189
+    },
+    {
+      "epoch": 0.15445584798292855,
+      "grad_norm": 2.9257330894470215,
+      "learning_rate": 1.919926502589331e-05,
+      "loss": 1.3387,
+      "step": 190
+    },
+    {
+      "epoch": 0.1552687734986282,
+      "grad_norm": 2.5029993057250977,
+      "learning_rate": 1.9188907997124666e-05,
+      "loss": 1.3892,
+      "step": 191
+    },
+    {
+      "epoch": 0.1560816990143278,
+      "grad_norm": 2.6917388439178467,
+      "learning_rate": 1.9178487247366652e-05,
+      "loss": 1.3946,
+      "step": 192
+    },
+    {
+      "epoch": 0.15689462453002745,
+      "grad_norm": 2.7038626670837402,
+      "learning_rate": 1.916800284888253e-05,
+      "loss": 1.4082,
+      "step": 193
+    },
+    {
+      "epoch": 0.15770755004572706,
+      "grad_norm": 2.586545467376709,
+      "learning_rate": 1.915745487437694e-05,
+      "loss": 1.3431,
+      "step": 194
+    },
+    {
+      "epoch": 0.15852047556142668,
+      "grad_norm": 3.043938159942627,
+      "learning_rate": 1.9146843396995396e-05,
+      "loss": 1.3967,
+      "step": 195
+    },
+    {
+      "epoch": 0.15933340107712632,
+      "grad_norm": 2.80709171295166,
+      "learning_rate": 1.9136168490323772e-05,
+      "loss": 1.3617,
+      "step": 196
+    },
+    {
+      "epoch": 0.16014632659282593,
+      "grad_norm": 5.03334903717041,
+      "learning_rate": 1.9125430228387794e-05,
+      "loss": 1.4326,
+      "step": 197
+    },
+    {
+      "epoch": 0.16095925210852555,
+      "grad_norm": 4.717489719390869,
+      "learning_rate": 1.9114628685652535e-05,
+      "loss": 1.3459,
+      "step": 198
+    },
+    {
+      "epoch": 0.1617721776242252,
+      "grad_norm": 3.0668435096740723,
+      "learning_rate": 1.9103763937021887e-05,
+      "loss": 1.3763,
+      "step": 199
+    },
+    {
+      "epoch": 0.1625851031399248,
+      "grad_norm": 2.712122678756714,
+      "learning_rate": 1.909283605783805e-05,
+      "loss": 1.3319,
+      "step": 200
+    },
+    {
+      "epoch": 0.16339802865562442,
+      "grad_norm": 2.7631924152374268,
+      "learning_rate": 1.9081845123881002e-05,
+      "loss": 1.3641,
+      "step": 201
+    },
+    {
+      "epoch": 0.16421095417132406,
+      "grad_norm": 3.499955654144287,
+      "learning_rate": 1.9070791211367984e-05,
+      "loss": 1.3259,
+      "step": 202
+    },
+    {
+      "epoch": 0.16502387968702367,
+      "grad_norm": 2.913755416870117,
+      "learning_rate": 1.9059674396952963e-05,
+      "loss": 1.3386,
+      "step": 203
+    },
+    {
+      "epoch": 0.16583680520272331,
+      "grad_norm": 2.5671772956848145,
+      "learning_rate": 1.90484947577261e-05,
+      "loss": 1.3301,
+      "step": 204
+    },
+    {
+      "epoch": 0.16664973071842293,
+      "grad_norm": 3.3566508293151855,
+      "learning_rate": 1.903725237121322e-05,
+      "loss": 1.3375,
+      "step": 205
+    },
+    {
+      "epoch": 0.16746265623412254,
+      "grad_norm": 2.3617210388183594,
+      "learning_rate": 1.902594731537527e-05,
+      "loss": 1.4476,
+      "step": 206
+    },
+    {
+      "epoch": 0.16827558174982218,
+      "grad_norm": 2.8202669620513916,
+      "learning_rate": 1.901457966860779e-05,
+      "loss": 1.334,
+      "step": 207
+    },
+    {
+      "epoch": 0.1690885072655218,
+      "grad_norm": 2.5990843772888184,
+      "learning_rate": 1.9003149509740347e-05,
+      "loss": 1.4321,
+      "step": 208
+    },
+    {
+      "epoch": 0.1699014327812214,
+      "grad_norm": 5.9826507568359375,
+      "learning_rate": 1.899165691803601e-05,
+      "loss": 1.4338,
+      "step": 209
+    },
+    {
+      "epoch": 0.17071435829692105,
+      "grad_norm": 3.9570019245147705,
+      "learning_rate": 1.8980101973190787e-05,
+      "loss": 1.3265,
+      "step": 210
+    },
+    {
+      "epoch": 0.17152728381262067,
+      "grad_norm": 2.8985307216644287,
+      "learning_rate": 1.896848475533309e-05,
+      "loss": 1.3297,
+      "step": 211
+    },
+    {
+      "epoch": 0.17234020932832028,
+      "grad_norm": 3.2575559616088867,
+      "learning_rate": 1.8956805345023145e-05,
+      "loss": 1.4086,
+      "step": 212
+    },
+    {
+      "epoch": 0.17315313484401992,
+      "grad_norm": 3.264796733856201,
+      "learning_rate": 1.894506382325248e-05,
+      "loss": 1.391,
+      "step": 213
+    },
+    {
+      "epoch": 0.17396606035971954,
+      "grad_norm": 2.767975330352783,
+      "learning_rate": 1.8933260271443313e-05,
+      "loss": 1.3731,
+      "step": 214
+    },
+    {
+      "epoch": 0.17477898587541915,
+      "grad_norm": 2.3556087017059326,
+      "learning_rate": 1.8921394771448032e-05,
+      "loss": 1.3288,
+      "step": 215
+    },
+    {
+      "epoch": 0.1755919113911188,
+      "grad_norm": 4.253211975097656,
+      "learning_rate": 1.89094674055486e-05,
+      "loss": 1.3776,
+      "step": 216
+    },
+    {
+      "epoch": 0.1764048369068184,
+      "grad_norm": 3.0681605339050293,
+      "learning_rate": 1.889747825645599e-05,
+      "loss": 1.4169,
+      "step": 217
+    },
+    {
+      "epoch": 0.17721776242251805,
+      "grad_norm": 2.3741588592529297,
+      "learning_rate": 1.8885427407309627e-05,
+      "loss": 1.3392,
+      "step": 218
+    },
+    {
+      "epoch": 0.17803068793821766,
+      "grad_norm": 2.968780279159546,
+      "learning_rate": 1.887331494167678e-05,
+      "loss": 1.4019,
+      "step": 219
+    },
+    {
+      "epoch": 0.17884361345391728,
+      "grad_norm": 2.3684914112091064,
+      "learning_rate": 1.8861140943552014e-05,
+      "loss": 1.3599,
+      "step": 220
+    },
+    {
+      "epoch": 0.17965653896961692,
+      "grad_norm": 3.0405993461608887,
+      "learning_rate": 1.884890549735659e-05,
+      "loss": 1.4245,
+      "step": 221
+    },
+    {
+      "epoch": 0.18046946448531653,
+      "grad_norm": 3.397047281265259,
+      "learning_rate": 1.8836608687937883e-05,
+      "loss": 1.392,
+      "step": 222
+    },
+    {
+      "epoch": 0.18128239000101615,
+      "grad_norm": 2.4693644046783447,
+      "learning_rate": 1.8824250600568798e-05,
+      "loss": 1.3726,
+      "step": 223
+    },
+    {
+      "epoch": 0.1820953155167158,
+      "grad_norm": 6.75039005279541,
+      "learning_rate": 1.8811831320947177e-05,
+      "loss": 1.3473,
+      "step": 224
+    },
+    {
+      "epoch": 0.1829082410324154,
+      "grad_norm": 2.922574758529663,
+      "learning_rate": 1.879935093519519e-05,
+      "loss": 1.4221,
+      "step": 225
+    },
+    {
+      "epoch": 0.18372116654811502,
+      "grad_norm": 5.6719136238098145,
+      "learning_rate": 1.878680952985877e-05,
+      "loss": 1.3844,
+      "step": 226
+    },
+    {
+      "epoch": 0.18453409206381466,
+      "grad_norm": 2.6967201232910156,
+      "learning_rate": 1.8774207191906976e-05,
+      "loss": 1.344,
+      "step": 227
+    },
+    {
+      "epoch": 0.18534701757951427,
+      "grad_norm": 3.049881935119629,
+      "learning_rate": 1.8761544008731426e-05,
+      "loss": 1.3912,
+      "step": 228
+    },
+    {
+      "epoch": 0.18615994309521391,
+      "grad_norm": 3.1408843994140625,
+      "learning_rate": 1.874882006814565e-05,
+      "loss": 1.4048,
+      "step": 229
+    },
+    {
+      "epoch": 0.18697286861091353,
+      "grad_norm": 2.6653666496276855,
+      "learning_rate": 1.8736035458384528e-05,
+      "loss": 1.3844,
+      "step": 230
+    },
+    {
+      "epoch": 0.18778579412661314,
+      "grad_norm": 2.6866488456726074,
+      "learning_rate": 1.8723190268103634e-05,
+      "loss": 1.3586,
+      "step": 231
+    },
+    {
+      "epoch": 0.18859871964231278,
+      "grad_norm": 3.2653231620788574,
+      "learning_rate": 1.8710284586378645e-05,
+      "loss": 1.3856,
+      "step": 232
+    },
+    {
+      "epoch": 0.1894116451580124,
+      "grad_norm": 2.841388463973999,
+      "learning_rate": 1.8697318502704734e-05,
+      "loss": 1.3868,
+      "step": 233
+    },
+    {
+      "epoch": 0.190224570673712,
+      "grad_norm": 2.797558307647705,
+      "learning_rate": 1.8684292106995916e-05,
+      "loss": 1.3885,
+      "step": 234
+    },
+    {
+      "epoch": 0.19103749618941165,
+      "grad_norm": 2.915003776550293,
+      "learning_rate": 1.8671205489584453e-05,
+      "loss": 1.3434,
+      "step": 235
+    },
+    {
+      "epoch": 0.19185042170511127,
+      "grad_norm": 3.2142281532287598,
+      "learning_rate": 1.865805874122021e-05,
+      "loss": 1.3975,
+      "step": 236
+    },
+    {
+      "epoch": 0.19266334722081088,
+      "grad_norm": 3.0831453800201416,
+      "learning_rate": 1.8644851953070045e-05,
+      "loss": 1.367,
+      "step": 237
+    },
+    {
+      "epoch": 0.19347627273651052,
+      "grad_norm": 3.2555181980133057,
+      "learning_rate": 1.863158521671716e-05,
+      "loss": 1.33,
+      "step": 238
+    },
+    {
+      "epoch": 0.19428919825221014,
+      "grad_norm": 2.8768310546875,
+      "learning_rate": 1.8618258624160465e-05,
+      "loss": 1.3867,
+      "step": 239
+    },
+    {
+      "epoch": 0.19510212376790975,
+      "grad_norm": 2.9737942218780518,
+      "learning_rate": 1.8604872267813954e-05,
+      "loss": 1.3726,
+      "step": 240
+    },
+    {
+      "epoch": 0.1959150492836094,
+      "grad_norm": 2.5942904949188232,
+      "learning_rate": 1.859142624050605e-05,
+      "loss": 1.3704,
+      "step": 241
+    },
+    {
+      "epoch": 0.196727974799309,
+      "grad_norm": 2.6901443004608154,
+      "learning_rate": 1.8577920635478976e-05,
+      "loss": 1.3523,
+      "step": 242
+    },
+    {
+      "epoch": 0.19754090031500865,
+      "grad_norm": 2.4508392810821533,
+      "learning_rate": 1.8564355546388094e-05,
+      "loss": 1.3758,
+      "step": 243
+    },
+    {
+      "epoch": 0.19835382583070826,
+      "grad_norm": 2.3041279315948486,
+      "learning_rate": 1.855073106730126e-05,
+      "loss": 1.3491,
+      "step": 244
+    },
+    {
+      "epoch": 0.19916675134640788,
+      "grad_norm": 2.8388736248016357,
+      "learning_rate": 1.8537047292698175e-05,
+      "loss": 1.3578,
+      "step": 245
+    },
+    {
+      "epoch": 0.19997967686210752,
+      "grad_norm": 3.058314085006714,
+      "learning_rate": 1.852330431746973e-05,
+      "loss": 1.3547,
+      "step": 246
+    },
+    {
+      "epoch": 0.20079260237780713,
+      "grad_norm": 2.881788492202759,
+      "learning_rate": 1.8509502236917353e-05,
+      "loss": 1.3823,
+      "step": 247
+    },
+    {
+      "epoch": 0.20160552789350675,
+      "grad_norm": 2.623408794403076,
+      "learning_rate": 1.8495641146752322e-05,
+      "loss": 1.4516,
+      "step": 248
+    },
+    {
+      "epoch": 0.2024184534092064,
+      "grad_norm": 2.662614345550537,
+      "learning_rate": 1.848172114309513e-05,
+      "loss": 1.3924,
+      "step": 249
+    },
+    {
+      "epoch": 0.203231378924906,
+      "grad_norm": 2.520263671875,
+      "learning_rate": 1.8467742322474822e-05,
+      "loss": 1.4097,
+      "step": 250
+    },
+    {
+      "epoch": 0.20404430444060562,
+      "grad_norm": 4.465703964233398,
+      "learning_rate": 1.845370478182829e-05,
+      "loss": 1.3645,
+      "step": 251
+    },
+    {
+      "epoch": 0.20485722995630526,
+      "grad_norm": 2.5109176635742188,
+      "learning_rate": 1.8439608618499637e-05,
+      "loss": 1.3238,
+      "step": 252
+    },
+    {
+      "epoch": 0.20567015547200487,
+      "grad_norm": 2.703659772872925,
+      "learning_rate": 1.842545393023949e-05,
+      "loss": 1.4027,
+      "step": 253
+    },
+    {
+      "epoch": 0.20648308098770451,
+      "grad_norm": 3.483933448791504,
+      "learning_rate": 1.841124081520431e-05,
+      "loss": 1.4167,
+      "step": 254
+    },
+    {
+      "epoch": 0.20729600650340413,
+      "grad_norm": 2.7172889709472656,
+      "learning_rate": 1.8396969371955724e-05,
+      "loss": 1.3017,
+      "step": 255
+    },
+    {
+      "epoch": 0.20810893201910374,
+      "grad_norm": 2.512045383453369,
+      "learning_rate": 1.838263969945985e-05,
+      "loss": 1.4112,
+      "step": 256
+    },
+    {
+      "epoch": 0.20892185753480338,
+      "grad_norm": 2.4449141025543213,
+      "learning_rate": 1.836825189708659e-05,
+      "loss": 1.3396,
+      "step": 257
+    },
+    {
+      "epoch": 0.209734783050503,
+      "grad_norm": 2.9280951023101807,
+      "learning_rate": 1.8353806064608953e-05,
+      "loss": 1.3461,
+      "step": 258
+    },
+    {
+      "epoch": 0.2105477085662026,
+      "grad_norm": 3.962769031524658,
+      "learning_rate": 1.833930230220236e-05,
+      "loss": 1.3347,
+      "step": 259
+    },
+    {
+      "epoch": 0.21136063408190225,
+      "grad_norm": 3.3168771266937256,
+      "learning_rate": 1.8324740710443955e-05,
+      "loss": 1.3264,
+      "step": 260
+    },
+    {
+      "epoch": 0.21217355959760187,
+      "grad_norm": 2.754786252975464,
+      "learning_rate": 1.831012139031189e-05,
+      "loss": 1.3859,
+      "step": 261
+    },
+    {
+      "epoch": 0.21298648511330148,
+      "grad_norm": 2.5179426670074463,
+      "learning_rate": 1.829544444318466e-05,
+      "loss": 1.3653,
+      "step": 262
+    },
+    {
+      "epoch": 0.21379941062900112,
+      "grad_norm": 2.9228906631469727,
+      "learning_rate": 1.8280709970840352e-05,
+      "loss": 1.3929,
+      "step": 263
+    },
+    {
+      "epoch": 0.21461233614470074,
+      "grad_norm": 2.732806921005249,
+      "learning_rate": 1.8265918075455985e-05,
+      "loss": 1.3197,
+      "step": 264
+    },
+    {
+      "epoch": 0.21542526166040035,
+      "grad_norm": 2.7236287593841553,
+      "learning_rate": 1.8251068859606777e-05,
+      "loss": 1.3156,
+      "step": 265
+    },
+    {
+      "epoch": 0.2162381871761,
+      "grad_norm": 3.677654504776001,
+      "learning_rate": 1.823616242626542e-05,
+      "loss": 1.3565,
+      "step": 266
+    },
+    {
+      "epoch": 0.2170511126917996,
+      "grad_norm": 2.4574098587036133,
+      "learning_rate": 1.8221198878801415e-05,
+      "loss": 1.3802,
+      "step": 267
+    },
+    {
+      "epoch": 0.21786403820749925,
+      "grad_norm": 3.2601144313812256,
+      "learning_rate": 1.8206178320980295e-05,
+      "loss": 1.3606,
+      "step": 268
+    },
+    {
+      "epoch": 0.21867696372319886,
+      "grad_norm": 2.4183156490325928,
+      "learning_rate": 1.819110085696295e-05,
+      "loss": 1.3327,
+      "step": 269
+    },
+    {
+      "epoch": 0.21948988923889848,
+      "grad_norm": 2.6820755004882812,
+      "learning_rate": 1.817596659130489e-05,
+      "loss": 1.3676,
+      "step": 270
+    },
+    {
+      "epoch": 0.22030281475459812,
+      "grad_norm": 2.619580030441284,
+      "learning_rate": 1.816077562895551e-05,
+      "loss": 1.408,
+      "step": 271
+    },
+    {
+      "epoch": 0.22111574027029773,
+      "grad_norm": 2.4499645233154297,
+      "learning_rate": 1.814552807525738e-05,
+      "loss": 1.3445,
+      "step": 272
+    },
+    {
+      "epoch": 0.22192866578599735,
+      "grad_norm": 2.5966873168945312,
+      "learning_rate": 1.81302240359455e-05,
+      "loss": 1.3354,
+      "step": 273
+    },
+    {
+      "epoch": 0.222741591301697,
+      "grad_norm": 8.227926254272461,
+      "learning_rate": 1.8114863617146576e-05,
+      "loss": 1.3495,
+      "step": 274
+    },
+    {
+      "epoch": 0.2235545168173966,
+      "grad_norm": 5.334491729736328,
+      "learning_rate": 1.8099446925378278e-05,
+      "loss": 1.3845,
+      "step": 275
+    },
+    {
+      "epoch": 0.22436744233309622,
+      "grad_norm": 2.436473846435547,
+      "learning_rate": 1.8083974067548506e-05,
+      "loss": 1.3152,
+      "step": 276
+    },
+    {
+      "epoch": 0.22518036784879586,
+      "grad_norm": 2.4906110763549805,
+      "learning_rate": 1.806844515095465e-05,
+      "loss": 1.3213,
+      "step": 277
+    },
+    {
+      "epoch": 0.22599329336449547,
+      "grad_norm": 2.627547264099121,
+      "learning_rate": 1.8052860283282832e-05,
+      "loss": 1.3394,
+      "step": 278
+    },
+    {
+      "epoch": 0.22680621888019512,
+      "grad_norm": 3.9034616947174072,
+      "learning_rate": 1.8037219572607177e-05,
+      "loss": 1.2956,
+      "step": 279
+    },
+    {
+      "epoch": 0.22761914439589473,
+      "grad_norm": 2.9307639598846436,
+      "learning_rate": 1.8021523127389066e-05,
+      "loss": 1.3507,
+      "step": 280
+    },
+    {
+      "epoch": 0.22843206991159434,
+      "grad_norm": 2.6711225509643555,
+      "learning_rate": 1.800577105647635e-05,
+      "loss": 1.4043,
+      "step": 281
+    },
+    {
+      "epoch": 0.22924499542729398,
+      "grad_norm": 2.9251246452331543,
+      "learning_rate": 1.7989963469102643e-05,
+      "loss": 1.3424,
+      "step": 282
+    },
+    {
+      "epoch": 0.2300579209429936,
+      "grad_norm": 2.2818679809570312,
+      "learning_rate": 1.797410047488653e-05,
+      "loss": 1.334,
+      "step": 283
+    },
+    {
+      "epoch": 0.2308708464586932,
+      "grad_norm": 2.6961264610290527,
+      "learning_rate": 1.7958182183830816e-05,
+      "loss": 1.3411,
+      "step": 284
+    },
+    {
+      "epoch": 0.23168377197439285,
+      "grad_norm": 2.5082268714904785,
+      "learning_rate": 1.794220870632177e-05,
+      "loss": 1.3815,
+      "step": 285
+    },
+    {
+      "epoch": 0.23249669749009247,
+      "grad_norm": 2.6569674015045166,
+      "learning_rate": 1.7926180153128358e-05,
+      "loss": 1.4037,
+      "step": 286
+    },
+    {
+      "epoch": 0.23330962300579208,
+      "grad_norm": 2.559483289718628,
+      "learning_rate": 1.791009663540146e-05,
+      "loss": 1.333,
+      "step": 287
+    },
+    {
+      "epoch": 0.23412254852149172,
+      "grad_norm": 2.6982040405273438,
+      "learning_rate": 1.789395826467312e-05,
+      "loss": 1.4168,
+      "step": 288
+    },
+    {
+      "epoch": 0.23493547403719134,
+      "grad_norm": 2.414900541305542,
+      "learning_rate": 1.7877765152855757e-05,
+      "loss": 1.3583,
+      "step": 289
+    },
+    {
+      "epoch": 0.23574839955289095,
+      "grad_norm": 2.465045928955078,
+      "learning_rate": 1.78615174122414e-05,
+      "loss": 1.44,
+      "step": 290
+    },
+    {
+      "epoch": 0.2365613250685906,
+      "grad_norm": 2.306795597076416,
+      "learning_rate": 1.78452151555009e-05,
+      "loss": 1.3215,
+      "step": 291
+    },
+    {
+      "epoch": 0.2373742505842902,
+      "grad_norm": 2.6841700077056885,
+      "learning_rate": 1.7828858495683162e-05,
+      "loss": 1.351,
+      "step": 292
+    },
+    {
+      "epoch": 0.23818717609998985,
+      "grad_norm": 2.4231340885162354,
+      "learning_rate": 1.781244754621434e-05,
+      "loss": 1.3923,
+      "step": 293
+    },
+    {
+      "epoch": 0.23900010161568946,
+      "grad_norm": 2.8300161361694336,
+      "learning_rate": 1.779598242089707e-05,
+      "loss": 1.3876,
+      "step": 294
+    },
+    {
+      "epoch": 0.23981302713138908,
+      "grad_norm": 2.6287200450897217,
+      "learning_rate": 1.7779463233909677e-05,
+      "loss": 1.3609,
+      "step": 295
+    },
+    {
+      "epoch": 0.24062595264708872,
+      "grad_norm": 2.656332015991211,
+      "learning_rate": 1.7762890099805362e-05,
+      "loss": 1.3538,
+      "step": 296
+    },
+    {
+      "epoch": 0.24143887816278833,
+      "grad_norm": 2.5331099033355713,
+      "learning_rate": 1.774626313351145e-05,
+      "loss": 1.3154,
+      "step": 297
+    },
+    {
+      "epoch": 0.24225180367848795,
+      "grad_norm": 2.8881306648254395,
+      "learning_rate": 1.7729582450328547e-05,
+      "loss": 1.3561,
+      "step": 298
+    },
+    {
+      "epoch": 0.2430647291941876,
+      "grad_norm": 2.4491260051727295,
+      "learning_rate": 1.771284816592978e-05,
+      "loss": 1.3494,
+      "step": 299
+    },
+    {
+      "epoch": 0.2438776547098872,
+      "grad_norm": 2.8161392211914062,
+      "learning_rate": 1.7696060396359956e-05,
+      "loss": 1.3125,
+      "step": 300
+    },
+    {
+      "epoch": 0.24469058022558682,
+      "grad_norm": 2.788238048553467,
+      "learning_rate": 1.7679219258034798e-05,
+      "loss": 1.41,
+      "step": 301
+    },
+    {
+      "epoch": 0.24550350574128646,
+      "grad_norm": 3.0948519706726074,
+      "learning_rate": 1.7662324867740102e-05,
+      "loss": 1.4138,
+      "step": 302
+    },
+    {
+      "epoch": 0.24631643125698607,
+      "grad_norm": 3.617783308029175,
+      "learning_rate": 1.7645377342630956e-05,
+      "loss": 1.3995,
+      "step": 303
+    },
+    {
+      "epoch": 0.24712935677268572,
+      "grad_norm": 2.713531255722046,
+      "learning_rate": 1.76283768002309e-05,
+      "loss": 1.354,
+      "step": 304
+    },
+    {
+      "epoch": 0.24794228228838533,
+      "grad_norm": 3.9215407371520996,
+      "learning_rate": 1.7611323358431145e-05,
+      "loss": 1.3939,
+      "step": 305
+    },
+    {
+      "epoch": 0.24875520780408494,
+      "grad_norm": 3.519932508468628,
+      "learning_rate": 1.759421713548971e-05,
+      "loss": 1.3311,
+      "step": 306
+    },
+    {
+      "epoch": 0.24956813331978459,
+      "grad_norm": 3.0680055618286133,
+      "learning_rate": 1.757705825003065e-05,
+      "loss": 1.4131,
+      "step": 307
+    },
+    {
+      "epoch": 0.25038105883548417,
+      "grad_norm": 2.456533908843994,
+      "learning_rate": 1.7559846821043205e-05,
+      "loss": 1.3132,
+      "step": 308
+    },
+    {
+      "epoch": 0.25119398435118384,
+      "grad_norm": 2.6937081813812256,
+      "learning_rate": 1.754258296788097e-05,
+      "loss": 1.3041,
+      "step": 309
+    },
+    {
+      "epoch": 0.25200690986688346,
+      "grad_norm": 5.319806098937988,
+      "learning_rate": 1.7525266810261096e-05,
+      "loss": 1.3544,
+      "step": 310
+    },
+    {
+      "epoch": 0.25281983538258307,
+      "grad_norm": 2.9595742225646973,
+      "learning_rate": 1.7507898468263422e-05,
+      "loss": 1.3528,
+      "step": 311
+    },
+    {
+      "epoch": 0.2536327608982827,
+      "grad_norm": 4.085862636566162,
+      "learning_rate": 1.7490478062329686e-05,
+      "loss": 1.3314,
+      "step": 312
+    },
+    {
+      "epoch": 0.2544456864139823,
+      "grad_norm": 2.4585909843444824,
+      "learning_rate": 1.7473005713262644e-05,
+      "loss": 1.3622,
+      "step": 313
+    },
+    {
+      "epoch": 0.25525861192968197,
+      "grad_norm": 2.4798450469970703,
+      "learning_rate": 1.7455481542225272e-05,
+      "loss": 1.3804,
+      "step": 314
+    },
+    {
+      "epoch": 0.2560715374453816,
+      "grad_norm": 2.686068534851074,
+      "learning_rate": 1.7437905670739893e-05,
+      "loss": 1.2945,
+      "step": 315
+    },
+    {
+      "epoch": 0.2568844629610812,
+      "grad_norm": 2.7424585819244385,
+      "learning_rate": 1.7420278220687366e-05,
+      "loss": 1.3561,
+      "step": 316
+    },
+    {
+      "epoch": 0.2576973884767808,
+      "grad_norm": 2.964237928390503,
+      "learning_rate": 1.7402599314306207e-05,
+      "loss": 1.3701,
+      "step": 317
+    },
+    {
+      "epoch": 0.2585103139924804,
+      "grad_norm": 2.7983458042144775,
+      "learning_rate": 1.7384869074191777e-05,
+      "loss": 1.3536,
+      "step": 318
+    },
+    {
+      "epoch": 0.25932323950818004,
+      "grad_norm": 2.6008524894714355,
+      "learning_rate": 1.7367087623295394e-05,
+      "loss": 1.3394,
+      "step": 319
+    },
+    {
+      "epoch": 0.2601361650238797,
+      "grad_norm": 2.4116249084472656,
+      "learning_rate": 1.7349255084923517e-05,
+      "loss": 1.3785,
+      "step": 320
+    },
+    {
+      "epoch": 0.2609490905395793,
+      "grad_norm": 2.9649388790130615,
+      "learning_rate": 1.7331371582736864e-05,
+      "loss": 1.3779,
+      "step": 321
+    },
+    {
+      "epoch": 0.26176201605527893,
+      "grad_norm": 2.692847490310669,
+      "learning_rate": 1.731343724074957e-05,
+      "loss": 1.3715,
+      "step": 322
+    },
+    {
+      "epoch": 0.26257494157097855,
+      "grad_norm": 2.6246955394744873,
+      "learning_rate": 1.7295452183328317e-05,
+      "loss": 1.3856,
+      "step": 323
+    },
+    {
+      "epoch": 0.26338786708667816,
+      "grad_norm": 2.822334051132202,
+      "learning_rate": 1.7277416535191478e-05,
+      "loss": 1.3289,
+      "step": 324
+    },
+    {
+      "epoch": 0.26420079260237783,
+      "grad_norm": 2.703158378601074,
+      "learning_rate": 1.7259330421408247e-05,
+      "loss": 1.3447,
+      "step": 325
+    },
+    {
+      "epoch": 0.26501371811807745,
+      "grad_norm": 2.5357322692871094,
+      "learning_rate": 1.7241193967397784e-05,
+      "loss": 1.3414,
+      "step": 326
+    },
+    {
+      "epoch": 0.26582664363377706,
+      "grad_norm": 2.7839202880859375,
+      "learning_rate": 1.7223007298928322e-05,
+      "loss": 1.3725,
+      "step": 327
+    },
+    {
+      "epoch": 0.2666395691494767,
+      "grad_norm": 2.6645684242248535,
+      "learning_rate": 1.7204770542116326e-05,
+      "loss": 1.3163,
+      "step": 328
+    },
+    {
+      "epoch": 0.2674524946651763,
+      "grad_norm": 4.677945137023926,
+      "learning_rate": 1.7186483823425582e-05,
+      "loss": 1.3583,
+      "step": 329
+    },
+    {
+      "epoch": 0.2682654201808759,
+      "grad_norm": 2.948094367980957,
+      "learning_rate": 1.7168147269666357e-05,
+      "loss": 1.3643,
+      "step": 330
+    },
+    {
+      "epoch": 0.26907834569657557,
+      "grad_norm": 2.5047991275787354,
+      "learning_rate": 1.714976100799449e-05,
+      "loss": 1.3542,
+      "step": 331
+    },
+    {
+      "epoch": 0.2698912712122752,
+      "grad_norm": 2.680239677429199,
+      "learning_rate": 1.713132516591053e-05,
+      "loss": 1.3204,
+      "step": 332
+    },
+    {
+      "epoch": 0.2707041967279748,
+      "grad_norm": 2.703165054321289,
+      "learning_rate": 1.7112839871258838e-05,
+      "loss": 1.3467,
+      "step": 333
+    },
+    {
+      "epoch": 0.2715171222436744,
+      "grad_norm": 2.5855846405029297,
+      "learning_rate": 1.7094305252226713e-05,
+      "loss": 1.3807,
+      "step": 334
+    },
+    {
+      "epoch": 0.272330047759374,
+      "grad_norm": 2.8401761054992676,
+      "learning_rate": 1.7075721437343488e-05,
+      "loss": 1.4032,
+      "step": 335
+    },
+    {
+      "epoch": 0.2731429732750737,
+      "grad_norm": 2.727287530899048,
+      "learning_rate": 1.705708855547966e-05,
+      "loss": 1.3416,
+      "step": 336
+    },
+    {
+      "epoch": 0.2739558987907733,
+      "grad_norm": 2.9767589569091797,
+      "learning_rate": 1.7038406735845967e-05,
+      "loss": 1.3062,
+      "step": 337
+    },
+    {
+      "epoch": 0.2747688243064729,
+      "grad_norm": 2.6532137393951416,
+      "learning_rate": 1.7019676107992523e-05,
+      "loss": 1.3717,
+      "step": 338
+    },
+    {
+      "epoch": 0.27558174982217254,
+      "grad_norm": 5.618951797485352,
+      "learning_rate": 1.70008968018079e-05,
+      "loss": 1.4021,
+      "step": 339
+    },
+    {
+      "epoch": 0.27639467533787215,
+      "grad_norm": 2.75219464302063,
+      "learning_rate": 1.6982068947518235e-05,
+      "loss": 1.3345,
+      "step": 340
+    },
+    {
+      "epoch": 0.27720760085357177,
+      "grad_norm": 2.7771074771881104,
+      "learning_rate": 1.6963192675686312e-05,
+      "loss": 1.3613,
+      "step": 341
+    },
+    {
+      "epoch": 0.27802052636927144,
+      "grad_norm": 2.4822003841400146,
+      "learning_rate": 1.694426811721069e-05,
+      "loss": 1.3465,
+      "step": 342
+    },
+    {
+      "epoch": 0.27883345188497105,
+      "grad_norm": 2.684894323348999,
+      "learning_rate": 1.6925295403324758e-05,
+      "loss": 1.337,
+      "step": 343
+    },
+    {
+      "epoch": 0.27964637740067066,
+      "grad_norm": 2.804255962371826,
+      "learning_rate": 1.6906274665595854e-05,
+      "loss": 1.2862,
+      "step": 344
+    },
+    {
+      "epoch": 0.2804593029163703,
+      "grad_norm": 2.7327306270599365,
+      "learning_rate": 1.688720603592432e-05,
+      "loss": 1.3826,
+      "step": 345
+    },
+    {
+      "epoch": 0.2812722284320699,
+      "grad_norm": 3.0967769622802734,
+      "learning_rate": 1.6868089646542632e-05,
+      "loss": 1.3406,
+      "step": 346
+    },
+    {
+      "epoch": 0.28208515394776956,
+      "grad_norm": 2.4972376823425293,
+      "learning_rate": 1.6848925630014445e-05,
+      "loss": 1.3315,
+      "step": 347
+    },
+    {
+      "epoch": 0.2828980794634692,
+      "grad_norm": 22.60991668701172,
+      "learning_rate": 1.6829714119233688e-05,
+      "loss": 1.3325,
+      "step": 348
+    },
+    {
+      "epoch": 0.2837110049791688,
+      "grad_norm": 3.207625389099121,
+      "learning_rate": 1.6810455247423634e-05,
+      "loss": 1.3926,
+      "step": 349
+    },
+    {
+      "epoch": 0.2845239304948684,
+      "grad_norm": 2.6568946838378906,
+      "learning_rate": 1.6791149148136003e-05,
+      "loss": 1.3464,
+      "step": 350
+    },
+    {
+      "epoch": 0.285336856010568,
+      "grad_norm": 2.9483156204223633,
+      "learning_rate": 1.677179595525e-05,
+      "loss": 1.2875,
+      "step": 351
+    },
+    {
+      "epoch": 0.28614978152626763,
+      "grad_norm": 2.841442584991455,
+      "learning_rate": 1.675239580297141e-05,
+      "loss": 1.3441,
+      "step": 352
+    },
+    {
+      "epoch": 0.2869627070419673,
+      "grad_norm": 3.3877551555633545,
+      "learning_rate": 1.6732948825831657e-05,
+      "loss": 1.3662,
+      "step": 353
+    },
+    {
+      "epoch": 0.2877756325576669,
+      "grad_norm": 2.9442946910858154,
+      "learning_rate": 1.671345515868688e-05,
+      "loss": 1.3075,
+      "step": 354
+    },
+    {
+      "epoch": 0.28858855807336653,
+      "grad_norm": 2.672950029373169,
+      "learning_rate": 1.6693914936716983e-05,
+      "loss": 1.2982,
+      "step": 355
+    },
+    {
+      "epoch": 0.28940148358906614,
+      "grad_norm": 2.7699198722839355,
+      "learning_rate": 1.6674328295424723e-05,
+      "loss": 1.3331,
+      "step": 356
+    },
+    {
+      "epoch": 0.29021440910476576,
+      "grad_norm": 2.578444719314575,
+      "learning_rate": 1.6654695370634738e-05,
+      "loss": 1.3768,
+      "step": 357
+    },
+    {
+      "epoch": 0.29102733462046537,
+      "grad_norm": 2.748466968536377,
+      "learning_rate": 1.6635016298492628e-05,
+      "loss": 1.3108,
+      "step": 358
+    },
+    {
+      "epoch": 0.29184026013616504,
+      "grad_norm": 2.818321943283081,
+      "learning_rate": 1.6615291215464005e-05,
+      "loss": 1.2586,
+      "step": 359
+    },
+    {
+      "epoch": 0.29265318565186466,
+      "grad_norm": 3.6742396354675293,
+      "learning_rate": 1.6595520258333545e-05,
+      "loss": 1.3112,
+      "step": 360
+    },
+    {
+      "epoch": 0.29346611116756427,
+      "grad_norm": 2.999140977859497,
+      "learning_rate": 1.657570356420404e-05,
+      "loss": 1.2923,
+      "step": 361
+    },
+    {
+      "epoch": 0.2942790366832639,
+      "grad_norm": 2.704463481903076,
+      "learning_rate": 1.6555841270495456e-05,
+      "loss": 1.3329,
+      "step": 362
+    },
+    {
+      "epoch": 0.2950919621989635,
+      "grad_norm": 3.2639801502227783,
+      "learning_rate": 1.6535933514943955e-05,
+      "loss": 1.3215,
+      "step": 363
+    },
+    {
+      "epoch": 0.29590488771466317,
+      "grad_norm": 3.2200841903686523,
+      "learning_rate": 1.6515980435600965e-05,
+      "loss": 1.3792,
+      "step": 364
+    },
+    {
+      "epoch": 0.2967178132303628,
+      "grad_norm": 2.9226245880126953,
+      "learning_rate": 1.6495982170832224e-05,
+      "loss": 1.3565,
+      "step": 365
+    },
+    {
+      "epoch": 0.2975307387460624,
+      "grad_norm": 3.096405029296875,
+      "learning_rate": 1.6475938859316795e-05,
+      "loss": 1.3857,
+      "step": 366
+    },
+    {
+      "epoch": 0.298343664261762,
+      "grad_norm": 2.7694365978240967,
+      "learning_rate": 1.6455850640046134e-05,
+      "loss": 1.3782,
+      "step": 367
+    },
+    {
+      "epoch": 0.2991565897774616,
+      "grad_norm": 3.011751890182495,
+      "learning_rate": 1.6435717652323097e-05,
+      "loss": 1.3426,
+      "step": 368
+    },
+    {
+      "epoch": 0.29996951529316124,
+      "grad_norm": 2.7828853130340576,
+      "learning_rate": 1.6415540035761008e-05,
+      "loss": 1.3429,
+      "step": 369
+    },
+    {
+      "epoch": 0.3007824408088609,
+      "grad_norm": 2.5543785095214844,
+      "learning_rate": 1.639531793028265e-05,
+      "loss": 1.3768,
+      "step": 370
+    },
+    {
+      "epoch": 0.3015953663245605,
+      "grad_norm": 2.8462271690368652,
+      "learning_rate": 1.637505147611934e-05,
+      "loss": 1.3203,
+      "step": 371
+    },
+    {
+      "epoch": 0.30240829184026013,
+      "grad_norm": 2.404257297515869,
+      "learning_rate": 1.6354740813809917e-05,
+      "loss": 1.3693,
+      "step": 372
+    },
+    {
+      "epoch": 0.30322121735595975,
+      "grad_norm": 2.674553394317627,
+      "learning_rate": 1.6334386084199787e-05,
+      "loss": 1.3518,
+      "step": 373
+    },
+    {
+      "epoch": 0.30403414287165936,
+      "grad_norm": 2.4954397678375244,
+      "learning_rate": 1.631398742843995e-05,
+      "loss": 1.3669,
+      "step": 374
+    },
+    {
+      "epoch": 0.30484706838735903,
+      "grad_norm": 3.333721876144409,
+      "learning_rate": 1.629354498798601e-05,
+      "loss": 1.3358,
+      "step": 375
+    },
+    {
+      "epoch": 0.30565999390305865,
+      "grad_norm": 2.859560966491699,
+      "learning_rate": 1.627305890459719e-05,
+      "loss": 1.3334,
+      "step": 376
+    },
+    {
+      "epoch": 0.30647291941875826,
+      "grad_norm": 2.8346803188323975,
+      "learning_rate": 1.625252932033538e-05,
+      "loss": 1.3366,
+      "step": 377
+    },
+    {
+      "epoch": 0.3072858449344579,
+      "grad_norm": 2.64909029006958,
+      "learning_rate": 1.6231956377564095e-05,
+      "loss": 1.3398,
+      "step": 378
+    },
+    {
+      "epoch": 0.3080987704501575,
+      "grad_norm": 3.935067653656006,
+      "learning_rate": 1.621134021894756e-05,
+      "loss": 1.2953,
+      "step": 379
+    },
+    {
+      "epoch": 0.3089116959658571,
+      "grad_norm": 5.056494235992432,
+      "learning_rate": 1.619068098744965e-05,
+      "loss": 1.3245,
+      "step": 380
+    },
+    {
+      "epoch": 0.30972462148155677,
+      "grad_norm": 2.9668800830841064,
+      "learning_rate": 1.6169978826332955e-05,
+      "loss": 1.3199,
+      "step": 381
+    },
+    {
+      "epoch": 0.3105375469972564,
+      "grad_norm": 2.6101276874542236,
+      "learning_rate": 1.6149233879157747e-05,
+      "loss": 1.3317,
+      "step": 382
+    },
+    {
+      "epoch": 0.311350472512956,
+      "grad_norm": 2.677374839782715,
+      "learning_rate": 1.6128446289781012e-05,
+      "loss": 1.304,
+      "step": 383
+    },
+    {
+      "epoch": 0.3121633980286556,
+      "grad_norm": 4.049331188201904,
+      "learning_rate": 1.610761620235543e-05,
+      "loss": 1.3241,
+      "step": 384
+    },
+    {
+      "epoch": 0.31297632354435523,
+      "grad_norm": 2.566908836364746,
+      "learning_rate": 1.60867437613284e-05,
+      "loss": 1.3392,
+      "step": 385
+    },
+    {
+      "epoch": 0.3137892490600549,
+      "grad_norm": 2.550367832183838,
+      "learning_rate": 1.6065829111441e-05,
+      "loss": 1.3274,
+      "step": 386
+    },
+    {
+      "epoch": 0.3146021745757545,
+      "grad_norm": 4.543491363525391,
+      "learning_rate": 1.6044872397727037e-05,
+      "loss": 1.2993,
+      "step": 387
+    },
+    {
+      "epoch": 0.3154151000914541,
+      "grad_norm": 2.8900489807128906,
+      "learning_rate": 1.6023873765511993e-05,
+      "loss": 1.3274,
+      "step": 388
+    },
+    {
+      "epoch": 0.31622802560715374,
+      "grad_norm": 2.4930450916290283,
+      "learning_rate": 1.6002833360412044e-05,
+      "loss": 1.3074,
+      "step": 389
+    },
+    {
+      "epoch": 0.31704095112285335,
+      "grad_norm": 3.0221235752105713,
+      "learning_rate": 1.5981751328333036e-05,
+      "loss": 1.3077,
+      "step": 390
+    },
+    {
+      "epoch": 0.31785387663855297,
+      "grad_norm": 3.0569851398468018,
+      "learning_rate": 1.5960627815469486e-05,
+      "loss": 1.3705,
+      "step": 391
+    },
+    {
+      "epoch": 0.31866680215425264,
+      "grad_norm": 7.261632442474365,
+      "learning_rate": 1.5939462968303554e-05,
+      "loss": 1.3564,
+      "step": 392
+    },
+    {
+      "epoch": 0.31947972766995225,
+      "grad_norm": 3.0555789470672607,
+      "learning_rate": 1.5918256933604047e-05,
+      "loss": 1.3451,
+      "step": 393
+    },
+    {
+      "epoch": 0.32029265318565187,
+      "grad_norm": 3.360779047012329,
+      "learning_rate": 1.589700985842538e-05,
+      "loss": 1.2764,
+      "step": 394
+    },
+    {
+      "epoch": 0.3211055787013515,
+      "grad_norm": 2.9022507667541504,
+      "learning_rate": 1.5875721890106574e-05,
+      "loss": 1.3424,
+      "step": 395
+    },
+    {
+      "epoch": 0.3219185042170511,
+      "grad_norm": 5.119380474090576,
+      "learning_rate": 1.5854393176270205e-05,
+      "loss": 1.3392,
+      "step": 396
+    },
+    {
+      "epoch": 0.32273142973275076,
+      "grad_norm": 2.7554409503936768,
+      "learning_rate": 1.5833023864821427e-05,
+      "loss": 1.3762,
+      "step": 397
+    },
+    {
+      "epoch": 0.3235443552484504,
+      "grad_norm": 2.553323984146118,
+      "learning_rate": 1.5811614103946905e-05,
+      "loss": 1.3066,
+      "step": 398
+    },
+    {
+      "epoch": 0.32435728076415,
+      "grad_norm": 3.514381170272827,
+      "learning_rate": 1.5790164042113805e-05,
+      "loss": 1.3575,
+      "step": 399
+    },
+    {
+      "epoch": 0.3251702062798496,
+      "grad_norm": 2.89054012298584,
+      "learning_rate": 1.576867382806877e-05,
+      "loss": 1.3106,
+      "step": 400
+    },
+    {
+      "epoch": 0.3259831317955492,
+      "grad_norm": 2.9955763816833496,
+      "learning_rate": 1.5747143610836873e-05,
+      "loss": 1.3634,
+      "step": 401
+    },
+    {
+      "epoch": 0.32679605731124883,
+      "grad_norm": 3.175438404083252,
+      "learning_rate": 1.5725573539720592e-05,
+      "loss": 1.2876,
+      "step": 402
+    },
+    {
+      "epoch": 0.3276089828269485,
+      "grad_norm": 2.6269116401672363,
+      "learning_rate": 1.570396376429877e-05,
+      "loss": 1.342,
+      "step": 403
+    },
+    {
+      "epoch": 0.3284219083426481,
+      "grad_norm": 2.900568962097168,
+      "learning_rate": 1.5682314434425593e-05,
+      "loss": 1.3133,
+      "step": 404
+    },
+    {
+      "epoch": 0.32923483385834773,
+      "grad_norm": 2.6711323261260986,
+      "learning_rate": 1.5660625700229526e-05,
+      "loss": 1.2702,
+      "step": 405
+    },
+    {
+      "epoch": 0.33004775937404734,
+      "grad_norm": 2.8045928478240967,
+      "learning_rate": 1.5638897712112303e-05,
+      "loss": 1.3336,
+      "step": 406
+    },
+    {
+      "epoch": 0.33086068488974696,
+      "grad_norm": 2.9632303714752197,
+      "learning_rate": 1.561713062074785e-05,
+      "loss": 1.3546,
+      "step": 407
+    },
+    {
+      "epoch": 0.33167361040544663,
+      "grad_norm": 2.5156984329223633,
+      "learning_rate": 1.5595324577081265e-05,
+      "loss": 1.3587,
+      "step": 408
+    },
+    {
+      "epoch": 0.33248653592114624,
+      "grad_norm": 2.6634364128112793,
+      "learning_rate": 1.5573479732327758e-05,
+      "loss": 1.3317,
+      "step": 409
+    },
+    {
+      "epoch": 0.33329946143684586,
+      "grad_norm": 4.38008451461792,
+      "learning_rate": 1.555159623797161e-05,
+      "loss": 1.3078,
+      "step": 410
+    },
+    {
+      "epoch": 0.33411238695254547,
+      "grad_norm": 3.089078903198242,
+      "learning_rate": 1.552967424576512e-05,
+      "loss": 1.328,
+      "step": 411
+    },
+    {
+      "epoch": 0.3349253124682451,
+      "grad_norm": 2.9011247158050537,
+      "learning_rate": 1.5507713907727557e-05,
+      "loss": 1.349,
+      "step": 412
+    },
+    {
+      "epoch": 0.3357382379839447,
+      "grad_norm": 2.431152582168579,
+      "learning_rate": 1.5485715376144087e-05,
+      "loss": 1.383,
+      "step": 413
+    },
+    {
+      "epoch": 0.33655116349964437,
+      "grad_norm": 2.6097633838653564,
+      "learning_rate": 1.5463678803564753e-05,
+      "loss": 1.3414,
+      "step": 414
+    },
+    {
+      "epoch": 0.337364089015344,
+      "grad_norm": 2.9973533153533936,
+      "learning_rate": 1.5441604342803374e-05,
+      "loss": 1.3359,
+      "step": 415
+    },
+    {
+      "epoch": 0.3381770145310436,
+      "grad_norm": 2.849950075149536,
+      "learning_rate": 1.5419492146936518e-05,
+      "loss": 1.3378,
+      "step": 416
+    },
+    {
+      "epoch": 0.3389899400467432,
+      "grad_norm": 2.600947856903076,
+      "learning_rate": 1.5397342369302425e-05,
+      "loss": 1.3411,
+      "step": 417
+    },
+    {
+      "epoch": 0.3398028655624428,
+      "grad_norm": 2.946190595626831,
+      "learning_rate": 1.5375155163499953e-05,
+      "loss": 1.2981,
+      "step": 418
+    },
+    {
+      "epoch": 0.34061579107814244,
+      "grad_norm": 3.5300893783569336,
+      "learning_rate": 1.5352930683387502e-05,
+      "loss": 1.3717,
+      "step": 419
+    },
+    {
+      "epoch": 0.3414287165938421,
+      "grad_norm": 2.342288017272949,
+      "learning_rate": 1.5330669083081956e-05,
+      "loss": 1.2734,
+      "step": 420
+    },
+    {
+      "epoch": 0.3422416421095417,
+      "grad_norm": 3.7037856578826904,
+      "learning_rate": 1.5308370516957617e-05,
+      "loss": 1.3402,
+      "step": 421
+    },
+    {
+      "epoch": 0.34305456762524134,
+      "grad_norm": 2.5814309120178223,
+      "learning_rate": 1.528603513964511e-05,
+      "loss": 1.3207,
+      "step": 422
+    },
+    {
+      "epoch": 0.34386749314094095,
+      "grad_norm": 2.4542317390441895,
+      "learning_rate": 1.5263663106030347e-05,
+      "loss": 1.3257,
+      "step": 423
+    },
+    {
+      "epoch": 0.34468041865664056,
+      "grad_norm": 2.689870595932007,
+      "learning_rate": 1.5241254571253433e-05,
+      "loss": 1.3105,
+      "step": 424
+    },
+    {
+      "epoch": 0.34549334417234023,
+      "grad_norm": 2.900061845779419,
+      "learning_rate": 1.5218809690707583e-05,
+      "loss": 1.3113,
+      "step": 425
+    },
+    {
+      "epoch": 0.34630626968803985,
+      "grad_norm": 2.7165238857269287,
+      "learning_rate": 1.5196328620038059e-05,
+      "loss": 1.335,
+      "step": 426
+    },
+    {
+      "epoch": 0.34711919520373946,
+      "grad_norm": 2.3893747329711914,
+      "learning_rate": 1.5173811515141083e-05,
+      "loss": 1.3062,
+      "step": 427
+    },
+    {
+      "epoch": 0.3479321207194391,
+      "grad_norm": 2.568575143814087,
+      "learning_rate": 1.5151258532162771e-05,
+      "loss": 1.3338,
+      "step": 428
+    },
+    {
+      "epoch": 0.3487450462351387,
+      "grad_norm": 3.406301736831665,
+      "learning_rate": 1.5128669827498024e-05,
+      "loss": 1.3189,
+      "step": 429
+    },
+    {
+      "epoch": 0.3495579717508383,
+      "grad_norm": 2.752307653427124,
+      "learning_rate": 1.5106045557789453e-05,
+      "loss": 1.331,
+      "step": 430
+    },
+    {
+      "epoch": 0.350370897266538,
+      "grad_norm": 2.570742130279541,
+      "learning_rate": 1.5083385879926309e-05,
+      "loss": 1.2887,
+      "step": 431
+    },
+    {
+      "epoch": 0.3511838227822376,
+      "grad_norm": 2.4754555225372314,
+      "learning_rate": 1.5060690951043385e-05,
+      "loss": 1.3432,
+      "step": 432
+    },
+    {
+      "epoch": 0.3519967482979372,
+      "grad_norm": 3.853609561920166,
+      "learning_rate": 1.5037960928519902e-05,
+      "loss": 1.3625,
+      "step": 433
+    },
+    {
+      "epoch": 0.3528096738136368,
+      "grad_norm": 2.6506130695343018,
+      "learning_rate": 1.501519596997847e-05,
+      "loss": 1.2797,
+      "step": 434
+    },
+    {
+      "epoch": 0.35362259932933643,
+      "grad_norm": 2.8529601097106934,
+      "learning_rate": 1.499239623328394e-05,
+      "loss": 1.2868,
+      "step": 435
+    },
+    {
+      "epoch": 0.3544355248450361,
+      "grad_norm": 4.091727256774902,
+      "learning_rate": 1.4969561876542348e-05,
+      "loss": 1.2648,
+      "step": 436
+    },
+    {
+      "epoch": 0.3552484503607357,
+      "grad_norm": 2.5217483043670654,
+      "learning_rate": 1.4946693058099802e-05,
+      "loss": 1.2792,
+      "step": 437
+    },
+    {
+      "epoch": 0.3560613758764353,
+      "grad_norm": 3.035297155380249,
+      "learning_rate": 1.4923789936541378e-05,
+      "loss": 1.3267,
+      "step": 438
+    },
+    {
+      "epoch": 0.35687430139213494,
+      "grad_norm": 4.371755599975586,
+      "learning_rate": 1.4900852670690044e-05,
+      "loss": 1.3114,
+      "step": 439
+    },
+    {
+      "epoch": 0.35768722690783455,
+      "grad_norm": 2.904101610183716,
+      "learning_rate": 1.487788141960553e-05,
+      "loss": 1.3716,
+      "step": 440
+    },
+    {
+      "epoch": 0.35850015242353417,
+      "grad_norm": 2.663241147994995,
+      "learning_rate": 1.4854876342583246e-05,
+      "loss": 1.3269,
+      "step": 441
+    },
+    {
+      "epoch": 0.35931307793923384,
+      "grad_norm": 2.626646041870117,
+      "learning_rate": 1.4831837599153165e-05,
+      "loss": 1.3077,
+      "step": 442
+    },
+    {
+      "epoch": 0.36012600345493345,
+      "grad_norm": 2.8876073360443115,
+      "learning_rate": 1.4808765349078729e-05,
+      "loss": 1.2807,
+      "step": 443
+    },
+    {
+      "epoch": 0.36093892897063307,
+      "grad_norm": 2.5428106784820557,
+      "learning_rate": 1.4785659752355724e-05,
+      "loss": 1.3242,
+      "step": 444
+    },
+    {
+      "epoch": 0.3617518544863327,
+      "grad_norm": 2.7515244483947754,
+      "learning_rate": 1.4762520969211186e-05,
+      "loss": 1.3356,
+      "step": 445
+    },
+    {
+      "epoch": 0.3625647800020323,
+      "grad_norm": 2.771684408187866,
+      "learning_rate": 1.4739349160102285e-05,
+      "loss": 1.3255,
+      "step": 446
+    },
+    {
+      "epoch": 0.36337770551773196,
+      "grad_norm": 2.7270543575286865,
+      "learning_rate": 1.4716144485715209e-05,
+      "loss": 1.2797,
+      "step": 447
+    },
+    {
+      "epoch": 0.3641906310334316,
+      "grad_norm": 3.5211868286132812,
+      "learning_rate": 1.4692907106964051e-05,
+      "loss": 1.3098,
+      "step": 448
+    },
+    {
+      "epoch": 0.3650035565491312,
+      "grad_norm": 5.923196315765381,
+      "learning_rate": 1.4669637184989696e-05,
+      "loss": 1.3212,
+      "step": 449
+    },
+    {
+      "epoch": 0.3658164820648308,
+      "grad_norm": 2.50697922706604,
+      "learning_rate": 1.4646334881158704e-05,
+      "loss": 1.3195,
+      "step": 450
+    },
+    {
+      "epoch": 0.3666294075805304,
+      "grad_norm": 3.3721578121185303,
+      "learning_rate": 1.4623000357062184e-05,
+      "loss": 1.2747,
+      "step": 451
+    },
+    {
+      "epoch": 0.36744233309623003,
+      "grad_norm": 2.429243803024292,
+      "learning_rate": 1.459963377451468e-05,
+      "loss": 1.3122,
+      "step": 452
+    },
+    {
+      "epoch": 0.3682552586119297,
+      "grad_norm": 4.240250587463379,
+      "learning_rate": 1.457623529555305e-05,
+      "loss": 1.3447,
+      "step": 453
+    },
+    {
+      "epoch": 0.3690681841276293,
+      "grad_norm": 2.631667137145996,
+      "learning_rate": 1.4552805082435333e-05,
+      "loss": 1.3171,
+      "step": 454
+    },
+    {
+      "epoch": 0.36988110964332893,
+      "grad_norm": 2.906388521194458,
+      "learning_rate": 1.4529343297639638e-05,
+      "loss": 1.3193,
+      "step": 455
+    },
+    {
+      "epoch": 0.37069403515902855,
+      "grad_norm": 3.047884464263916,
+      "learning_rate": 1.4505850103863007e-05,
+      "loss": 1.3181,
+      "step": 456
+    },
+    {
+      "epoch": 0.37150696067472816,
+      "grad_norm": 2.3922433853149414,
+      "learning_rate": 1.448232566402028e-05,
+      "loss": 1.3203,
+      "step": 457
+    },
+    {
+      "epoch": 0.37231988619042783,
+      "grad_norm": 3.278813123703003,
+      "learning_rate": 1.4458770141242992e-05,
+      "loss": 1.3309,
+      "step": 458
+    },
+    {
+      "epoch": 0.37313281170612744,
+      "grad_norm": 2.7148866653442383,
+      "learning_rate": 1.4435183698878212e-05,
+      "loss": 1.3408,
+      "step": 459
+    },
+    {
+      "epoch": 0.37394573722182706,
+      "grad_norm": 2.913823366165161,
+      "learning_rate": 1.4411566500487425e-05,
+      "loss": 1.3426,
+      "step": 460
+    },
+    {
+      "epoch": 0.37475866273752667,
+      "grad_norm": 2.435643196105957,
+      "learning_rate": 1.4387918709845395e-05,
+      "loss": 1.3357,
+      "step": 461
+    },
+    {
+      "epoch": 0.3755715882532263,
+      "grad_norm": 2.6099560260772705,
+      "learning_rate": 1.4364240490939032e-05,
+      "loss": 1.3013,
+      "step": 462
+    },
+    {
+      "epoch": 0.3763845137689259,
+      "grad_norm": 2.7896599769592285,
+      "learning_rate": 1.4340532007966252e-05,
+      "loss": 1.3284,
+      "step": 463
+    },
+    {
+      "epoch": 0.37719743928462557,
+      "grad_norm": 2.857205867767334,
+      "learning_rate": 1.4316793425334836e-05,
+      "loss": 1.2926,
+      "step": 464
+    },
+    {
+      "epoch": 0.3780103648003252,
+      "grad_norm": 2.4580750465393066,
+      "learning_rate": 1.4293024907661295e-05,
+      "loss": 1.3926,
+      "step": 465
+    },
+    {
+      "epoch": 0.3788232903160248,
+      "grad_norm": 2.6340065002441406,
+      "learning_rate": 1.4269226619769727e-05,
+      "loss": 1.3315,
+      "step": 466
+    },
+    {
+      "epoch": 0.3796362158317244,
+      "grad_norm": 3.416398525238037,
+      "learning_rate": 1.424539872669067e-05,
+      "loss": 1.2822,
+      "step": 467
+    },
+    {
+      "epoch": 0.380449141347424,
+      "grad_norm": 2.4222054481506348,
+      "learning_rate": 1.4221541393659966e-05,
+      "loss": 1.2894,
+      "step": 468
+    },
+    {
+      "epoch": 0.38126206686312364,
+      "grad_norm": 2.797074794769287,
+      "learning_rate": 1.4197654786117604e-05,
+      "loss": 1.3519,
+      "step": 469
+    },
+    {
+      "epoch": 0.3820749923788233,
+      "grad_norm": 2.563831329345703,
+      "learning_rate": 1.4173739069706586e-05,
+      "loss": 1.3474,
+      "step": 470
+    },
+    {
+      "epoch": 0.3828879178945229,
+      "grad_norm": 2.4004971981048584,
+      "learning_rate": 1.414979441027176e-05,
+      "loss": 1.3007,
+      "step": 471
+    },
+    {
+      "epoch": 0.38370084341022254,
+      "grad_norm": 2.532390594482422,
+      "learning_rate": 1.4125820973858693e-05,
+      "loss": 1.2613,
+      "step": 472
+    },
+    {
+      "epoch": 0.38451376892592215,
+      "grad_norm": 2.5733683109283447,
+      "learning_rate": 1.41018189267125e-05,
+      "loss": 1.3212,
+      "step": 473
+    },
+    {
+      "epoch": 0.38532669444162176,
+      "grad_norm": 2.710106134414673,
+      "learning_rate": 1.4077788435276701e-05,
+      "loss": 1.3235,
+      "step": 474
+    },
+    {
+      "epoch": 0.38613961995732143,
+      "grad_norm": 2.996795892715454,
+      "learning_rate": 1.4053729666192067e-05,
+      "loss": 1.3722,
+      "step": 475
+    },
+    {
+      "epoch": 0.38695254547302105,
+      "grad_norm": 2.4392545223236084,
+      "learning_rate": 1.4029642786295452e-05,
+      "loss": 1.3706,
+      "step": 476
+    },
+    {
+      "epoch": 0.38776547098872066,
+      "grad_norm": 2.6843369007110596,
+      "learning_rate": 1.400552796261866e-05,
+      "loss": 1.3382,
+      "step": 477
+    },
+    {
+      "epoch": 0.3885783965044203,
+      "grad_norm": 2.405515193939209,
+      "learning_rate": 1.3981385362387268e-05,
+      "loss": 1.316,
+      "step": 478
+    },
+    {
+      "epoch": 0.3893913220201199,
+      "grad_norm": 2.425203800201416,
+      "learning_rate": 1.3957215153019463e-05,
+      "loss": 1.3578,
+      "step": 479
+    },
+    {
+      "epoch": 0.3902042475358195,
+      "grad_norm": 2.5134634971618652,
+      "learning_rate": 1.3933017502124897e-05,
+      "loss": 1.3531,
+      "step": 480
+    },
+    {
+      "epoch": 0.3910171730515192,
+      "grad_norm": 2.4274141788482666,
+      "learning_rate": 1.3908792577503514e-05,
+      "loss": 1.3705,
+      "step": 481
+    },
+    {
+      "epoch": 0.3918300985672188,
+      "grad_norm": 2.881443500518799,
+      "learning_rate": 1.3884540547144393e-05,
+      "loss": 1.3196,
+      "step": 482
+    },
+    {
+      "epoch": 0.3926430240829184,
+      "grad_norm": 2.5505170822143555,
+      "learning_rate": 1.3860261579224574e-05,
+      "loss": 1.3221,
+      "step": 483
+    },
+    {
+      "epoch": 0.393455949598618,
+      "grad_norm": 2.5604939460754395,
+      "learning_rate": 1.3835955842107897e-05,
+      "loss": 1.2565,
+      "step": 484
+    },
+    {
+      "epoch": 0.39426887511431763,
+      "grad_norm": 2.8203351497650146,
+      "learning_rate": 1.3811623504343845e-05,
+      "loss": 1.323,
+      "step": 485
+    },
+    {
+      "epoch": 0.3950818006300173,
+      "grad_norm": 3.9116978645324707,
+      "learning_rate": 1.378726473466635e-05,
+      "loss": 1.3188,
+      "step": 486
+    },
+    {
+      "epoch": 0.3958947261457169,
+      "grad_norm": 2.918548822402954,
+      "learning_rate": 1.3762879701992642e-05,
+      "loss": 1.337,
+      "step": 487
+    },
+    {
+      "epoch": 0.3967076516614165,
+      "grad_norm": 3.048039674758911,
+      "learning_rate": 1.373846857542208e-05,
+      "loss": 1.3379,
+      "step": 488
+    },
+    {
+      "epoch": 0.39752057717711614,
+      "grad_norm": 2.6825406551361084,
+      "learning_rate": 1.3714031524234965e-05,
+      "loss": 1.3096,
+      "step": 489
+    },
+    {
+      "epoch": 0.39833350269281576,
+      "grad_norm": 2.5955066680908203,
+      "learning_rate": 1.3689568717891381e-05,
+      "loss": 1.2947,
+      "step": 490
+    },
+    {
+      "epoch": 0.39914642820851537,
+      "grad_norm": 2.5204849243164062,
+      "learning_rate": 1.3665080326029997e-05,
+      "loss": 1.2852,
+      "step": 491
+    },
+    {
+      "epoch": 0.39995935372421504,
+      "grad_norm": 3.158151865005493,
+      "learning_rate": 1.364056651846693e-05,
+      "loss": 1.3323,
+      "step": 492
+    },
+    {
+      "epoch": 0.40077227923991465,
+      "grad_norm": 2.787951946258545,
+      "learning_rate": 1.3616027465194525e-05,
+      "loss": 1.325,
+      "step": 493
+    },
+    {
+      "epoch": 0.40158520475561427,
+      "grad_norm": 3.462423324584961,
+      "learning_rate": 1.35914633363802e-05,
+      "loss": 1.2689,
+      "step": 494
+    },
+    {
+      "epoch": 0.4023981302713139,
+      "grad_norm": 3.3612263202667236,
+      "learning_rate": 1.356687430236526e-05,
+      "loss": 1.2846,
+      "step": 495
+    },
+    {
+      "epoch": 0.4032110557870135,
+      "grad_norm": 2.521135091781616,
+      "learning_rate": 1.3542260533663723e-05,
+      "loss": 1.2845,
+      "step": 496
+    },
+    {
+      "epoch": 0.40402398130271316,
+      "grad_norm": 2.702359914779663,
+      "learning_rate": 1.351762220096112e-05,
+      "loss": 1.2982,
+      "step": 497
+    },
+    {
+      "epoch": 0.4048369068184128,
+      "grad_norm": 2.928270101547241,
+      "learning_rate": 1.3492959475113332e-05,
+      "loss": 1.2878,
+      "step": 498
+    },
+    {
+      "epoch": 0.4056498323341124,
+      "grad_norm": 2.491701126098633,
+      "learning_rate": 1.3468272527145388e-05,
+      "loss": 1.2913,
+      "step": 499
+    },
+    {
+      "epoch": 0.406462757849812,
+      "grad_norm": 2.8777735233306885,
+      "learning_rate": 1.3443561528250295e-05,
+      "loss": 1.328,
+      "step": 500
+    },
+    {
+      "epoch": 0.4072756833655116,
+      "grad_norm": 3.4918212890625,
+      "learning_rate": 1.3418826649787834e-05,
+      "loss": 1.3415,
+      "step": 501
+    },
+    {
+      "epoch": 0.40808860888121123,
+      "grad_norm": 2.6940505504608154,
+      "learning_rate": 1.3394068063283387e-05,
+      "loss": 1.3017,
+      "step": 502
+    },
+    {
+      "epoch": 0.4089015343969109,
+      "grad_norm": 3.9722023010253906,
+      "learning_rate": 1.3369285940426737e-05,
+      "loss": 1.3161,
+      "step": 503
+    },
+    {
+      "epoch": 0.4097144599126105,
+      "grad_norm": 2.6105010509490967,
+      "learning_rate": 1.334448045307088e-05,
+      "loss": 1.2853,
+      "step": 504
+    },
+    {
+      "epoch": 0.41052738542831013,
+      "grad_norm": 4.094304084777832,
+      "learning_rate": 1.331965177323084e-05,
+      "loss": 1.3059,
+      "step": 505
+    },
+    {
+      "epoch": 0.41134031094400975,
+      "grad_norm": 2.5570600032806396,
+      "learning_rate": 1.3294800073082464e-05,
+      "loss": 1.2957,
+      "step": 506
+    },
+    {
+      "epoch": 0.41215323645970936,
+      "grad_norm": 2.60870099067688,
+      "learning_rate": 1.3269925524961237e-05,
+      "loss": 1.2887,
+      "step": 507
+    },
+    {
+      "epoch": 0.41296616197540903,
+      "grad_norm": 2.3958325386047363,
+      "learning_rate": 1.3245028301361086e-05,
+      "loss": 1.3207,
+      "step": 508
+    },
+    {
+      "epoch": 0.41377908749110864,
+      "grad_norm": 2.718470811843872,
+      "learning_rate": 1.3220108574933185e-05,
+      "loss": 1.2884,
+      "step": 509
+    },
+    {
+      "epoch": 0.41459201300680826,
+      "grad_norm": 2.9990408420562744,
+      "learning_rate": 1.3195166518484748e-05,
+      "loss": 1.3104,
+      "step": 510
+    },
+    {
+      "epoch": 0.41540493852250787,
+      "grad_norm": 3.256333589553833,
+      "learning_rate": 1.317020230497784e-05,
+      "loss": 1.2586,
+      "step": 511
+    },
+    {
+      "epoch": 0.4162178640382075,
+      "grad_norm": 3.0497708320617676,
+      "learning_rate": 1.3145216107528178e-05,
+      "loss": 1.2946,
+      "step": 512
+    },
+    {
+      "epoch": 0.4170307895539071,
+      "grad_norm": 2.6696412563323975,
+      "learning_rate": 1.3120208099403926e-05,
+      "loss": 1.3413,
+      "step": 513
+    },
+    {
+      "epoch": 0.41784371506960677,
+      "grad_norm": 2.592937469482422,
+      "learning_rate": 1.3095178454024496e-05,
+      "loss": 1.2827,
+      "step": 514
+    },
+    {
+      "epoch": 0.4186566405853064,
+      "grad_norm": 2.450669288635254,
+      "learning_rate": 1.3070127344959348e-05,
+      "loss": 1.2505,
+      "step": 515
+    },
+    {
+      "epoch": 0.419469566101006,
+      "grad_norm": 4.529777526855469,
+      "learning_rate": 1.3045054945926775e-05,
+      "loss": 1.3001,
+      "step": 516
+    },
+    {
+      "epoch": 0.4202824916167056,
+      "grad_norm": 3.2491648197174072,
+      "learning_rate": 1.3019961430792711e-05,
+      "loss": 1.2932,
+      "step": 517
+    },
+    {
+      "epoch": 0.4210954171324052,
+      "grad_norm": 3.3505818843841553,
+      "learning_rate": 1.2994846973569524e-05,
+      "loss": 1.3516,
+      "step": 518
+    },
+    {
+      "epoch": 0.4219083426481049,
+      "grad_norm": 3.5476715564727783,
+      "learning_rate": 1.2969711748414804e-05,
+      "loss": 1.2834,
+      "step": 519
+    },
+    {
+      "epoch": 0.4227212681638045,
+      "grad_norm": 2.738903522491455,
+      "learning_rate": 1.2944555929630152e-05,
+      "loss": 1.2978,
+      "step": 520
+    },
+    {
+      "epoch": 0.4235341936795041,
+      "grad_norm": 2.5854766368865967,
+      "learning_rate": 1.2919379691659979e-05,
+      "loss": 1.293,
+      "step": 521
+    },
+    {
+      "epoch": 0.42434711919520374,
+      "grad_norm": 3.76955246925354,
+      "learning_rate": 1.2894183209090304e-05,
+      "loss": 1.2517,
+      "step": 522
+    },
+    {
+      "epoch": 0.42516004471090335,
+      "grad_norm": 2.566361904144287,
+      "learning_rate": 1.2868966656647522e-05,
+      "loss": 1.3295,
+      "step": 523
+    },
+    {
+      "epoch": 0.42597297022660296,
+      "grad_norm": 2.7477164268493652,
+      "learning_rate": 1.2843730209197203e-05,
+      "loss": 1.3067,
+      "step": 524
+    },
+    {
+      "epoch": 0.42678589574230263,
+      "grad_norm": 3.0560967922210693,
+      "learning_rate": 1.2818474041742885e-05,
+      "loss": 1.2951,
+      "step": 525
+    },
+    {
+      "epoch": 0.42759882125800225,
+      "grad_norm": 2.9634625911712646,
+      "learning_rate": 1.2793198329424858e-05,
+      "loss": 1.268,
+      "step": 526
+    },
+    {
+      "epoch": 0.42841174677370186,
+      "grad_norm": 2.8108301162719727,
+      "learning_rate": 1.2767903247518945e-05,
+      "loss": 1.3319,
+      "step": 527
+    },
+    {
+      "epoch": 0.4292246722894015,
+      "grad_norm": 3.85799241065979,
+      "learning_rate": 1.2742588971435276e-05,
+      "loss": 1.3764,
+      "step": 528
+    },
+    {
+      "epoch": 0.4300375978051011,
+      "grad_norm": 2.564434766769409,
+      "learning_rate": 1.2717255676717106e-05,
+      "loss": 1.2854,
+      "step": 529
+    },
+    {
+      "epoch": 0.4308505233208007,
+      "grad_norm": 5.098544597625732,
+      "learning_rate": 1.2691903539039563e-05,
+      "loss": 1.3143,
+      "step": 530
+    },
+    {
+      "epoch": 0.4316634488365004,
+      "grad_norm": 7.195343017578125,
+      "learning_rate": 1.2666532734208437e-05,
+      "loss": 1.3026,
+      "step": 531
+    },
+    {
+      "epoch": 0.4324763743522,
+      "grad_norm": 2.743298053741455,
+      "learning_rate": 1.264114343815898e-05,
+      "loss": 1.3124,
+      "step": 532
+    },
+    {
+      "epoch": 0.4332892998678996,
+      "grad_norm": 3.183859348297119,
+      "learning_rate": 1.2615735826954664e-05,
+      "loss": 1.3132,
+      "step": 533
+    },
+    {
+      "epoch": 0.4341022253835992,
+      "grad_norm": 7.095142364501953,
+      "learning_rate": 1.2590310076785974e-05,
+      "loss": 1.2599,
+      "step": 534
+    },
+    {
+      "epoch": 0.43491515089929883,
+      "grad_norm": 2.91894268989563,
+      "learning_rate": 1.256486636396917e-05,
+      "loss": 1.3251,
+      "step": 535
+    },
+    {
+      "epoch": 0.4357280764149985,
+      "grad_norm": 2.931509494781494,
+      "learning_rate": 1.2539404864945087e-05,
+      "loss": 1.3347,
+      "step": 536
+    },
+    {
+      "epoch": 0.4365410019306981,
+      "grad_norm": 2.4552268981933594,
+      "learning_rate": 1.2513925756277894e-05,
+      "loss": 1.3469,
+      "step": 537
+    },
+    {
+      "epoch": 0.43735392744639773,
+      "grad_norm": 2.846196174621582,
+      "learning_rate": 1.2488429214653871e-05,
+      "loss": 1.2654,
+      "step": 538
+    },
+    {
+      "epoch": 0.43816685296209734,
+      "grad_norm": 3.494403600692749,
+      "learning_rate": 1.24629154168802e-05,
+      "loss": 1.2688,
+      "step": 539
+    },
+    {
+      "epoch": 0.43897977847779696,
+      "grad_norm": 3.00067138671875,
+      "learning_rate": 1.2437384539883715e-05,
+      "loss": 1.2865,
+      "step": 540
+    },
+    {
+      "epoch": 0.43979270399349657,
+      "grad_norm": 3.0412096977233887,
+      "learning_rate": 1.2411836760709686e-05,
+      "loss": 1.269,
+      "step": 541
+    },
+    {
+      "epoch": 0.44060562950919624,
+      "grad_norm": 2.3580715656280518,
+      "learning_rate": 1.2386272256520606e-05,
+      "loss": 1.2752,
+      "step": 542
+    },
+    {
+      "epoch": 0.44141855502489585,
+      "grad_norm": 9.030720710754395,
+      "learning_rate": 1.2360691204594937e-05,
+      "loss": 1.3074,
+      "step": 543
+    },
+    {
+      "epoch": 0.44223148054059547,
+      "grad_norm": 3.970172882080078,
+      "learning_rate": 1.2335093782325889e-05,
+      "loss": 1.3117,
+      "step": 544
+    },
+    {
+      "epoch": 0.4430444060562951,
+      "grad_norm": 2.8179943561553955,
+      "learning_rate": 1.2309480167220203e-05,
+      "loss": 1.3196,
+      "step": 545
+    },
+    {
+      "epoch": 0.4438573315719947,
+      "grad_norm": 2.9376232624053955,
+      "learning_rate": 1.2283850536896907e-05,
+      "loss": 1.2614,
+      "step": 546
+    },
+    {
+      "epoch": 0.44467025708769436,
+      "grad_norm": 2.811709403991699,
+      "learning_rate": 1.2258205069086082e-05,
+      "loss": 1.2666,
+      "step": 547
+    },
+    {
+      "epoch": 0.445483182603394,
+      "grad_norm": 3.060638427734375,
+      "learning_rate": 1.2232543941627641e-05,
+      "loss": 1.2891,
+      "step": 548
+    },
+    {
+      "epoch": 0.4462961081190936,
+      "grad_norm": 2.581530809402466,
+      "learning_rate": 1.2206867332470091e-05,
+      "loss": 1.2875,
+      "step": 549
+    },
+    {
+      "epoch": 0.4471090336347932,
+      "grad_norm": 2.588129997253418,
+      "learning_rate": 1.2181175419669293e-05,
+      "loss": 1.2964,
+      "step": 550
+    },
+    {
+      "epoch": 0.4479219591504928,
+      "grad_norm": 3.0943429470062256,
+      "learning_rate": 1.215546838138723e-05,
+      "loss": 1.29,
+      "step": 551
+    },
+    {
+      "epoch": 0.44873488466619244,
+      "grad_norm": 2.960190534591675,
+      "learning_rate": 1.212974639589078e-05,
+      "loss": 1.2812,
+      "step": 552
+    },
+    {
+      "epoch": 0.4495478101818921,
+      "grad_norm": 2.7364282608032227,
+      "learning_rate": 1.2104009641550472e-05,
+      "loss": 1.2783,
+      "step": 553
+    },
+    {
+      "epoch": 0.4503607356975917,
+      "grad_norm": 2.509277105331421,
+      "learning_rate": 1.2078258296839245e-05,
+      "loss": 1.2859,
+      "step": 554
+    },
+    {
+      "epoch": 0.45117366121329133,
+      "grad_norm": 2.769371747970581,
+      "learning_rate": 1.2052492540331218e-05,
+      "loss": 1.2866,
+      "step": 555
+    },
+    {
+      "epoch": 0.45198658672899095,
+      "grad_norm": 3.057968854904175,
+      "learning_rate": 1.2026712550700457e-05,
+      "loss": 1.3051,
+      "step": 556
+    },
+    {
+      "epoch": 0.45279951224469056,
+      "grad_norm": 3.4182374477386475,
+      "learning_rate": 1.200091850671972e-05,
+      "loss": 1.3266,
+      "step": 557
+    },
+    {
+      "epoch": 0.45361243776039023,
+      "grad_norm": 2.6871426105499268,
+      "learning_rate": 1.1975110587259222e-05,
+      "loss": 1.2596,
+      "step": 558
+    },
+    {
+      "epoch": 0.45442536327608984,
+      "grad_norm": 3.463675022125244,
+      "learning_rate": 1.1949288971285411e-05,
+      "loss": 1.2767,
+      "step": 559
+    },
+    {
+      "epoch": 0.45523828879178946,
+      "grad_norm": 2.8260090351104736,
+      "learning_rate": 1.1923453837859706e-05,
+      "loss": 1.2734,
+      "step": 560
+    },
+    {
+      "epoch": 0.4560512143074891,
+      "grad_norm": 2.6161341667175293,
+      "learning_rate": 1.1897605366137264e-05,
+      "loss": 1.2377,
+      "step": 561
+    },
+    {
+      "epoch": 0.4568641398231887,
+      "grad_norm": 2.847534418106079,
+      "learning_rate": 1.1871743735365735e-05,
+      "loss": 1.3128,
+      "step": 562
+    },
+    {
+      "epoch": 0.4576770653388883,
+      "grad_norm": 3.116063117980957,
+      "learning_rate": 1.1845869124884027e-05,
+      "loss": 1.3114,
+      "step": 563
+    },
+    {
+      "epoch": 0.45848999085458797,
+      "grad_norm": 3.2849061489105225,
+      "learning_rate": 1.1819981714121054e-05,
+      "loss": 1.2761,
+      "step": 564
+    },
+    {
+      "epoch": 0.4593029163702876,
+      "grad_norm": 2.484531879425049,
+      "learning_rate": 1.1794081682594491e-05,
+      "loss": 1.2978,
+      "step": 565
+    },
+    {
+      "epoch": 0.4601158418859872,
+      "grad_norm": 3.111940383911133,
+      "learning_rate": 1.176816920990954e-05,
+      "loss": 1.2928,
+      "step": 566
+    },
+    {
+      "epoch": 0.4609287674016868,
+      "grad_norm": 3.063422918319702,
+      "learning_rate": 1.174224447575767e-05,
+      "loss": 1.3137,
+      "step": 567
+    },
+    {
+      "epoch": 0.4617416929173864,
+      "grad_norm": 4.031757831573486,
+      "learning_rate": 1.171630765991538e-05,
+      "loss": 1.2986,
+      "step": 568
+    },
+    {
+      "epoch": 0.4625546184330861,
+      "grad_norm": 2.650336980819702,
+      "learning_rate": 1.169035894224295e-05,
+      "loss": 1.3328,
+      "step": 569
+    },
+    {
+      "epoch": 0.4633675439487857,
+      "grad_norm": 2.574526309967041,
+      "learning_rate": 1.1664398502683194e-05,
+      "loss": 1.3078,
+      "step": 570
+    },
+    {
+      "epoch": 0.4641804694644853,
+      "grad_norm": 2.3674449920654297,
+      "learning_rate": 1.1638426521260211e-05,
+      "loss": 1.2819,
+      "step": 571
+    },
+    {
+      "epoch": 0.46499339498018494,
+      "grad_norm": 2.8870980739593506,
+      "learning_rate": 1.1612443178078138e-05,
+      "loss": 1.2661,
+      "step": 572
+    },
+    {
+      "epoch": 0.46580632049588455,
+      "grad_norm": 2.4961047172546387,
+      "learning_rate": 1.1586448653319908e-05,
+      "loss": 1.3042,
+      "step": 573
+    },
+    {
+      "epoch": 0.46661924601158417,
+      "grad_norm": 2.6196508407592773,
+      "learning_rate": 1.156044312724598e-05,
+      "loss": 1.2306,
+      "step": 574
+    },
+    {
+      "epoch": 0.46743217152728384,
+      "grad_norm": 2.7249913215637207,
+      "learning_rate": 1.153442678019311e-05,
+      "loss": 1.3095,
+      "step": 575
+    },
+    {
+      "epoch": 0.46824509704298345,
+      "grad_norm": 2.9108643531799316,
+      "learning_rate": 1.1508399792573095e-05,
+      "loss": 1.2513,
+      "step": 576
+    },
+    {
+      "epoch": 0.46905802255868306,
+      "grad_norm": 2.7690494060516357,
+      "learning_rate": 1.1482362344871514e-05,
+      "loss": 1.3445,
+      "step": 577
+    },
+    {
+      "epoch": 0.4698709480743827,
+      "grad_norm": 3.629122734069824,
+      "learning_rate": 1.1456314617646482e-05,
+      "loss": 1.2616,
+      "step": 578
+    },
+    {
+      "epoch": 0.4706838735900823,
+      "grad_norm": 2.6831417083740234,
+      "learning_rate": 1.1430256791527406e-05,
+      "loss": 1.2786,
+      "step": 579
+    },
+    {
+      "epoch": 0.4714967991057819,
+      "grad_norm": 2.5316171646118164,
+      "learning_rate": 1.1404189047213716e-05,
+      "loss": 1.3195,
+      "step": 580
+    },
+    {
+      "epoch": 0.4723097246214816,
+      "grad_norm": 4.602120399475098,
+      "learning_rate": 1.137811156547362e-05,
+      "loss": 1.2378,
+      "step": 581
+    },
+    {
+      "epoch": 0.4731226501371812,
+      "grad_norm": 2.5073766708374023,
+      "learning_rate": 1.1352024527142855e-05,
+      "loss": 1.2426,
+      "step": 582
+    },
+    {
+      "epoch": 0.4739355756528808,
+      "grad_norm": 2.5561444759368896,
+      "learning_rate": 1.1325928113123431e-05,
+      "loss": 1.318,
+      "step": 583
+    },
+    {
+      "epoch": 0.4747485011685804,
+      "grad_norm": 2.8386447429656982,
+      "learning_rate": 1.129982250438237e-05,
+      "loss": 1.2529,
+      "step": 584
+    },
+    {
+      "epoch": 0.47556142668428003,
+      "grad_norm": 2.3654778003692627,
+      "learning_rate": 1.1273707881950445e-05,
+      "loss": 1.2822,
+      "step": 585
+    },
+    {
+      "epoch": 0.4763743521999797,
+      "grad_norm": 3.125446081161499,
+      "learning_rate": 1.1247584426920962e-05,
+      "loss": 1.3588,
+      "step": 586
+    },
+    {
+      "epoch": 0.4771872777156793,
+      "grad_norm": 3.600827217102051,
+      "learning_rate": 1.1221452320448449e-05,
+      "loss": 1.3023,
+      "step": 587
+    },
+    {
+      "epoch": 0.47800020323137893,
+      "grad_norm": 3.858783483505249,
+      "learning_rate": 1.1195311743747445e-05,
+      "loss": 1.2784,
+      "step": 588
+    },
+    {
+      "epoch": 0.47881312874707854,
+      "grad_norm": 2.841679334640503,
+      "learning_rate": 1.116916287809122e-05,
+      "loss": 1.3084,
+      "step": 589
+    },
+    {
+      "epoch": 0.47962605426277816,
+      "grad_norm": 2.9722323417663574,
+      "learning_rate": 1.1143005904810527e-05,
+      "loss": 1.2983,
+      "step": 590
+    },
+    {
+      "epoch": 0.48043897977847777,
+      "grad_norm": 2.560037136077881,
+      "learning_rate": 1.1116841005292339e-05,
+      "loss": 1.3175,
+      "step": 591
+    },
+    {
+      "epoch": 0.48125190529417744,
+      "grad_norm": 3.1770455837249756,
+      "learning_rate": 1.1090668360978589e-05,
+      "loss": 1.2603,
+      "step": 592
+    },
+    {
+      "epoch": 0.48206483080987705,
+      "grad_norm": 2.4485607147216797,
+      "learning_rate": 1.106448815336493e-05,
+      "loss": 1.2792,
+      "step": 593
+    },
+    {
+      "epoch": 0.48287775632557667,
+      "grad_norm": 3.7001748085021973,
+      "learning_rate": 1.1038300563999455e-05,
+      "loss": 1.2846,
+      "step": 594
+    },
+    {
+      "epoch": 0.4836906818412763,
+      "grad_norm": 2.6942710876464844,
+      "learning_rate": 1.1012105774481446e-05,
+      "loss": 1.2864,
+      "step": 595
+    },
+    {
+      "epoch": 0.4845036073569759,
+      "grad_norm": 2.5104377269744873,
+      "learning_rate": 1.0985903966460115e-05,
+      "loss": 1.256,
+      "step": 596
+    },
+    {
+      "epoch": 0.48531653287267557,
+      "grad_norm": 2.4864704608917236,
+      "learning_rate": 1.0959695321633346e-05,
+      "loss": 1.2838,
+      "step": 597
+    },
+    {
+      "epoch": 0.4861294583883752,
+      "grad_norm": 3.2645606994628906,
+      "learning_rate": 1.0933480021746432e-05,
+      "loss": 1.2966,
+      "step": 598
+    },
+    {
+      "epoch": 0.4869423839040748,
+      "grad_norm": 28.041383743286133,
+      "learning_rate": 1.0907258248590816e-05,
+      "loss": 1.2513,
+      "step": 599
+    },
+    {
+      "epoch": 0.4877553094197744,
+      "grad_norm": 2.736785888671875,
+      "learning_rate": 1.0881030184002827e-05,
+      "loss": 1.3217,
+      "step": 600
+    },
+    {
+      "epoch": 0.488568234935474,
+      "grad_norm": 4.294330596923828,
+      "learning_rate": 1.0854796009862434e-05,
+      "loss": 1.3007,
+      "step": 601
+    },
+    {
+      "epoch": 0.48938116045117364,
+      "grad_norm": 2.629371404647827,
+      "learning_rate": 1.0828555908091958e-05,
+      "loss": 1.2884,
+      "step": 602
+    },
+    {
+      "epoch": 0.4901940859668733,
+      "grad_norm": 3.166304588317871,
+      "learning_rate": 1.0802310060654832e-05,
+      "loss": 1.3127,
+      "step": 603
+    },
+    {
+      "epoch": 0.4910070114825729,
+      "grad_norm": 2.5344200134277344,
+      "learning_rate": 1.0776058649554336e-05,
+      "loss": 1.249,
+      "step": 604
+    },
+    {
+      "epoch": 0.49181993699827253,
+      "grad_norm": 3.2902913093566895,
+      "learning_rate": 1.0749801856832325e-05,
+      "loss": 1.2341,
+      "step": 605
+    },
+    {
+      "epoch": 0.49263286251397215,
+      "grad_norm": 2.5863964557647705,
+      "learning_rate": 1.0723539864567983e-05,
+      "loss": 1.3534,
+      "step": 606
+    },
+    {
+      "epoch": 0.49344578802967176,
+      "grad_norm": 3.1407294273376465,
+      "learning_rate": 1.0697272854876537e-05,
+      "loss": 1.2452,
+      "step": 607
+    },
+    {
+      "epoch": 0.49425871354537143,
+      "grad_norm": 2.339702844619751,
+      "learning_rate": 1.0671001009908015e-05,
+      "loss": 1.2597,
+      "step": 608
+    },
+    {
+      "epoch": 0.49507163906107104,
+      "grad_norm": 2.5861027240753174,
+      "learning_rate": 1.0644724511845976e-05,
+      "loss": 1.304,
+      "step": 609
+    },
+    {
+      "epoch": 0.49588456457677066,
+      "grad_norm": 2.6124143600463867,
+      "learning_rate": 1.0618443542906251e-05,
+      "loss": 1.2333,
+      "step": 610
+    },
+    {
+      "epoch": 0.4966974900924703,
+      "grad_norm": 2.53468918800354,
+      "learning_rate": 1.059215828533566e-05,
+      "loss": 1.2587,
+      "step": 611
+    },
+    {
+      "epoch": 0.4975104156081699,
+      "grad_norm": 5.205654621124268,
+      "learning_rate": 1.0565868921410776e-05,
+      "loss": 1.2758,
+      "step": 612
+    },
+    {
+      "epoch": 0.4983233411238695,
+      "grad_norm": 3.3307433128356934,
+      "learning_rate": 1.0539575633436645e-05,
+      "loss": 1.3197,
+      "step": 613
+    },
+    {
+      "epoch": 0.49913626663956917,
+      "grad_norm": 2.4654664993286133,
+      "learning_rate": 1.0513278603745523e-05,
+      "loss": 1.2733,
+      "step": 614
+    },
+    {
+      "epoch": 0.4999491921552688,
+      "grad_norm": 2.5150272846221924,
+      "learning_rate": 1.0486978014695606e-05,
+      "loss": 1.2841,
+      "step": 615
+    },
+    {
+      "epoch": 0.5007621176709683,
+      "grad_norm": 2.660186767578125,
+      "learning_rate": 1.0460674048669783e-05,
+      "loss": 1.3007,
+      "step": 616
+    },
+    {
+      "epoch": 0.501575043186668,
+      "grad_norm": 2.7415716648101807,
+      "learning_rate": 1.0434366888074363e-05,
+      "loss": 1.2974,
+      "step": 617
+    },
+    {
+      "epoch": 0.5023879687023677,
+      "grad_norm": 2.479142427444458,
+      "learning_rate": 1.0408056715337797e-05,
+      "loss": 1.301,
+      "step": 618
+    },
+    {
+      "epoch": 0.5032008942180672,
+      "grad_norm": 2.4590210914611816,
+      "learning_rate": 1.0381743712909424e-05,
+      "loss": 1.2253,
+      "step": 619
+    },
+    {
+      "epoch": 0.5040138197337669,
+      "grad_norm": 2.4704954624176025,
+      "learning_rate": 1.0355428063258224e-05,
+      "loss": 1.1927,
+      "step": 620
+    },
+    {
+      "epoch": 0.5048267452494665,
+      "grad_norm": 3.5037641525268555,
+      "learning_rate": 1.0329109948871512e-05,
+      "loss": 1.2727,
+      "step": 621
+    },
+    {
+      "epoch": 0.5056396707651661,
+      "grad_norm": 2.6537327766418457,
+      "learning_rate": 1.0302789552253702e-05,
+      "loss": 1.2295,
+      "step": 622
+    },
+    {
+      "epoch": 0.5064525962808658,
+      "grad_norm": 3.4443886280059814,
+      "learning_rate": 1.0276467055925044e-05,
+      "loss": 1.2403,
+      "step": 623
+    },
+    {
+      "epoch": 0.5072655217965654,
+      "grad_norm": 4.377493858337402,
+      "learning_rate": 1.0250142642420335e-05,
+      "loss": 1.2667,
+      "step": 624
+    },
+    {
+      "epoch": 0.508078447312265,
+      "grad_norm": 2.712472677230835,
+      "learning_rate": 1.0223816494287675e-05,
+      "loss": 1.3323,
+      "step": 625
+    },
+    {
+      "epoch": 0.5088913728279646,
+      "grad_norm": 2.922093152999878,
+      "learning_rate": 1.0197488794087188e-05,
+      "loss": 1.2713,
+      "step": 626
+    },
+    {
+      "epoch": 0.5097042983436643,
+      "grad_norm": 11.951809883117676,
+      "learning_rate": 1.0171159724389766e-05,
+      "loss": 1.2997,
+      "step": 627
+    },
+    {
+      "epoch": 0.5105172238593639,
+      "grad_norm": 2.5700554847717285,
+      "learning_rate": 1.0144829467775794e-05,
+      "loss": 1.261,
+      "step": 628
+    },
+    {
+      "epoch": 0.5113301493750635,
+      "grad_norm": 2.6800413131713867,
+      "learning_rate": 1.0118498206833886e-05,
+      "loss": 1.3292,
+      "step": 629
+    },
+    {
+      "epoch": 0.5121430748907632,
+      "grad_norm": 4.24453592300415,
+      "learning_rate": 1.0092166124159628e-05,
+      "loss": 1.3281,
+      "step": 630
+    },
+    {
+      "epoch": 0.5129560004064627,
+      "grad_norm": 2.7513749599456787,
+      "learning_rate": 1.0065833402354302e-05,
+      "loss": 1.2944,
+      "step": 631
+    },
+    {
+      "epoch": 0.5137689259221624,
+      "grad_norm": 2.610588788986206,
+      "learning_rate": 1.003950022402361e-05,
+      "loss": 1.3129,
+      "step": 632
+    },
+    {
+      "epoch": 0.5145818514378621,
+      "grad_norm": 2.949564218521118,
+      "learning_rate": 1.0013166771776441e-05,
+      "loss": 1.2961,
+      "step": 633
+    },
+    {
+      "epoch": 0.5153947769535616,
+      "grad_norm": 2.5617198944091797,
+      "learning_rate": 9.986833228223562e-06,
+      "loss": 1.2898,
+      "step": 634
+    },
+    {
+      "epoch": 0.5162077024692613,
+      "grad_norm": 2.779733896255493,
+      "learning_rate": 9.96049977597639e-06,
+      "loss": 1.2988,
+      "step": 635
+    },
+    {
+      "epoch": 0.5170206279849608,
+      "grad_norm": 2.8505136966705322,
+      "learning_rate": 9.934166597645703e-06,
+      "loss": 1.2652,
+      "step": 636
+    },
+    {
+      "epoch": 0.5178335535006605,
+      "grad_norm": 2.847262144088745,
+      "learning_rate": 9.907833875840374e-06,
+      "loss": 1.3076,
+      "step": 637
+    },
+    {
+      "epoch": 0.5186464790163601,
+      "grad_norm": 4.957255840301514,
+      "learning_rate": 9.881501793166117e-06,
+      "loss": 1.214,
+      "step": 638
+    },
+    {
+      "epoch": 0.5194594045320597,
+      "grad_norm": 2.7829556465148926,
+      "learning_rate": 9.85517053222421e-06,
+      "loss": 1.2379,
+      "step": 639
+    },
+    {
+      "epoch": 0.5202723300477594,
+      "grad_norm": 2.7060935497283936,
+      "learning_rate": 9.82884027561024e-06,
+      "loss": 1.3016,
+      "step": 640
+    },
+    {
+      "epoch": 0.521085255563459,
+      "grad_norm": 6.336554527282715,
+      "learning_rate": 9.802511205912815e-06,
+      "loss": 1.269,
+      "step": 641
+    },
+    {
+      "epoch": 0.5218981810791586,
+      "grad_norm": 3.0378448963165283,
+      "learning_rate": 9.776183505712327e-06,
+      "loss": 1.317,
+      "step": 642
+    },
+    {
+      "epoch": 0.5227111065948582,
+      "grad_norm": 5.806065082550049,
+      "learning_rate": 9.749857357579667e-06,
+      "loss": 1.3165,
+      "step": 643
+    },
+    {
+      "epoch": 0.5235240321105579,
+      "grad_norm": 2.7738869190216064,
+      "learning_rate": 9.723532944074961e-06,
+      "loss": 1.2835,
+      "step": 644
+    },
+    {
+      "epoch": 0.5243369576262575,
+      "grad_norm": 2.6603453159332275,
+      "learning_rate": 9.6972104477463e-06,
+      "loss": 1.2673,
+      "step": 645
+    },
+    {
+      "epoch": 0.5251498831419571,
+      "grad_norm": 2.9316189289093018,
+      "learning_rate": 9.670890051128493e-06,
+      "loss": 1.249,
+      "step": 646
+    },
+    {
+      "epoch": 0.5259628086576568,
+      "grad_norm": 2.8541407585144043,
+      "learning_rate": 9.644571936741778e-06,
+      "loss": 1.2835,
+      "step": 647
+    },
+    {
+      "epoch": 0.5267757341733563,
+      "grad_norm": 2.6935575008392334,
+      "learning_rate": 9.618256287090576e-06,
+      "loss": 1.2859,
+      "step": 648
+    },
+    {
+      "epoch": 0.527588659689056,
+      "grad_norm": 3.057039260864258,
+      "learning_rate": 9.591943284662206e-06,
+      "loss": 1.2538,
+      "step": 649
+    },
+    {
+      "epoch": 0.5284015852047557,
+      "grad_norm": 3.2430379390716553,
+      "learning_rate": 9.56563311192564e-06,
+      "loss": 1.294,
+      "step": 650
+    },
+    {
+      "epoch": 0.5292145107204552,
+      "grad_norm": 2.378072500228882,
+      "learning_rate": 9.53932595133022e-06,
+      "loss": 1.2793,
+      "step": 651
+    },
+    {
+      "epoch": 0.5300274362361549,
+      "grad_norm": 3.2185440063476562,
+      "learning_rate": 9.513021985304399e-06,
+      "loss": 1.2868,
+      "step": 652
+    },
+    {
+      "epoch": 0.5308403617518545,
+      "grad_norm": 3.272632122039795,
+      "learning_rate": 9.486721396254484e-06,
+      "loss": 1.2128,
+      "step": 653
+    },
+    {
+      "epoch": 0.5316532872675541,
+      "grad_norm": 3.163884401321411,
+      "learning_rate": 9.460424366563355e-06,
+      "loss": 1.2962,
+      "step": 654
+    },
+    {
+      "epoch": 0.5324662127832538,
+      "grad_norm": 3.096857786178589,
+      "learning_rate": 9.434131078589224e-06,
+      "loss": 1.2575,
+      "step": 655
+    },
+    {
+      "epoch": 0.5332791382989533,
+      "grad_norm": 2.711069107055664,
+      "learning_rate": 9.407841714664343e-06,
+      "loss": 1.2969,
+      "step": 656
+    },
+    {
+      "epoch": 0.534092063814653,
+      "grad_norm": 4.4655866622924805,
+      "learning_rate": 9.381556457093752e-06,
+      "loss": 1.2229,
+      "step": 657
+    },
+    {
+      "epoch": 0.5349049893303526,
+      "grad_norm": 2.7365305423736572,
+      "learning_rate": 9.355275488154025e-06,
+      "loss": 1.285,
+      "step": 658
+    },
+    {
+      "epoch": 0.5357179148460522,
+      "grad_norm": 3.4264895915985107,
+      "learning_rate": 9.32899899009199e-06,
+      "loss": 1.3222,
+      "step": 659
+    },
+    {
+      "epoch": 0.5365308403617518,
+      "grad_norm": 2.9572296142578125,
+      "learning_rate": 9.30272714512347e-06,
+      "loss": 1.2771,
+      "step": 660
+    },
+    {
+      "epoch": 0.5373437658774515,
+      "grad_norm": 3.124464988708496,
+      "learning_rate": 9.276460135432019e-06,
+      "loss": 1.2362,
+      "step": 661
+    },
+    {
+      "epoch": 0.5381566913931511,
+      "grad_norm": 3.484861373901367,
+      "learning_rate": 9.250198143167675e-06,
+      "loss": 1.2624,
+      "step": 662
+    },
+    {
+      "epoch": 0.5389696169088507,
+      "grad_norm": 3.191455602645874,
+      "learning_rate": 9.223941350445666e-06,
+      "loss": 1.3271,
+      "step": 663
+    },
+    {
+      "epoch": 0.5397825424245504,
+      "grad_norm": 3.055478572845459,
+      "learning_rate": 9.19768993934517e-06,
+      "loss": 1.2476,
+      "step": 664
+    },
+    {
+      "epoch": 0.5405954679402499,
+      "grad_norm": 2.8661985397338867,
+      "learning_rate": 9.171444091908046e-06,
+      "loss": 1.2575,
+      "step": 665
+    },
+    {
+      "epoch": 0.5414083934559496,
+      "grad_norm": 3.042300224304199,
+      "learning_rate": 9.145203990137571e-06,
+      "loss": 1.2472,
+      "step": 666
+    },
+    {
+      "epoch": 0.5422213189716493,
+      "grad_norm": 3.324767827987671,
+      "learning_rate": 9.118969815997174e-06,
+      "loss": 1.2608,
+      "step": 667
+    },
+    {
+      "epoch": 0.5430342444873488,
+      "grad_norm": 2.8374948501586914,
+      "learning_rate": 9.092741751409186e-06,
+      "loss": 1.2865,
+      "step": 668
+    },
+    {
+      "epoch": 0.5438471700030485,
+      "grad_norm": 3.3593552112579346,
+      "learning_rate": 9.06651997825357e-06,
+      "loss": 1.2746,
+      "step": 669
+    },
+    {
+      "epoch": 0.544660095518748,
+      "grad_norm": 3.2432382106781006,
+      "learning_rate": 9.040304678366658e-06,
+      "loss": 1.2864,
+      "step": 670
+    },
+    {
+      "epoch": 0.5454730210344477,
+      "grad_norm": 2.890409469604492,
+      "learning_rate": 9.014096033539889e-06,
+      "loss": 1.2685,
+      "step": 671
+    },
+    {
+      "epoch": 0.5462859465501474,
+      "grad_norm": 3.0769150257110596,
+      "learning_rate": 8.987894225518556e-06,
+      "loss": 1.2701,
+      "step": 672
+    },
+    {
+      "epoch": 0.547098872065847,
+      "grad_norm": 3.453287363052368,
+      "learning_rate": 8.961699436000548e-06,
+      "loss": 1.2218,
+      "step": 673
+    },
+    {
+      "epoch": 0.5479117975815466,
+      "grad_norm": 3.1950011253356934,
+      "learning_rate": 8.93551184663507e-06,
+      "loss": 1.2267,
+      "step": 674
+    },
+    {
+      "epoch": 0.5487247230972462,
+      "grad_norm": 3.445006847381592,
+      "learning_rate": 8.909331639021414e-06,
+      "loss": 1.283,
+      "step": 675
+    },
+    {
+      "epoch": 0.5495376486129459,
+      "grad_norm": 2.5453741550445557,
+      "learning_rate": 8.883158994707666e-06,
+      "loss": 1.3102,
+      "step": 676
+    },
+    {
+      "epoch": 0.5503505741286454,
+      "grad_norm": 4.167499542236328,
+      "learning_rate": 8.856994095189477e-06,
+      "loss": 1.2881,
+      "step": 677
+    },
+    {
+      "epoch": 0.5511634996443451,
+      "grad_norm": 2.6888363361358643,
+      "learning_rate": 8.830837121908783e-06,
+      "loss": 1.2332,
+      "step": 678
+    },
+    {
+      "epoch": 0.5519764251600447,
+      "grad_norm": 2.9484667778015137,
+      "learning_rate": 8.804688256252557e-06,
+      "loss": 1.2676,
+      "step": 679
+    },
+    {
+      "epoch": 0.5527893506757443,
+      "grad_norm": 2.5477519035339355,
+      "learning_rate": 8.778547679551555e-06,
+      "loss": 1.2956,
+      "step": 680
+    },
+    {
+      "epoch": 0.553602276191444,
+      "grad_norm": 2.3307385444641113,
+      "learning_rate": 8.75241557307904e-06,
+      "loss": 1.3021,
+      "step": 681
+    },
+    {
+      "epoch": 0.5544152017071435,
+      "grad_norm": 3.1104202270507812,
+      "learning_rate": 8.726292118049555e-06,
+      "loss": 1.2861,
+      "step": 682
+    },
+    {
+      "epoch": 0.5552281272228432,
+      "grad_norm": 3.2731287479400635,
+      "learning_rate": 8.700177495617635e-06,
+      "loss": 1.33,
+      "step": 683
+    },
+    {
+      "epoch": 0.5560410527385429,
+      "grad_norm": 2.923478364944458,
+      "learning_rate": 8.674071886876572e-06,
+      "loss": 1.2946,
+      "step": 684
+    },
+    {
+      "epoch": 0.5568539782542424,
+      "grad_norm": 3.1030538082122803,
+      "learning_rate": 8.647975472857148e-06,
+      "loss": 1.2481,
+      "step": 685
+    },
+    {
+      "epoch": 0.5576669037699421,
+      "grad_norm": 2.6904759407043457,
+      "learning_rate": 8.621888434526382e-06,
+      "loss": 1.2637,
+      "step": 686
+    },
+    {
+      "epoch": 0.5584798292856417,
+      "grad_norm": 3.6781442165374756,
+      "learning_rate": 8.595810952786289e-06,
+      "loss": 1.2875,
+      "step": 687
+    },
+    {
+      "epoch": 0.5592927548013413,
+      "grad_norm": 4.897818565368652,
+      "learning_rate": 8.569743208472594e-06,
+      "loss": 1.2804,
+      "step": 688
+    },
+    {
+      "epoch": 0.560105680317041,
+      "grad_norm": 2.9090828895568848,
+      "learning_rate": 8.543685382353518e-06,
+      "loss": 1.2817,
+      "step": 689
+    },
+    {
+      "epoch": 0.5609186058327406,
+      "grad_norm": 3.3284378051757812,
+      "learning_rate": 8.51763765512849e-06,
+      "loss": 1.2928,
+      "step": 690
+    },
+    {
+      "epoch": 0.5617315313484402,
+      "grad_norm": 3.440209150314331,
+      "learning_rate": 8.491600207426907e-06,
+      "loss": 1.2667,
+      "step": 691
+    },
+    {
+      "epoch": 0.5625444568641398,
+      "grad_norm": 3.1297762393951416,
+      "learning_rate": 8.465573219806893e-06,
+      "loss": 1.2752,
+      "step": 692
+    },
+    {
+      "epoch": 0.5633573823798395,
+      "grad_norm": 3.460277795791626,
+      "learning_rate": 8.439556872754025e-06,
+      "loss": 1.2611,
+      "step": 693
+    },
+    {
+      "epoch": 0.5641703078955391,
+      "grad_norm": 2.6390557289123535,
+      "learning_rate": 8.413551346680095e-06,
+      "loss": 1.2339,
+      "step": 694
+    },
+    {
+      "epoch": 0.5649832334112387,
+      "grad_norm": 2.365945339202881,
+      "learning_rate": 8.38755682192186e-06,
+      "loss": 1.2333,
+      "step": 695
+    },
+    {
+      "epoch": 0.5657961589269384,
+      "grad_norm": 3.140129804611206,
+      "learning_rate": 8.36157347873979e-06,
+      "loss": 1.2614,
+      "step": 696
+    },
+    {
+      "epoch": 0.5666090844426379,
+      "grad_norm": 4.027166366577148,
+      "learning_rate": 8.335601497316809e-06,
+      "loss": 1.263,
+      "step": 697
+    },
+    {
+      "epoch": 0.5674220099583376,
+      "grad_norm": 2.6872942447662354,
+      "learning_rate": 8.309641057757052e-06,
+      "loss": 1.2479,
+      "step": 698
+    },
+    {
+      "epoch": 0.5682349354740371,
+      "grad_norm": 2.575493574142456,
+      "learning_rate": 8.283692340084623e-06,
+      "loss": 1.2818,
+      "step": 699
+    },
+    {
+      "epoch": 0.5690478609897368,
+      "grad_norm": 2.6429176330566406,
+      "learning_rate": 8.257755524242333e-06,
+      "loss": 1.2921,
+      "step": 700
+    },
+    {
+      "epoch": 0.5698607865054365,
+      "grad_norm": 4.695654392242432,
+      "learning_rate": 8.231830790090461e-06,
+      "loss": 1.2046,
+      "step": 701
+    },
+    {
+      "epoch": 0.570673712021136,
+      "grad_norm": 2.4642715454101562,
+      "learning_rate": 8.205918317405508e-06,
+      "loss": 1.3013,
+      "step": 702
+    },
+    {
+      "epoch": 0.5714866375368357,
+      "grad_norm": 2.567474842071533,
+      "learning_rate": 8.18001828587895e-06,
+      "loss": 1.3458,
+      "step": 703
+    },
+    {
+      "epoch": 0.5722995630525353,
+      "grad_norm": 2.934668779373169,
+      "learning_rate": 8.154130875115978e-06,
+      "loss": 1.2804,
+      "step": 704
+    },
+    {
+      "epoch": 0.5731124885682349,
+      "grad_norm": 2.669285297393799,
+      "learning_rate": 8.12825626463427e-06,
+      "loss": 1.2329,
+      "step": 705
+    },
+    {
+      "epoch": 0.5739254140839346,
+      "grad_norm": 2.7390220165252686,
+      "learning_rate": 8.102394633862743e-06,
+      "loss": 1.2177,
+      "step": 706
+    },
+    {
+      "epoch": 0.5747383395996342,
+      "grad_norm": 3.19964861869812,
+      "learning_rate": 8.0765461621403e-06,
+      "loss": 1.2625,
+      "step": 707
+    },
+    {
+      "epoch": 0.5755512651153338,
+      "grad_norm": 2.753469705581665,
+      "learning_rate": 8.050711028714589e-06,
+      "loss": 1.2357,
+      "step": 708
+    },
+    {
+      "epoch": 0.5763641906310334,
+      "grad_norm": 3.3288702964782715,
+      "learning_rate": 8.02488941274078e-06,
+      "loss": 1.217,
+      "step": 709
+    },
+    {
+      "epoch": 0.5771771161467331,
+      "grad_norm": 2.808100700378418,
+      "learning_rate": 7.999081493280283e-06,
+      "loss": 1.3156,
+      "step": 710
+    },
+    {
+      "epoch": 0.5779900416624327,
+      "grad_norm": 2.8736870288848877,
+      "learning_rate": 7.973287449299545e-06,
+      "loss": 1.3122,
+      "step": 711
+    },
+    {
+      "epoch": 0.5788029671781323,
+      "grad_norm": 6.863023281097412,
+      "learning_rate": 7.947507459668784e-06,
+      "loss": 1.2218,
+      "step": 712
+    },
+    {
+      "epoch": 0.579615892693832,
+      "grad_norm": 4.454842567443848,
+      "learning_rate": 7.921741703160758e-06,
+      "loss": 1.1918,
+      "step": 713
+    },
+    {
+      "epoch": 0.5804288182095315,
+      "grad_norm": 2.4465959072113037,
+      "learning_rate": 7.895990358449533e-06,
+      "loss": 1.2705,
+      "step": 714
+    },
+    {
+      "epoch": 0.5812417437252312,
+      "grad_norm": 3.3625428676605225,
+      "learning_rate": 7.87025360410922e-06,
+      "loss": 1.2644,
+      "step": 715
+    },
+    {
+      "epoch": 0.5820546692409307,
+      "grad_norm": 2.846947431564331,
+      "learning_rate": 7.844531618612772e-06,
+      "loss": 1.2612,
+      "step": 716
+    },
+    {
+      "epoch": 0.5828675947566304,
+      "grad_norm": 3.332118034362793,
+      "learning_rate": 7.81882458033071e-06,
+      "loss": 1.2597,
+      "step": 717
+    },
+    {
+      "epoch": 0.5836805202723301,
+      "grad_norm": 2.646106719970703,
+      "learning_rate": 7.79313266752991e-06,
+      "loss": 1.2613,
+      "step": 718
+    },
+    {
+      "epoch": 0.5844934457880296,
+      "grad_norm": 2.8592135906219482,
+      "learning_rate": 7.767456058372362e-06,
+      "loss": 1.282,
+      "step": 719
+    },
+    {
+      "epoch": 0.5853063713037293,
+      "grad_norm": 2.748481035232544,
+      "learning_rate": 7.741794930913922e-06,
+      "loss": 1.2869,
+      "step": 720
+    },
+    {
+      "epoch": 0.5861192968194289,
+      "grad_norm": 2.8134074211120605,
+      "learning_rate": 7.7161494631031e-06,
+      "loss": 1.3079,
+      "step": 721
+    },
+    {
+      "epoch": 0.5869322223351285,
+      "grad_norm": 3.059119939804077,
+      "learning_rate": 7.690519832779799e-06,
+      "loss": 1.2705,
+      "step": 722
+    },
+    {
+      "epoch": 0.5877451478508282,
+      "grad_norm": 2.6439130306243896,
+      "learning_rate": 7.664906217674115e-06,
+      "loss": 1.2413,
+      "step": 723
+    },
+    {
+      "epoch": 0.5885580733665278,
+      "grad_norm": 2.812056303024292,
+      "learning_rate": 7.639308795405066e-06,
+      "loss": 1.2543,
+      "step": 724
+    },
+    {
+      "epoch": 0.5893709988822274,
+      "grad_norm": 3.2603330612182617,
+      "learning_rate": 7.613727743479395e-06,
+      "loss": 1.2442,
+      "step": 725
+    },
+    {
+      "epoch": 0.590183924397927,
+      "grad_norm": 2.544433116912842,
+      "learning_rate": 7.588163239290316e-06,
+      "loss": 1.3034,
+      "step": 726
+    },
+    {
+      "epoch": 0.5909968499136267,
+      "grad_norm": 4.0246262550354,
+      "learning_rate": 7.562615460116289e-06,
+      "loss": 1.3188,
+      "step": 727
+    },
+    {
+      "epoch": 0.5918097754293263,
+      "grad_norm": 4.249239444732666,
+      "learning_rate": 7.537084583119802e-06,
+      "loss": 1.3091,
+      "step": 728
+    },
+    {
+      "epoch": 0.5926227009450259,
+      "grad_norm": 2.7686362266540527,
+      "learning_rate": 7.511570785346129e-06,
+      "loss": 1.2449,
+      "step": 729
+    },
+    {
+      "epoch": 0.5934356264607256,
+      "grad_norm": 2.8529245853424072,
+      "learning_rate": 7.486074243722109e-06,
+      "loss": 1.2392,
+      "step": 730
+    },
+    {
+      "epoch": 0.5942485519764251,
+      "grad_norm": 3.073486328125,
+      "learning_rate": 7.460595135054916e-06,
+      "loss": 1.2848,
+      "step": 731
+    },
+    {
+      "epoch": 0.5950614774921248,
+      "grad_norm": 3.365366220474243,
+      "learning_rate": 7.435133636030831e-06,
+      "loss": 1.2912,
+      "step": 732
+    },
+    {
+      "epoch": 0.5958744030078245,
+      "grad_norm": 2.4938106536865234,
+      "learning_rate": 7.4096899232140295e-06,
+      "loss": 1.2965,
+      "step": 733
+    },
+    {
+      "epoch": 0.596687328523524,
+      "grad_norm": 2.9927473068237305,
+      "learning_rate": 7.384264173045339e-06,
+      "loss": 1.2748,
+      "step": 734
+    },
+    {
+      "epoch": 0.5975002540392237,
+      "grad_norm": 7.3427205085754395,
+      "learning_rate": 7.358856561841021e-06,
+      "loss": 1.2457,
+      "step": 735
+    },
+    {
+      "epoch": 0.5983131795549232,
+      "grad_norm": 3.274311065673828,
+      "learning_rate": 7.333467265791563e-06,
+      "loss": 1.2225,
+      "step": 736
+    },
+    {
+      "epoch": 0.5991261050706229,
+      "grad_norm": 4.503856658935547,
+      "learning_rate": 7.308096460960441e-06,
+      "loss": 1.2603,
+      "step": 737
+    },
+    {
+      "epoch": 0.5999390305863225,
+      "grad_norm": 3.6017913818359375,
+      "learning_rate": 7.282744323282895e-06,
+      "loss": 1.2278,
+      "step": 738
+    },
+    {
+      "epoch": 0.6007519561020221,
+      "grad_norm": 3.0930585861206055,
+      "learning_rate": 7.2574110285647244e-06,
+      "loss": 1.2649,
+      "step": 739
+    },
+    {
+      "epoch": 0.6015648816177218,
+      "grad_norm": 2.6793737411499023,
+      "learning_rate": 7.232096752481061e-06,
+      "loss": 1.215,
+      "step": 740
+    },
+    {
+      "epoch": 0.6023778071334214,
+      "grad_norm": 3.0066819190979004,
+      "learning_rate": 7.206801670575145e-06,
+      "loss": 1.2953,
+      "step": 741
+    },
+    {
+      "epoch": 0.603190732649121,
+      "grad_norm": 3.2586004734039307,
+      "learning_rate": 7.181525958257116e-06,
+      "loss": 1.1988,
+      "step": 742
+    },
+    {
+      "epoch": 0.6040036581648206,
+      "grad_norm": 3.186267375946045,
+      "learning_rate": 7.156269790802801e-06,
+      "loss": 1.2425,
+      "step": 743
+    },
+    {
+      "epoch": 0.6048165836805203,
+      "grad_norm": 3.919509172439575,
+      "learning_rate": 7.131033343352483e-06,
+      "loss": 1.3432,
+      "step": 744
+    },
+    {
+      "epoch": 0.6056295091962199,
+      "grad_norm": 3.8313186168670654,
+      "learning_rate": 7.105816790909699e-06,
+      "loss": 1.2491,
+      "step": 745
+    },
+    {
+      "epoch": 0.6064424347119195,
+      "grad_norm": 2.7689011096954346,
+      "learning_rate": 7.080620308340024e-06,
+      "loss": 1.2673,
+      "step": 746
+    },
+    {
+      "epoch": 0.6072553602276192,
+      "grad_norm": 4.105691909790039,
+      "learning_rate": 7.055444070369852e-06,
+      "loss": 1.2688,
+      "step": 747
+    },
+    {
+      "epoch": 0.6080682857433187,
+      "grad_norm": 3.336580276489258,
+      "learning_rate": 7.0302882515852025e-06,
+      "loss": 1.2613,
+      "step": 748
+    },
+    {
+      "epoch": 0.6088812112590184,
+      "grad_norm": 3.7272021770477295,
+      "learning_rate": 7.005153026430476e-06,
+      "loss": 1.1882,
+      "step": 749
+    },
+    {
+      "epoch": 0.6096941367747181,
+      "grad_norm": 4.220558166503906,
+      "learning_rate": 6.980038569207291e-06,
+      "loss": 1.1853,
+      "step": 750
+    },
+    {
+      "epoch": 0.6105070622904176,
+      "grad_norm": 2.8943638801574707,
+      "learning_rate": 6.954945054073228e-06,
+      "loss": 1.2408,
+      "step": 751
+    },
+    {
+      "epoch": 0.6113199878061173,
+      "grad_norm": 2.740449905395508,
+      "learning_rate": 6.929872655040655e-06,
+      "loss": 1.2233,
+      "step": 752
+    },
+    {
+      "epoch": 0.6121329133218169,
+      "grad_norm": 3.1293320655822754,
+      "learning_rate": 6.904821545975507e-06,
+      "loss": 1.2362,
+      "step": 753
+    },
+    {
+      "epoch": 0.6129458388375165,
+      "grad_norm": 2.9130334854125977,
+      "learning_rate": 6.879791900596077e-06,
+      "loss": 1.2525,
+      "step": 754
+    },
+    {
+      "epoch": 0.6137587643532162,
+      "grad_norm": 2.6800663471221924,
+      "learning_rate": 6.854783892471823e-06,
+      "loss": 1.2811,
+      "step": 755
+    },
+    {
+      "epoch": 0.6145716898689157,
+      "grad_norm": 2.7140908241271973,
+      "learning_rate": 6.829797695022163e-06,
+      "loss": 1.2693,
+      "step": 756
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 2.687870740890503,
+      "learning_rate": 6.804833481515256e-06,
+      "loss": 1.2124,
+      "step": 757
+    },
+    {
+      "epoch": 0.616197540900315,
+      "grad_norm": 3.170487880706787,
+      "learning_rate": 6.7798914250668154e-06,
+      "loss": 1.2373,
+      "step": 758
+    },
+    {
+      "epoch": 0.6170104664160146,
+      "grad_norm": 2.6142961978912354,
+      "learning_rate": 6.7549716986389146e-06,
+      "loss": 1.2527,
+      "step": 759
+    },
+    {
+      "epoch": 0.6178233919317142,
+      "grad_norm": 3.4092085361480713,
+      "learning_rate": 6.730074475038766e-06,
+      "loss": 1.2401,
+      "step": 760
+    },
+    {
+      "epoch": 0.6186363174474139,
+      "grad_norm": 3.256838083267212,
+      "learning_rate": 6.7051999269175405e-06,
+      "loss": 1.1863,
+      "step": 761
+    },
+    {
+      "epoch": 0.6194492429631135,
+      "grad_norm": 2.8312947750091553,
+      "learning_rate": 6.680348226769162e-06,
+      "loss": 1.241,
+      "step": 762
+    },
+    {
+      "epoch": 0.6202621684788131,
+      "grad_norm": 2.799750804901123,
+      "learning_rate": 6.655519546929121e-06,
+      "loss": 1.2601,
+      "step": 763
+    },
+    {
+      "epoch": 0.6210750939945128,
+      "grad_norm": 3.188913106918335,
+      "learning_rate": 6.630714059573267e-06,
+      "loss": 1.2719,
+      "step": 764
+    },
+    {
+      "epoch": 0.6218880195102123,
+      "grad_norm": 5.547321796417236,
+      "learning_rate": 6.6059319367166165e-06,
+      "loss": 1.2307,
+      "step": 765
+    },
+    {
+      "epoch": 0.622700945025912,
+      "grad_norm": 3.2380361557006836,
+      "learning_rate": 6.581173350212169e-06,
+      "loss": 1.2125,
+      "step": 766
+    },
+    {
+      "epoch": 0.6235138705416117,
+      "grad_norm": 2.61883282661438,
+      "learning_rate": 6.55643847174971e-06,
+      "loss": 1.2556,
+      "step": 767
+    },
+    {
+      "epoch": 0.6243267960573112,
+      "grad_norm": 3.0079920291900635,
+      "learning_rate": 6.531727472854617e-06,
+      "loss": 1.2761,
+      "step": 768
+    },
+    {
+      "epoch": 0.6251397215730109,
+      "grad_norm": 3.995910882949829,
+      "learning_rate": 6.507040524886672e-06,
+      "loss": 1.302,
+      "step": 769
+    },
+    {
+      "epoch": 0.6259526470887105,
+      "grad_norm": 2.7787578105926514,
+      "learning_rate": 6.482377799038882e-06,
+      "loss": 1.2249,
+      "step": 770
+    },
+    {
+      "epoch": 0.6267655726044101,
+      "grad_norm": 3.6458895206451416,
+      "learning_rate": 6.45773946633628e-06,
+      "loss": 1.2833,
+      "step": 771
+    },
+    {
+      "epoch": 0.6275784981201098,
+      "grad_norm": 2.9308435916900635,
+      "learning_rate": 6.4331256976347434e-06,
+      "loss": 1.309,
+      "step": 772
+    },
+    {
+      "epoch": 0.6283914236358094,
+      "grad_norm": 3.7917234897613525,
+      "learning_rate": 6.408536663619803e-06,
+      "loss": 1.2996,
+      "step": 773
+    },
+    {
+      "epoch": 0.629204349151509,
+      "grad_norm": 8.85531234741211,
+      "learning_rate": 6.383972534805478e-06,
+      "loss": 1.2499,
+      "step": 774
+    },
+    {
+      "epoch": 0.6300172746672086,
+      "grad_norm": 4.16661262512207,
+      "learning_rate": 6.359433481533074e-06,
+      "loss": 1.1928,
+      "step": 775
+    },
+    {
+      "epoch": 0.6308302001829083,
+      "grad_norm": 3.6679298877716064,
+      "learning_rate": 6.3349196739700024e-06,
+      "loss": 1.2917,
+      "step": 776
+    },
+    {
+      "epoch": 0.6316431256986078,
+      "grad_norm": 3.2031593322753906,
+      "learning_rate": 6.310431282108622e-06,
+      "loss": 1.2926,
+      "step": 777
+    },
+    {
+      "epoch": 0.6324560512143075,
+      "grad_norm": 2.7538363933563232,
+      "learning_rate": 6.2859684757650365e-06,
+      "loss": 1.2634,
+      "step": 778
+    },
+    {
+      "epoch": 0.6332689767300071,
+      "grad_norm": 3.4906575679779053,
+      "learning_rate": 6.261531424577923e-06,
+      "loss": 1.2711,
+      "step": 779
+    },
+    {
+      "epoch": 0.6340819022457067,
+      "grad_norm": 3.4287617206573486,
+      "learning_rate": 6.2371202980073596e-06,
+      "loss": 1.2412,
+      "step": 780
+    },
+    {
+      "epoch": 0.6348948277614064,
+      "grad_norm": 3.5826241970062256,
+      "learning_rate": 6.212735265333655e-06,
+      "loss": 1.1782,
+      "step": 781
+    },
+    {
+      "epoch": 0.6357077532771059,
+      "grad_norm": 3.369983673095703,
+      "learning_rate": 6.188376495656156e-06,
+      "loss": 1.2628,
+      "step": 782
+    },
+    {
+      "epoch": 0.6365206787928056,
+      "grad_norm": 3.6163413524627686,
+      "learning_rate": 6.164044157892102e-06,
+      "loss": 1.3304,
+      "step": 783
+    },
+    {
+      "epoch": 0.6373336043085053,
+      "grad_norm": 2.6903252601623535,
+      "learning_rate": 6.13973842077543e-06,
+      "loss": 1.2458,
+      "step": 784
+    },
+    {
+      "epoch": 0.6381465298242048,
+      "grad_norm": 3.919074296951294,
+      "learning_rate": 6.11545945285561e-06,
+      "loss": 1.253,
+      "step": 785
+    },
+    {
+      "epoch": 0.6389594553399045,
+      "grad_norm": 2.9155240058898926,
+      "learning_rate": 6.091207422496489e-06,
+      "loss": 1.2661,
+      "step": 786
+    },
+    {
+      "epoch": 0.6397723808556041,
+      "grad_norm": 3.2426347732543945,
+      "learning_rate": 6.066982497875109e-06,
+      "loss": 1.2556,
+      "step": 787
+    },
+    {
+      "epoch": 0.6405853063713037,
+      "grad_norm": 3.078899383544922,
+      "learning_rate": 6.042784846980542e-06,
+      "loss": 1.2572,
+      "step": 788
+    },
+    {
+      "epoch": 0.6413982318870034,
+      "grad_norm": 3.3044381141662598,
+      "learning_rate": 6.018614637612733e-06,
+      "loss": 1.2301,
+      "step": 789
+    },
+    {
+      "epoch": 0.642211157402703,
+      "grad_norm": 2.8474955558776855,
+      "learning_rate": 5.99447203738134e-06,
+      "loss": 1.2042,
+      "step": 790
+    },
+    {
+      "epoch": 0.6430240829184026,
+      "grad_norm": 2.9787845611572266,
+      "learning_rate": 5.9703572137045495e-06,
+      "loss": 1.2608,
+      "step": 791
+    },
+    {
+      "epoch": 0.6438370084341022,
+      "grad_norm": 3.380209445953369,
+      "learning_rate": 5.946270333807937e-06,
+      "loss": 1.2973,
+      "step": 792
+    },
+    {
+      "epoch": 0.6446499339498019,
+      "grad_norm": 2.81736421585083,
+      "learning_rate": 5.922211564723302e-06,
+      "loss": 1.2791,
+      "step": 793
+    },
+    {
+      "epoch": 0.6454628594655015,
+      "grad_norm": 2.9054102897644043,
+      "learning_rate": 5.898181073287504e-06,
+      "loss": 1.2692,
+      "step": 794
+    },
+    {
+      "epoch": 0.6462757849812011,
+      "grad_norm": 3.2480154037475586,
+      "learning_rate": 5.87417902614131e-06,
+      "loss": 1.311,
+      "step": 795
+    },
+    {
+      "epoch": 0.6470887104969008,
+      "grad_norm": 2.8822832107543945,
+      "learning_rate": 5.850205589728239e-06,
+      "loss": 1.2528,
+      "step": 796
+    },
+    {
+      "epoch": 0.6479016360126003,
+      "grad_norm": 2.8832008838653564,
+      "learning_rate": 5.826260930293417e-06,
+      "loss": 1.2631,
+      "step": 797
+    },
+    {
+      "epoch": 0.6487145615283,
+      "grad_norm": 3.547271490097046,
+      "learning_rate": 5.802345213882396e-06,
+      "loss": 1.2543,
+      "step": 798
+    },
+    {
+      "epoch": 0.6495274870439995,
+      "grad_norm": 9.93248176574707,
+      "learning_rate": 5.778458606340037e-06,
+      "loss": 1.3218,
+      "step": 799
+    },
+    {
+      "epoch": 0.6503404125596992,
+      "grad_norm": 4.664019584655762,
+      "learning_rate": 5.754601273309333e-06,
+      "loss": 1.2487,
+      "step": 800
+    },
+    {
+      "epoch": 0.6511533380753989,
+      "grad_norm": 3.191390037536621,
+      "learning_rate": 5.730773380230276e-06,
+      "loss": 1.1966,
+      "step": 801
+    },
+    {
+      "epoch": 0.6519662635910984,
+      "grad_norm": 3.228309392929077,
+      "learning_rate": 5.70697509233871e-06,
+      "loss": 1.2556,
+      "step": 802
+    },
+    {
+      "epoch": 0.6527791891067981,
+      "grad_norm": 3.1456098556518555,
+      "learning_rate": 5.683206574665165e-06,
+      "loss": 1.2308,
+      "step": 803
+    },
+    {
+      "epoch": 0.6535921146224977,
+      "grad_norm": 2.800039052963257,
+      "learning_rate": 5.6594679920337514e-06,
+      "loss": 1.2599,
+      "step": 804
+    },
+    {
+      "epoch": 0.6544050401381973,
+      "grad_norm": 2.9048550128936768,
+      "learning_rate": 5.635759509060969e-06,
+      "loss": 1.2707,
+      "step": 805
+    },
+    {
+      "epoch": 0.655217965653897,
+      "grad_norm": 4.015383720397949,
+      "learning_rate": 5.612081290154607e-06,
+      "loss": 1.1853,
+      "step": 806
+    },
+    {
+      "epoch": 0.6560308911695966,
+      "grad_norm": 2.6166458129882812,
+      "learning_rate": 5.58843349951258e-06,
+      "loss": 1.2589,
+      "step": 807
+    },
+    {
+      "epoch": 0.6568438166852962,
+      "grad_norm": 4.735121726989746,
+      "learning_rate": 5.564816301121792e-06,
+      "loss": 1.2395,
+      "step": 808
+    },
+    {
+      "epoch": 0.6576567422009958,
+      "grad_norm": 3.5069589614868164,
+      "learning_rate": 5.541229858757011e-06,
+      "loss": 1.2888,
+      "step": 809
+    },
+    {
+      "epoch": 0.6584696677166955,
+      "grad_norm": 2.354539394378662,
+      "learning_rate": 5.517674335979721e-06,
+      "loss": 1.1898,
+      "step": 810
+    },
+    {
+      "epoch": 0.6592825932323951,
+      "grad_norm": 3.2337725162506104,
+      "learning_rate": 5.494149896136998e-06,
+      "loss": 1.311,
+      "step": 811
+    },
+    {
+      "epoch": 0.6600955187480947,
+      "grad_norm": 2.6511757373809814,
+      "learning_rate": 5.470656702360367e-06,
+      "loss": 1.2788,
+      "step": 812
+    },
+    {
+      "epoch": 0.6609084442637944,
+      "grad_norm": 3.772780179977417,
+      "learning_rate": 5.447194917564671e-06,
+      "loss": 1.2211,
+      "step": 813
+    },
+    {
+      "epoch": 0.6617213697794939,
+      "grad_norm": 2.540316581726074,
+      "learning_rate": 5.423764704446954e-06,
+      "loss": 1.2647,
+      "step": 814
+    },
+    {
+      "epoch": 0.6625342952951936,
+      "grad_norm": 2.792747735977173,
+      "learning_rate": 5.400366225485326e-06,
+      "loss": 1.2184,
+      "step": 815
+    },
+    {
+      "epoch": 0.6633472208108933,
+      "grad_norm": 3.32261061668396,
+      "learning_rate": 5.376999642937817e-06,
+      "loss": 1.2727,
+      "step": 816
+    },
+    {
+      "epoch": 0.6641601463265928,
+      "grad_norm": 4.128072738647461,
+      "learning_rate": 5.353665118841296e-06,
+      "loss": 1.2718,
+      "step": 817
+    },
+    {
+      "epoch": 0.6649730718422925,
+      "grad_norm": 2.9913909435272217,
+      "learning_rate": 5.330362815010306e-06,
+      "loss": 1.2698,
+      "step": 818
+    },
+    {
+      "epoch": 0.665785997357992,
+      "grad_norm": 2.9993457794189453,
+      "learning_rate": 5.307092893035951e-06,
+      "loss": 1.2447,
+      "step": 819
+    },
+    {
+      "epoch": 0.6665989228736917,
+      "grad_norm": 2.801236629486084,
+      "learning_rate": 5.2838555142847925e-06,
+      "loss": 1.209,
+      "step": 820
+    },
+    {
+      "epoch": 0.6674118483893913,
+      "grad_norm": 3.982821464538574,
+      "learning_rate": 5.260650839897719e-06,
+      "loss": 1.3099,
+      "step": 821
+    },
+    {
+      "epoch": 0.6682247739050909,
+      "grad_norm": 2.9553382396698,
+      "learning_rate": 5.237479030788817e-06,
+      "loss": 1.2652,
+      "step": 822
+    },
+    {
+      "epoch": 0.6690376994207906,
+      "grad_norm": 3.233414888381958,
+      "learning_rate": 5.214340247644278e-06,
+      "loss": 1.2256,
+      "step": 823
+    },
+    {
+      "epoch": 0.6698506249364902,
+      "grad_norm": 3.1418299674987793,
+      "learning_rate": 5.191234650921273e-06,
+      "loss": 1.2225,
+      "step": 824
+    },
+    {
+      "epoch": 0.6706635504521898,
+      "grad_norm": 2.8071773052215576,
+      "learning_rate": 5.168162400846835e-06,
+      "loss": 1.3381,
+      "step": 825
+    },
+    {
+      "epoch": 0.6714764759678894,
+      "grad_norm": 3.2606897354125977,
+      "learning_rate": 5.145123657416759e-06,
+      "loss": 1.2671,
+      "step": 826
+    },
+    {
+      "epoch": 0.6722894014835891,
+      "grad_norm": 2.5103461742401123,
+      "learning_rate": 5.122118580394473e-06,
+      "loss": 1.2349,
+      "step": 827
+    },
+    {
+      "epoch": 0.6731023269992887,
+      "grad_norm": 2.882448196411133,
+      "learning_rate": 5.099147329309959e-06,
+      "loss": 1.2466,
+      "step": 828
+    },
+    {
+      "epoch": 0.6739152525149883,
+      "grad_norm": 3.0320730209350586,
+      "learning_rate": 5.076210063458622e-06,
+      "loss": 1.2157,
+      "step": 829
+    },
+    {
+      "epoch": 0.674728178030688,
+      "grad_norm": 3.285125970840454,
+      "learning_rate": 5.0533069419002e-06,
+      "loss": 1.3087,
+      "step": 830
+    },
+    {
+      "epoch": 0.6755411035463875,
+      "grad_norm": 3.9807510375976562,
+      "learning_rate": 5.030438123457655e-06,
+      "loss": 1.2153,
+      "step": 831
+    },
+    {
+      "epoch": 0.6763540290620872,
+      "grad_norm": 3.12975811958313,
+      "learning_rate": 5.007603766716063e-06,
+      "loss": 1.2064,
+      "step": 832
+    },
+    {
+      "epoch": 0.6771669545777869,
+      "grad_norm": 2.9132258892059326,
+      "learning_rate": 4.984804030021533e-06,
+      "loss": 1.2132,
+      "step": 833
+    },
+    {
+      "epoch": 0.6779798800934864,
+      "grad_norm": 2.872042417526245,
+      "learning_rate": 4.962039071480102e-06,
+      "loss": 1.2618,
+      "step": 834
+    },
+    {
+      "epoch": 0.6787928056091861,
+      "grad_norm": 3.7190613746643066,
+      "learning_rate": 4.939309048956622e-06,
+      "loss": 1.2482,
+      "step": 835
+    },
+    {
+      "epoch": 0.6796057311248856,
+      "grad_norm": 5.171625137329102,
+      "learning_rate": 4.9166141200736885e-06,
+      "loss": 1.2848,
+      "step": 836
+    },
+    {
+      "epoch": 0.6804186566405853,
+      "grad_norm": 3.5912961959838867,
+      "learning_rate": 4.89395444221055e-06,
+      "loss": 1.2525,
+      "step": 837
+    },
+    {
+      "epoch": 0.6812315821562849,
+      "grad_norm": 3.9113729000091553,
+      "learning_rate": 4.871330172501979e-06,
+      "loss": 1.2444,
+      "step": 838
+    },
+    {
+      "epoch": 0.6820445076719845,
+      "grad_norm": 5.135432720184326,
+      "learning_rate": 4.848741467837228e-06,
+      "loss": 1.2189,
+      "step": 839
+    },
+    {
+      "epoch": 0.6828574331876842,
+      "grad_norm": 3.0934841632843018,
+      "learning_rate": 4.826188484858918e-06,
+      "loss": 1.2357,
+      "step": 840
+    },
+    {
+      "epoch": 0.6836703587033838,
+      "grad_norm": 3.951188325881958,
+      "learning_rate": 4.803671379961945e-06,
+      "loss": 1.2539,
+      "step": 841
+    },
+    {
+      "epoch": 0.6844832842190834,
+      "grad_norm": 6.205260753631592,
+      "learning_rate": 4.781190309292421e-06,
+      "loss": 1.2537,
+      "step": 842
+    },
+    {
+      "epoch": 0.685296209734783,
+      "grad_norm": 4.493546485900879,
+      "learning_rate": 4.758745428746569e-06,
+      "loss": 1.252,
+      "step": 843
+    },
+    {
+      "epoch": 0.6861091352504827,
+      "grad_norm": 4.0202436447143555,
+      "learning_rate": 4.736336893969652e-06,
+      "loss": 1.1887,
+      "step": 844
+    },
+    {
+      "epoch": 0.6869220607661823,
+      "grad_norm": 2.65285587310791,
+      "learning_rate": 4.7139648603548925e-06,
+      "loss": 1.2612,
+      "step": 845
+    },
+    {
+      "epoch": 0.6877349862818819,
+      "grad_norm": 3.629551410675049,
+      "learning_rate": 4.691629483042387e-06,
+      "loss": 1.2411,
+      "step": 846
+    },
+    {
+      "epoch": 0.6885479117975816,
+      "grad_norm": 3.20709228515625,
+      "learning_rate": 4.669330916918043e-06,
+      "loss": 1.1949,
+      "step": 847
+    },
+    {
+      "epoch": 0.6893608373132811,
+      "grad_norm": 3.19427752494812,
+      "learning_rate": 4.647069316612502e-06,
+      "loss": 1.2134,
+      "step": 848
+    },
+    {
+      "epoch": 0.6901737628289808,
+      "grad_norm": 3.6364243030548096,
+      "learning_rate": 4.624844836500052e-06,
+      "loss": 1.2915,
+      "step": 849
+    },
+    {
+      "epoch": 0.6909866883446805,
+      "grad_norm": 3.5689237117767334,
+      "learning_rate": 4.60265763069758e-06,
+      "loss": 1.2234,
+      "step": 850
+    },
+    {
+      "epoch": 0.69179961386038,
+      "grad_norm": 3.1175014972686768,
+      "learning_rate": 4.580507853063487e-06,
+      "loss": 1.1833,
+      "step": 851
+    },
+    {
+      "epoch": 0.6926125393760797,
+      "grad_norm": 2.945756196975708,
+      "learning_rate": 4.5583956571966295e-06,
+      "loss": 1.2231,
+      "step": 852
+    },
+    {
+      "epoch": 0.6934254648917793,
+      "grad_norm": 4.729986667633057,
+      "learning_rate": 4.5363211964352524e-06,
+      "loss": 1.2578,
+      "step": 853
+    },
+    {
+      "epoch": 0.6942383904074789,
+      "grad_norm": 2.7775003910064697,
+      "learning_rate": 4.514284623855915e-06,
+      "loss": 1.2678,
+      "step": 854
+    },
+    {
+      "epoch": 0.6950513159231786,
+      "grad_norm": 4.027686595916748,
+      "learning_rate": 4.4922860922724466e-06,
+      "loss": 1.1692,
+      "step": 855
+    },
+    {
+      "epoch": 0.6958642414388782,
+      "grad_norm": 3.3442118167877197,
+      "learning_rate": 4.470325754234881e-06,
+      "loss": 1.2515,
+      "step": 856
+    },
+    {
+      "epoch": 0.6966771669545778,
+      "grad_norm": 3.197281837463379,
+      "learning_rate": 4.448403762028391e-06,
+      "loss": 1.2789,
+      "step": 857
+    },
+    {
+      "epoch": 0.6974900924702774,
+      "grad_norm": 3.1467063426971436,
+      "learning_rate": 4.426520267672244e-06,
+      "loss": 1.2498,
+      "step": 858
+    },
+    {
+      "epoch": 0.698303017985977,
+      "grad_norm": 8.657835960388184,
+      "learning_rate": 4.40467542291874e-06,
+      "loss": 1.2149,
+      "step": 859
+    },
+    {
+      "epoch": 0.6991159435016766,
+      "grad_norm": 5.045658111572266,
+      "learning_rate": 4.382869379252152e-06,
+      "loss": 1.2143,
+      "step": 860
+    },
+    {
+      "epoch": 0.6999288690173763,
+      "grad_norm": 3.543026924133301,
+      "learning_rate": 4.361102287887698e-06,
+      "loss": 1.2727,
+      "step": 861
+    },
+    {
+      "epoch": 0.700741794533076,
+      "grad_norm": 3.2592012882232666,
+      "learning_rate": 4.339374299770477e-06,
+      "loss": 1.2528,
+      "step": 862
+    },
+    {
+      "epoch": 0.7015547200487755,
+      "grad_norm": 3.284749984741211,
+      "learning_rate": 4.31768556557441e-06,
+      "loss": 1.1814,
+      "step": 863
+    },
+    {
+      "epoch": 0.7023676455644752,
+      "grad_norm": 2.9172427654266357,
+      "learning_rate": 4.296036235701235e-06,
+      "loss": 1.2536,
+      "step": 864
+    },
+    {
+      "epoch": 0.7031805710801747,
+      "grad_norm": 8.07040023803711,
+      "learning_rate": 4.274426460279412e-06,
+      "loss": 1.2113,
+      "step": 865
+    },
+    {
+      "epoch": 0.7039934965958744,
+      "grad_norm": 3.0349769592285156,
+      "learning_rate": 4.252856389163128e-06,
+      "loss": 1.2279,
+      "step": 866
+    },
+    {
+      "epoch": 0.7048064221115741,
+      "grad_norm": 2.7983269691467285,
+      "learning_rate": 4.231326171931231e-06,
+      "loss": 1.2585,
+      "step": 867
+    },
+    {
+      "epoch": 0.7056193476272736,
+      "grad_norm": 3.153099775314331,
+      "learning_rate": 4.209835957886196e-06,
+      "loss": 1.2576,
+      "step": 868
+    },
+    {
+      "epoch": 0.7064322731429733,
+      "grad_norm": 3.4303712844848633,
+      "learning_rate": 4.188385896053098e-06,
+      "loss": 1.2569,
+      "step": 869
+    },
+    {
+      "epoch": 0.7072451986586729,
+      "grad_norm": 3.310842990875244,
+      "learning_rate": 4.166976135178575e-06,
+      "loss": 1.2162,
+      "step": 870
+    },
+    {
+      "epoch": 0.7080581241743725,
+      "grad_norm": 3.982365846633911,
+      "learning_rate": 4.1456068237297964e-06,
+      "loss": 1.2409,
+      "step": 871
+    },
+    {
+      "epoch": 0.7088710496900722,
+      "grad_norm": 3.0641191005706787,
+      "learning_rate": 4.124278109893432e-06,
+      "loss": 1.2563,
+      "step": 872
+    },
+    {
+      "epoch": 0.7096839752057718,
+      "grad_norm": 2.9682273864746094,
+      "learning_rate": 4.10299014157462e-06,
+      "loss": 1.1857,
+      "step": 873
+    },
+    {
+      "epoch": 0.7104969007214714,
+      "grad_norm": 6.076914310455322,
+      "learning_rate": 4.0817430663959536e-06,
+      "loss": 1.2108,
+      "step": 874
+    },
+    {
+      "epoch": 0.711309826237171,
+      "grad_norm": 8.528678894042969,
+      "learning_rate": 4.06053703169645e-06,
+      "loss": 1.2185,
+      "step": 875
+    },
+    {
+      "epoch": 0.7121227517528707,
+      "grad_norm": 3.4424145221710205,
+      "learning_rate": 4.039372184530521e-06,
+      "loss": 1.2461,
+      "step": 876
+    },
+    {
+      "epoch": 0.7129356772685703,
+      "grad_norm": 3.1624224185943604,
+      "learning_rate": 4.0182486716669656e-06,
+      "loss": 1.2282,
+      "step": 877
+    },
+    {
+      "epoch": 0.7137486027842699,
+      "grad_norm": 4.986435890197754,
+      "learning_rate": 3.9971666395879605e-06,
+      "loss": 1.2048,
+      "step": 878
+    },
+    {
+      "epoch": 0.7145615282999696,
+      "grad_norm": 3.537174701690674,
+      "learning_rate": 3.9761262344880096e-06,
+      "loss": 1.2752,
+      "step": 879
+    },
+    {
+      "epoch": 0.7153744538156691,
+      "grad_norm": 2.7389779090881348,
+      "learning_rate": 3.9551276022729644e-06,
+      "loss": 1.2434,
+      "step": 880
+    },
+    {
+      "epoch": 0.7161873793313688,
+      "grad_norm": 3.5238423347473145,
+      "learning_rate": 3.9341708885590034e-06,
+      "loss": 1.2409,
+      "step": 881
+    },
+    {
+      "epoch": 0.7170003048470683,
+      "grad_norm": 3.9080941677093506,
+      "learning_rate": 3.913256238671607e-06,
+      "loss": 1.2019,
+      "step": 882
+    },
+    {
+      "epoch": 0.717813230362768,
+      "grad_norm": 4.038003921508789,
+      "learning_rate": 3.89238379764457e-06,
+      "loss": 1.2212,
+      "step": 883
+    },
+    {
+      "epoch": 0.7186261558784677,
+      "grad_norm": 3.344622850418091,
+      "learning_rate": 3.871553710218988e-06,
+      "loss": 1.2067,
+      "step": 884
+    },
+    {
+      "epoch": 0.7194390813941672,
+      "grad_norm": 3.5090816020965576,
+      "learning_rate": 3.850766120842252e-06,
+      "loss": 1.2171,
+      "step": 885
+    },
+    {
+      "epoch": 0.7202520069098669,
+      "grad_norm": 3.003899335861206,
+      "learning_rate": 3.830021173667048e-06,
+      "loss": 1.2371,
+      "step": 886
+    },
+    {
+      "epoch": 0.7210649324255665,
+      "grad_norm": 3.3116228580474854,
+      "learning_rate": 3.809319012550352e-06,
+      "loss": 1.2123,
+      "step": 887
+    },
+    {
+      "epoch": 0.7218778579412661,
+      "grad_norm": 3.532245397567749,
+      "learning_rate": 3.788659781052444e-06,
+      "loss": 1.2629,
+      "step": 888
+    },
+    {
+      "epoch": 0.7226907834569658,
+      "grad_norm": 4.061065196990967,
+      "learning_rate": 3.7680436224359084e-06,
+      "loss": 1.174,
+      "step": 889
+    },
+    {
+      "epoch": 0.7235037089726654,
+      "grad_norm": 3.3992788791656494,
+      "learning_rate": 3.747470679664624e-06,
+      "loss": 1.2209,
+      "step": 890
+    },
+    {
+      "epoch": 0.724316634488365,
+      "grad_norm": 3.4010937213897705,
+      "learning_rate": 3.7269410954028107e-06,
+      "loss": 1.2426,
+      "step": 891
+    },
+    {
+      "epoch": 0.7251295600040646,
+      "grad_norm": 2.854327917098999,
+      "learning_rate": 3.706455012013994e-06,
+      "loss": 1.1932,
+      "step": 892
+    },
+    {
+      "epoch": 0.7259424855197643,
+      "grad_norm": 3.451002836227417,
+      "learning_rate": 3.6860125715600513e-06,
+      "loss": 1.253,
+      "step": 893
+    },
+    {
+      "epoch": 0.7267554110354639,
+      "grad_norm": 3.123344898223877,
+      "learning_rate": 3.665613915800217e-06,
+      "loss": 1.2187,
+      "step": 894
+    },
+    {
+      "epoch": 0.7275683365511635,
+      "grad_norm": 3.021973133087158,
+      "learning_rate": 3.6452591861900886e-06,
+      "loss": 1.2165,
+      "step": 895
+    },
+    {
+      "epoch": 0.7283812620668632,
+      "grad_norm": 3.234985589981079,
+      "learning_rate": 3.6249485238806637e-06,
+      "loss": 1.212,
+      "step": 896
+    },
+    {
+      "epoch": 0.7291941875825627,
+      "grad_norm": 3.7146785259246826,
+      "learning_rate": 3.6046820697173514e-06,
+      "loss": 1.2697,
+      "step": 897
+    },
+    {
+      "epoch": 0.7300071130982624,
+      "grad_norm": 3.134507417678833,
+      "learning_rate": 3.5844599642389965e-06,
+      "loss": 1.2433,
+      "step": 898
+    },
+    {
+      "epoch": 0.7308200386139619,
+      "grad_norm": 2.9155194759368896,
+      "learning_rate": 3.564282347676903e-06,
+      "loss": 1.2403,
+      "step": 899
+    },
+    {
+      "epoch": 0.7316329641296616,
+      "grad_norm": 3.148232936859131,
+      "learning_rate": 3.54414935995387e-06,
+      "loss": 1.2575,
+      "step": 900
+    },
+    {
+      "epoch": 0.7324458896453613,
+      "grad_norm": 2.685274124145508,
+      "learning_rate": 3.524061140683206e-06,
+      "loss": 1.2124,
+      "step": 901
+    },
+    {
+      "epoch": 0.7332588151610608,
+      "grad_norm": 3.4557571411132812,
+      "learning_rate": 3.5040178291677816e-06,
+      "loss": 1.2105,
+      "step": 902
+    },
+    {
+      "epoch": 0.7340717406767605,
+      "grad_norm": 2.8230202198028564,
+      "learning_rate": 3.4840195643990383e-06,
+      "loss": 1.1745,
+      "step": 903
+    },
+    {
+      "epoch": 0.7348846661924601,
+      "grad_norm": 3.311697483062744,
+      "learning_rate": 3.464066485056048e-06,
+      "loss": 1.222,
+      "step": 904
+    },
+    {
+      "epoch": 0.7356975917081597,
+      "grad_norm": 3.2953929901123047,
+      "learning_rate": 3.444158729504549e-06,
+      "loss": 1.2688,
+      "step": 905
+    },
+    {
+      "epoch": 0.7365105172238594,
+      "grad_norm": 3.3319778442382812,
+      "learning_rate": 3.4242964357959597e-06,
+      "loss": 1.2539,
+      "step": 906
+    },
+    {
+      "epoch": 0.737323442739559,
+      "grad_norm": 3.124361753463745,
+      "learning_rate": 3.4044797416664564e-06,
+      "loss": 1.2527,
+      "step": 907
+    },
+    {
+      "epoch": 0.7381363682552586,
+      "grad_norm": 2.9690327644348145,
+      "learning_rate": 3.3847087845359996e-06,
+      "loss": 1.2722,
+      "step": 908
+    },
+    {
+      "epoch": 0.7389492937709582,
+      "grad_norm": 5.119561672210693,
+      "learning_rate": 3.364983701507376e-06,
+      "loss": 1.2233,
+      "step": 909
+    },
+    {
+      "epoch": 0.7397622192866579,
+      "grad_norm": 2.818423271179199,
+      "learning_rate": 3.3453046293652657e-06,
+      "loss": 1.2438,
+      "step": 910
+    },
+    {
+      "epoch": 0.7405751448023575,
+      "grad_norm": 3.0988523960113525,
+      "learning_rate": 3.3256717045752794e-06,
+      "loss": 1.223,
+      "step": 911
+    },
+    {
+      "epoch": 0.7413880703180571,
+      "grad_norm": 3.082066297531128,
+      "learning_rate": 3.3060850632830167e-06,
+      "loss": 1.244,
+      "step": 912
+    },
+    {
+      "epoch": 0.7422009958337568,
+      "grad_norm": 2.944265127182007,
+      "learning_rate": 3.286544841313126e-06,
+      "loss": 1.2308,
+      "step": 913
+    },
+    {
+      "epoch": 0.7430139213494563,
+      "grad_norm": 3.608762502670288,
+      "learning_rate": 3.2670511741683475e-06,
+      "loss": 1.2018,
+      "step": 914
+    },
+    {
+      "epoch": 0.743826846865156,
+      "grad_norm": 3.958385705947876,
+      "learning_rate": 3.2476041970285945e-06,
+      "loss": 1.2136,
+      "step": 915
+    },
+    {
+      "epoch": 0.7446397723808557,
+      "grad_norm": 2.9133267402648926,
+      "learning_rate": 3.2282040447500063e-06,
+      "loss": 1.2649,
+      "step": 916
+    },
+    {
+      "epoch": 0.7454526978965552,
+      "grad_norm": 3.8698244094848633,
+      "learning_rate": 3.208850851863998e-06,
+      "loss": 1.2265,
+      "step": 917
+    },
+    {
+      "epoch": 0.7462656234122549,
+      "grad_norm": 4.550247669219971,
+      "learning_rate": 3.189544752576369e-06,
+      "loss": 1.2046,
+      "step": 918
+    },
+    {
+      "epoch": 0.7470785489279544,
+      "grad_norm": 2.9886014461517334,
+      "learning_rate": 3.1702858807663175e-06,
+      "loss": 1.2812,
+      "step": 919
+    },
+    {
+      "epoch": 0.7478914744436541,
+      "grad_norm": 3.3736209869384766,
+      "learning_rate": 3.151074369985556e-06,
+      "loss": 1.2482,
+      "step": 920
+    },
+    {
+      "epoch": 0.7487043999593537,
+      "grad_norm": 2.7061290740966797,
+      "learning_rate": 3.131910353457369e-06,
+      "loss": 1.2474,
+      "step": 921
+    },
+    {
+      "epoch": 0.7495173254750533,
+      "grad_norm": 4.058886528015137,
+      "learning_rate": 3.112793964075681e-06,
+      "loss": 1.1897,
+      "step": 922
+    },
+    {
+      "epoch": 0.750330250990753,
+      "grad_norm": 3.3311798572540283,
+      "learning_rate": 3.0937253344041507e-06,
+      "loss": 1.2129,
+      "step": 923
+    },
+    {
+      "epoch": 0.7511431765064526,
+      "grad_norm": 3.2716569900512695,
+      "learning_rate": 3.074704596675242e-06,
+      "loss": 1.1763,
+      "step": 924
+    },
+    {
+      "epoch": 0.7519561020221522,
+      "grad_norm": 3.360356569290161,
+      "learning_rate": 3.055731882789311e-06,
+      "loss": 1.2771,
+      "step": 925
+    },
+    {
+      "epoch": 0.7527690275378518,
+      "grad_norm": 3.9494638442993164,
+      "learning_rate": 3.0368073243136874e-06,
+      "loss": 1.2551,
+      "step": 926
+    },
+    {
+      "epoch": 0.7535819530535515,
+      "grad_norm": 3.3180434703826904,
+      "learning_rate": 3.0179310524817707e-06,
+      "loss": 1.245,
+      "step": 927
+    },
+    {
+      "epoch": 0.7543948785692511,
+      "grad_norm": 4.963752746582031,
+      "learning_rate": 2.9991031981921026e-06,
+      "loss": 1.2266,
+      "step": 928
+    },
+    {
+      "epoch": 0.7552078040849507,
+      "grad_norm": 3.1220555305480957,
+      "learning_rate": 2.9803238920074784e-06,
+      "loss": 1.2057,
+      "step": 929
+    },
+    {
+      "epoch": 0.7560207296006504,
+      "grad_norm": 2.8764801025390625,
+      "learning_rate": 2.961593264154038e-06,
+      "loss": 1.2157,
+      "step": 930
+    },
+    {
+      "epoch": 0.7568336551163499,
+      "grad_norm": 2.682791233062744,
+      "learning_rate": 2.9429114445203423e-06,
+      "loss": 1.1899,
+      "step": 931
+    },
+    {
+      "epoch": 0.7576465806320496,
+      "grad_norm": 5.8080878257751465,
+      "learning_rate": 2.924278562656514e-06,
+      "loss": 1.1661,
+      "step": 932
+    },
+    {
+      "epoch": 0.7584595061477493,
+      "grad_norm": 3.5146303176879883,
+      "learning_rate": 2.90569474777329e-06,
+      "loss": 1.2712,
+      "step": 933
+    },
+    {
+      "epoch": 0.7592724316634488,
+      "grad_norm": 3.092174530029297,
+      "learning_rate": 2.8871601287411634e-06,
+      "loss": 1.2297,
+      "step": 934
+    },
+    {
+      "epoch": 0.7600853571791485,
+      "grad_norm": 2.807847499847412,
+      "learning_rate": 2.8686748340894744e-06,
+      "loss": 1.2369,
+      "step": 935
+    },
+    {
+      "epoch": 0.760898282694848,
+      "grad_norm": 2.8753178119659424,
+      "learning_rate": 2.850238992005514e-06,
+      "loss": 1.2812,
+      "step": 936
+    },
+    {
+      "epoch": 0.7617112082105477,
+      "grad_norm": 4.227181434631348,
+      "learning_rate": 2.8318527303336465e-06,
+      "loss": 1.2143,
+      "step": 937
+    },
+    {
+      "epoch": 0.7625241337262473,
+      "grad_norm": 3.921201229095459,
+      "learning_rate": 2.81351617657442e-06,
+      "loss": 1.2446,
+      "step": 938
+    },
+    {
+      "epoch": 0.763337059241947,
+      "grad_norm": 3.164557695388794,
+      "learning_rate": 2.795229457883678e-06,
+      "loss": 1.2085,
+      "step": 939
+    },
+    {
+      "epoch": 0.7641499847576466,
+      "grad_norm": 3.0904717445373535,
+      "learning_rate": 2.7769927010716814e-06,
+      "loss": 1.2436,
+      "step": 940
+    },
+    {
+      "epoch": 0.7649629102733462,
+      "grad_norm": 9.615850448608398,
+      "learning_rate": 2.7588060326022205e-06,
+      "loss": 1.2179,
+      "step": 941
+    },
+    {
+      "epoch": 0.7657758357890458,
+      "grad_norm": 7.9210357666015625,
+      "learning_rate": 2.740669578591755e-06,
+      "loss": 1.1704,
+      "step": 942
+    },
+    {
+      "epoch": 0.7665887613047454,
+      "grad_norm": 3.03359055519104,
+      "learning_rate": 2.7225834648085282e-06,
+      "loss": 1.1919,
+      "step": 943
+    },
+    {
+      "epoch": 0.7674016868204451,
+      "grad_norm": 3.331894636154175,
+      "learning_rate": 2.7045478166716843e-06,
+      "loss": 1.2297,
+      "step": 944
+    },
+    {
+      "epoch": 0.7682146123361447,
+      "grad_norm": 2.9995782375335693,
+      "learning_rate": 2.6865627592504295e-06,
+      "loss": 1.1936,
+      "step": 945
+    },
+    {
+      "epoch": 0.7690275378518443,
+      "grad_norm": 11.267196655273438,
+      "learning_rate": 2.668628417263137e-06,
+      "loss": 1.2385,
+      "step": 946
+    },
+    {
+      "epoch": 0.769840463367544,
+      "grad_norm": 4.058920383453369,
+      "learning_rate": 2.6507449150764852e-06,
+      "loss": 1.2078,
+      "step": 947
+    },
+    {
+      "epoch": 0.7706533888832435,
+      "grad_norm": 2.8774616718292236,
+      "learning_rate": 2.632912376704607e-06,
+      "loss": 1.2585,
+      "step": 948
+    },
+    {
+      "epoch": 0.7714663143989432,
+      "grad_norm": 3.4053540229797363,
+      "learning_rate": 2.615130925808228e-06,
+      "loss": 1.2739,
+      "step": 949
+    },
+    {
+      "epoch": 0.7722792399146429,
+      "grad_norm": 3.0022501945495605,
+      "learning_rate": 2.597400685693795e-06,
+      "loss": 1.2136,
+      "step": 950
+    },
+    {
+      "epoch": 0.7730921654303424,
+      "grad_norm": 3.6466481685638428,
+      "learning_rate": 2.5797217793126373e-06,
+      "loss": 1.3104,
+      "step": 951
+    },
+    {
+      "epoch": 0.7739050909460421,
+      "grad_norm": 4.021648406982422,
+      "learning_rate": 2.5620943292601074e-06,
+      "loss": 1.2621,
+      "step": 952
+    },
+    {
+      "epoch": 0.7747180164617417,
+      "grad_norm": 2.996817111968994,
+      "learning_rate": 2.5445184577747305e-06,
+      "loss": 1.2194,
+      "step": 953
+    },
+    {
+      "epoch": 0.7755309419774413,
+      "grad_norm": 3.8881189823150635,
+      "learning_rate": 2.52699428673736e-06,
+      "loss": 1.2516,
+      "step": 954
+    },
+    {
+      "epoch": 0.776343867493141,
+      "grad_norm": 3.279557228088379,
+      "learning_rate": 2.5095219376703183e-06,
+      "loss": 1.2116,
+      "step": 955
+    },
+    {
+      "epoch": 0.7771567930088406,
+      "grad_norm": 3.1030569076538086,
+      "learning_rate": 2.4921015317365794e-06,
+      "loss": 1.2902,
+      "step": 956
+    },
+    {
+      "epoch": 0.7779697185245402,
+      "grad_norm": 3.7724967002868652,
+      "learning_rate": 2.4747331897389103e-06,
+      "loss": 1.2783,
+      "step": 957
+    },
+    {
+      "epoch": 0.7787826440402398,
+      "grad_norm": 2.808138132095337,
+      "learning_rate": 2.4574170321190305e-06,
+      "loss": 1.2191,
+      "step": 958
+    },
+    {
+      "epoch": 0.7795955695559394,
+      "grad_norm": 2.6033871173858643,
+      "learning_rate": 2.440153178956798e-06,
+      "loss": 1.2282,
+      "step": 959
+    },
+    {
+      "epoch": 0.780408495071639,
+      "grad_norm": 2.870957612991333,
+      "learning_rate": 2.42294174996935e-06,
+      "loss": 1.2118,
+      "step": 960
+    },
+    {
+      "epoch": 0.7812214205873387,
+      "grad_norm": 2.913543462753296,
+      "learning_rate": 2.40578286451029e-06,
+      "loss": 1.2352,
+      "step": 961
+    },
+    {
+      "epoch": 0.7820343461030383,
+      "grad_norm": 3.7069716453552246,
+      "learning_rate": 2.38867664156886e-06,
+      "loss": 1.2218,
+      "step": 962
+    },
+    {
+      "epoch": 0.7828472716187379,
+      "grad_norm": 4.073693752288818,
+      "learning_rate": 2.3716231997691007e-06,
+      "loss": 1.1997,
+      "step": 963
+    },
+    {
+      "epoch": 0.7836601971344376,
+      "grad_norm": 2.7815756797790527,
+      "learning_rate": 2.3546226573690444e-06,
+      "loss": 1.1898,
+      "step": 964
+    },
+    {
+      "epoch": 0.7844731226501371,
+      "grad_norm": 3.2033910751342773,
+      "learning_rate": 2.3376751322599e-06,
+      "loss": 1.2575,
+      "step": 965
+    },
+    {
+      "epoch": 0.7852860481658368,
+      "grad_norm": 2.805227518081665,
+      "learning_rate": 2.320780741965206e-06,
+      "loss": 1.221,
+      "step": 966
+    },
+    {
+      "epoch": 0.7860989736815365,
+      "grad_norm": 2.747638463973999,
+      "learning_rate": 2.3039396036400463e-06,
+      "loss": 1.2199,
+      "step": 967
+    },
+    {
+      "epoch": 0.786911899197236,
+      "grad_norm": 2.758178234100342,
+      "learning_rate": 2.287151834070226e-06,
+      "loss": 1.1847,
+      "step": 968
+    },
+    {
+      "epoch": 0.7877248247129357,
+      "grad_norm": 3.467595338821411,
+      "learning_rate": 2.2704175496714552e-06,
+      "loss": 1.2456,
+      "step": 969
+    },
+    {
+      "epoch": 0.7885377502286353,
+      "grad_norm": 5.487158298492432,
+      "learning_rate": 2.2537368664885527e-06,
+      "loss": 1.2061,
+      "step": 970
+    },
+    {
+      "epoch": 0.7893506757443349,
+      "grad_norm": 3.063075542449951,
+      "learning_rate": 2.2371099001946385e-06,
+      "loss": 1.264,
+      "step": 971
+    },
+    {
+      "epoch": 0.7901636012600346,
+      "grad_norm": 2.6598317623138428,
+      "learning_rate": 2.2205367660903267e-06,
+      "loss": 1.1971,
+      "step": 972
+    },
+    {
+      "epoch": 0.7909765267757342,
+      "grad_norm": 3.249379873275757,
+      "learning_rate": 2.2040175791029305e-06,
+      "loss": 1.2442,
+      "step": 973
+    },
+    {
+      "epoch": 0.7917894522914338,
+      "grad_norm": 3.2312817573547363,
+      "learning_rate": 2.187552453785662e-06,
+      "loss": 1.1871,
+      "step": 974
+    },
+    {
+      "epoch": 0.7926023778071334,
+      "grad_norm": 3.060171604156494,
+      "learning_rate": 2.1711415043168395e-06,
+      "loss": 1.2198,
+      "step": 975
+    },
+    {
+      "epoch": 0.793415303322833,
+      "grad_norm": 3.2674033641815186,
+      "learning_rate": 2.1547848444991025e-06,
+      "loss": 1.2343,
+      "step": 976
+    },
+    {
+      "epoch": 0.7942282288385327,
+      "grad_norm": 3.822357654571533,
+      "learning_rate": 2.138482587758605e-06,
+      "loss": 1.1876,
+      "step": 977
+    },
+    {
+      "epoch": 0.7950411543542323,
+      "grad_norm": 3.4773342609405518,
+      "learning_rate": 2.1222348471442477e-06,
+      "loss": 1.1976,
+      "step": 978
+    },
+    {
+      "epoch": 0.795854079869932,
+      "grad_norm": 3.8379478454589844,
+      "learning_rate": 2.1060417353268845e-06,
+      "loss": 1.198,
+      "step": 979
+    },
+    {
+      "epoch": 0.7966670053856315,
+      "grad_norm": 4.963233470916748,
+      "learning_rate": 2.0899033645985423e-06,
+      "loss": 1.2991,
+      "step": 980
+    },
+    {
+      "epoch": 0.7974799309013312,
+      "grad_norm": 3.4560701847076416,
+      "learning_rate": 2.073819846871646e-06,
+      "loss": 1.1936,
+      "step": 981
+    },
+    {
+      "epoch": 0.7982928564170307,
+      "grad_norm": 2.69124698638916,
+      "learning_rate": 2.0577912936782317e-06,
+      "loss": 1.1708,
+      "step": 982
+    },
+    {
+      "epoch": 0.7991057819327304,
+      "grad_norm": 2.973618268966675,
+      "learning_rate": 2.041817816169187e-06,
+      "loss": 1.2535,
+      "step": 983
+    },
+    {
+      "epoch": 0.7999187074484301,
+      "grad_norm": 3.1709506511688232,
+      "learning_rate": 2.025899525113474e-06,
+      "loss": 1.2015,
+      "step": 984
+    },
+    {
+      "epoch": 0.8007316329641296,
+      "grad_norm": 2.750272274017334,
+      "learning_rate": 2.010036530897359e-06,
+      "loss": 1.2677,
+      "step": 985
+    },
+    {
+      "epoch": 0.8015445584798293,
+      "grad_norm": 2.7218148708343506,
+      "learning_rate": 1.9942289435236506e-06,
+      "loss": 1.2679,
+      "step": 986
+    },
+    {
+      "epoch": 0.8023574839955289,
+      "grad_norm": 3.0237209796905518,
+      "learning_rate": 1.978476872610939e-06,
+      "loss": 1.2425,
+      "step": 987
+    },
+    {
+      "epoch": 0.8031704095112285,
+      "grad_norm": 4.8593363761901855,
+      "learning_rate": 1.962780427392823e-06,
+      "loss": 1.2754,
+      "step": 988
+    },
+    {
+      "epoch": 0.8039833350269282,
+      "grad_norm": 4.2402544021606445,
+      "learning_rate": 1.9471397167171714e-06,
+      "loss": 1.1841,
+      "step": 989
+    },
+    {
+      "epoch": 0.8047962605426278,
+      "grad_norm": 2.8616418838500977,
+      "learning_rate": 1.931554849045355e-06,
+      "loss": 1.1712,
+      "step": 990
+    },
+    {
+      "epoch": 0.8056091860583274,
+      "grad_norm": 3.0303030014038086,
+      "learning_rate": 1.916025932451493e-06,
+      "loss": 1.2217,
+      "step": 991
+    },
+    {
+      "epoch": 0.806422111574027,
+      "grad_norm": 3.096165180206299,
+      "learning_rate": 1.9005530746217238e-06,
+      "loss": 1.1515,
+      "step": 992
+    },
+    {
+      "epoch": 0.8072350370897267,
+      "grad_norm": 5.142411231994629,
+      "learning_rate": 1.8851363828534253e-06,
+      "loss": 1.167,
+      "step": 993
+    },
+    {
+      "epoch": 0.8080479626054263,
+      "grad_norm": 3.1720876693725586,
+      "learning_rate": 1.869775964054501e-06,
+      "loss": 1.1896,
+      "step": 994
+    },
+    {
+      "epoch": 0.8088608881211259,
+      "grad_norm": 3.833009719848633,
+      "learning_rate": 1.8544719247426224e-06,
+      "loss": 1.2517,
+      "step": 995
+    },
+    {
+      "epoch": 0.8096738136368256,
+      "grad_norm": 3.188974618911743,
+      "learning_rate": 1.8392243710444911e-06,
+      "loss": 1.2795,
+      "step": 996
+    },
+    {
+      "epoch": 0.8104867391525251,
+      "grad_norm": 3.601663589477539,
+      "learning_rate": 1.8240334086951117e-06,
+      "loss": 1.2366,
+      "step": 997
+    },
+    {
+      "epoch": 0.8112996646682248,
+      "grad_norm": 3.1258544921875,
+      "learning_rate": 1.8088991430370506e-06,
+      "loss": 1.2002,
+      "step": 998
+    },
+    {
+      "epoch": 0.8121125901839243,
+      "grad_norm": 2.71299409866333,
+      "learning_rate": 1.7938216790197071e-06,
+      "loss": 1.2609,
+      "step": 999
+    },
+    {
+      "epoch": 0.812925515699624,
+      "grad_norm": 3.2866601943969727,
+      "learning_rate": 1.77880112119859e-06,
+      "loss": 1.2571,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8137384412153237,
+      "grad_norm": 3.1053292751312256,
+      "learning_rate": 1.7638375737345804e-06,
+      "loss": 1.2316,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8145513667310232,
+      "grad_norm": 2.839862823486328,
+      "learning_rate": 1.7489311403932274e-06,
+      "loss": 1.2464,
+      "step": 1002
+    },
+    {
+      "epoch": 0.8153642922467229,
+      "grad_norm": 2.750040292739868,
+      "learning_rate": 1.7340819245440166e-06,
+      "loss": 1.2639,
+      "step": 1003
+    },
+    {
+      "epoch": 0.8161772177624225,
+      "grad_norm": 3.918286085128784,
+      "learning_rate": 1.7192900291596493e-06,
+      "loss": 1.2379,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8169901432781221,
+      "grad_norm": 3.579942226409912,
+      "learning_rate": 1.7045555568153415e-06,
+      "loss": 1.1943,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8178030687938218,
+      "grad_norm": 3.2873690128326416,
+      "learning_rate": 1.6898786096881104e-06,
+      "loss": 1.2457,
+      "step": 1006
+    },
+    {
+      "epoch": 0.8186159943095214,
+      "grad_norm": 2.721126079559326,
+      "learning_rate": 1.6752592895560493e-06,
+      "loss": 1.2681,
+      "step": 1007
+    },
+    {
+      "epoch": 0.819428919825221,
+      "grad_norm": 2.9273929595947266,
+      "learning_rate": 1.6606976977976408e-06,
+      "loss": 1.1985,
+      "step": 1008
+    },
+    {
+      "epoch": 0.8202418453409206,
+      "grad_norm": 3.6816606521606445,
+      "learning_rate": 1.6461939353910494e-06,
+      "loss": 1.2128,
+      "step": 1009
+    },
+    {
+      "epoch": 0.8210547708566203,
+      "grad_norm": 2.8991682529449463,
+      "learning_rate": 1.631748102913412e-06,
+      "loss": 1.224,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8218676963723199,
+      "grad_norm": 3.2517406940460205,
+      "learning_rate": 1.6173603005401505e-06,
+      "loss": 1.1936,
+      "step": 1011
+    },
+    {
+      "epoch": 0.8226806218880195,
+      "grad_norm": 3.0502426624298096,
+      "learning_rate": 1.6030306280442764e-06,
+      "loss": 1.2555,
+      "step": 1012
+    },
+    {
+      "epoch": 0.8234935474037192,
+      "grad_norm": 3.2694664001464844,
+      "learning_rate": 1.588759184795694e-06,
+      "loss": 1.2643,
+      "step": 1013
+    },
+    {
+      "epoch": 0.8243064729194187,
+      "grad_norm": 2.9429259300231934,
+      "learning_rate": 1.574546069760514e-06,
+      "loss": 1.2221,
+      "step": 1014
+    },
+    {
+      "epoch": 0.8251193984351184,
+      "grad_norm": 3.2481369972229004,
+      "learning_rate": 1.5603913815003634e-06,
+      "loss": 1.1949,
+      "step": 1015
+    },
+    {
+      "epoch": 0.8259323239508181,
+      "grad_norm": 3.006603717803955,
+      "learning_rate": 1.5462952181717117e-06,
+      "loss": 1.1593,
+      "step": 1016
+    },
+    {
+      "epoch": 0.8267452494665176,
+      "grad_norm": 2.8126094341278076,
+      "learning_rate": 1.532257677525183e-06,
+      "loss": 1.2094,
+      "step": 1017
+    },
+    {
+      "epoch": 0.8275581749822173,
+      "grad_norm": 3.258910894393921,
+      "learning_rate": 1.5182788569048689e-06,
+      "loss": 1.1524,
+      "step": 1018
+    },
+    {
+      "epoch": 0.8283711004979168,
+      "grad_norm": 3.097121477127075,
+      "learning_rate": 1.5043588532476827e-06,
+      "loss": 1.2063,
+      "step": 1019
+    },
+    {
+      "epoch": 0.8291840260136165,
+      "grad_norm": 3.5429606437683105,
+      "learning_rate": 1.49049776308265e-06,
+      "loss": 1.1579,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8299969515293161,
+      "grad_norm": 3.0676991939544678,
+      "learning_rate": 1.476695682530268e-06,
+      "loss": 1.2063,
+      "step": 1021
+    },
+    {
+      "epoch": 0.8308098770450157,
+      "grad_norm": 3.191493272781372,
+      "learning_rate": 1.4629527073018267e-06,
+      "loss": 1.2724,
+      "step": 1022
+    },
+    {
+      "epoch": 0.8316228025607154,
+      "grad_norm": 4.181521415710449,
+      "learning_rate": 1.449268932698743e-06,
+      "loss": 1.2627,
+      "step": 1023
+    },
+    {
+      "epoch": 0.832435728076415,
+      "grad_norm": 3.7330870628356934,
+      "learning_rate": 1.4356444536119085e-06,
+      "loss": 1.1875,
+      "step": 1024
+    },
+    {
+      "epoch": 0.8332486535921146,
+      "grad_norm": 3.5213124752044678,
+      "learning_rate": 1.422079364521024e-06,
+      "loss": 1.2345,
+      "step": 1025
+    },
+    {
+      "epoch": 0.8340615791078142,
+      "grad_norm": 3.672848701477051,
+      "learning_rate": 1.4085737594939497e-06,
+      "loss": 1.2451,
+      "step": 1026
+    },
+    {
+      "epoch": 0.8348745046235139,
+      "grad_norm": 3.2613043785095215,
+      "learning_rate": 1.3951277321860468e-06,
+      "loss": 1.261,
+      "step": 1027
+    },
+    {
+      "epoch": 0.8356874301392135,
+      "grad_norm": 3.1444427967071533,
+      "learning_rate": 1.381741375839537e-06,
+      "loss": 1.2205,
+      "step": 1028
+    },
+    {
+      "epoch": 0.8365003556549131,
+      "grad_norm": 3.7306652069091797,
+      "learning_rate": 1.3684147832828409e-06,
+      "loss": 1.2343,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8373132811706128,
+      "grad_norm": 3.6698615550994873,
+      "learning_rate": 1.355148046929956e-06,
+      "loss": 1.2195,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8381262066863123,
+      "grad_norm": 4.807132244110107,
+      "learning_rate": 1.3419412587797908e-06,
+      "loss": 1.1946,
+      "step": 1031
+    },
+    {
+      "epoch": 0.838939132202012,
+      "grad_norm": 3.0877437591552734,
+      "learning_rate": 1.3287945104155487e-06,
+      "loss": 1.1901,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8397520577177117,
+      "grad_norm": 6.123032093048096,
+      "learning_rate": 1.3157078930040856e-06,
+      "loss": 1.2338,
+      "step": 1033
+    },
+    {
+      "epoch": 0.8405649832334112,
+      "grad_norm": 3.8207807540893555,
+      "learning_rate": 1.3026814972952674e-06,
+      "loss": 1.2064,
+      "step": 1034
+    },
+    {
+      "epoch": 0.8413779087491109,
+      "grad_norm": 3.591054916381836,
+      "learning_rate": 1.2897154136213542e-06,
+      "loss": 1.248,
+      "step": 1035
+    },
+    {
+      "epoch": 0.8421908342648105,
+      "grad_norm": 3.14103364944458,
+      "learning_rate": 1.2768097318963701e-06,
+      "loss": 1.2247,
+      "step": 1036
+    },
+    {
+      "epoch": 0.8430037597805101,
+      "grad_norm": 3.2605819702148438,
+      "learning_rate": 1.2639645416154744e-06,
+      "loss": 1.2265,
+      "step": 1037
+    },
+    {
+      "epoch": 0.8438166852962098,
+      "grad_norm": 3.2860848903656006,
+      "learning_rate": 1.2511799318543493e-06,
+      "loss": 1.2083,
+      "step": 1038
+    },
+    {
+      "epoch": 0.8446296108119093,
+      "grad_norm": 3.6271586418151855,
+      "learning_rate": 1.2384559912685768e-06,
+      "loss": 1.2562,
+      "step": 1039
+    },
+    {
+      "epoch": 0.845442536327609,
+      "grad_norm": 3.0439271926879883,
+      "learning_rate": 1.2257928080930236e-06,
+      "loss": 1.1838,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8462554618433086,
+      "grad_norm": 2.9285664558410645,
+      "learning_rate": 1.2131904701412345e-06,
+      "loss": 1.2271,
+      "step": 1041
+    },
+    {
+      "epoch": 0.8470683873590082,
+      "grad_norm": 4.422233581542969,
+      "learning_rate": 1.2006490648048118e-06,
+      "loss": 1.2218,
+      "step": 1042
+    },
+    {
+      "epoch": 0.8478813128747078,
+      "grad_norm": 3.193469524383545,
+      "learning_rate": 1.1881686790528279e-06,
+      "loss": 1.2167,
+      "step": 1043
+    },
+    {
+      "epoch": 0.8486942383904075,
+      "grad_norm": 2.9041225910186768,
+      "learning_rate": 1.1757493994312052e-06,
+      "loss": 1.1652,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8495071639061071,
+      "grad_norm": 2.902376890182495,
+      "learning_rate": 1.1633913120621188e-06,
+      "loss": 1.209,
+      "step": 1045
+    },
+    {
+      "epoch": 0.8503200894218067,
+      "grad_norm": 2.7561545372009277,
+      "learning_rate": 1.151094502643414e-06,
+      "loss": 1.2105,
+      "step": 1046
+    },
+    {
+      "epoch": 0.8511330149375064,
+      "grad_norm": 3.4532971382141113,
+      "learning_rate": 1.1388590564479895e-06,
+      "loss": 1.2457,
+      "step": 1047
+    },
+    {
+      "epoch": 0.8519459404532059,
+      "grad_norm": 4.540160179138184,
+      "learning_rate": 1.1266850583232224e-06,
+      "loss": 1.1941,
+      "step": 1048
+    },
+    {
+      "epoch": 0.8527588659689056,
+      "grad_norm": 2.99617075920105,
+      "learning_rate": 1.1145725926903772e-06,
+      "loss": 1.2138,
+      "step": 1049
+    },
+    {
+      "epoch": 0.8535717914846053,
+      "grad_norm": 3.2309064865112305,
+      "learning_rate": 1.1025217435440116e-06,
+      "loss": 1.2373,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8543847170003048,
+      "grad_norm": 2.7454960346221924,
+      "learning_rate": 1.0905325944514034e-06,
+      "loss": 1.2473,
+      "step": 1051
+    },
+    {
+      "epoch": 0.8551976425160045,
+      "grad_norm": 8.090238571166992,
+      "learning_rate": 1.078605228551971e-06,
+      "loss": 1.2342,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8560105680317041,
+      "grad_norm": 3.7213146686553955,
+      "learning_rate": 1.0667397285566893e-06,
+      "loss": 1.2232,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8568234935474037,
+      "grad_norm": 3.4427578449249268,
+      "learning_rate": 1.0549361767475241e-06,
+      "loss": 1.2474,
+      "step": 1054
+    },
+    {
+      "epoch": 0.8576364190631034,
+      "grad_norm": 3.212726593017578,
+      "learning_rate": 1.0431946549768567e-06,
+      "loss": 1.2727,
+      "step": 1055
+    },
+    {
+      "epoch": 0.858449344578803,
+      "grad_norm": 3.895224094390869,
+      "learning_rate": 1.0315152446669142e-06,
+      "loss": 1.2451,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8592622700945026,
+      "grad_norm": 2.8261964321136475,
+      "learning_rate": 1.019898026809214e-06,
+      "loss": 1.2416,
+      "step": 1057
+    },
+    {
+      "epoch": 0.8600751956102022,
+      "grad_norm": 3.2146759033203125,
+      "learning_rate": 1.0083430819639962e-06,
+      "loss": 1.2258,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8608881211259019,
+      "grad_norm": 5.239031791687012,
+      "learning_rate": 9.968504902596566e-07,
+      "loss": 1.2089,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8617010466416014,
+      "grad_norm": 3.1487622261047363,
+      "learning_rate": 9.85420331392214e-07,
+      "loss": 1.2445,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8625139721573011,
+      "grad_norm": 3.5642974376678467,
+      "learning_rate": 9.74052684624731e-07,
+      "loss": 1.2724,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8633268976730007,
+      "grad_norm": 3.3064541816711426,
+      "learning_rate": 9.62747628786782e-07,
+      "loss": 1.235,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8641398231887003,
+      "grad_norm": 2.7583703994750977,
+      "learning_rate": 9.515052422739035e-07,
+      "loss": 1.1864,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8649527487044,
+      "grad_norm": 2.8002755641937256,
+      "learning_rate": 9.403256030470386e-07,
+      "loss": 1.1888,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8657656742200995,
+      "grad_norm": 4.0211710929870605,
+      "learning_rate": 9.292087886320166e-07,
+      "loss": 1.2513,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8665785997357992,
+      "grad_norm": 3.937668561935425,
+      "learning_rate": 9.181548761189996e-07,
+      "loss": 1.2111,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8673915252514989,
+      "grad_norm": 31.291566848754883,
+      "learning_rate": 9.071639421619527e-07,
+      "loss": 1.2234,
+      "step": 1067
+    },
+    {
+      "epoch": 0.8682044507671984,
+      "grad_norm": 4.150018692016602,
+      "learning_rate": 8.962360629781164e-07,
+      "loss": 1.2205,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8690173762828981,
+      "grad_norm": 2.8017213344573975,
+      "learning_rate": 8.853713143474685e-07,
+      "loss": 1.27,
+      "step": 1069
+    },
+    {
+      "epoch": 0.8698303017985977,
+      "grad_norm": 2.9798476696014404,
+      "learning_rate": 8.745697716122081e-07,
+      "loss": 1.2169,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8706432273142973,
+      "grad_norm": 4.344991683959961,
+      "learning_rate": 8.638315096762318e-07,
+      "loss": 1.2217,
+      "step": 1071
+    },
+    {
+      "epoch": 0.871456152829997,
+      "grad_norm": 2.9421257972717285,
+      "learning_rate": 8.531566030046035e-07,
+      "loss": 1.2399,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8722690783456966,
+      "grad_norm": 3.4676921367645264,
+      "learning_rate": 8.425451256230588e-07,
+      "loss": 1.1957,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8730820038613962,
+      "grad_norm": 3.2855141162872314,
+      "learning_rate": 8.319971511174718e-07,
+      "loss": 1.2399,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8738949293770958,
+      "grad_norm": 2.990471839904785,
+      "learning_rate": 8.215127526333499e-07,
+      "loss": 1.2787,
+      "step": 1075
+    },
+    {
+      "epoch": 0.8747078548927955,
+      "grad_norm": 3.183928966522217,
+      "learning_rate": 8.110920028753355e-07,
+      "loss": 1.1831,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8755207804084951,
+      "grad_norm": 2.8277997970581055,
+      "learning_rate": 8.007349741066939e-07,
+      "loss": 1.248,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8763337059241947,
+      "grad_norm": 2.7392983436584473,
+      "learning_rate": 7.904417381488083e-07,
+      "loss": 1.23,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8771466314398944,
+      "grad_norm": 5.617170333862305,
+      "learning_rate": 7.802123663806938e-07,
+      "loss": 1.2267,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8779595569555939,
+      "grad_norm": 2.906653642654419,
+      "learning_rate": 7.700469297384927e-07,
+      "loss": 1.2245,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8787724824712936,
+      "grad_norm": 2.7728428840637207,
+      "learning_rate": 7.599454987149868e-07,
+      "loss": 1.2131,
+      "step": 1081
+    },
+    {
+      "epoch": 0.8795854079869931,
+      "grad_norm": 2.683861017227173,
+      "learning_rate": 7.499081433591071e-07,
+      "loss": 1.1936,
+      "step": 1082
+    },
+    {
+      "epoch": 0.8803983335026928,
+      "grad_norm": 2.6362993717193604,
+      "learning_rate": 7.399349332754458e-07,
+      "loss": 1.2169,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8812112590183925,
+      "grad_norm": 3.3068742752075195,
+      "learning_rate": 7.300259376237795e-07,
+      "loss": 1.2098,
+      "step": 1084
+    },
+    {
+      "epoch": 0.882024184534092,
+      "grad_norm": 2.825416326522827,
+      "learning_rate": 7.201812251185869e-07,
+      "loss": 1.2543,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8828371100497917,
+      "grad_norm": 3.172919750213623,
+      "learning_rate": 7.104008640285642e-07,
+      "loss": 1.1768,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8836500355654913,
+      "grad_norm": 3.052677869796753,
+      "learning_rate": 7.006849221761736e-07,
+      "loss": 1.2068,
+      "step": 1087
+    },
+    {
+      "epoch": 0.8844629610811909,
+      "grad_norm": 2.8510589599609375,
+      "learning_rate": 6.910334669371433e-07,
+      "loss": 1.2043,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8852758865968906,
+      "grad_norm": 3.4369497299194336,
+      "learning_rate": 6.814465652400237e-07,
+      "loss": 1.2467,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8860888121125902,
+      "grad_norm": 2.667567491531372,
+      "learning_rate": 6.719242835657147e-07,
+      "loss": 1.2594,
+      "step": 1090
+    },
+    {
+      "epoch": 0.8869017376282898,
+      "grad_norm": 2.983642816543579,
+      "learning_rate": 6.62466687947001e-07,
+      "loss": 1.2199,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8877146631439894,
+      "grad_norm": 3.583439350128174,
+      "learning_rate": 6.530738439681017e-07,
+      "loss": 1.1827,
+      "step": 1092
+    },
+    {
+      "epoch": 0.8885275886596891,
+      "grad_norm": 4.706247806549072,
+      "learning_rate": 6.437458167642164e-07,
+      "loss": 1.2292,
+      "step": 1093
+    },
+    {
+      "epoch": 0.8893405141753887,
+      "grad_norm": 4.394626140594482,
+      "learning_rate": 6.344826710210584e-07,
+      "loss": 1.2975,
+      "step": 1094
+    },
+    {
+      "epoch": 0.8901534396910883,
+      "grad_norm": 4.5692572593688965,
+      "learning_rate": 6.252844709744255e-07,
+      "loss": 1.1853,
+      "step": 1095
+    },
+    {
+      "epoch": 0.890966365206788,
+      "grad_norm": 3.4114434719085693,
+      "learning_rate": 6.161512804097436e-07,
+      "loss": 1.2067,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8917792907224875,
+      "grad_norm": 7.298144340515137,
+      "learning_rate": 6.070831626616236e-07,
+      "loss": 1.2149,
+      "step": 1097
+    },
+    {
+      "epoch": 0.8925922162381872,
+      "grad_norm": 2.7437572479248047,
+      "learning_rate": 5.980801806134318e-07,
+      "loss": 1.2002,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8934051417538869,
+      "grad_norm": 3.101397752761841,
+      "learning_rate": 5.891423966968413e-07,
+      "loss": 1.2594,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8942180672695864,
+      "grad_norm": 3.186479091644287,
+      "learning_rate": 5.80269872891408e-07,
+      "loss": 1.1895,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8950309927852861,
+      "grad_norm": 3.5605878829956055,
+      "learning_rate": 5.714626707241411e-07,
+      "loss": 1.1804,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8958439183009856,
+      "grad_norm": 3.0213913917541504,
+      "learning_rate": 5.627208512690641e-07,
+      "loss": 1.2619,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8966568438166853,
+      "grad_norm": 3.0476791858673096,
+      "learning_rate": 5.5404447514681e-07,
+      "loss": 1.1429,
+      "step": 1103
+    },
+    {
+      "epoch": 0.8974697693323849,
+      "grad_norm": 2.9802823066711426,
+      "learning_rate": 5.45433602524188e-07,
+      "loss": 1.2353,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8982826948480845,
+      "grad_norm": 3.168029308319092,
+      "learning_rate": 5.368882931137675e-07,
+      "loss": 1.1771,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8990956203637842,
+      "grad_norm": 2.8624963760375977,
+      "learning_rate": 5.284086061734672e-07,
+      "loss": 1.1929,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8999085458794838,
+      "grad_norm": 3.3826193809509277,
+      "learning_rate": 5.199946005061462e-07,
+      "loss": 1.1379,
+      "step": 1107
+    },
+    {
+      "epoch": 0.9007214713951834,
+      "grad_norm": 3.2084782123565674,
+      "learning_rate": 5.116463344591893e-07,
+      "loss": 1.1694,
+      "step": 1108
+    },
+    {
+      "epoch": 0.901534396910883,
+      "grad_norm": 3.6624932289123535,
+      "learning_rate": 5.033638659241102e-07,
+      "loss": 1.219,
+      "step": 1109
+    },
+    {
+      "epoch": 0.9023473224265827,
+      "grad_norm": 3.2314536571502686,
+      "learning_rate": 4.951472523361401e-07,
+      "loss": 1.2457,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9031602479422823,
+      "grad_norm": 3.1179494857788086,
+      "learning_rate": 4.869965506738416e-07,
+      "loss": 1.232,
+      "step": 1111
+    },
+    {
+      "epoch": 0.9039731734579819,
+      "grad_norm": 2.875725030899048,
+      "learning_rate": 4.789118174587071e-07,
+      "loss": 1.2515,
+      "step": 1112
+    },
+    {
+      "epoch": 0.9047860989736816,
+      "grad_norm": 2.5742199420928955,
+      "learning_rate": 4.7089310875475856e-07,
+      "loss": 1.2554,
+      "step": 1113
+    },
+    {
+      "epoch": 0.9055990244893811,
+      "grad_norm": 3.2250759601593018,
+      "learning_rate": 4.6294048016817917e-07,
+      "loss": 1.2281,
+      "step": 1114
+    },
+    {
+      "epoch": 0.9064119500050808,
+      "grad_norm": 2.866562843322754,
+      "learning_rate": 4.550539868469106e-07,
+      "loss": 1.2559,
+      "step": 1115
+    },
+    {
+      "epoch": 0.9072248755207805,
+      "grad_norm": 2.9703938961029053,
+      "learning_rate": 4.4723368348027375e-07,
+      "loss": 1.307,
+      "step": 1116
+    },
+    {
+      "epoch": 0.90803780103648,
+      "grad_norm": 3.0078420639038086,
+      "learning_rate": 4.394796242985933e-07,
+      "loss": 1.2285,
+      "step": 1117
+    },
+    {
+      "epoch": 0.9088507265521797,
+      "grad_norm": 3.0581750869750977,
+      "learning_rate": 4.317918630728235e-07,
+      "loss": 1.1751,
+      "step": 1118
+    },
+    {
+      "epoch": 0.9096636520678792,
+      "grad_norm": 4.224788188934326,
+      "learning_rate": 4.241704531141633e-07,
+      "loss": 1.155,
+      "step": 1119
+    },
+    {
+      "epoch": 0.9104765775835789,
+      "grad_norm": 3.2800920009613037,
+      "learning_rate": 4.166154472737061e-07,
+      "loss": 1.199,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9112895030992785,
+      "grad_norm": 5.579473495483398,
+      "learning_rate": 4.091268979420537e-07,
+      "loss": 1.1558,
+      "step": 1121
+    },
+    {
+      "epoch": 0.9121024286149781,
+      "grad_norm": 3.660987615585327,
+      "learning_rate": 4.0170485704896453e-07,
+      "loss": 1.2258,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9129153541306778,
+      "grad_norm": 11.064430236816406,
+      "learning_rate": 3.943493760629924e-07,
+      "loss": 1.1699,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9137282796463774,
+      "grad_norm": 4.9747138023376465,
+      "learning_rate": 3.8706050599112363e-07,
+      "loss": 1.2415,
+      "step": 1124
+    },
+    {
+      "epoch": 0.914541205162077,
+      "grad_norm": 3.7896888256073,
+      "learning_rate": 3.798382973784298e-07,
+      "loss": 1.2221,
+      "step": 1125
+    },
+    {
+      "epoch": 0.9153541306777766,
+      "grad_norm": 3.383769989013672,
+      "learning_rate": 3.7268280030771655e-07,
+      "loss": 1.196,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9161670561934763,
+      "grad_norm": 3.491272211074829,
+      "learning_rate": 3.655940643991718e-07,
+      "loss": 1.1786,
+      "step": 1127
+    },
+    {
+      "epoch": 0.9169799817091759,
+      "grad_norm": 3.1759097576141357,
+      "learning_rate": 3.585721388100283e-07,
+      "loss": 1.1696,
+      "step": 1128
+    },
+    {
+      "epoch": 0.9177929072248755,
+      "grad_norm": 2.7568089962005615,
+      "learning_rate": 3.516170722342127e-07,
+      "loss": 1.1703,
+      "step": 1129
+    },
+    {
+      "epoch": 0.9186058327405752,
+      "grad_norm": 2.992725372314453,
+      "learning_rate": 3.4472891290201927e-07,
+      "loss": 1.1739,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9194187582562747,
+      "grad_norm": 4.317306041717529,
+      "learning_rate": 3.3790770857976995e-07,
+      "loss": 1.184,
+      "step": 1131
+    },
+    {
+      "epoch": 0.9202316837719744,
+      "grad_norm": 3.9048075675964355,
+      "learning_rate": 3.3115350656948043e-07,
+      "loss": 1.2651,
+      "step": 1132
+    },
+    {
+      "epoch": 0.9210446092876741,
+      "grad_norm": 3.3990674018859863,
+      "learning_rate": 3.2446635370853686e-07,
+      "loss": 1.205,
+      "step": 1133
+    },
+    {
+      "epoch": 0.9218575348033736,
+      "grad_norm": 4.0517754554748535,
+      "learning_rate": 3.1784629636937404e-07,
+      "loss": 1.1996,
+      "step": 1134
+    },
+    {
+      "epoch": 0.9226704603190733,
+      "grad_norm": 3.340564489364624,
+      "learning_rate": 3.1129338045914004e-07,
+      "loss": 1.2215,
+      "step": 1135
+    },
+    {
+      "epoch": 0.9234833858347729,
+      "grad_norm": 3.5760183334350586,
+      "learning_rate": 3.0480765141939316e-07,
+      "loss": 1.2191,
+      "step": 1136
+    },
+    {
+      "epoch": 0.9242963113504725,
+      "grad_norm": 2.8496994972229004,
+      "learning_rate": 2.9838915422578e-07,
+      "loss": 1.2217,
+      "step": 1137
+    },
+    {
+      "epoch": 0.9251092368661722,
+      "grad_norm": 3.025475025177002,
+      "learning_rate": 2.920379333877221e-07,
+      "loss": 1.2332,
+      "step": 1138
+    },
+    {
+      "epoch": 0.9259221623818717,
+      "grad_norm": 4.238699436187744,
+      "learning_rate": 2.8575403294811123e-07,
+      "loss": 1.2223,
+      "step": 1139
+    },
+    {
+      "epoch": 0.9267350878975714,
+      "grad_norm": 2.9650015830993652,
+      "learning_rate": 2.795374964830022e-07,
+      "loss": 1.2149,
+      "step": 1140
+    },
+    {
+      "epoch": 0.927548013413271,
+      "grad_norm": 2.731064796447754,
+      "learning_rate": 2.733883671013082e-07,
+      "loss": 1.2116,
+      "step": 1141
+    },
+    {
+      "epoch": 0.9283609389289706,
+      "grad_norm": 4.153676986694336,
+      "learning_rate": 2.673066874445096e-07,
+      "loss": 1.1189,
+      "step": 1142
+    },
+    {
+      "epoch": 0.9291738644446702,
+      "grad_norm": 3.843541383743286,
+      "learning_rate": 2.612924996863453e-07,
+      "loss": 1.1933,
+      "step": 1143
+    },
+    {
+      "epoch": 0.9299867899603699,
+      "grad_norm": 3.0720019340515137,
+      "learning_rate": 2.5534584553253526e-07,
+      "loss": 1.1859,
+      "step": 1144
+    },
+    {
+      "epoch": 0.9307997154760695,
+      "grad_norm": 3.4368112087249756,
+      "learning_rate": 2.494667662204797e-07,
+      "loss": 1.22,
+      "step": 1145
+    },
+    {
+      "epoch": 0.9316126409917691,
+      "grad_norm": 2.524754285812378,
+      "learning_rate": 2.436553025189758e-07,
+      "loss": 1.2561,
+      "step": 1146
+    },
+    {
+      "epoch": 0.9324255665074688,
+      "grad_norm": 3.2625484466552734,
+      "learning_rate": 2.3791149472794373e-07,
+      "loss": 1.2026,
+      "step": 1147
+    },
+    {
+      "epoch": 0.9332384920231683,
+      "grad_norm": 3.4842891693115234,
+      "learning_rate": 2.3223538267813317e-07,
+      "loss": 1.234,
+      "step": 1148
+    },
+    {
+      "epoch": 0.934051417538868,
+      "grad_norm": 2.9896857738494873,
+      "learning_rate": 2.2662700573085505e-07,
+      "loss": 1.2008,
+      "step": 1149
+    },
+    {
+      "epoch": 0.9348643430545677,
+      "grad_norm": 3.3465092182159424,
+      "learning_rate": 2.2108640277771153e-07,
+      "loss": 1.2392,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9356772685702672,
+      "grad_norm": 2.6980130672454834,
+      "learning_rate": 2.156136122403174e-07,
+      "loss": 1.2083,
+      "step": 1151
+    },
+    {
+      "epoch": 0.9364901940859669,
+      "grad_norm": 3.4942784309387207,
+      "learning_rate": 2.1020867207004026e-07,
+      "loss": 1.2232,
+      "step": 1152
+    },
+    {
+      "epoch": 0.9373031196016665,
+      "grad_norm": 2.874210834503174,
+      "learning_rate": 2.048716197477374e-07,
+      "loss": 1.2447,
+      "step": 1153
+    },
+    {
+      "epoch": 0.9381160451173661,
+      "grad_norm": 3.429757833480835,
+      "learning_rate": 1.996024922834905e-07,
+      "loss": 1.1562,
+      "step": 1154
+    },
+    {
+      "epoch": 0.9389289706330658,
+      "grad_norm": 2.96549654006958,
+      "learning_rate": 1.9440132621635687e-07,
+      "loss": 1.2543,
+      "step": 1155
+    },
+    {
+      "epoch": 0.9397418961487654,
+      "grad_norm": 3.1660540103912354,
+      "learning_rate": 1.8926815761410867e-07,
+      "loss": 1.1931,
+      "step": 1156
+    },
+    {
+      "epoch": 0.940554821664465,
+      "grad_norm": 2.848574161529541,
+      "learning_rate": 1.8420302207298623e-07,
+      "loss": 1.1837,
+      "step": 1157
+    },
+    {
+      "epoch": 0.9413677471801646,
+      "grad_norm": 4.005343437194824,
+      "learning_rate": 1.792059547174507e-07,
+      "loss": 1.2423,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9421806726958643,
+      "grad_norm": 2.7809975147247314,
+      "learning_rate": 1.7427699019994415e-07,
+      "loss": 1.1665,
+      "step": 1159
+    },
+    {
+      "epoch": 0.9429935982115638,
+      "grad_norm": 4.211681365966797,
+      "learning_rate": 1.6941616270063854e-07,
+      "loss": 1.2526,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9438065237272635,
+      "grad_norm": 4.117452144622803,
+      "learning_rate": 1.6462350592721498e-07,
+      "loss": 1.1957,
+      "step": 1161
+    },
+    {
+      "epoch": 0.9446194492429631,
+      "grad_norm": 2.9959964752197266,
+      "learning_rate": 1.5989905311461274e-07,
+      "loss": 1.2342,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9454323747586627,
+      "grad_norm": 3.091280460357666,
+      "learning_rate": 1.5524283702481158e-07,
+      "loss": 1.2168,
+      "step": 1163
+    },
+    {
+      "epoch": 0.9462453002743624,
+      "grad_norm": 4.000481128692627,
+      "learning_rate": 1.5065488994659983e-07,
+      "loss": 1.2206,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9470582257900619,
+      "grad_norm": 3.2974343299865723,
+      "learning_rate": 1.461352436953478e-07,
+      "loss": 1.1955,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9478711513057616,
+      "grad_norm": 3.589606285095215,
+      "learning_rate": 1.4168392961279254e-07,
+      "loss": 1.1277,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9486840768214613,
+      "grad_norm": 3.071859121322632,
+      "learning_rate": 1.3730097856681668e-07,
+      "loss": 1.1837,
+      "step": 1167
+    },
+    {
+      "epoch": 0.9494970023371608,
+      "grad_norm": 3.4584462642669678,
+      "learning_rate": 1.329864209512377e-07,
+      "loss": 1.249,
+      "step": 1168
+    },
+    {
+      "epoch": 0.9503099278528605,
+      "grad_norm": 4.1693434715271,
+      "learning_rate": 1.2874028668559247e-07,
+      "loss": 1.2234,
+      "step": 1169
+    },
+    {
+      "epoch": 0.9511228533685601,
+      "grad_norm": 3.1776278018951416,
+      "learning_rate": 1.245626052149318e-07,
+      "loss": 1.2047,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9519357788842597,
+      "grad_norm": 3.347137689590454,
+      "learning_rate": 1.2045340550961958e-07,
+      "loss": 1.2995,
+      "step": 1171
+    },
+    {
+      "epoch": 0.9527487043999594,
+      "grad_norm": 3.2806451320648193,
+      "learning_rate": 1.164127160651285e-07,
+      "loss": 1.1546,
+      "step": 1172
+    },
+    {
+      "epoch": 0.953561629915659,
+      "grad_norm": 4.498492240905762,
+      "learning_rate": 1.1244056490184008e-07,
+      "loss": 1.2469,
+      "step": 1173
+    },
+    {
+      "epoch": 0.9543745554313586,
+      "grad_norm": 3.0195493698120117,
+      "learning_rate": 1.0853697956485942e-07,
+      "loss": 1.2373,
+      "step": 1174
+    },
+    {
+      "epoch": 0.9551874809470582,
+      "grad_norm": 4.176177501678467,
+      "learning_rate": 1.0470198712381086e-07,
+      "loss": 1.2486,
+      "step": 1175
+    },
+    {
+      "epoch": 0.9560004064627579,
+      "grad_norm": 3.222987413406372,
+      "learning_rate": 1.009356141726614e-07,
+      "loss": 1.1905,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9568133319784575,
+      "grad_norm": 2.6555376052856445,
+      "learning_rate": 9.723788682953539e-08,
+      "loss": 1.1666,
+      "step": 1177
+    },
+    {
+      "epoch": 0.9576262574941571,
+      "grad_norm": 4.015134334564209,
+      "learning_rate": 9.360883073652238e-08,
+      "loss": 1.2675,
+      "step": 1178
+    },
+    {
+      "epoch": 0.9584391830098568,
+      "grad_norm": 3.029994487762451,
+      "learning_rate": 9.004847105951509e-08,
+      "loss": 1.1977,
+      "step": 1179
+    },
+    {
+      "epoch": 0.9592521085255563,
+      "grad_norm": 2.7363007068634033,
+      "learning_rate": 8.655683248802282e-08,
+      "loss": 1.2359,
+      "step": 1180
+    },
+    {
+      "epoch": 0.960065034041256,
+      "grad_norm": 4.360199451446533,
+      "learning_rate": 8.313393923500613e-08,
+      "loss": 1.2099,
+      "step": 1181
+    },
+    {
+      "epoch": 0.9608779595569555,
+      "grad_norm": 2.9082043170928955,
+      "learning_rate": 7.977981503670795e-08,
+      "loss": 1.2632,
+      "step": 1182
+    },
+    {
+      "epoch": 0.9616908850726552,
+      "grad_norm": 3.0049242973327637,
+      "learning_rate": 7.64944831524872e-08,
+      "loss": 1.2128,
+      "step": 1183
+    },
+    {
+      "epoch": 0.9625038105883549,
+      "grad_norm": 2.9180142879486084,
+      "learning_rate": 7.327796636465767e-08,
+      "loss": 1.2075,
+      "step": 1184
+    },
+    {
+      "epoch": 0.9633167361040544,
+      "grad_norm": 2.8545587062835693,
+      "learning_rate": 7.01302869783338e-08,
+      "loss": 1.1809,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9641296616197541,
+      "grad_norm": 3.2359890937805176,
+      "learning_rate": 6.705146682127184e-08,
+      "loss": 1.2404,
+      "step": 1186
+    },
+    {
+      "epoch": 0.9649425871354537,
+      "grad_norm": 7.442730903625488,
+      "learning_rate": 6.404152724371892e-08,
+      "loss": 1.2081,
+      "step": 1187
+    },
+    {
+      "epoch": 0.9657555126511533,
+      "grad_norm": 2.9155330657958984,
+      "learning_rate": 6.110048911826871e-08,
+      "loss": 1.1837,
+      "step": 1188
+    },
+    {
+      "epoch": 0.966568438166853,
+      "grad_norm": 5.689270496368408,
+      "learning_rate": 5.82283728397115e-08,
+      "loss": 1.2039,
+      "step": 1189
+    },
+    {
+      "epoch": 0.9673813636825526,
+      "grad_norm": 2.791161060333252,
+      "learning_rate": 5.542519832489546e-08,
+      "loss": 1.2032,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9681942891982522,
+      "grad_norm": 3.127793312072754,
+      "learning_rate": 5.269098501259007e-08,
+      "loss": 1.2016,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9690072147139518,
+      "grad_norm": 2.8209614753723145,
+      "learning_rate": 5.002575186334735e-08,
+      "loss": 1.1624,
+      "step": 1192
+    },
+    {
+      "epoch": 0.9698201402296515,
+      "grad_norm": 3.3611080646514893,
+      "learning_rate": 4.742951735937418e-08,
+      "loss": 1.2068,
+      "step": 1193
+    },
+    {
+      "epoch": 0.9706330657453511,
+      "grad_norm": 5.118293285369873,
+      "learning_rate": 4.490229950440239e-08,
+      "loss": 1.2398,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9714459912610507,
+      "grad_norm": 9.395883560180664,
+      "learning_rate": 4.2444115823562226e-08,
+      "loss": 1.3143,
+      "step": 1195
+    },
+    {
+      "epoch": 0.9722589167767504,
+      "grad_norm": 3.1017065048217773,
+      "learning_rate": 4.005498336326463e-08,
+      "loss": 1.1918,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9730718422924499,
+      "grad_norm": 3.226966142654419,
+      "learning_rate": 3.773491869108137e-08,
+      "loss": 1.2046,
+      "step": 1197
+    },
+    {
+      "epoch": 0.9738847678081496,
+      "grad_norm": 3.233693838119507,
+      "learning_rate": 3.548393789562732e-08,
+      "loss": 1.2325,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9746976933238493,
+      "grad_norm": 3.159299612045288,
+      "learning_rate": 3.3302056586453916e-08,
+      "loss": 1.1693,
+      "step": 1199
+    },
+    {
+      "epoch": 0.9755106188395488,
+      "grad_norm": 2.7059924602508545,
+      "learning_rate": 3.118928989393699e-08,
+      "loss": 1.2422,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9763235443552485,
+      "grad_norm": 3.511061668395996,
+      "learning_rate": 2.9145652469174666e-08,
+      "loss": 1.2184,
+      "step": 1201
+    },
+    {
+      "epoch": 0.977136469870948,
+      "grad_norm": 4.077070236206055,
+      "learning_rate": 2.7171158483882963e-08,
+      "loss": 1.2309,
+      "step": 1202
+    },
+    {
+      "epoch": 0.9779493953866477,
+      "grad_norm": 3.434537887573242,
+      "learning_rate": 2.5265821630298116e-08,
+      "loss": 1.1943,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9787623209023473,
+      "grad_norm": 3.698641300201416,
+      "learning_rate": 2.3429655121085525e-08,
+      "loss": 1.2671,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9795752464180469,
+      "grad_norm": 6.674719333648682,
+      "learning_rate": 2.1662671689242076e-08,
+      "loss": 1.1961,
+      "step": 1205
+    },
+    {
+      "epoch": 0.9803881719337466,
+      "grad_norm": 4.9146952629089355,
+      "learning_rate": 1.996488358801174e-08,
+      "loss": 1.2345,
+      "step": 1206
+    },
+    {
+      "epoch": 0.9812010974494462,
+      "grad_norm": 2.7147114276885986,
+      "learning_rate": 1.8336302590798992e-08,
+      "loss": 1.2118,
+      "step": 1207
+    },
+    {
+      "epoch": 0.9820140229651458,
+      "grad_norm": 2.809692859649658,
+      "learning_rate": 1.677693999109109e-08,
+      "loss": 1.2162,
+      "step": 1208
+    },
+    {
+      "epoch": 0.9828269484808454,
+      "grad_norm": 3.857846975326538,
+      "learning_rate": 1.5286806602372583e-08,
+      "loss": 1.1792,
+      "step": 1209
+    },
+    {
+      "epoch": 0.9836398739965451,
+      "grad_norm": 3.8911325931549072,
+      "learning_rate": 1.3865912758054267e-08,
+      "loss": 1.2332,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9844527995122447,
+      "grad_norm": 3.5572190284729004,
+      "learning_rate": 1.2514268311405452e-08,
+      "loss": 1.2174,
+      "step": 1211
+    },
+    {
+      "epoch": 0.9852657250279443,
+      "grad_norm": 3.22208833694458,
+      "learning_rate": 1.1231882635477364e-08,
+      "loss": 1.2146,
+      "step": 1212
+    },
+    {
+      "epoch": 0.986078650543644,
+      "grad_norm": 4.469923973083496,
+      "learning_rate": 1.0018764623045407e-08,
+      "loss": 1.2168,
+      "step": 1213
+    },
+    {
+      "epoch": 0.9868915760593435,
+      "grad_norm": 3.1559510231018066,
+      "learning_rate": 8.874922686541442e-09,
+      "loss": 1.2074,
+      "step": 1214
+    },
+    {
+      "epoch": 0.9877045015750432,
+      "grad_norm": 2.6890878677368164,
+      "learning_rate": 7.800364758002721e-09,
+      "loss": 1.2358,
+      "step": 1215
+    },
+    {
+      "epoch": 0.9885174270907429,
+      "grad_norm": 3.4091622829437256,
+      "learning_rate": 6.795098289008595e-09,
+      "loss": 1.2484,
+      "step": 1216
+    },
+    {
+      "epoch": 0.9893303526064424,
+      "grad_norm": 3.0762569904327393,
+      "learning_rate": 5.859130250636113e-09,
+      "loss": 1.1787,
+      "step": 1217
+    },
+    {
+      "epoch": 0.9901432781221421,
+      "grad_norm": 2.616163492202759,
+      "learning_rate": 4.992467133406731e-09,
+      "loss": 1.2092,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9909562036378416,
+      "grad_norm": 3.0248591899871826,
+      "learning_rate": 4.195114947244117e-09,
+      "loss": 1.1998,
+      "step": 1219
+    },
+    {
+      "epoch": 0.9917691291535413,
+      "grad_norm": 5.664068698883057,
+      "learning_rate": 3.4670792214297476e-09,
+      "loss": 1.2539,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9925820546692409,
+      "grad_norm": 3.449087619781494,
+      "learning_rate": 2.808365004569602e-09,
+      "loss": 1.2463,
+      "step": 1221
+    },
+    {
+      "epoch": 0.9933949801849405,
+      "grad_norm": 2.958399534225464,
+      "learning_rate": 2.2189768645519693e-09,
+      "loss": 1.2076,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9942079057006402,
+      "grad_norm": 3.4361188411712646,
+      "learning_rate": 1.6989188885219165e-09,
+      "loss": 1.2436,
+      "step": 1223
+    },
+    {
+      "epoch": 0.9950208312163398,
+      "grad_norm": 3.0529403686523438,
+      "learning_rate": 1.2481946828502011e-09,
+      "loss": 1.1955,
+      "step": 1224
+    },
+    {
+      "epoch": 0.9958337567320394,
+      "grad_norm": 3.090090274810791,
+      "learning_rate": 8.668073731088467e-10,
+      "loss": 1.1455,
+      "step": 1225
+    },
+    {
+      "epoch": 0.996646682247739,
+      "grad_norm": 3.2662580013275146,
+      "learning_rate": 5.547596040489378e-10,
+      "loss": 1.2283,
+      "step": 1226
+    },
+    {
+      "epoch": 0.9974596077634387,
+      "grad_norm": 2.7874884605407715,
+      "learning_rate": 3.1205353958285724e-10,
+      "loss": 1.2011,
+      "step": 1227
+    },
+    {
+      "epoch": 0.9982725332791383,
+      "grad_norm": 2.9483141899108887,
+      "learning_rate": 1.3869086276985243e-10,
+      "loss": 1.272,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9990854587948379,
+      "grad_norm": 3.550588607788086,
+      "learning_rate": 3.467277580271322e-11,
+      "loss": 1.1665,
+      "step": 1229
+    },
+    {
+      "epoch": 0.9998983843105376,
+      "grad_norm": 3.500861406326294,
+      "learning_rate": 0.0,
+      "loss": 1.2382,
+      "step": 1230
+    },
+    {
+      "epoch": 0.9998983843105376,
+      "step": 1230,
+      "total_flos": 3.1215366383127757e+18,
+      "train_loss": 1.3087712280149382,
+      "train_runtime": 25084.8125,
+      "train_samples_per_second": 6.277,
+      "train_steps_per_second": 0.049
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1230,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 7975,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.1215366383127757e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}