diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24653 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 3517,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002843652434581183,
+      "grad_norm": 49.12633514404297,
+      "learning_rate": 0.0,
+      "loss": 13.2624,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005687304869162366,
+      "grad_norm": 52.63550567626953,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 12.8107,
+      "step": 2
+    },
+    {
+      "epoch": 0.000853095730374355,
+      "grad_norm": 51.16219711303711,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 12.7371,
+      "step": 3
+    },
+    {
+      "epoch": 0.0011374609738324733,
+      "grad_norm": 48.81122970581055,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 11.5423,
+      "step": 4
+    },
+    {
+      "epoch": 0.0014218262172905917,
+      "grad_norm": 49.11295700073242,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 11.0671,
+      "step": 5
+    },
+    {
+      "epoch": 0.00170619146074871,
+      "grad_norm": 43.88680648803711,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 10.293,
+      "step": 6
+    },
+    {
+      "epoch": 0.0019905567042068284,
+      "grad_norm": 41.188690185546875,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 9.5132,
+      "step": 7
+    },
+    {
+      "epoch": 0.0022749219476649466,
+      "grad_norm": 34.85954284667969,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 8.9321,
+      "step": 8
+    },
+    {
+      "epoch": 0.002559287191123065,
+      "grad_norm": 40.25858688354492,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 12.101,
+      "step": 9
+    },
+    {
+      "epoch": 0.0028436524345811834,
+      "grad_norm": 37.3657112121582,
+      "learning_rate": 1e-05,
+      "loss": 10.7739,
+      "step": 10
+    },
+    {
+      "epoch": 0.0031280176780393017,
+      "grad_norm": 31.502487182617188,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 10.0982,
+      "step": 11
+    },
+    {
+      "epoch": 0.00341238292149742,
+      "grad_norm": 33.45932388305664,
+      "learning_rate": 1.2222222222222224e-05,
+      "loss": 9.332,
+      "step": 12
+    },
+    {
+      "epoch": 0.0036967481649555385,
+      "grad_norm": 27.317989349365234,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 8.051,
+      "step": 13
+    },
+    {
+      "epoch": 0.003981113408413657,
+      "grad_norm": 19.718740463256836,
+      "learning_rate": 1.4444444444444446e-05,
+      "loss": 7.3297,
+      "step": 14
+    },
+    {
+      "epoch": 0.004265478651871775,
+      "grad_norm": 18.006954193115234,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 6.4547,
+      "step": 15
+    },
+    {
+      "epoch": 0.004549843895329893,
+      "grad_norm": 14.854341506958008,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 6.0132,
+      "step": 16
+    },
+    {
+      "epoch": 0.004834209138788011,
+      "grad_norm": 24.93276596069336,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 9.1995,
+      "step": 17
+    },
+    {
+      "epoch": 0.00511857438224613,
+      "grad_norm": 22.159881591796875,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 8.4067,
+      "step": 18
+    },
+    {
+      "epoch": 0.005402939625704249,
+      "grad_norm": 17.916210174560547,
+      "learning_rate": 2e-05,
+      "loss": 7.7151,
+      "step": 19
+    },
+    {
+      "epoch": 0.005687304869162367,
+      "grad_norm": 15.477251052856445,
+      "learning_rate": 2.1111111111111114e-05,
+      "loss": 7.1676,
+      "step": 20
+    },
+    {
+      "epoch": 0.005971670112620485,
+      "grad_norm": 15.920812606811523,
+      "learning_rate": 2.2222222222222227e-05,
+      "loss": 6.4084,
+      "step": 21
+    },
+    {
+      "epoch": 0.006256035356078603,
+      "grad_norm": 15.113493919372559,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 6.1668,
+      "step": 22
+    },
+    {
+      "epoch": 0.0065404005995367215,
+      "grad_norm": 11.510760307312012,
+      "learning_rate": 2.444444444444445e-05,
+      "loss": 5.4598,
+      "step": 23
+    },
+    {
+      "epoch": 0.00682476584299484,
+      "grad_norm": 7.77994966506958,
+      "learning_rate": 2.5555555555555554e-05,
+      "loss": 5.2792,
+      "step": 24
+    },
+    {
+      "epoch": 0.007109131086452958,
+      "grad_norm": 12.673455238342285,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 7.6026,
+      "step": 25
+    },
+    {
+      "epoch": 0.007393496329911077,
+      "grad_norm": 13.792389869689941,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 7.2058,
+      "step": 26
+    },
+    {
+      "epoch": 0.007677861573369195,
+      "grad_norm": 10.524365425109863,
+      "learning_rate": 2.888888888888889e-05,
+      "loss": 6.3227,
+      "step": 27
+    },
+    {
+      "epoch": 0.007962226816827313,
+      "grad_norm": 105.05674743652344,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 6.1134,
+      "step": 28
+    },
+    {
+      "epoch": 0.008246592060285432,
+      "grad_norm": 13.615738868713379,
+      "learning_rate": 3.111111111111112e-05,
+      "loss": 5.7431,
+      "step": 29
+    },
+    {
+      "epoch": 0.00853095730374355,
+      "grad_norm": 17.219816207885742,
+      "learning_rate": 3.222222222222223e-05,
+      "loss": 5.403,
+      "step": 30
+    },
+    {
+      "epoch": 0.008815322547201668,
+      "grad_norm": 9.333234786987305,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 4.8142,
+      "step": 31
+    },
+    {
+      "epoch": 0.009099687790659786,
+      "grad_norm": 86.27250671386719,
+      "learning_rate": 3.444444444444445e-05,
+      "loss": 4.7489,
+      "step": 32
+    },
+    {
+      "epoch": 0.009384053034117905,
+      "grad_norm": 18.35126304626465,
+      "learning_rate": 3.555555555555555e-05,
+      "loss": 6.8241,
+      "step": 33
+    },
+    {
+      "epoch": 0.009668418277576023,
+      "grad_norm": 22.33354377746582,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 6.552,
+      "step": 34
+    },
+    {
+      "epoch": 0.009952783521034141,
+      "grad_norm": 21.162761688232422,
+      "learning_rate": 3.777777777777778e-05,
+      "loss": 6.3146,
+      "step": 35
+    },
+    {
+      "epoch": 0.01023714876449226,
+      "grad_norm": 18.522459030151367,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 5.7251,
+      "step": 36
+    },
+    {
+      "epoch": 0.010521514007950379,
+      "grad_norm": 19.82053565979004,
+      "learning_rate": 4e-05,
+      "loss": 5.5137,
+      "step": 37
+    },
+    {
+      "epoch": 0.010805879251408497,
+      "grad_norm": 16.579193115234375,
+      "learning_rate": 3.998850904912382e-05,
+      "loss": 5.1559,
+      "step": 38
+    },
+    {
+      "epoch": 0.011090244494866616,
+      "grad_norm": 6.53575325012207,
+      "learning_rate": 3.997701809824763e-05,
+      "loss": 4.8303,
+      "step": 39
+    },
+    {
+      "epoch": 0.011374609738324734,
+      "grad_norm": 16.461362838745117,
+      "learning_rate": 3.996552714737145e-05,
+      "loss": 4.6216,
+      "step": 40
+    },
+    {
+      "epoch": 0.011658974981782852,
+      "grad_norm": 24.829219818115234,
+      "learning_rate": 3.995403619649526e-05,
+      "loss": 6.8729,
+      "step": 41
+    },
+    {
+      "epoch": 0.01194334022524097,
+      "grad_norm": 23.324426651000977,
+      "learning_rate": 3.994254524561908e-05,
+      "loss": 6.3201,
+      "step": 42
+    },
+    {
+      "epoch": 0.012227705468699088,
+      "grad_norm": 20.28281021118164,
+      "learning_rate": 3.993105429474289e-05,
+      "loss": 5.8256,
+      "step": 43
+    },
+    {
+      "epoch": 0.012512070712157207,
+      "grad_norm": 15.08171272277832,
+      "learning_rate": 3.991956334386671e-05,
+      "loss": 5.3353,
+      "step": 44
+    },
+    {
+      "epoch": 0.012796435955615325,
+      "grad_norm": 9.521646499633789,
+      "learning_rate": 3.990807239299052e-05,
+      "loss": 5.1606,
+      "step": 45
+    },
+    {
+      "epoch": 0.013080801199073443,
+      "grad_norm": 14.245501518249512,
+      "learning_rate": 3.989658144211434e-05,
+      "loss": 4.9048,
+      "step": 46
+    },
+    {
+      "epoch": 0.013365166442531561,
+      "grad_norm": 16.89498519897461,
+      "learning_rate": 3.9885090491238155e-05,
+      "loss": 4.4536,
+      "step": 47
+    },
+    {
+      "epoch": 0.01364953168598968,
+      "grad_norm": 11.991838455200195,
+      "learning_rate": 3.9873599540361966e-05,
+      "loss": 4.1927,
+      "step": 48
+    },
+    {
+      "epoch": 0.013933896929447798,
+      "grad_norm": 14.163195610046387,
+      "learning_rate": 3.9862108589485784e-05,
+      "loss": 6.2989,
+      "step": 49
+    },
+    {
+      "epoch": 0.014218262172905916,
+      "grad_norm": 11.865116119384766,
+      "learning_rate": 3.9850617638609595e-05,
+      "loss": 5.6554,
+      "step": 50
+    },
+    {
+      "epoch": 0.014502627416364034,
+      "grad_norm": 8.582511901855469,
+      "learning_rate": 3.983912668773341e-05,
+      "loss": 5.5022,
+      "step": 51
+    },
+    {
+      "epoch": 0.014786992659822154,
+      "grad_norm": 9.730727195739746,
+      "learning_rate": 3.9827635736857224e-05,
+      "loss": 4.9733,
+      "step": 52
+    },
+    {
+      "epoch": 0.015071357903280272,
+      "grad_norm": 11.268597602844238,
+      "learning_rate": 3.981614478598104e-05,
+      "loss": 4.6759,
+      "step": 53
+    },
+    {
+      "epoch": 0.01535572314673839,
+      "grad_norm": 13.982292175292969,
+      "learning_rate": 3.980465383510485e-05,
+      "loss": 4.7606,
+      "step": 54
+    },
+    {
+      "epoch": 0.015640088390196507,
+      "grad_norm": 8.659544944763184,
+      "learning_rate": 3.979316288422867e-05,
+      "loss": 4.394,
+      "step": 55
+    },
+    {
+      "epoch": 0.015924453633654627,
+      "grad_norm": 5.434732437133789,
+      "learning_rate": 3.978167193335249e-05,
+      "loss": 3.9735,
+      "step": 56
+    },
+    {
+      "epoch": 0.016208818877112743,
+      "grad_norm": 5.0773773193359375,
+      "learning_rate": 3.977018098247631e-05,
+      "loss": 5.9472,
+      "step": 57
+    },
+    {
+      "epoch": 0.016493184120570863,
+      "grad_norm": 6.471149444580078,
+      "learning_rate": 3.975869003160012e-05,
+      "loss": 5.2329,
+      "step": 58
+    },
+    {
+      "epoch": 0.016777549364028983,
+      "grad_norm": 8.178523063659668,
+      "learning_rate": 3.9747199080723936e-05,
+      "loss": 5.1182,
+      "step": 59
+    },
+    {
+      "epoch": 0.0170619146074871,
+      "grad_norm": 10.304596900939941,
+      "learning_rate": 3.973570812984775e-05,
+      "loss": 4.8543,
+      "step": 60
+    },
+    {
+      "epoch": 0.01734627985094522,
+      "grad_norm": 9.378915786743164,
+      "learning_rate": 3.9724217178971565e-05,
+      "loss": 4.4329,
+      "step": 61
+    },
+    {
+      "epoch": 0.017630645094403336,
+      "grad_norm": 7.112333297729492,
+      "learning_rate": 3.971272622809538e-05,
+      "loss": 4.1529,
+      "step": 62
+    },
+    {
+      "epoch": 0.017915010337861456,
+      "grad_norm": 6.022056579589844,
+      "learning_rate": 3.9701235277219194e-05,
+      "loss": 4.0711,
+      "step": 63
+    },
+    {
+      "epoch": 0.018199375581319573,
+      "grad_norm": 6.005387783050537,
+      "learning_rate": 3.968974432634301e-05,
+      "loss": 3.7158,
+      "step": 64
+    },
+    {
+      "epoch": 0.018483740824777693,
+      "grad_norm": 11.265110969543457,
+      "learning_rate": 3.967825337546682e-05,
+      "loss": 5.7308,
+      "step": 65
+    },
+    {
+      "epoch": 0.01876810606823581,
+      "grad_norm": 15.99368953704834,
+      "learning_rate": 3.966676242459064e-05,
+      "loss": 5.2962,
+      "step": 66
+    },
+    {
+      "epoch": 0.01905247131169393,
+      "grad_norm": 18.645151138305664,
+      "learning_rate": 3.965527147371445e-05,
+      "loss": 5.2347,
+      "step": 67
+    },
+    {
+      "epoch": 0.019336836555152045,
+      "grad_norm": 16.649398803710938,
+      "learning_rate": 3.964378052283827e-05,
+      "loss": 4.9301,
+      "step": 68
+    },
+    {
+      "epoch": 0.019621201798610165,
+      "grad_norm": 12.267605781555176,
+      "learning_rate": 3.963228957196208e-05,
+      "loss": 4.6981,
+      "step": 69
+    },
+    {
+      "epoch": 0.019905567042068282,
+      "grad_norm": 8.383939743041992,
+      "learning_rate": 3.96207986210859e-05,
+      "loss": 4.4809,
+      "step": 70
+    },
+    {
+      "epoch": 0.020189932285526402,
+      "grad_norm": 8.573983192443848,
+      "learning_rate": 3.9609307670209717e-05,
+      "loss": 3.8532,
+      "step": 71
+    },
+    {
+      "epoch": 0.02047429752898452,
+      "grad_norm": 11.06430435180664,
+      "learning_rate": 3.959781671933353e-05,
+      "loss": 4.1749,
+      "step": 72
+    },
+    {
+      "epoch": 0.02075866277244264,
+      "grad_norm": 10.86804485321045,
+      "learning_rate": 3.9586325768457346e-05,
+      "loss": 5.4128,
+      "step": 73
+    },
+    {
+      "epoch": 0.021043028015900758,
+      "grad_norm": 10.805957794189453,
+      "learning_rate": 3.957483481758116e-05,
+      "loss": 4.873,
+      "step": 74
+    },
+    {
+      "epoch": 0.021327393259358875,
+      "grad_norm": 8.829920768737793,
+      "learning_rate": 3.9563343866704975e-05,
+      "loss": 4.7101,
+      "step": 75
+    },
+    {
+      "epoch": 0.021611758502816995,
+      "grad_norm": 7.231428146362305,
+      "learning_rate": 3.9551852915828786e-05,
+      "loss": 4.696,
+      "step": 76
+    },
+    {
+      "epoch": 0.02189612374627511,
+      "grad_norm": 5.561997413635254,
+      "learning_rate": 3.9540361964952603e-05,
+      "loss": 4.1954,
+      "step": 77
+    },
+    {
+      "epoch": 0.02218048898973323,
+      "grad_norm": 7.825724124908447,
+      "learning_rate": 3.9528871014076415e-05,
+      "loss": 3.9524,
+      "step": 78
+    },
+    {
+      "epoch": 0.022464854233191348,
+      "grad_norm": 7.717831611633301,
+      "learning_rate": 3.951738006320023e-05,
+      "loss": 3.5277,
+      "step": 79
+    },
+    {
+      "epoch": 0.022749219476649468,
+      "grad_norm": 6.9925618171691895,
+      "learning_rate": 3.950588911232405e-05,
+      "loss": 3.3675,
+      "step": 80
+    },
+    {
+      "epoch": 0.023033584720107584,
+      "grad_norm": 6.320919513702393,
+      "learning_rate": 3.949439816144786e-05,
+      "loss": 5.2571,
+      "step": 81
+    },
+    {
+      "epoch": 0.023317949963565704,
+      "grad_norm": 5.361174583435059,
+      "learning_rate": 3.948290721057168e-05,
+      "loss": 4.7185,
+      "step": 82
+    },
+    {
+      "epoch": 0.02360231520702382,
+      "grad_norm": 4.8023362159729,
+      "learning_rate": 3.947141625969549e-05,
+      "loss": 4.4968,
+      "step": 83
+    },
+    {
+      "epoch": 0.02388668045048194,
+      "grad_norm": 6.406538963317871,
+      "learning_rate": 3.945992530881931e-05,
+      "loss": 4.3588,
+      "step": 84
+    },
+    {
+      "epoch": 0.024171045693940057,
+      "grad_norm": 9.289493560791016,
+      "learning_rate": 3.944843435794312e-05,
+      "loss": 4.2662,
+      "step": 85
+    },
+    {
+      "epoch": 0.024455410937398177,
+      "grad_norm": 9.779614448547363,
+      "learning_rate": 3.943694340706694e-05,
+      "loss": 4.169,
+      "step": 86
+    },
+    {
+      "epoch": 0.024739776180856293,
+      "grad_norm": 8.45078182220459,
+      "learning_rate": 3.942545245619075e-05,
+      "loss": 3.7212,
+      "step": 87
+    },
+    {
+      "epoch": 0.025024141424314413,
+      "grad_norm": 7.9592084884643555,
+      "learning_rate": 3.9413961505314566e-05,
+      "loss": 3.7354,
+      "step": 88
+    },
+    {
+      "epoch": 0.025308506667772533,
+      "grad_norm": 6.674322128295898,
+      "learning_rate": 3.9402470554438384e-05,
+      "loss": 4.912,
+      "step": 89
+    },
+    {
+      "epoch": 0.02559287191123065,
+      "grad_norm": 5.126534461975098,
+      "learning_rate": 3.9390979603562195e-05,
+      "loss": 4.7649,
+      "step": 90
+    },
+    {
+      "epoch": 0.02587723715468877,
+      "grad_norm": 7.692286014556885,
+      "learning_rate": 3.937948865268601e-05,
+      "loss": 4.4216,
+      "step": 91
+    },
+    {
+      "epoch": 0.026161602398146886,
+      "grad_norm": 13.10142993927002,
+      "learning_rate": 3.9367997701809824e-05,
+      "loss": 4.4418,
+      "step": 92
+    },
+    {
+      "epoch": 0.026445967641605006,
+      "grad_norm": 13.905843734741211,
+      "learning_rate": 3.935650675093364e-05,
+      "loss": 4.2435,
+      "step": 93
+    },
+    {
+      "epoch": 0.026730332885063122,
+      "grad_norm": 9.100831985473633,
+      "learning_rate": 3.934501580005746e-05,
+      "loss": 3.7519,
+      "step": 94
+    },
+    {
+      "epoch": 0.027014698128521242,
+      "grad_norm": 6.695760726928711,
+      "learning_rate": 3.933352484918128e-05,
+      "loss": 3.3665,
+      "step": 95
+    },
+    {
+      "epoch": 0.02729906337197936,
+      "grad_norm": 3.8529748916625977,
+      "learning_rate": 3.932203389830509e-05,
+      "loss": 3.6497,
+      "step": 96
+    },
+    {
+      "epoch": 0.02758342861543748,
+      "grad_norm": 7.971911907196045,
+      "learning_rate": 3.931054294742891e-05,
+      "loss": 4.8833,
+      "step": 97
+    },
+    {
+      "epoch": 0.027867793858895595,
+      "grad_norm": 9.789183616638184,
+      "learning_rate": 3.929905199655272e-05,
+      "loss": 4.6882,
+      "step": 98
+    },
+    {
+      "epoch": 0.028152159102353715,
+      "grad_norm": 9.128085136413574,
+      "learning_rate": 3.9287561045676536e-05,
+      "loss": 4.4283,
+      "step": 99
+    },
+    {
+      "epoch": 0.028436524345811832,
+      "grad_norm": 7.517631530761719,
+      "learning_rate": 3.927607009480035e-05,
+      "loss": 4.0737,
+      "step": 100
+    },
+    {
+      "epoch": 0.02872088958926995,
+      "grad_norm": 8.111870765686035,
+      "learning_rate": 3.9264579143924165e-05,
+      "loss": 3.9202,
+      "step": 101
+    },
+    {
+      "epoch": 0.029005254832728068,
+      "grad_norm": 6.428286075592041,
+      "learning_rate": 3.9253088193047976e-05,
+      "loss": 3.8114,
+      "step": 102
+    },
+    {
+      "epoch": 0.029289620076186188,
+      "grad_norm": 3.650003671646118,
+      "learning_rate": 3.9241597242171794e-05,
+      "loss": 3.4783,
+      "step": 103
+    },
+    {
+      "epoch": 0.029573985319644308,
+      "grad_norm": 3.3745474815368652,
+      "learning_rate": 3.923010629129561e-05,
+      "loss": 3.2149,
+      "step": 104
+    },
+    {
+      "epoch": 0.029858350563102425,
+      "grad_norm": 6.395346164703369,
+      "learning_rate": 3.921861534041942e-05,
+      "loss": 4.6489,
+      "step": 105
+    },
+    {
+      "epoch": 0.030142715806560545,
+      "grad_norm": 4.820883750915527,
+      "learning_rate": 3.920712438954324e-05,
+      "loss": 4.141,
+      "step": 106
+    },
+    {
+      "epoch": 0.03042708105001866,
+      "grad_norm": 5.000969409942627,
+      "learning_rate": 3.919563343866705e-05,
+      "loss": 4.0283,
+      "step": 107
+    },
+    {
+      "epoch": 0.03071144629347678,
+      "grad_norm": 6.004735469818115,
+      "learning_rate": 3.918414248779087e-05,
+      "loss": 4.0384,
+      "step": 108
+    },
+    {
+      "epoch": 0.030995811536934897,
+      "grad_norm": 4.477464199066162,
+      "learning_rate": 3.917265153691468e-05,
+      "loss": 3.8376,
+      "step": 109
+    },
+    {
+      "epoch": 0.031280176780393014,
+      "grad_norm": 3.1061179637908936,
+      "learning_rate": 3.91611605860385e-05,
+      "loss": 3.7448,
+      "step": 110
+    },
+    {
+      "epoch": 0.03156454202385114,
+      "grad_norm": 4.4171600341796875,
+      "learning_rate": 3.914966963516231e-05,
+      "loss": 3.4045,
+      "step": 111
+    },
+    {
+      "epoch": 0.031848907267309254,
+      "grad_norm": 5.526491641998291,
+      "learning_rate": 3.913817868428613e-05,
+      "loss": 3.2656,
+      "step": 112
+    },
+    {
+      "epoch": 0.03213327251076737,
+      "grad_norm": 8.430256843566895,
+      "learning_rate": 3.9126687733409946e-05,
+      "loss": 4.759,
+      "step": 113
+    },
+    {
+      "epoch": 0.03241763775422549,
+      "grad_norm": 7.523831367492676,
+      "learning_rate": 3.911519678253376e-05,
+      "loss": 4.3822,
+      "step": 114
+    },
+    {
+      "epoch": 0.03270200299768361,
+      "grad_norm": 6.166808605194092,
+      "learning_rate": 3.9103705831657575e-05,
+      "loss": 3.8537,
+      "step": 115
+    },
+    {
+      "epoch": 0.03298636824114173,
+      "grad_norm": 4.141753196716309,
+      "learning_rate": 3.9092214880781386e-05,
+      "loss": 3.9163,
+      "step": 116
+    },
+    {
+      "epoch": 0.03327073348459984,
+      "grad_norm": 4.088106632232666,
+      "learning_rate": 3.9080723929905204e-05,
+      "loss": 3.7164,
+      "step": 117
+    },
+    {
+      "epoch": 0.03355509872805797,
+      "grad_norm": 7.6691107749938965,
+      "learning_rate": 3.9069232979029015e-05,
+      "loss": 3.4528,
+      "step": 118
+    },
+    {
+      "epoch": 0.03383946397151608,
+      "grad_norm": 8.339359283447266,
+      "learning_rate": 3.905774202815283e-05,
+      "loss": 3.0847,
+      "step": 119
+    },
+    {
+      "epoch": 0.0341238292149742,
+      "grad_norm": 5.894288063049316,
+      "learning_rate": 3.9046251077276644e-05,
+      "loss": 3.1057,
+      "step": 120
+    },
+    {
+      "epoch": 0.034408194458432316,
+      "grad_norm": 4.960434436798096,
+      "learning_rate": 3.903476012640046e-05,
+      "loss": 4.8553,
+      "step": 121
+    },
+    {
+      "epoch": 0.03469255970189044,
+      "grad_norm": 4.705537796020508,
+      "learning_rate": 3.902326917552428e-05,
+      "loss": 4.2162,
+      "step": 122
+    },
+    {
+      "epoch": 0.034976924945348556,
+      "grad_norm": 3.077615737915039,
+      "learning_rate": 3.901177822464809e-05,
+      "loss": 4.2706,
+      "step": 123
+    },
+    {
+      "epoch": 0.03526129018880667,
+      "grad_norm": 3.1837899684906006,
+      "learning_rate": 3.900028727377191e-05,
+      "loss": 3.9014,
+      "step": 124
+    },
+    {
+      "epoch": 0.03554565543226479,
+      "grad_norm": 6.796784400939941,
+      "learning_rate": 3.898879632289572e-05,
+      "loss": 3.7483,
+      "step": 125
+    },
+    {
+      "epoch": 0.03583002067572291,
+      "grad_norm": 6.224754810333252,
+      "learning_rate": 3.897730537201954e-05,
+      "loss": 3.6634,
+      "step": 126
+    },
+    {
+      "epoch": 0.03611438591918103,
+      "grad_norm": 4.740405559539795,
+      "learning_rate": 3.896581442114335e-05,
+      "loss": 3.378,
+      "step": 127
+    },
+    {
+      "epoch": 0.036398751162639145,
+      "grad_norm": 4.660193920135498,
+      "learning_rate": 3.8954323470267167e-05,
+      "loss": 3.2241,
+      "step": 128
+    },
+    {
+      "epoch": 0.03668311640609726,
+      "grad_norm": 6.38066291809082,
+      "learning_rate": 3.894283251939098e-05,
+      "loss": 4.5271,
+      "step": 129
+    },
+    {
+      "epoch": 0.036967481649555385,
+      "grad_norm": 4.577174186706543,
+      "learning_rate": 3.89313415685148e-05,
+      "loss": 4.356,
+      "step": 130
+    },
+    {
+      "epoch": 0.0372518468930135,
+      "grad_norm": 4.075042247772217,
+      "learning_rate": 3.8919850617638613e-05,
+      "loss": 3.8877,
+      "step": 131
+    },
+    {
+      "epoch": 0.03753621213647162,
+      "grad_norm": 5.604481220245361,
+      "learning_rate": 3.890835966676243e-05,
+      "loss": 3.8834,
+      "step": 132
+    },
+    {
+      "epoch": 0.03782057737992974,
+      "grad_norm": 5.59628438949585,
+      "learning_rate": 3.889686871588624e-05,
+      "loss": 3.7806,
+      "step": 133
+    },
+    {
+      "epoch": 0.03810494262338786,
+      "grad_norm": 5.5282206535339355,
+      "learning_rate": 3.888537776501006e-05,
+      "loss": 3.4985,
+      "step": 134
+    },
+    {
+      "epoch": 0.038389307866845974,
+      "grad_norm": 5.6480865478515625,
+      "learning_rate": 3.887388681413387e-05,
+      "loss": 3.4003,
+      "step": 135
+    },
+    {
+      "epoch": 0.03867367311030409,
+      "grad_norm": 3.6527063846588135,
+      "learning_rate": 3.886239586325769e-05,
+      "loss": 2.9528,
+      "step": 136
+    },
+    {
+      "epoch": 0.038958038353762214,
+      "grad_norm": 4.994736671447754,
+      "learning_rate": 3.885090491238151e-05,
+      "loss": 4.6074,
+      "step": 137
+    },
+    {
+      "epoch": 0.03924240359722033,
+      "grad_norm": 5.265069484710693,
+      "learning_rate": 3.883941396150532e-05,
+      "loss": 3.9333,
+      "step": 138
+    },
+    {
+      "epoch": 0.03952676884067845,
+      "grad_norm": 5.489358901977539,
+      "learning_rate": 3.8827923010629136e-05,
+      "loss": 3.9805,
+      "step": 139
+    },
+    {
+      "epoch": 0.039811134084136564,
+      "grad_norm": 5.382353782653809,
+      "learning_rate": 3.881643205975295e-05,
+      "loss": 3.9738,
+      "step": 140
+    },
+    {
+      "epoch": 0.04009549932759469,
+      "grad_norm": 6.055698394775391,
+      "learning_rate": 3.8804941108876765e-05,
+      "loss": 3.621,
+      "step": 141
+    },
+    {
+      "epoch": 0.040379864571052804,
+      "grad_norm": 5.292023181915283,
+      "learning_rate": 3.8793450158000576e-05,
+      "loss": 3.5441,
+      "step": 142
+    },
+    {
+      "epoch": 0.04066422981451092,
+      "grad_norm": 3.348857879638672,
+      "learning_rate": 3.8781959207124394e-05,
+      "loss": 2.9929,
+      "step": 143
+    },
+    {
+      "epoch": 0.04094859505796904,
+      "grad_norm": 3.1222856044769287,
+      "learning_rate": 3.8770468256248205e-05,
+      "loss": 2.8722,
+      "step": 144
+    },
+    {
+      "epoch": 0.04123296030142716,
+      "grad_norm": 8.240327835083008,
+      "learning_rate": 3.875897730537202e-05,
+      "loss": 4.6144,
+      "step": 145
+    },
+    {
+      "epoch": 0.04151732554488528,
+      "grad_norm": 7.63242769241333,
+      "learning_rate": 3.874748635449584e-05,
+      "loss": 4.1743,
+      "step": 146
+    },
+    {
+      "epoch": 0.04180169078834339,
+      "grad_norm": 6.464611530303955,
+      "learning_rate": 3.873599540361965e-05,
+      "loss": 3.898,
+      "step": 147
+    },
+    {
+      "epoch": 0.042086056031801516,
+      "grad_norm": 6.321760177612305,
+      "learning_rate": 3.872450445274347e-05,
+      "loss": 3.8015,
+      "step": 148
+    },
+    {
+      "epoch": 0.04237042127525963,
+      "grad_norm": 4.688291549682617,
+      "learning_rate": 3.871301350186728e-05,
+      "loss": 3.5913,
+      "step": 149
+    },
+    {
+      "epoch": 0.04265478651871775,
+      "grad_norm": 3.5380592346191406,
+      "learning_rate": 3.87015225509911e-05,
+      "loss": 3.2963,
+      "step": 150
+    },
+    {
+      "epoch": 0.042939151762175866,
+      "grad_norm": 3.186795473098755,
+      "learning_rate": 3.869003160011491e-05,
+      "loss": 3.1665,
+      "step": 151
+    },
+    {
+      "epoch": 0.04322351700563399,
+      "grad_norm": 3.351107597351074,
+      "learning_rate": 3.867854064923873e-05,
+      "loss": 3.0804,
+      "step": 152
+    },
+    {
+      "epoch": 0.043507882249092106,
+      "grad_norm": 6.562431812286377,
+      "learning_rate": 3.866704969836254e-05,
+      "loss": 4.5577,
+      "step": 153
+    },
+    {
+      "epoch": 0.04379224749255022,
+      "grad_norm": 5.137665748596191,
+      "learning_rate": 3.865555874748636e-05,
+      "loss": 3.9937,
+      "step": 154
+    },
+    {
+      "epoch": 0.04407661273600834,
+      "grad_norm": 4.792855262756348,
+      "learning_rate": 3.8644067796610175e-05,
+      "loss": 3.6608,
+      "step": 155
+    },
+    {
+      "epoch": 0.04436097797946646,
+      "grad_norm": 4.835788726806641,
+      "learning_rate": 3.8632576845733986e-05,
+      "loss": 3.7309,
+      "step": 156
+    },
+    {
+      "epoch": 0.04464534322292458,
+      "grad_norm": 4.597768306732178,
+      "learning_rate": 3.8621085894857804e-05,
+      "loss": 3.348,
+      "step": 157
+    },
+    {
+      "epoch": 0.044929708466382695,
+      "grad_norm": 4.271462440490723,
+      "learning_rate": 3.8609594943981615e-05,
+      "loss": 3.3393,
+      "step": 158
+    },
+    {
+      "epoch": 0.04521407370984081,
+      "grad_norm": 3.781461477279663,
+      "learning_rate": 3.859810399310543e-05,
+      "loss": 3.0811,
+      "step": 159
+    },
+    {
+      "epoch": 0.045498438953298935,
+      "grad_norm": 3.5072243213653564,
+      "learning_rate": 3.8586613042229244e-05,
+      "loss": 2.8874,
+      "step": 160
+    },
+    {
+      "epoch": 0.04578280419675705,
+      "grad_norm": 4.078586101531982,
+      "learning_rate": 3.857512209135306e-05,
+      "loss": 4.3578,
+      "step": 161
+    },
+    {
+      "epoch": 0.04606716944021517,
+      "grad_norm": 4.5630784034729,
+      "learning_rate": 3.856363114047687e-05,
+      "loss": 4.0361,
+      "step": 162
+    },
+    {
+      "epoch": 0.04635153468367329,
+      "grad_norm": 4.149646282196045,
+      "learning_rate": 3.855214018960069e-05,
+      "loss": 3.7167,
+      "step": 163
+    },
+    {
+      "epoch": 0.04663589992713141,
+      "grad_norm": 3.7590112686157227,
+      "learning_rate": 3.854064923872451e-05,
+      "loss": 3.3838,
+      "step": 164
+    },
+    {
+      "epoch": 0.046920265170589524,
+      "grad_norm": 4.184499263763428,
+      "learning_rate": 3.852915828784832e-05,
+      "loss": 3.5132,
+      "step": 165
+    },
+    {
+      "epoch": 0.04720463041404764,
+      "grad_norm": 4.422394275665283,
+      "learning_rate": 3.851766733697214e-05,
+      "loss": 3.2853,
+      "step": 166
+    },
+    {
+      "epoch": 0.047488995657505764,
+      "grad_norm": 4.050603866577148,
+      "learning_rate": 3.8506176386095956e-05,
+      "loss": 2.9437,
+      "step": 167
+    },
+    {
+      "epoch": 0.04777336090096388,
+      "grad_norm": 4.047832489013672,
+      "learning_rate": 3.849468543521977e-05,
+      "loss": 2.949,
+      "step": 168
+    },
+    {
+      "epoch": 0.048057726144422,
+      "grad_norm": 4.343886375427246,
+      "learning_rate": 3.8483194484343585e-05,
+      "loss": 4.5678,
+      "step": 169
+    },
+    {
+      "epoch": 0.048342091387880114,
+      "grad_norm": 3.9210119247436523,
+      "learning_rate": 3.84717035334674e-05,
+      "loss": 3.8641,
+      "step": 170
+    },
+    {
+      "epoch": 0.04862645663133824,
+      "grad_norm": 3.6741321086883545,
+      "learning_rate": 3.8460212582591214e-05,
+      "loss": 3.498,
+      "step": 171
+    },
+    {
+      "epoch": 0.048910821874796354,
+      "grad_norm": 3.9248976707458496,
+      "learning_rate": 3.844872163171503e-05,
+      "loss": 3.6691,
+      "step": 172
+    },
+    {
+      "epoch": 0.04919518711825447,
+      "grad_norm": 4.195116996765137,
+      "learning_rate": 3.843723068083884e-05,
+      "loss": 3.4649,
+      "step": 173
+    },
+    {
+      "epoch": 0.049479552361712587,
+      "grad_norm": 4.228196144104004,
+      "learning_rate": 3.842573972996266e-05,
+      "loss": 3.1212,
+      "step": 174
+    },
+    {
+      "epoch": 0.04976391760517071,
+      "grad_norm": 4.049994468688965,
+      "learning_rate": 3.841424877908647e-05,
+      "loss": 3.144,
+      "step": 175
+    },
+    {
+      "epoch": 0.050048282848628826,
+      "grad_norm": 3.637624502182007,
+      "learning_rate": 3.840275782821029e-05,
+      "loss": 2.8251,
+      "step": 176
+    },
+    {
+      "epoch": 0.05033264809208694,
+      "grad_norm": 3.211629867553711,
+      "learning_rate": 3.83912668773341e-05,
+      "loss": 4.115,
+      "step": 177
+    },
+    {
+      "epoch": 0.050617013335545066,
+      "grad_norm": 3.2752835750579834,
+      "learning_rate": 3.837977592645792e-05,
+      "loss": 3.8892,
+      "step": 178
+    },
+    {
+      "epoch": 0.05090137857900318,
+      "grad_norm": 3.31851863861084,
+      "learning_rate": 3.8368284975581736e-05,
+      "loss": 3.4402,
+      "step": 179
+    },
+    {
+      "epoch": 0.0511857438224613,
+      "grad_norm": 3.9213197231292725,
+      "learning_rate": 3.835679402470555e-05,
+      "loss": 3.3855,
+      "step": 180
+    },
+    {
+      "epoch": 0.051470109065919416,
+      "grad_norm": 4.550943851470947,
+      "learning_rate": 3.8345303073829365e-05,
+      "loss": 3.3385,
+      "step": 181
+    },
+    {
+      "epoch": 0.05175447430937754,
+      "grad_norm": 3.618424654006958,
+      "learning_rate": 3.8333812122953177e-05,
+      "loss": 3.0774,
+      "step": 182
+    },
+    {
+      "epoch": 0.052038839552835656,
+      "grad_norm": 2.8778865337371826,
+      "learning_rate": 3.8322321172076994e-05,
+      "loss": 2.9035,
+      "step": 183
+    },
+    {
+      "epoch": 0.05232320479629377,
+      "grad_norm": 2.7407846450805664,
+      "learning_rate": 3.8310830221200806e-05,
+      "loss": 2.8378,
+      "step": 184
+    },
+    {
+      "epoch": 0.05260757003975189,
+      "grad_norm": 3.778500556945801,
+      "learning_rate": 3.8299339270324623e-05,
+      "loss": 4.2712,
+      "step": 185
+    },
+    {
+      "epoch": 0.05289193528321001,
+      "grad_norm": 4.52739143371582,
+      "learning_rate": 3.8287848319448435e-05,
+      "loss": 3.698,
+      "step": 186
+    },
+    {
+      "epoch": 0.05317630052666813,
+      "grad_norm": 4.739940166473389,
+      "learning_rate": 3.827635736857225e-05,
+      "loss": 3.7459,
+      "step": 187
+    },
+    {
+      "epoch": 0.053460665770126245,
+      "grad_norm": 4.952041149139404,
+      "learning_rate": 3.826486641769607e-05,
+      "loss": 3.355,
+      "step": 188
+    },
+    {
+      "epoch": 0.05374503101358436,
+      "grad_norm": 3.8098363876342773,
+      "learning_rate": 3.825337546681988e-05,
+      "loss": 3.2634,
+      "step": 189
+    },
+    {
+      "epoch": 0.054029396257042485,
+      "grad_norm": 3.0846543312072754,
+      "learning_rate": 3.82418845159437e-05,
+      "loss": 3.0904,
+      "step": 190
+    },
+    {
+      "epoch": 0.0543137615005006,
+      "grad_norm": 2.909727096557617,
+      "learning_rate": 3.823039356506751e-05,
+      "loss": 2.9777,
+      "step": 191
+    },
+    {
+      "epoch": 0.05459812674395872,
+      "grad_norm": 3.1065726280212402,
+      "learning_rate": 3.821890261419133e-05,
+      "loss": 2.7359,
+      "step": 192
+    },
+    {
+      "epoch": 0.05488249198741684,
+      "grad_norm": 7.011332988739014,
+      "learning_rate": 3.820741166331514e-05,
+      "loss": 4.007,
+      "step": 193
+    },
+    {
+      "epoch": 0.05516685723087496,
+      "grad_norm": 5.220023155212402,
+      "learning_rate": 3.819592071243896e-05,
+      "loss": 3.6895,
+      "step": 194
+    },
+    {
+      "epoch": 0.055451222474333074,
+      "grad_norm": 4.016128063201904,
+      "learning_rate": 3.818442976156277e-05,
+      "loss": 3.6332,
+      "step": 195
+    },
+    {
+      "epoch": 0.05573558771779119,
+      "grad_norm": 3.669994831085205,
+      "learning_rate": 3.8172938810686586e-05,
+      "loss": 3.395,
+      "step": 196
+    },
+    {
+      "epoch": 0.056019952961249314,
+      "grad_norm": 3.2430341243743896,
+      "learning_rate": 3.8161447859810404e-05,
+      "loss": 3.4718,
+      "step": 197
+    },
+    {
+      "epoch": 0.05630431820470743,
+      "grad_norm": 2.657214403152466,
+      "learning_rate": 3.8149956908934215e-05,
+      "loss": 3.0473,
+      "step": 198
+    },
+    {
+      "epoch": 0.05658868344816555,
+      "grad_norm": 2.9551098346710205,
+      "learning_rate": 3.813846595805803e-05,
+      "loss": 2.7037,
+      "step": 199
+    },
+    {
+      "epoch": 0.056873048691623664,
+      "grad_norm": 2.8400704860687256,
+      "learning_rate": 3.8126975007181844e-05,
+      "loss": 2.6874,
+      "step": 200
+    },
+    {
+      "epoch": 0.05715741393508179,
+      "grad_norm": 3.2741611003875732,
+      "learning_rate": 3.811548405630566e-05,
+      "loss": 4.1424,
+      "step": 201
+    },
+    {
+      "epoch": 0.0574417791785399,
+      "grad_norm": 3.464838981628418,
+      "learning_rate": 3.810399310542947e-05,
+      "loss": 3.7012,
+      "step": 202
+    },
+    {
+      "epoch": 0.05772614442199802,
+      "grad_norm": 2.831378698348999,
+      "learning_rate": 3.809250215455329e-05,
+      "loss": 3.4572,
+      "step": 203
+    },
+    {
+      "epoch": 0.058010509665456136,
+      "grad_norm": 2.67508864402771,
+      "learning_rate": 3.808101120367711e-05,
+      "loss": 3.5624,
+      "step": 204
+    },
+    {
+      "epoch": 0.05829487490891426,
+      "grad_norm": 2.527092218399048,
+      "learning_rate": 3.806952025280093e-05,
+      "loss": 3.2684,
+      "step": 205
+    },
+    {
+      "epoch": 0.058579240152372376,
+      "grad_norm": 3.604074478149414,
+      "learning_rate": 3.805802930192474e-05,
+      "loss": 3.1068,
+      "step": 206
+    },
+    {
+      "epoch": 0.05886360539583049,
+      "grad_norm": 3.566804885864258,
+      "learning_rate": 3.8046538351048556e-05,
+      "loss": 2.9472,
+      "step": 207
+    },
+    {
+      "epoch": 0.059147970639288616,
+      "grad_norm": 3.518846273422241,
+      "learning_rate": 3.803504740017237e-05,
+      "loss": 2.7991,
+      "step": 208
+    },
+    {
+      "epoch": 0.05943233588274673,
+      "grad_norm": 5.438695430755615,
+      "learning_rate": 3.8023556449296185e-05,
+      "loss": 4.3441,
+      "step": 209
+    },
+    {
+      "epoch": 0.05971670112620485,
+      "grad_norm": 4.187833786010742,
+      "learning_rate": 3.8012065498419996e-05,
+      "loss": 3.7446,
+      "step": 210
+    },
+    {
+      "epoch": 0.060001066369662966,
+      "grad_norm": 2.9121296405792236,
+      "learning_rate": 3.8000574547543814e-05,
+      "loss": 3.4443,
+      "step": 211
+    },
+    {
+      "epoch": 0.06028543161312109,
+      "grad_norm": 3.0015816688537598,
+      "learning_rate": 3.798908359666763e-05,
+      "loss": 3.2636,
+      "step": 212
+    },
+    {
+      "epoch": 0.060569796856579206,
+      "grad_norm": 4.640799045562744,
+      "learning_rate": 3.797759264579144e-05,
+      "loss": 3.5289,
+      "step": 213
+    },
+    {
+      "epoch": 0.06085416210003732,
+      "grad_norm": 4.131928443908691,
+      "learning_rate": 3.796610169491526e-05,
+      "loss": 3.0966,
+      "step": 214
+    },
+    {
+      "epoch": 0.06113852734349544,
+      "grad_norm": 3.303802251815796,
+      "learning_rate": 3.795461074403907e-05,
+      "loss": 2.6751,
+      "step": 215
+    },
+    {
+      "epoch": 0.06142289258695356,
+      "grad_norm": 3.197770595550537,
+      "learning_rate": 3.794311979316289e-05,
+      "loss": 2.8337,
+      "step": 216
+    },
+    {
+      "epoch": 0.06170725783041168,
+      "grad_norm": 3.2084145545959473,
+      "learning_rate": 3.79316288422867e-05,
+      "loss": 3.9611,
+      "step": 217
+    },
+    {
+      "epoch": 0.061991623073869795,
+      "grad_norm": 2.7938461303710938,
+      "learning_rate": 3.792013789141052e-05,
+      "loss": 3.6372,
+      "step": 218
+    },
+    {
+      "epoch": 0.06227598831732791,
+      "grad_norm": 3.456489324569702,
+      "learning_rate": 3.790864694053433e-05,
+      "loss": 3.4849,
+      "step": 219
+    },
+    {
+      "epoch": 0.06256035356078603,
+      "grad_norm": 4.275204658508301,
+      "learning_rate": 3.789715598965815e-05,
+      "loss": 3.2791,
+      "step": 220
+    },
+    {
+      "epoch": 0.06284471880424415,
+      "grad_norm": 3.960286855697632,
+      "learning_rate": 3.7885665038781966e-05,
+      "loss": 3.3154,
+      "step": 221
+    },
+    {
+      "epoch": 0.06312908404770227,
+      "grad_norm": 3.113219738006592,
+      "learning_rate": 3.787417408790578e-05,
+      "loss": 3.0051,
+      "step": 222
+    },
+    {
+      "epoch": 0.06341344929116038,
+      "grad_norm": 2.509889841079712,
+      "learning_rate": 3.7862683137029595e-05,
+      "loss": 2.7279,
+      "step": 223
+    },
+    {
+      "epoch": 0.06369781453461851,
+      "grad_norm": 2.478565216064453,
+      "learning_rate": 3.7851192186153406e-05,
+      "loss": 2.6428,
+      "step": 224
+    },
+    {
+      "epoch": 0.06398217977807663,
+      "grad_norm": 4.031311988830566,
+      "learning_rate": 3.7839701235277224e-05,
+      "loss": 3.8377,
+      "step": 225
+    },
+    {
+      "epoch": 0.06426654502153474,
+      "grad_norm": 4.466932773590088,
+      "learning_rate": 3.7828210284401035e-05,
+      "loss": 3.516,
+      "step": 226
+    },
+    {
+      "epoch": 0.06455091026499286,
+      "grad_norm": 4.715795040130615,
+      "learning_rate": 3.781671933352485e-05,
+      "loss": 3.545,
+      "step": 227
+    },
+    {
+      "epoch": 0.06483527550845097,
+      "grad_norm": 3.541882276535034,
+      "learning_rate": 3.7805228382648664e-05,
+      "loss": 3.2869,
+      "step": 228
+    },
+    {
+      "epoch": 0.0651196407519091,
+      "grad_norm": 3.242234945297241,
+      "learning_rate": 3.779373743177248e-05,
+      "loss": 3.0996,
+      "step": 229
+    },
+    {
+      "epoch": 0.06540400599536722,
+      "grad_norm": 2.6109092235565186,
+      "learning_rate": 3.77822464808963e-05,
+      "loss": 2.9875,
+      "step": 230
+    },
+    {
+      "epoch": 0.06568837123882533,
+      "grad_norm": 2.639467239379883,
+      "learning_rate": 3.777075553002011e-05,
+      "loss": 2.8205,
+      "step": 231
+    },
+    {
+      "epoch": 0.06597273648228345,
+      "grad_norm": 2.852081775665283,
+      "learning_rate": 3.775926457914393e-05,
+      "loss": 2.8712,
+      "step": 232
+    },
+    {
+      "epoch": 0.06625710172574158,
+      "grad_norm": 4.4990339279174805,
+      "learning_rate": 3.774777362826774e-05,
+      "loss": 3.8808,
+      "step": 233
+    },
+    {
+      "epoch": 0.06654146696919969,
+      "grad_norm": 5.339641571044922,
+      "learning_rate": 3.773628267739156e-05,
+      "loss": 3.8167,
+      "step": 234
+    },
+    {
+      "epoch": 0.06682583221265781,
+      "grad_norm": 3.1748464107513428,
+      "learning_rate": 3.772479172651537e-05,
+      "loss": 3.5924,
+      "step": 235
+    },
+    {
+      "epoch": 0.06711019745611593,
+      "grad_norm": 3.4898841381073,
+      "learning_rate": 3.7713300775639187e-05,
+      "loss": 3.2489,
+      "step": 236
+    },
+    {
+      "epoch": 0.06739456269957404,
+      "grad_norm": 3.2079555988311768,
+      "learning_rate": 3.7701809824763e-05,
+      "loss": 3.1664,
+      "step": 237
+    },
+    {
+      "epoch": 0.06767892794303217,
+      "grad_norm": 2.7962253093719482,
+      "learning_rate": 3.7690318873886816e-05,
+      "loss": 2.8651,
+      "step": 238
+    },
+    {
+      "epoch": 0.06796329318649028,
+      "grad_norm": 3.1500935554504395,
+      "learning_rate": 3.767882792301063e-05,
+      "loss": 2.7629,
+      "step": 239
+    },
+    {
+      "epoch": 0.0682476584299484,
+      "grad_norm": 2.5772945880889893,
+      "learning_rate": 3.7667336972134444e-05,
+      "loss": 2.6508,
+      "step": 240
+    },
+    {
+      "epoch": 0.06853202367340652,
+      "grad_norm": 4.086240291595459,
+      "learning_rate": 3.765584602125826e-05,
+      "loss": 3.8843,
+      "step": 241
+    },
+    {
+      "epoch": 0.06881638891686463,
+      "grad_norm": 2.856458902359009,
+      "learning_rate": 3.764435507038208e-05,
+      "loss": 3.4318,
+      "step": 242
+    },
+    {
+      "epoch": 0.06910075416032276,
+      "grad_norm": 3.179042100906372,
+      "learning_rate": 3.763286411950589e-05,
+      "loss": 3.0556,
+      "step": 243
+    },
+    {
+      "epoch": 0.06938511940378088,
+      "grad_norm": 3.2401700019836426,
+      "learning_rate": 3.762137316862971e-05,
+      "loss": 3.3399,
+      "step": 244
+    },
+    {
+      "epoch": 0.06966948464723899,
+      "grad_norm": 2.8616445064544678,
+      "learning_rate": 3.760988221775353e-05,
+      "loss": 3.0333,
+      "step": 245
+    },
+    {
+      "epoch": 0.06995384989069711,
+      "grad_norm": 2.7594053745269775,
+      "learning_rate": 3.759839126687734e-05,
+      "loss": 2.9015,
+      "step": 246
+    },
+    {
+      "epoch": 0.07023821513415522,
+      "grad_norm": 2.3787291049957275,
+      "learning_rate": 3.7586900316001156e-05,
+      "loss": 2.6074,
+      "step": 247
+    },
+    {
+      "epoch": 0.07052258037761334,
+      "grad_norm": 2.604524612426758,
+      "learning_rate": 3.757540936512497e-05,
+      "loss": 2.5631,
+      "step": 248
+    },
+    {
+      "epoch": 0.07080694562107147,
+      "grad_norm": 5.74310302734375,
+      "learning_rate": 3.7563918414248785e-05,
+      "loss": 3.9635,
+      "step": 249
+    },
+    {
+      "epoch": 0.07109131086452958,
+      "grad_norm": 4.676276206970215,
+      "learning_rate": 3.7552427463372596e-05,
+      "loss": 3.4952,
+      "step": 250
+    },
+    {
+      "epoch": 0.0713756761079877,
+      "grad_norm": 3.3065385818481445,
+      "learning_rate": 3.7540936512496414e-05,
+      "loss": 3.2999,
+      "step": 251
+    },
+    {
+      "epoch": 0.07166004135144582,
+      "grad_norm": 2.9583752155303955,
+      "learning_rate": 3.7529445561620225e-05,
+      "loss": 3.2314,
+      "step": 252
+    },
+    {
+      "epoch": 0.07194440659490393,
+      "grad_norm": 3.083742380142212,
+      "learning_rate": 3.751795461074404e-05,
+      "loss": 3.0786,
+      "step": 253
+    },
+    {
+      "epoch": 0.07222877183836206,
+      "grad_norm": 3.241976261138916,
+      "learning_rate": 3.750646365986786e-05,
+      "loss": 2.8807,
+      "step": 254
+    },
+    {
+      "epoch": 0.07251313708182018,
+      "grad_norm": 2.7543532848358154,
+      "learning_rate": 3.749497270899167e-05,
+      "loss": 2.8752,
+      "step": 255
+    },
+    {
+      "epoch": 0.07279750232527829,
+      "grad_norm": 2.6907711029052734,
+      "learning_rate": 3.748348175811549e-05,
+      "loss": 2.7082,
+      "step": 256
+    },
+    {
+      "epoch": 0.07308186756873641,
+      "grad_norm": 3.050748586654663,
+      "learning_rate": 3.74719908072393e-05,
+      "loss": 3.8712,
+      "step": 257
+    },
+    {
+      "epoch": 0.07336623281219452,
+      "grad_norm": 3.0048818588256836,
+      "learning_rate": 3.746049985636312e-05,
+      "loss": 3.4344,
+      "step": 258
+    },
+    {
+      "epoch": 0.07365059805565265,
+      "grad_norm": 2.973074197769165,
+      "learning_rate": 3.744900890548693e-05,
+      "loss": 3.1367,
+      "step": 259
+    },
+    {
+      "epoch": 0.07393496329911077,
+      "grad_norm": 3.0169665813446045,
+      "learning_rate": 3.743751795461075e-05,
+      "loss": 3.4544,
+      "step": 260
+    },
+    {
+      "epoch": 0.07421932854256888,
+      "grad_norm": 2.8069064617156982,
+      "learning_rate": 3.742602700373456e-05,
+      "loss": 3.1697,
+      "step": 261
+    },
+    {
+      "epoch": 0.074503693786027,
+      "grad_norm": 2.524655818939209,
+      "learning_rate": 3.741453605285838e-05,
+      "loss": 2.9794,
+      "step": 262
+    },
+    {
+      "epoch": 0.07478805902948513,
+      "grad_norm": 2.5447611808776855,
+      "learning_rate": 3.7403045101982195e-05,
+      "loss": 2.5633,
+      "step": 263
+    },
+    {
+      "epoch": 0.07507242427294324,
+      "grad_norm": 3.1086103916168213,
+      "learning_rate": 3.7391554151106006e-05,
+      "loss": 2.6274,
+      "step": 264
+    },
+    {
+      "epoch": 0.07535678951640136,
+      "grad_norm": 3.7446553707122803,
+      "learning_rate": 3.7380063200229824e-05,
+      "loss": 4.0418,
+      "step": 265
+    },
+    {
+      "epoch": 0.07564115475985948,
+      "grad_norm": 3.5950560569763184,
+      "learning_rate": 3.7368572249353635e-05,
+      "loss": 3.3891,
+      "step": 266
+    },
+    {
+      "epoch": 0.07592552000331759,
+      "grad_norm": 3.037245273590088,
+      "learning_rate": 3.735708129847745e-05,
+      "loss": 3.389,
+      "step": 267
+    },
+    {
+      "epoch": 0.07620988524677572,
+      "grad_norm": 2.6339914798736572,
+      "learning_rate": 3.7345590347601264e-05,
+      "loss": 3.0648,
+      "step": 268
+    },
+    {
+      "epoch": 0.07649425049023383,
+      "grad_norm": 2.379084587097168,
+      "learning_rate": 3.733409939672508e-05,
+      "loss": 2.9169,
+      "step": 269
+    },
+    {
+      "epoch": 0.07677861573369195,
+      "grad_norm": 2.9946842193603516,
+      "learning_rate": 3.732260844584889e-05,
+      "loss": 2.7304,
+      "step": 270
+    },
+    {
+      "epoch": 0.07706298097715007,
+      "grad_norm": 3.1141395568847656,
+      "learning_rate": 3.731111749497271e-05,
+      "loss": 2.7589,
+      "step": 271
+    },
+    {
+      "epoch": 0.07734734622060818,
+      "grad_norm": 2.931774854660034,
+      "learning_rate": 3.729962654409653e-05,
+      "loss": 2.61,
+      "step": 272
+    },
+    {
+      "epoch": 0.0776317114640663,
+      "grad_norm": 4.752538681030273,
+      "learning_rate": 3.728813559322034e-05,
+      "loss": 3.82,
+      "step": 273
+    },
+    {
+      "epoch": 0.07791607670752443,
+      "grad_norm": 3.4177722930908203,
+      "learning_rate": 3.727664464234416e-05,
+      "loss": 3.3525,
+      "step": 274
+    },
+    {
+      "epoch": 0.07820044195098254,
+      "grad_norm": 2.749371290206909,
+      "learning_rate": 3.726515369146797e-05,
+      "loss": 3.1633,
+      "step": 275
+    },
+    {
+      "epoch": 0.07848480719444066,
+      "grad_norm": 3.0202836990356445,
+      "learning_rate": 3.725366274059179e-05,
+      "loss": 3.2241,
+      "step": 276
+    },
+    {
+      "epoch": 0.07876917243789877,
+      "grad_norm": 3.951889753341675,
+      "learning_rate": 3.72421717897156e-05,
+      "loss": 3.1181,
+      "step": 277
+    },
+    {
+      "epoch": 0.0790535376813569,
+      "grad_norm": 3.7062735557556152,
+      "learning_rate": 3.723068083883942e-05,
+      "loss": 2.924,
+      "step": 278
+    },
+    {
+      "epoch": 0.07933790292481502,
+      "grad_norm": 2.9112234115600586,
+      "learning_rate": 3.7219189887963234e-05,
+      "loss": 2.6434,
+      "step": 279
+    },
+    {
+      "epoch": 0.07962226816827313,
+      "grad_norm": 2.5431931018829346,
+      "learning_rate": 3.720769893708705e-05,
+      "loss": 2.5729,
+      "step": 280
+    },
+    {
+      "epoch": 0.07990663341173125,
+      "grad_norm": 3.293665647506714,
+      "learning_rate": 3.719620798621086e-05,
+      "loss": 3.6881,
+      "step": 281
+    },
+    {
+      "epoch": 0.08019099865518937,
+      "grad_norm": 3.0450286865234375,
+      "learning_rate": 3.718471703533468e-05,
+      "loss": 3.6725,
+      "step": 282
+    },
+    {
+      "epoch": 0.08047536389864748,
+      "grad_norm": 2.8695693016052246,
+      "learning_rate": 3.717322608445849e-05,
+      "loss": 3.4356,
+      "step": 283
+    },
+    {
+      "epoch": 0.08075972914210561,
+      "grad_norm": 2.7940287590026855,
+      "learning_rate": 3.716173513358231e-05,
+      "loss": 3.1586,
+      "step": 284
+    },
+    {
+      "epoch": 0.08104409438556373,
+      "grad_norm": 3.127352476119995,
+      "learning_rate": 3.715024418270612e-05,
+      "loss": 3.0592,
+      "step": 285
+    },
+    {
+      "epoch": 0.08132845962902184,
+      "grad_norm": 2.77877140045166,
+      "learning_rate": 3.713875323182994e-05,
+      "loss": 2.6598,
+      "step": 286
+    },
+    {
+      "epoch": 0.08161282487247996,
+      "grad_norm": 2.7355434894561768,
+      "learning_rate": 3.7127262280953756e-05,
+      "loss": 2.5461,
+      "step": 287
+    },
+    {
+      "epoch": 0.08189719011593807,
+      "grad_norm": 2.800222873687744,
+      "learning_rate": 3.711577133007757e-05,
+      "loss": 2.6196,
+      "step": 288
+    },
+    {
+      "epoch": 0.0821815553593962,
+      "grad_norm": 3.4335219860076904,
+      "learning_rate": 3.7104280379201385e-05,
+      "loss": 3.6921,
+      "step": 289
+    },
+    {
+      "epoch": 0.08246592060285432,
+      "grad_norm": 3.0871481895446777,
+      "learning_rate": 3.7092789428325196e-05,
+      "loss": 3.2295,
+      "step": 290
+    },
+    {
+      "epoch": 0.08275028584631243,
+      "grad_norm": 3.372642993927002,
+      "learning_rate": 3.7081298477449014e-05,
+      "loss": 3.5505,
+      "step": 291
+    },
+    {
+      "epoch": 0.08303465108977055,
+      "grad_norm": 2.830824613571167,
+      "learning_rate": 3.7069807526572825e-05,
+      "loss": 3.1795,
+      "step": 292
+    },
+    {
+      "epoch": 0.08331901633322868,
+      "grad_norm": 2.850874423980713,
+      "learning_rate": 3.705831657569664e-05,
+      "loss": 2.9981,
+      "step": 293
+    },
+    {
+      "epoch": 0.08360338157668679,
+      "grad_norm": 2.798980712890625,
+      "learning_rate": 3.7046825624820454e-05,
+      "loss": 2.7752,
+      "step": 294
+    },
+    {
+      "epoch": 0.08388774682014491,
+      "grad_norm": 2.3986728191375732,
+      "learning_rate": 3.703533467394427e-05,
+      "loss": 2.4156,
+      "step": 295
+    },
+    {
+      "epoch": 0.08417211206360303,
+      "grad_norm": 2.329164505004883,
+      "learning_rate": 3.702384372306809e-05,
+      "loss": 2.5282,
+      "step": 296
+    },
+    {
+      "epoch": 0.08445647730706114,
+      "grad_norm": 3.428501844406128,
+      "learning_rate": 3.70123527721919e-05,
+      "loss": 3.8649,
+      "step": 297
+    },
+    {
+      "epoch": 0.08474084255051927,
+      "grad_norm": 3.2161340713500977,
+      "learning_rate": 3.700086182131572e-05,
+      "loss": 3.4523,
+      "step": 298
+    },
+    {
+      "epoch": 0.08502520779397738,
+      "grad_norm": 2.801135778427124,
+      "learning_rate": 3.698937087043953e-05,
+      "loss": 3.3231,
+      "step": 299
+    },
+    {
+      "epoch": 0.0853095730374355,
+      "grad_norm": 2.886110544204712,
+      "learning_rate": 3.697787991956335e-05,
+      "loss": 2.918,
+      "step": 300
+    },
+    {
+      "epoch": 0.08559393828089362,
+      "grad_norm": 3.050361156463623,
+      "learning_rate": 3.696638896868716e-05,
+      "loss": 2.9013,
+      "step": 301
+    },
+    {
+      "epoch": 0.08587830352435173,
+      "grad_norm": 2.4915997982025146,
+      "learning_rate": 3.695489801781098e-05,
+      "loss": 2.9062,
+      "step": 302
+    },
+    {
+      "epoch": 0.08616266876780986,
+      "grad_norm": 2.6376094818115234,
+      "learning_rate": 3.694340706693479e-05,
+      "loss": 2.8681,
+      "step": 303
+    },
+    {
+      "epoch": 0.08644703401126798,
+      "grad_norm": 2.602095603942871,
+      "learning_rate": 3.6931916116058606e-05,
+      "loss": 2.6519,
+      "step": 304
+    },
+    {
+      "epoch": 0.08673139925472609,
+      "grad_norm": 3.274116277694702,
+      "learning_rate": 3.6920425165182424e-05,
+      "loss": 3.812,
+      "step": 305
+    },
+    {
+      "epoch": 0.08701576449818421,
+      "grad_norm": 2.873872756958008,
+      "learning_rate": 3.6908934214306235e-05,
+      "loss": 3.3963,
+      "step": 306
+    },
+    {
+      "epoch": 0.08730012974164232,
+      "grad_norm": 2.432584047317505,
+      "learning_rate": 3.689744326343005e-05,
+      "loss": 3.2271,
+      "step": 307
+    },
+    {
+      "epoch": 0.08758449498510044,
+      "grad_norm": 2.374149799346924,
+      "learning_rate": 3.6885952312553864e-05,
+      "loss": 3.4667,
+      "step": 308
+    },
+    {
+      "epoch": 0.08786886022855857,
+      "grad_norm": 2.5668742656707764,
+      "learning_rate": 3.687446136167768e-05,
+      "loss": 2.9856,
+      "step": 309
+    },
+    {
+      "epoch": 0.08815322547201668,
+      "grad_norm": 2.4312264919281006,
+      "learning_rate": 3.686297041080149e-05,
+      "loss": 2.6518,
+      "step": 310
+    },
+    {
+      "epoch": 0.0884375907154748,
+      "grad_norm": 2.417278528213501,
+      "learning_rate": 3.685147945992531e-05,
+      "loss": 2.5552,
+      "step": 311
+    },
+    {
+      "epoch": 0.08872195595893292,
+      "grad_norm": 2.399350166320801,
+      "learning_rate": 3.683998850904912e-05,
+      "loss": 2.5939,
+      "step": 312
+    },
+    {
+      "epoch": 0.08900632120239103,
+      "grad_norm": 4.476291656494141,
+      "learning_rate": 3.682849755817294e-05,
+      "loss": 3.7203,
+      "step": 313
+    },
+    {
+      "epoch": 0.08929068644584916,
+      "grad_norm": 2.964276075363159,
+      "learning_rate": 3.681700660729676e-05,
+      "loss": 3.3704,
+      "step": 314
+    },
+    {
+      "epoch": 0.08957505168930728,
+      "grad_norm": 2.778978109359741,
+      "learning_rate": 3.6805515656420576e-05,
+      "loss": 3.3189,
+      "step": 315
+    },
+    {
+      "epoch": 0.08985941693276539,
+      "grad_norm": 3.131481647491455,
+      "learning_rate": 3.679402470554439e-05,
+      "loss": 3.2355,
+      "step": 316
+    },
+    {
+      "epoch": 0.09014378217622351,
+      "grad_norm": 3.234072208404541,
+      "learning_rate": 3.6782533754668205e-05,
+      "loss": 2.8635,
+      "step": 317
+    },
+    {
+      "epoch": 0.09042814741968162,
+      "grad_norm": 3.1541082859039307,
+      "learning_rate": 3.6771042803792016e-05,
+      "loss": 2.7849,
+      "step": 318
+    },
+    {
+      "epoch": 0.09071251266313975,
+      "grad_norm": 2.8236806392669678,
+      "learning_rate": 3.6759551852915834e-05,
+      "loss": 2.7041,
+      "step": 319
+    },
+    {
+      "epoch": 0.09099687790659787,
+      "grad_norm": 2.4180235862731934,
+      "learning_rate": 3.674806090203965e-05,
+      "loss": 2.4594,
+      "step": 320
+    },
+    {
+      "epoch": 0.09128124315005598,
+      "grad_norm": 4.0737080574035645,
+      "learning_rate": 3.673656995116346e-05,
+      "loss": 3.5109,
+      "step": 321
+    },
+    {
+      "epoch": 0.0915656083935141,
+      "grad_norm": 4.438284873962402,
+      "learning_rate": 3.672507900028728e-05,
+      "loss": 3.2919,
+      "step": 322
+    },
+    {
+      "epoch": 0.09184997363697223,
+      "grad_norm": 4.133691310882568,
+      "learning_rate": 3.671358804941109e-05,
+      "loss": 3.0562,
+      "step": 323
+    },
+    {
+      "epoch": 0.09213433888043034,
+      "grad_norm": 4.211462020874023,
+      "learning_rate": 3.670209709853491e-05,
+      "loss": 3.2076,
+      "step": 324
+    },
+    {
+      "epoch": 0.09241870412388846,
+      "grad_norm": 3.6107709407806396,
+      "learning_rate": 3.669060614765872e-05,
+      "loss": 3.003,
+      "step": 325
+    },
+    {
+      "epoch": 0.09270306936734658,
+      "grad_norm": 2.793802261352539,
+      "learning_rate": 3.667911519678254e-05,
+      "loss": 2.8852,
+      "step": 326
+    },
+    {
+      "epoch": 0.09298743461080469,
+      "grad_norm": 2.2556843757629395,
+      "learning_rate": 3.666762424590635e-05,
+      "loss": 2.5742,
+      "step": 327
+    },
+    {
+      "epoch": 0.09327179985426282,
+      "grad_norm": 2.088768243789673,
+      "learning_rate": 3.665613329503017e-05,
+      "loss": 2.5716,
+      "step": 328
+    },
+    {
+      "epoch": 0.09355616509772093,
+      "grad_norm": 4.9838056564331055,
+      "learning_rate": 3.6644642344153986e-05,
+      "loss": 3.7839,
+      "step": 329
+    },
+    {
+      "epoch": 0.09384053034117905,
+      "grad_norm": 3.877955436706543,
+      "learning_rate": 3.66331513932778e-05,
+      "loss": 3.2162,
+      "step": 330
+    },
+    {
+      "epoch": 0.09412489558463717,
+      "grad_norm": 3.442711353302002,
+      "learning_rate": 3.6621660442401615e-05,
+      "loss": 3.2394,
+      "step": 331
+    },
+    {
+      "epoch": 0.09440926082809528,
+      "grad_norm": 3.3739242553710938,
+      "learning_rate": 3.6610169491525426e-05,
+      "loss": 3.1204,
+      "step": 332
+    },
+    {
+      "epoch": 0.0946936260715534,
+      "grad_norm": 3.4325954914093018,
+      "learning_rate": 3.6598678540649244e-05,
+      "loss": 2.7708,
+      "step": 333
+    },
+    {
+      "epoch": 0.09497799131501153,
+      "grad_norm": 2.4555609226226807,
+      "learning_rate": 3.6587187589773055e-05,
+      "loss": 2.5941,
+      "step": 334
+    },
+    {
+      "epoch": 0.09526235655846964,
+      "grad_norm": 2.201622724533081,
+      "learning_rate": 3.657569663889687e-05,
+      "loss": 2.5229,
+      "step": 335
+    },
+    {
+      "epoch": 0.09554672180192776,
+      "grad_norm": 2.3770065307617188,
+      "learning_rate": 3.6564205688020684e-05,
+      "loss": 2.4801,
+      "step": 336
+    },
+    {
+      "epoch": 0.09583108704538587,
+      "grad_norm": 3.16601300239563,
+      "learning_rate": 3.65527147371445e-05,
+      "loss": 3.9019,
+      "step": 337
+    },
+    {
+      "epoch": 0.096115452288844,
+      "grad_norm": 2.69059419631958,
+      "learning_rate": 3.654122378626832e-05,
+      "loss": 3.2074,
+      "step": 338
+    },
+    {
+      "epoch": 0.09639981753230212,
+      "grad_norm": 2.622850179672241,
+      "learning_rate": 3.652973283539213e-05,
+      "loss": 3.1565,
+      "step": 339
+    },
+    {
+      "epoch": 0.09668418277576023,
+      "grad_norm": 2.384047746658325,
+      "learning_rate": 3.651824188451595e-05,
+      "loss": 2.9815,
+      "step": 340
+    },
+    {
+      "epoch": 0.09696854801921835,
+      "grad_norm": 2.255505084991455,
+      "learning_rate": 3.650675093363976e-05,
+      "loss": 3.0731,
+      "step": 341
+    },
+    {
+      "epoch": 0.09725291326267647,
+      "grad_norm": 2.502235174179077,
+      "learning_rate": 3.649525998276358e-05,
+      "loss": 2.795,
+      "step": 342
+    },
+    {
+      "epoch": 0.09753727850613458,
+      "grad_norm": 2.6988747119903564,
+      "learning_rate": 3.648376903188739e-05,
+      "loss": 2.6843,
+      "step": 343
+    },
+    {
+      "epoch": 0.09782164374959271,
+      "grad_norm": 2.856539011001587,
+      "learning_rate": 3.6472278081011206e-05,
+      "loss": 2.4656,
+      "step": 344
+    },
+    {
+      "epoch": 0.09810600899305083,
+      "grad_norm": 3.2413103580474854,
+      "learning_rate": 3.646078713013502e-05,
+      "loss": 3.8422,
+      "step": 345
+    },
+    {
+      "epoch": 0.09839037423650894,
+      "grad_norm": 2.434211254119873,
+      "learning_rate": 3.6449296179258835e-05,
+      "loss": 3.0491,
+      "step": 346
+    },
+    {
+      "epoch": 0.09867473947996706,
+      "grad_norm": 2.2801740169525146,
+      "learning_rate": 3.643780522838265e-05,
+      "loss": 3.0232,
+      "step": 347
+    },
+    {
+      "epoch": 0.09895910472342517,
+      "grad_norm": 2.3498308658599854,
+      "learning_rate": 3.6426314277506464e-05,
+      "loss": 2.9397,
+      "step": 348
+    },
+    {
+      "epoch": 0.0992434699668833,
+      "grad_norm": 2.8017773628234863,
+      "learning_rate": 3.641482332663028e-05,
+      "loss": 2.8337,
+      "step": 349
+    },
+    {
+      "epoch": 0.09952783521034142,
+      "grad_norm": 2.93746018409729,
+      "learning_rate": 3.6403332375754093e-05,
+      "loss": 2.7192,
+      "step": 350
+    },
+    {
+      "epoch": 0.09981220045379953,
+      "grad_norm": 3.0276105403900146,
+      "learning_rate": 3.639184142487791e-05,
+      "loss": 2.3286,
+      "step": 351
+    },
+    {
+      "epoch": 0.10009656569725765,
+      "grad_norm": 2.725423812866211,
+      "learning_rate": 3.638035047400173e-05,
+      "loss": 2.298,
+      "step": 352
+    },
+    {
+      "epoch": 0.10038093094071578,
+      "grad_norm": 2.9816744327545166,
+      "learning_rate": 3.636885952312554e-05,
+      "loss": 3.6773,
+      "step": 353
+    },
+    {
+      "epoch": 0.10066529618417389,
+      "grad_norm": 2.4625086784362793,
+      "learning_rate": 3.635736857224936e-05,
+      "loss": 3.2648,
+      "step": 354
+    },
+    {
+      "epoch": 0.10094966142763201,
+      "grad_norm": 2.325002670288086,
+      "learning_rate": 3.6345877621373176e-05,
+      "loss": 2.9206,
+      "step": 355
+    },
+    {
+      "epoch": 0.10123402667109013,
+      "grad_norm": 2.643895149230957,
+      "learning_rate": 3.633438667049699e-05,
+      "loss": 2.6485,
+      "step": 356
+    },
+    {
+      "epoch": 0.10151839191454824,
+      "grad_norm": 2.9441092014312744,
+      "learning_rate": 3.6322895719620805e-05,
+      "loss": 2.8929,
+      "step": 357
+    },
+    {
+      "epoch": 0.10180275715800637,
+      "grad_norm": 2.6018552780151367,
+      "learning_rate": 3.6311404768744616e-05,
+      "loss": 2.4757,
+      "step": 358
+    },
+    {
+      "epoch": 0.10208712240146448,
+      "grad_norm": 2.31733775138855,
+      "learning_rate": 3.6299913817868434e-05,
+      "loss": 2.6008,
+      "step": 359
+    },
+    {
+      "epoch": 0.1023714876449226,
+      "grad_norm": 2.3203630447387695,
+      "learning_rate": 3.6288422866992245e-05,
+      "loss": 2.4726,
+      "step": 360
+    },
+    {
+      "epoch": 0.10265585288838072,
+      "grad_norm": 3.1516456604003906,
+      "learning_rate": 3.627693191611606e-05,
+      "loss": 3.8334,
+      "step": 361
+    },
+    {
+      "epoch": 0.10294021813183883,
+      "grad_norm": 2.976469039916992,
+      "learning_rate": 3.626544096523988e-05,
+      "loss": 3.2071,
+      "step": 362
+    },
+    {
+      "epoch": 0.10322458337529695,
+      "grad_norm": 2.7915210723876953,
+      "learning_rate": 3.625395001436369e-05,
+      "loss": 3.0369,
+      "step": 363
+    },
+    {
+      "epoch": 0.10350894861875508,
+      "grad_norm": 3.034956932067871,
+      "learning_rate": 3.624245906348751e-05,
+      "loss": 3.1236,
+      "step": 364
+    },
+    {
+      "epoch": 0.10379331386221319,
+      "grad_norm": 2.835688591003418,
+      "learning_rate": 3.623096811261132e-05,
+      "loss": 2.8524,
+      "step": 365
+    },
+    {
+      "epoch": 0.10407767910567131,
+      "grad_norm": 2.3832132816314697,
+      "learning_rate": 3.621947716173514e-05,
+      "loss": 2.4496,
+      "step": 366
+    },
+    {
+      "epoch": 0.10436204434912942,
+      "grad_norm": 2.339113473892212,
+      "learning_rate": 3.620798621085895e-05,
+      "loss": 2.4954,
+      "step": 367
+    },
+    {
+      "epoch": 0.10464640959258754,
+      "grad_norm": 2.407395362854004,
+      "learning_rate": 3.619649525998277e-05,
+      "loss": 2.568,
+      "step": 368
+    },
+    {
+      "epoch": 0.10493077483604567,
+      "grad_norm": 4.512346267700195,
+      "learning_rate": 3.618500430910658e-05,
+      "loss": 3.5987,
+      "step": 369
+    },
+    {
+      "epoch": 0.10521514007950378,
+      "grad_norm": 3.320215940475464,
+      "learning_rate": 3.61735133582304e-05,
+      "loss": 3.0956,
+      "step": 370
+    },
+    {
+      "epoch": 0.1054995053229619,
+      "grad_norm": 3.0713627338409424,
+      "learning_rate": 3.6162022407354215e-05,
+      "loss": 3.1864,
+      "step": 371
+    },
+    {
+      "epoch": 0.10578387056642002,
+      "grad_norm": 2.845768690109253,
+      "learning_rate": 3.6150531456478026e-05,
+      "loss": 3.0029,
+      "step": 372
+    },
+    {
+      "epoch": 0.10606823580987813,
+      "grad_norm": 2.5338187217712402,
+      "learning_rate": 3.6139040505601844e-05,
+      "loss": 2.6617,
+      "step": 373
+    },
+    {
+      "epoch": 0.10635260105333626,
+      "grad_norm": 2.5313353538513184,
+      "learning_rate": 3.6127549554725655e-05,
+      "loss": 2.5754,
+      "step": 374
+    },
+    {
+      "epoch": 0.10663696629679438,
+      "grad_norm": 2.485309362411499,
+      "learning_rate": 3.611605860384947e-05,
+      "loss": 2.4186,
+      "step": 375
+    },
+    {
+      "epoch": 0.10692133154025249,
+      "grad_norm": 2.544818878173828,
+      "learning_rate": 3.6104567652973284e-05,
+      "loss": 2.5501,
+      "step": 376
+    },
+    {
+      "epoch": 0.10720569678371061,
+      "grad_norm": 2.9218149185180664,
+      "learning_rate": 3.60930767020971e-05,
+      "loss": 3.7049,
+      "step": 377
+    },
+    {
+      "epoch": 0.10749006202716872,
+      "grad_norm": 2.9214773178100586,
+      "learning_rate": 3.608158575122091e-05,
+      "loss": 3.0877,
+      "step": 378
+    },
+    {
+      "epoch": 0.10777442727062685,
+      "grad_norm": 2.4883925914764404,
+      "learning_rate": 3.607009480034473e-05,
+      "loss": 3.1767,
+      "step": 379
+    },
+    {
+      "epoch": 0.10805879251408497,
+      "grad_norm": 2.176980972290039,
+      "learning_rate": 3.605860384946855e-05,
+      "loss": 2.9752,
+      "step": 380
+    },
+    {
+      "epoch": 0.10834315775754308,
+      "grad_norm": 2.442439317703247,
+      "learning_rate": 3.604711289859236e-05,
+      "loss": 3.0929,
+      "step": 381
+    },
+    {
+      "epoch": 0.1086275230010012,
+      "grad_norm": 2.6657745838165283,
+      "learning_rate": 3.603562194771618e-05,
+      "loss": 2.6189,
+      "step": 382
+    },
+    {
+      "epoch": 0.10891188824445933,
+      "grad_norm": 2.3688509464263916,
+      "learning_rate": 3.602413099683999e-05,
+      "loss": 2.5404,
+      "step": 383
+    },
+    {
+      "epoch": 0.10919625348791744,
+      "grad_norm": 2.235816717147827,
+      "learning_rate": 3.601264004596381e-05,
+      "loss": 2.4137,
+      "step": 384
+    },
+    {
+      "epoch": 0.10948061873137556,
+      "grad_norm": 3.8358049392700195,
+      "learning_rate": 3.600114909508762e-05,
+      "loss": 3.5268,
+      "step": 385
+    },
+    {
+      "epoch": 0.10976498397483368,
+      "grad_norm": 2.5932207107543945,
+      "learning_rate": 3.5989658144211436e-05,
+      "loss": 3.1315,
+      "step": 386
+    },
+    {
+      "epoch": 0.11004934921829179,
+      "grad_norm": 2.4010274410247803,
+      "learning_rate": 3.597816719333525e-05,
+      "loss": 3.072,
+      "step": 387
+    },
+    {
+      "epoch": 0.11033371446174992,
+      "grad_norm": 2.247072219848633,
+      "learning_rate": 3.5966676242459065e-05,
+      "loss": 2.5931,
+      "step": 388
+    },
+    {
+      "epoch": 0.11061807970520803,
+      "grad_norm": 3.017591953277588,
+      "learning_rate": 3.595518529158288e-05,
+      "loss": 2.8756,
+      "step": 389
+    },
+    {
+      "epoch": 0.11090244494866615,
+      "grad_norm": 2.5938799381256104,
+      "learning_rate": 3.59436943407067e-05,
+      "loss": 2.869,
+      "step": 390
+    },
+    {
+      "epoch": 0.11118681019212427,
+      "grad_norm": 2.039341926574707,
+      "learning_rate": 3.593220338983051e-05,
+      "loss": 2.4231,
+      "step": 391
+    },
+    {
+      "epoch": 0.11147117543558238,
+      "grad_norm": 2.2805583477020264,
+      "learning_rate": 3.592071243895433e-05,
+      "loss": 2.377,
+      "step": 392
+    },
+    {
+      "epoch": 0.1117555406790405,
+      "grad_norm": 2.4253814220428467,
+      "learning_rate": 3.590922148807814e-05,
+      "loss": 3.549,
+      "step": 393
+    },
+    {
+      "epoch": 0.11203990592249863,
+      "grad_norm": 2.2920830249786377,
+      "learning_rate": 3.589773053720196e-05,
+      "loss": 3.2533,
+      "step": 394
+    },
+    {
+      "epoch": 0.11232427116595674,
+      "grad_norm": 2.4961812496185303,
+      "learning_rate": 3.588623958632577e-05,
+      "loss": 2.8119,
+      "step": 395
+    },
+    {
+      "epoch": 0.11260863640941486,
+      "grad_norm": 2.3366754055023193,
+      "learning_rate": 3.587474863544959e-05,
+      "loss": 3.0371,
+      "step": 396
+    },
+    {
+      "epoch": 0.11289300165287297,
+      "grad_norm": 2.443265914916992,
+      "learning_rate": 3.5863257684573405e-05,
+      "loss": 2.731,
+      "step": 397
+    },
+    {
+      "epoch": 0.1131773668963311,
+      "grad_norm": 2.1697330474853516,
+      "learning_rate": 3.5851766733697216e-05,
+      "loss": 2.548,
+      "step": 398
+    },
+    {
+      "epoch": 0.11346173213978922,
+      "grad_norm": 2.0168380737304688,
+      "learning_rate": 3.5840275782821034e-05,
+      "loss": 2.4326,
+      "step": 399
+    },
+    {
+      "epoch": 0.11374609738324733,
+      "grad_norm": 2.136674642562866,
+      "learning_rate": 3.5828784831944845e-05,
+      "loss": 2.3445,
+      "step": 400
+    },
+    {
+      "epoch": 0.11403046262670545,
+      "grad_norm": 2.96239972114563,
+      "learning_rate": 3.581729388106866e-05,
+      "loss": 3.7026,
+      "step": 401
+    },
+    {
+      "epoch": 0.11431482787016357,
+      "grad_norm": 2.7048444747924805,
+      "learning_rate": 3.5805802930192474e-05,
+      "loss": 3.0997,
+      "step": 402
+    },
+    {
+      "epoch": 0.11459919311362168,
+      "grad_norm": 2.455439805984497,
+      "learning_rate": 3.579431197931629e-05,
+      "loss": 2.9647,
+      "step": 403
+    },
+    {
+      "epoch": 0.1148835583570798,
+      "grad_norm": 2.2661333084106445,
+      "learning_rate": 3.57828210284401e-05,
+      "loss": 2.7126,
+      "step": 404
+    },
+    {
+      "epoch": 0.11516792360053793,
+      "grad_norm": 2.03705096244812,
+      "learning_rate": 3.577133007756392e-05,
+      "loss": 2.755,
+      "step": 405
+    },
+    {
+      "epoch": 0.11545228884399604,
+      "grad_norm": 2.140838384628296,
+      "learning_rate": 3.575983912668774e-05,
+      "loss": 2.5794,
+      "step": 406
+    },
+    {
+      "epoch": 0.11573665408745416,
+      "grad_norm": 2.352357864379883,
+      "learning_rate": 3.574834817581155e-05,
+      "loss": 2.64,
+      "step": 407
+    },
+    {
+      "epoch": 0.11602101933091227,
+      "grad_norm": 2.3507237434387207,
+      "learning_rate": 3.573685722493537e-05,
+      "loss": 2.2045,
+      "step": 408
+    },
+    {
+      "epoch": 0.1163053845743704,
+      "grad_norm": 4.098755359649658,
+      "learning_rate": 3.572536627405918e-05,
+      "loss": 3.437,
+      "step": 409
+    },
+    {
+      "epoch": 0.11658974981782852,
+      "grad_norm": 2.8039045333862305,
+      "learning_rate": 3.5713875323183e-05,
+      "loss": 3.2256,
+      "step": 410
+    },
+    {
+      "epoch": 0.11687411506128663,
+      "grad_norm": 2.3482019901275635,
+      "learning_rate": 3.570238437230681e-05,
+      "loss": 2.9395,
+      "step": 411
+    },
+    {
+      "epoch": 0.11715848030474475,
+      "grad_norm": 2.0725553035736084,
+      "learning_rate": 3.5690893421430626e-05,
+      "loss": 2.8758,
+      "step": 412
+    },
+    {
+      "epoch": 0.11744284554820288,
+      "grad_norm": 2.3215219974517822,
+      "learning_rate": 3.567940247055444e-05,
+      "loss": 2.7822,
+      "step": 413
+    },
+    {
+      "epoch": 0.11772721079166099,
+      "grad_norm": 2.770766258239746,
+      "learning_rate": 3.5667911519678255e-05,
+      "loss": 2.5678,
+      "step": 414
+    },
+    {
+      "epoch": 0.11801157603511911,
+      "grad_norm": 2.5529537200927734,
+      "learning_rate": 3.565642056880207e-05,
+      "loss": 2.429,
+      "step": 415
+    },
+    {
+      "epoch": 0.11829594127857723,
+      "grad_norm": 2.2458651065826416,
+      "learning_rate": 3.5644929617925884e-05,
+      "loss": 2.1978,
+      "step": 416
+    },
+    {
+      "epoch": 0.11858030652203534,
+      "grad_norm": 2.636418581008911,
+      "learning_rate": 3.56334386670497e-05,
+      "loss": 3.3172,
+      "step": 417
+    },
+    {
+      "epoch": 0.11886467176549347,
+      "grad_norm": 2.1322054862976074,
+      "learning_rate": 3.562194771617351e-05,
+      "loss": 3.2257,
+      "step": 418
+    },
+    {
+      "epoch": 0.11914903700895157,
+      "grad_norm": 2.089116334915161,
+      "learning_rate": 3.561045676529733e-05,
+      "loss": 3.1797,
+      "step": 419
+    },
+    {
+      "epoch": 0.1194334022524097,
+      "grad_norm": 2.1360151767730713,
+      "learning_rate": 3.559896581442114e-05,
+      "loss": 2.8716,
+      "step": 420
+    },
+    {
+      "epoch": 0.11971776749586782,
+      "grad_norm": 2.4757323265075684,
+      "learning_rate": 3.558747486354496e-05,
+      "loss": 2.9043,
+      "step": 421
+    },
+    {
+      "epoch": 0.12000213273932593,
+      "grad_norm": 2.3627967834472656,
+      "learning_rate": 3.557598391266877e-05,
+      "loss": 2.4626,
+      "step": 422
+    },
+    {
+      "epoch": 0.12028649798278405,
+      "grad_norm": 2.1213998794555664,
+      "learning_rate": 3.556449296179259e-05,
+      "loss": 2.5062,
+      "step": 423
+    },
+    {
+      "epoch": 0.12057086322624218,
+      "grad_norm": 2.1772518157958984,
+      "learning_rate": 3.555300201091641e-05,
+      "loss": 2.3489,
+      "step": 424
+    },
+    {
+      "epoch": 0.12085522846970029,
+      "grad_norm": 3.624419689178467,
+      "learning_rate": 3.554151106004022e-05,
+      "loss": 3.6556,
+      "step": 425
+    },
+    {
+      "epoch": 0.12113959371315841,
+      "grad_norm": 2.4470715522766113,
+      "learning_rate": 3.5530020109164036e-05,
+      "loss": 3.1925,
+      "step": 426
+    },
+    {
+      "epoch": 0.12142395895661652,
+      "grad_norm": 2.2214372158050537,
+      "learning_rate": 3.5518529158287854e-05,
+      "loss": 2.9998,
+      "step": 427
+    },
+    {
+      "epoch": 0.12170832420007464,
+      "grad_norm": 2.347656011581421,
+      "learning_rate": 3.5507038207411665e-05,
+      "loss": 2.8243,
+      "step": 428
+    },
+    {
+      "epoch": 0.12199268944353277,
+      "grad_norm": 2.0817830562591553,
+      "learning_rate": 3.549554725653548e-05,
+      "loss": 2.8815,
+      "step": 429
+    },
+    {
+      "epoch": 0.12227705468699088,
+      "grad_norm": 1.9819862842559814,
+      "learning_rate": 3.54840563056593e-05,
+      "loss": 2.751,
+      "step": 430
+    },
+    {
+      "epoch": 0.122561419930449,
+      "grad_norm": 2.2460458278656006,
+      "learning_rate": 3.547256535478311e-05,
+      "loss": 2.6482,
+      "step": 431
+    },
+    {
+      "epoch": 0.12284578517390712,
+      "grad_norm": 2.3552284240722656,
+      "learning_rate": 3.546107440390693e-05,
+      "loss": 2.5644,
+      "step": 432
+    },
+    {
+      "epoch": 0.12313015041736523,
+      "grad_norm": 2.6936285495758057,
+      "learning_rate": 3.544958345303074e-05,
+      "loss": 3.5913,
+      "step": 433
+    },
+    {
+      "epoch": 0.12341451566082336,
+      "grad_norm": 2.459479570388794,
+      "learning_rate": 3.543809250215456e-05,
+      "loss": 2.9851,
+      "step": 434
+    },
+    {
+      "epoch": 0.12369888090428148,
+      "grad_norm": 2.364466667175293,
+      "learning_rate": 3.542660155127837e-05,
+      "loss": 2.9548,
+      "step": 435
+    },
+    {
+      "epoch": 0.12398324614773959,
+      "grad_norm": 2.216580629348755,
+      "learning_rate": 3.541511060040219e-05,
+      "loss": 2.9091,
+      "step": 436
+    },
+    {
+      "epoch": 0.12426761139119771,
+      "grad_norm": 2.2435667514801025,
+      "learning_rate": 3.5403619649526e-05,
+      "loss": 2.8564,
+      "step": 437
+    },
+    {
+      "epoch": 0.12455197663465582,
+      "grad_norm": 2.2413361072540283,
+      "learning_rate": 3.539212869864982e-05,
+      "loss": 2.586,
+      "step": 438
+    },
+    {
+      "epoch": 0.12483634187811395,
+      "grad_norm": 2.0291380882263184,
+      "learning_rate": 3.5380637747773635e-05,
+      "loss": 2.4712,
+      "step": 439
+    },
+    {
+      "epoch": 0.12512070712157206,
+      "grad_norm": 2.161757707595825,
+      "learning_rate": 3.5369146796897446e-05,
+      "loss": 2.3247,
+      "step": 440
+    },
+    {
+      "epoch": 0.12540507236503018,
+      "grad_norm": 2.918247938156128,
+      "learning_rate": 3.5357655846021264e-05,
+      "loss": 3.5365,
+      "step": 441
+    },
+    {
+      "epoch": 0.1256894376084883,
+      "grad_norm": 2.381178855895996,
+      "learning_rate": 3.5346164895145075e-05,
+      "loss": 3.15,
+      "step": 442
+    },
+    {
+      "epoch": 0.12597380285194643,
+      "grad_norm": 2.349804162979126,
+      "learning_rate": 3.533467394426889e-05,
+      "loss": 3.0955,
+      "step": 443
+    },
+    {
+      "epoch": 0.12625816809540455,
+      "grad_norm": 2.0576179027557373,
+      "learning_rate": 3.5323182993392704e-05,
+      "loss": 2.8447,
+      "step": 444
+    },
+    {
+      "epoch": 0.12654253333886264,
+      "grad_norm": 2.2097201347351074,
+      "learning_rate": 3.531169204251652e-05,
+      "loss": 2.6086,
+      "step": 445
+    },
+    {
+      "epoch": 0.12682689858232077,
+      "grad_norm": 2.347313642501831,
+      "learning_rate": 3.530020109164033e-05,
+      "loss": 2.5803,
+      "step": 446
+    },
+    {
+      "epoch": 0.1271112638257789,
+      "grad_norm": 2.059239625930786,
+      "learning_rate": 3.528871014076415e-05,
+      "loss": 2.5247,
+      "step": 447
+    },
+    {
+      "epoch": 0.12739562906923702,
+      "grad_norm": 2.106543779373169,
+      "learning_rate": 3.527721918988797e-05,
+      "loss": 2.1429,
+      "step": 448
+    },
+    {
+      "epoch": 0.12767999431269514,
+      "grad_norm": 4.565354824066162,
+      "learning_rate": 3.526572823901178e-05,
+      "loss": 3.4247,
+      "step": 449
+    },
+    {
+      "epoch": 0.12796435955615326,
+      "grad_norm": 3.1165966987609863,
+      "learning_rate": 3.52542372881356e-05,
+      "loss": 3.3897,
+      "step": 450
+    },
+    {
+      "epoch": 0.12824872479961136,
+      "grad_norm": 2.431389808654785,
+      "learning_rate": 3.524274633725941e-05,
+      "loss": 2.8458,
+      "step": 451
+    },
+    {
+      "epoch": 0.12853309004306948,
+      "grad_norm": 2.352386713027954,
+      "learning_rate": 3.5231255386383226e-05,
+      "loss": 2.8419,
+      "step": 452
+    },
+    {
+      "epoch": 0.1288174552865276,
+      "grad_norm": 2.195875406265259,
+      "learning_rate": 3.521976443550704e-05,
+      "loss": 2.7576,
+      "step": 453
+    },
+    {
+      "epoch": 0.12910182052998573,
+      "grad_norm": 2.407545566558838,
+      "learning_rate": 3.5208273484630855e-05,
+      "loss": 2.4436,
+      "step": 454
+    },
+    {
+      "epoch": 0.12938618577344385,
+      "grad_norm": 2.3552427291870117,
+      "learning_rate": 3.5196782533754666e-05,
+      "loss": 2.3348,
+      "step": 455
+    },
+    {
+      "epoch": 0.12967055101690195,
+      "grad_norm": 2.576530933380127,
+      "learning_rate": 3.5185291582878484e-05,
+      "loss": 2.3373,
+      "step": 456
+    },
+    {
+      "epoch": 0.12995491626036007,
+      "grad_norm": 2.8271255493164062,
+      "learning_rate": 3.51738006320023e-05,
+      "loss": 3.4477,
+      "step": 457
+    },
+    {
+      "epoch": 0.1302392815038182,
+      "grad_norm": 2.433382987976074,
+      "learning_rate": 3.516230968112611e-05,
+      "loss": 3.0918,
+      "step": 458
+    },
+    {
+      "epoch": 0.13052364674727632,
+      "grad_norm": 2.547974109649658,
+      "learning_rate": 3.515081873024993e-05,
+      "loss": 3.0768,
+      "step": 459
+    },
+    {
+      "epoch": 0.13080801199073444,
+      "grad_norm": 2.379182815551758,
+      "learning_rate": 3.513932777937374e-05,
+      "loss": 2.886,
+      "step": 460
+    },
+    {
+      "epoch": 0.13109237723419256,
+      "grad_norm": 2.6086502075195312,
+      "learning_rate": 3.512783682849756e-05,
+      "loss": 2.7411,
+      "step": 461
+    },
+    {
+      "epoch": 0.13137674247765066,
+      "grad_norm": 2.6192541122436523,
+      "learning_rate": 3.511634587762137e-05,
+      "loss": 2.6317,
+      "step": 462
+    },
+    {
+      "epoch": 0.13166110772110878,
+      "grad_norm": 2.5311689376831055,
+      "learning_rate": 3.5104854926745196e-05,
+      "loss": 2.4297,
+      "step": 463
+    },
+    {
+      "epoch": 0.1319454729645669,
+      "grad_norm": 1.9520777463912964,
+      "learning_rate": 3.509336397586901e-05,
+      "loss": 2.3321,
+      "step": 464
+    },
+    {
+      "epoch": 0.13222983820802503,
+      "grad_norm": 3.1615819931030273,
+      "learning_rate": 3.5081873024992825e-05,
+      "loss": 3.519,
+      "step": 465
+    },
+    {
+      "epoch": 0.13251420345148315,
+      "grad_norm": 2.8588624000549316,
+      "learning_rate": 3.5070382074116636e-05,
+      "loss": 3.0476,
+      "step": 466
+    },
+    {
+      "epoch": 0.13279856869494125,
+      "grad_norm": 2.8191323280334473,
+      "learning_rate": 3.5058891123240454e-05,
+      "loss": 2.92,
+      "step": 467
+    },
+    {
+      "epoch": 0.13308293393839937,
+      "grad_norm": 2.7357499599456787,
+      "learning_rate": 3.5047400172364265e-05,
+      "loss": 2.8323,
+      "step": 468
+    },
+    {
+      "epoch": 0.1333672991818575,
+      "grad_norm": 2.820420026779175,
+      "learning_rate": 3.503590922148808e-05,
+      "loss": 2.6254,
+      "step": 469
+    },
+    {
+      "epoch": 0.13365166442531562,
+      "grad_norm": 2.1745669841766357,
+      "learning_rate": 3.5024418270611894e-05,
+      "loss": 2.5728,
+      "step": 470
+    },
+    {
+      "epoch": 0.13393602966877374,
+      "grad_norm": 1.995119571685791,
+      "learning_rate": 3.501292731973571e-05,
+      "loss": 2.3181,
+      "step": 471
+    },
+    {
+      "epoch": 0.13422039491223187,
+      "grad_norm": 2.149378776550293,
+      "learning_rate": 3.500143636885953e-05,
+      "loss": 2.1623,
+      "step": 472
+    },
+    {
+      "epoch": 0.13450476015568996,
+      "grad_norm": 3.25516414642334,
+      "learning_rate": 3.498994541798334e-05,
+      "loss": 3.3434,
+      "step": 473
+    },
+    {
+      "epoch": 0.13478912539914809,
+      "grad_norm": 2.725456714630127,
+      "learning_rate": 3.497845446710716e-05,
+      "loss": 3.1657,
+      "step": 474
+    },
+    {
+      "epoch": 0.1350734906426062,
+      "grad_norm": 2.5315914154052734,
+      "learning_rate": 3.496696351623097e-05,
+      "loss": 3.0051,
+      "step": 475
+    },
+    {
+      "epoch": 0.13535785588606433,
+      "grad_norm": 2.51139760017395,
+      "learning_rate": 3.495547256535479e-05,
+      "loss": 2.8382,
+      "step": 476
+    },
+    {
+      "epoch": 0.13564222112952246,
+      "grad_norm": 2.4551467895507812,
+      "learning_rate": 3.49439816144786e-05,
+      "loss": 2.6955,
+      "step": 477
+    },
+    {
+      "epoch": 0.13592658637298055,
+      "grad_norm": 2.2206430435180664,
+      "learning_rate": 3.493249066360242e-05,
+      "loss": 2.5027,
+      "step": 478
+    },
+    {
+      "epoch": 0.13621095161643867,
+      "grad_norm": 2.4770166873931885,
+      "learning_rate": 3.492099971272623e-05,
+      "loss": 2.2923,
+      "step": 479
+    },
+    {
+      "epoch": 0.1364953168598968,
+      "grad_norm": 2.7834012508392334,
+      "learning_rate": 3.4909508761850046e-05,
+      "loss": 2.3686,
+      "step": 480
+    },
+    {
+      "epoch": 0.13677968210335492,
+      "grad_norm": 3.358705520629883,
+      "learning_rate": 3.4898017810973864e-05,
+      "loss": 3.3837,
+      "step": 481
+    },
+    {
+      "epoch": 0.13706404734681304,
+      "grad_norm": 2.3323278427124023,
+      "learning_rate": 3.4886526860097675e-05,
+      "loss": 2.8571,
+      "step": 482
+    },
+    {
+      "epoch": 0.13734841259027114,
+      "grad_norm": 2.330890417098999,
+      "learning_rate": 3.487503590922149e-05,
+      "loss": 2.7978,
+      "step": 483
+    },
+    {
+      "epoch": 0.13763277783372926,
+      "grad_norm": 2.012991428375244,
+      "learning_rate": 3.4863544958345304e-05,
+      "loss": 2.8301,
+      "step": 484
+    },
+    {
+      "epoch": 0.1379171430771874,
+      "grad_norm": 2.383934736251831,
+      "learning_rate": 3.485205400746912e-05,
+      "loss": 2.6773,
+      "step": 485
+    },
+    {
+      "epoch": 0.1382015083206455,
+      "grad_norm": 2.576904773712158,
+      "learning_rate": 3.484056305659293e-05,
+      "loss": 2.5425,
+      "step": 486
+    },
+    {
+      "epoch": 0.13848587356410363,
+      "grad_norm": 2.482586145401001,
+      "learning_rate": 3.482907210571675e-05,
+      "loss": 2.3212,
+      "step": 487
+    },
+    {
+      "epoch": 0.13877023880756176,
+      "grad_norm": 2.269639253616333,
+      "learning_rate": 3.481758115484056e-05,
+      "loss": 2.244,
+      "step": 488
+    },
+    {
+      "epoch": 0.13905460405101985,
+      "grad_norm": 2.263761281967163,
+      "learning_rate": 3.480609020396438e-05,
+      "loss": 3.6028,
+      "step": 489
+    },
+    {
+      "epoch": 0.13933896929447798,
+      "grad_norm": 2.271108388900757,
+      "learning_rate": 3.47945992530882e-05,
+      "loss": 3.1406,
+      "step": 490
+    },
+    {
+      "epoch": 0.1396233345379361,
+      "grad_norm": 2.075087547302246,
+      "learning_rate": 3.478310830221201e-05,
+      "loss": 2.6854,
+      "step": 491
+    },
+    {
+      "epoch": 0.13990769978139422,
+      "grad_norm": 2.018319845199585,
+      "learning_rate": 3.4771617351335827e-05,
+      "loss": 2.8959,
+      "step": 492
+    },
+    {
+      "epoch": 0.14019206502485235,
+      "grad_norm": 2.369109630584717,
+      "learning_rate": 3.476012640045964e-05,
+      "loss": 2.7621,
+      "step": 493
+    },
+    {
+      "epoch": 0.14047643026831044,
+      "grad_norm": 2.4382967948913574,
+      "learning_rate": 3.4748635449583456e-05,
+      "loss": 2.4256,
+      "step": 494
+    },
+    {
+      "epoch": 0.14076079551176857,
+      "grad_norm": 2.1346540451049805,
+      "learning_rate": 3.473714449870727e-05,
+      "loss": 2.4076,
+      "step": 495
+    },
+    {
+      "epoch": 0.1410451607552267,
+      "grad_norm": 2.194206953048706,
+      "learning_rate": 3.4725653547831085e-05,
+      "loss": 2.1643,
+      "step": 496
+    },
+    {
+      "epoch": 0.1413295259986848,
+      "grad_norm": 3.0039777755737305,
+      "learning_rate": 3.4714162596954896e-05,
+      "loss": 3.4954,
+      "step": 497
+    },
+    {
+      "epoch": 0.14161389124214294,
+      "grad_norm": 2.4674746990203857,
+      "learning_rate": 3.4702671646078714e-05,
+      "loss": 3.1891,
+      "step": 498
+    },
+    {
+      "epoch": 0.14189825648560106,
+      "grad_norm": 2.319490432739258,
+      "learning_rate": 3.469118069520253e-05,
+      "loss": 2.7782,
+      "step": 499
+    },
+    {
+      "epoch": 0.14218262172905916,
+      "grad_norm": 2.3277428150177,
+      "learning_rate": 3.467968974432635e-05,
+      "loss": 2.6973,
+      "step": 500
+    },
+    {
+      "epoch": 0.14246698697251728,
+      "grad_norm": 2.28387188911438,
+      "learning_rate": 3.466819879345016e-05,
+      "loss": 2.8054,
+      "step": 501
+    },
+    {
+      "epoch": 0.1427513522159754,
+      "grad_norm": 2.186497688293457,
+      "learning_rate": 3.465670784257398e-05,
+      "loss": 2.7287,
+      "step": 502
+    },
+    {
+      "epoch": 0.14303571745943353,
+      "grad_norm": 2.455045700073242,
+      "learning_rate": 3.464521689169779e-05,
+      "loss": 2.4847,
+      "step": 503
+    },
+    {
+      "epoch": 0.14332008270289165,
+      "grad_norm": 2.292936086654663,
+      "learning_rate": 3.463372594082161e-05,
+      "loss": 2.3205,
+      "step": 504
+    },
+    {
+      "epoch": 0.14360444794634974,
+      "grad_norm": 2.924741744995117,
+      "learning_rate": 3.4622234989945425e-05,
+      "loss": 3.244,
+      "step": 505
+    },
+    {
+      "epoch": 0.14388881318980787,
+      "grad_norm": 2.4970264434814453,
+      "learning_rate": 3.4610744039069236e-05,
+      "loss": 3.0944,
+      "step": 506
+    },
+    {
+      "epoch": 0.144173178433266,
+      "grad_norm": 2.490755319595337,
+      "learning_rate": 3.4599253088193054e-05,
+      "loss": 3.0244,
+      "step": 507
+    },
+    {
+      "epoch": 0.14445754367672412,
+      "grad_norm": 2.4618477821350098,
+      "learning_rate": 3.4587762137316865e-05,
+      "loss": 2.6005,
+      "step": 508
+    },
+    {
+      "epoch": 0.14474190892018224,
+      "grad_norm": 2.505988121032715,
+      "learning_rate": 3.457627118644068e-05,
+      "loss": 2.5832,
+      "step": 509
+    },
+    {
+      "epoch": 0.14502627416364036,
+      "grad_norm": 2.5353870391845703,
+      "learning_rate": 3.4564780235564494e-05,
+      "loss": 2.3534,
+      "step": 510
+    },
+    {
+      "epoch": 0.14531063940709846,
+      "grad_norm": 2.2250514030456543,
+      "learning_rate": 3.455328928468831e-05,
+      "loss": 2.2805,
+      "step": 511
+    },
+    {
+      "epoch": 0.14559500465055658,
+      "grad_norm": 2.35164475440979,
+      "learning_rate": 3.454179833381212e-05,
+      "loss": 2.214,
+      "step": 512
+    },
+    {
+      "epoch": 0.1458793698940147,
+      "grad_norm": 2.7433321475982666,
+      "learning_rate": 3.453030738293594e-05,
+      "loss": 3.4267,
+      "step": 513
+    },
+    {
+      "epoch": 0.14616373513747283,
+      "grad_norm": 2.7333459854125977,
+      "learning_rate": 3.451881643205976e-05,
+      "loss": 3.1427,
+      "step": 514
+    },
+    {
+      "epoch": 0.14644810038093095,
+      "grad_norm": 2.3346779346466064,
+      "learning_rate": 3.450732548118357e-05,
+      "loss": 2.7976,
+      "step": 515
+    },
+    {
+      "epoch": 0.14673246562438905,
+      "grad_norm": 2.188016176223755,
+      "learning_rate": 3.449583453030739e-05,
+      "loss": 2.7736,
+      "step": 516
+    },
+    {
+      "epoch": 0.14701683086784717,
+      "grad_norm": 2.3532261848449707,
+      "learning_rate": 3.44843435794312e-05,
+      "loss": 2.6379,
+      "step": 517
+    },
+    {
+      "epoch": 0.1473011961113053,
+      "grad_norm": 2.1835358142852783,
+      "learning_rate": 3.447285262855502e-05,
+      "loss": 2.2907,
+      "step": 518
+    },
+    {
+      "epoch": 0.14758556135476342,
+      "grad_norm": 1.9752788543701172,
+      "learning_rate": 3.446136167767883e-05,
+      "loss": 2.1688,
+      "step": 519
+    },
+    {
+      "epoch": 0.14786992659822154,
+      "grad_norm": 2.0870256423950195,
+      "learning_rate": 3.4449870726802646e-05,
+      "loss": 2.1826,
+      "step": 520
+    },
+    {
+      "epoch": 0.14815429184167966,
+      "grad_norm": 3.138823986053467,
+      "learning_rate": 3.443837977592646e-05,
+      "loss": 3.3426,
+      "step": 521
+    },
+    {
+      "epoch": 0.14843865708513776,
+      "grad_norm": 2.220222234725952,
+      "learning_rate": 3.4426888825050275e-05,
+      "loss": 2.9125,
+      "step": 522
+    },
+    {
+      "epoch": 0.14872302232859588,
+      "grad_norm": 2.0182180404663086,
+      "learning_rate": 3.441539787417409e-05,
+      "loss": 2.54,
+      "step": 523
+    },
+    {
+      "epoch": 0.149007387572054,
+      "grad_norm": 2.3546602725982666,
+      "learning_rate": 3.4403906923297904e-05,
+      "loss": 2.7855,
+      "step": 524
+    },
+    {
+      "epoch": 0.14929175281551213,
+      "grad_norm": 2.4429872035980225,
+      "learning_rate": 3.439241597242172e-05,
+      "loss": 2.6199,
+      "step": 525
+    },
+    {
+      "epoch": 0.14957611805897025,
+      "grad_norm": 2.0123140811920166,
+      "learning_rate": 3.438092502154553e-05,
+      "loss": 2.4555,
+      "step": 526
+    },
+    {
+      "epoch": 0.14986048330242835,
+      "grad_norm": 2.059921979904175,
+      "learning_rate": 3.436943407066935e-05,
+      "loss": 2.3095,
+      "step": 527
+    },
+    {
+      "epoch": 0.15014484854588647,
+      "grad_norm": 2.259676456451416,
+      "learning_rate": 3.435794311979316e-05,
+      "loss": 2.32,
+      "step": 528
+    },
+    {
+      "epoch": 0.1504292137893446,
+      "grad_norm": 2.4949564933776855,
+      "learning_rate": 3.434645216891698e-05,
+      "loss": 3.1353,
+      "step": 529
+    },
+    {
+      "epoch": 0.15071357903280272,
+      "grad_norm": 2.08988881111145,
+      "learning_rate": 3.433496121804079e-05,
+      "loss": 3.0615,
+      "step": 530
+    },
+    {
+      "epoch": 0.15099794427626084,
+      "grad_norm": 2.143328905105591,
+      "learning_rate": 3.432347026716461e-05,
+      "loss": 2.7796,
+      "step": 531
+    },
+    {
+      "epoch": 0.15128230951971897,
+      "grad_norm": 2.148768663406372,
+      "learning_rate": 3.431197931628843e-05,
+      "loss": 2.6519,
+      "step": 532
+    },
+    {
+      "epoch": 0.15156667476317706,
+      "grad_norm": 2.143799066543579,
+      "learning_rate": 3.430048836541224e-05,
+      "loss": 2.7152,
+      "step": 533
+    },
+    {
+      "epoch": 0.15185104000663519,
+      "grad_norm": 2.04683780670166,
+      "learning_rate": 3.4288997414536056e-05,
+      "loss": 2.3695,
+      "step": 534
+    },
+    {
+      "epoch": 0.1521354052500933,
+      "grad_norm": 2.0776238441467285,
+      "learning_rate": 3.427750646365987e-05,
+      "loss": 2.2031,
+      "step": 535
+    },
+    {
+      "epoch": 0.15241977049355143,
+      "grad_norm": 2.1074013710021973,
+      "learning_rate": 3.4266015512783685e-05,
+      "loss": 2.3794,
+      "step": 536
+    },
+    {
+      "epoch": 0.15270413573700956,
+      "grad_norm": 3.231952667236328,
+      "learning_rate": 3.42545245619075e-05,
+      "loss": 3.2961,
+      "step": 537
+    },
+    {
+      "epoch": 0.15298850098046765,
+      "grad_norm": 2.548482656478882,
+      "learning_rate": 3.424303361103132e-05,
+      "loss": 2.9752,
+      "step": 538
+    },
+    {
+      "epoch": 0.15327286622392577,
+      "grad_norm": 2.271571397781372,
+      "learning_rate": 3.423154266015513e-05,
+      "loss": 2.9183,
+      "step": 539
+    },
+    {
+      "epoch": 0.1535572314673839,
+      "grad_norm": 2.0318918228149414,
+      "learning_rate": 3.422005170927895e-05,
+      "loss": 2.6897,
+      "step": 540
+    },
+    {
+      "epoch": 0.15384159671084202,
+      "grad_norm": 2.172313690185547,
+      "learning_rate": 3.420856075840276e-05,
+      "loss": 2.6676,
+      "step": 541
+    },
+    {
+      "epoch": 0.15412596195430014,
+      "grad_norm": 1.9842435121536255,
+      "learning_rate": 3.419706980752658e-05,
+      "loss": 2.3229,
+      "step": 542
+    },
+    {
+      "epoch": 0.15441032719775824,
+      "grad_norm": 1.9959310293197632,
+      "learning_rate": 3.418557885665039e-05,
+      "loss": 2.1866,
+      "step": 543
+    },
+    {
+      "epoch": 0.15469469244121636,
+      "grad_norm": 2.2947375774383545,
+      "learning_rate": 3.417408790577421e-05,
+      "loss": 2.3691,
+      "step": 544
+    },
+    {
+      "epoch": 0.1549790576846745,
+      "grad_norm": 3.7838504314422607,
+      "learning_rate": 3.416259695489802e-05,
+      "loss": 3.3556,
+      "step": 545
+    },
+    {
+      "epoch": 0.1552634229281326,
+      "grad_norm": 2.482161045074463,
+      "learning_rate": 3.4151106004021837e-05,
+      "loss": 3.0544,
+      "step": 546
+    },
+    {
+      "epoch": 0.15554778817159073,
+      "grad_norm": 2.1925017833709717,
+      "learning_rate": 3.4139615053145654e-05,
+      "loss": 2.8413,
+      "step": 547
+    },
+    {
+      "epoch": 0.15583215341504886,
+      "grad_norm": 1.944459080696106,
+      "learning_rate": 3.4128124102269466e-05,
+      "loss": 2.7084,
+      "step": 548
+    },
+    {
+      "epoch": 0.15611651865850695,
+      "grad_norm": 1.926611065864563,
+      "learning_rate": 3.4116633151393283e-05,
+      "loss": 2.5604,
+      "step": 549
+    },
+    {
+      "epoch": 0.15640088390196508,
+      "grad_norm": 1.8744678497314453,
+      "learning_rate": 3.4105142200517095e-05,
+      "loss": 2.2331,
+      "step": 550
+    },
+    {
+      "epoch": 0.1566852491454232,
+      "grad_norm": 2.163674831390381,
+      "learning_rate": 3.409365124964091e-05,
+      "loss": 2.2382,
+      "step": 551
+    },
+    {
+      "epoch": 0.15696961438888132,
+      "grad_norm": 2.251648426055908,
+      "learning_rate": 3.4082160298764724e-05,
+      "loss": 2.0852,
+      "step": 552
+    },
+    {
+      "epoch": 0.15725397963233945,
+      "grad_norm": 2.419774055480957,
+      "learning_rate": 3.407066934788854e-05,
+      "loss": 3.393,
+      "step": 553
+    },
+    {
+      "epoch": 0.15753834487579754,
+      "grad_norm": 1.9137029647827148,
+      "learning_rate": 3.405917839701235e-05,
+      "loss": 3.2236,
+      "step": 554
+    },
+    {
+      "epoch": 0.15782271011925567,
+      "grad_norm": 1.9087616205215454,
+      "learning_rate": 3.404768744613617e-05,
+      "loss": 2.869,
+      "step": 555
+    },
+    {
+      "epoch": 0.1581070753627138,
+      "grad_norm": 1.9507688283920288,
+      "learning_rate": 3.403619649525999e-05,
+      "loss": 2.833,
+      "step": 556
+    },
+    {
+      "epoch": 0.1583914406061719,
+      "grad_norm": 2.030203342437744,
+      "learning_rate": 3.40247055443838e-05,
+      "loss": 2.6474,
+      "step": 557
+    },
+    {
+      "epoch": 0.15867580584963004,
+      "grad_norm": 2.2038793563842773,
+      "learning_rate": 3.401321459350762e-05,
+      "loss": 2.3888,
+      "step": 558
+    },
+    {
+      "epoch": 0.15896017109308816,
+      "grad_norm": 2.0459814071655273,
+      "learning_rate": 3.400172364263143e-05,
+      "loss": 2.2976,
+      "step": 559
+    },
+    {
+      "epoch": 0.15924453633654626,
+      "grad_norm": 2.089040517807007,
+      "learning_rate": 3.3990232691755246e-05,
+      "loss": 2.3395,
+      "step": 560
+    },
+    {
+      "epoch": 0.15952890158000438,
+      "grad_norm": 2.3878250122070312,
+      "learning_rate": 3.397874174087906e-05,
+      "loss": 3.3155,
+      "step": 561
+    },
+    {
+      "epoch": 0.1598132668234625,
+      "grad_norm": 2.1346888542175293,
+      "learning_rate": 3.3967250790002875e-05,
+      "loss": 3.1624,
+      "step": 562
+    },
+    {
+      "epoch": 0.16009763206692063,
+      "grad_norm": 2.1869750022888184,
+      "learning_rate": 3.3955759839126686e-05,
+      "loss": 2.6892,
+      "step": 563
+    },
+    {
+      "epoch": 0.16038199731037875,
+      "grad_norm": 2.045581102371216,
+      "learning_rate": 3.3944268888250504e-05,
+      "loss": 2.4891,
+      "step": 564
+    },
+    {
+      "epoch": 0.16066636255383684,
+      "grad_norm": 2.3556089401245117,
+      "learning_rate": 3.393277793737432e-05,
+      "loss": 2.7332,
+      "step": 565
+    },
+    {
+      "epoch": 0.16095072779729497,
+      "grad_norm": 2.4558780193328857,
+      "learning_rate": 3.392128698649813e-05,
+      "loss": 2.4167,
+      "step": 566
+    },
+    {
+      "epoch": 0.1612350930407531,
+      "grad_norm": 2.398458957672119,
+      "learning_rate": 3.390979603562195e-05,
+      "loss": 2.336,
+      "step": 567
+    },
+    {
+      "epoch": 0.16151945828421121,
+      "grad_norm": 2.256385564804077,
+      "learning_rate": 3.389830508474576e-05,
+      "loss": 2.3393,
+      "step": 568
+    },
+    {
+      "epoch": 0.16180382352766934,
+      "grad_norm": 2.4298534393310547,
+      "learning_rate": 3.388681413386958e-05,
+      "loss": 3.3351,
+      "step": 569
+    },
+    {
+      "epoch": 0.16208818877112746,
+      "grad_norm": 2.2216269969940186,
+      "learning_rate": 3.387532318299339e-05,
+      "loss": 3.1062,
+      "step": 570
+    },
+    {
+      "epoch": 0.16237255401458556,
+      "grad_norm": 2.279859781265259,
+      "learning_rate": 3.386383223211721e-05,
+      "loss": 2.7871,
+      "step": 571
+    },
+    {
+      "epoch": 0.16265691925804368,
+      "grad_norm": 2.4930570125579834,
+      "learning_rate": 3.385234128124102e-05,
+      "loss": 2.6784,
+      "step": 572
+    },
+    {
+      "epoch": 0.1629412845015018,
+      "grad_norm": 3.0474438667297363,
+      "learning_rate": 3.384085033036484e-05,
+      "loss": 2.7209,
+      "step": 573
+    },
+    {
+      "epoch": 0.16322564974495993,
+      "grad_norm": 2.276129961013794,
+      "learning_rate": 3.3829359379488656e-05,
+      "loss": 2.3874,
+      "step": 574
+    },
+    {
+      "epoch": 0.16351001498841805,
+      "grad_norm": 2.0157735347747803,
+      "learning_rate": 3.3817868428612474e-05,
+      "loss": 2.3545,
+      "step": 575
+    },
+    {
+      "epoch": 0.16379438023187615,
+      "grad_norm": 2.2889609336853027,
+      "learning_rate": 3.3806377477736285e-05,
+      "loss": 2.1654,
+      "step": 576
+    },
+    {
+      "epoch": 0.16407874547533427,
+      "grad_norm": 2.966151714324951,
+      "learning_rate": 3.37948865268601e-05,
+      "loss": 3.3948,
+      "step": 577
+    },
+    {
+      "epoch": 0.1643631107187924,
+      "grad_norm": 2.7245430946350098,
+      "learning_rate": 3.3783395575983914e-05,
+      "loss": 3.0326,
+      "step": 578
+    },
+    {
+      "epoch": 0.16464747596225052,
+      "grad_norm": 2.4125325679779053,
+      "learning_rate": 3.377190462510773e-05,
+      "loss": 2.7461,
+      "step": 579
+    },
+    {
+      "epoch": 0.16493184120570864,
+      "grad_norm": 2.4555015563964844,
+      "learning_rate": 3.376041367423155e-05,
+      "loss": 2.6876,
+      "step": 580
+    },
+    {
+      "epoch": 0.16521620644916676,
+      "grad_norm": 2.3780357837677,
+      "learning_rate": 3.374892272335536e-05,
+      "loss": 2.549,
+      "step": 581
+    },
+    {
+      "epoch": 0.16550057169262486,
+      "grad_norm": 1.9920421838760376,
+      "learning_rate": 3.373743177247918e-05,
+      "loss": 2.3843,
+      "step": 582
+    },
+    {
+      "epoch": 0.16578493693608298,
+      "grad_norm": 1.880908489227295,
+      "learning_rate": 3.372594082160299e-05,
+      "loss": 2.1699,
+      "step": 583
+    },
+    {
+      "epoch": 0.1660693021795411,
+      "grad_norm": 2.317730188369751,
+      "learning_rate": 3.371444987072681e-05,
+      "loss": 2.3338,
+      "step": 584
+    },
+    {
+      "epoch": 0.16635366742299923,
+      "grad_norm": 3.1581149101257324,
+      "learning_rate": 3.370295891985062e-05,
+      "loss": 3.3314,
+      "step": 585
+    },
+    {
+      "epoch": 0.16663803266645735,
+      "grad_norm": 2.4569809436798096,
+      "learning_rate": 3.369146796897444e-05,
+      "loss": 2.8549,
+      "step": 586
+    },
+    {
+      "epoch": 0.16692239790991545,
+      "grad_norm": 2.0873701572418213,
+      "learning_rate": 3.367997701809825e-05,
+      "loss": 2.6875,
+      "step": 587
+    },
+    {
+      "epoch": 0.16720676315337357,
+      "grad_norm": 2.1216537952423096,
+      "learning_rate": 3.3668486067222066e-05,
+      "loss": 2.7253,
+      "step": 588
+    },
+    {
+      "epoch": 0.1674911283968317,
+      "grad_norm": 1.9596827030181885,
+      "learning_rate": 3.3656995116345884e-05,
+      "loss": 2.4127,
+      "step": 589
+    },
+    {
+      "epoch": 0.16777549364028982,
+      "grad_norm": 2.244654417037964,
+      "learning_rate": 3.3645504165469695e-05,
+      "loss": 2.4795,
+      "step": 590
+    },
+    {
+      "epoch": 0.16805985888374794,
+      "grad_norm": 2.387822389602661,
+      "learning_rate": 3.363401321459351e-05,
+      "loss": 2.413,
+      "step": 591
+    },
+    {
+      "epoch": 0.16834422412720607,
+      "grad_norm": 2.2715797424316406,
+      "learning_rate": 3.3622522263717324e-05,
+      "loss": 2.3535,
+      "step": 592
+    },
+    {
+      "epoch": 0.16862858937066416,
+      "grad_norm": 3.2900567054748535,
+      "learning_rate": 3.361103131284114e-05,
+      "loss": 3.2949,
+      "step": 593
+    },
+    {
+      "epoch": 0.16891295461412228,
+      "grad_norm": 2.340487241744995,
+      "learning_rate": 3.359954036196495e-05,
+      "loss": 3.0242,
+      "step": 594
+    },
+    {
+      "epoch": 0.1691973198575804,
+      "grad_norm": 1.9996750354766846,
+      "learning_rate": 3.358804941108877e-05,
+      "loss": 2.7541,
+      "step": 595
+    },
+    {
+      "epoch": 0.16948168510103853,
+      "grad_norm": 1.8965973854064941,
+      "learning_rate": 3.357655846021258e-05,
+      "loss": 2.6425,
+      "step": 596
+    },
+    {
+      "epoch": 0.16976605034449666,
+      "grad_norm": 2.097822427749634,
+      "learning_rate": 3.35650675093364e-05,
+      "loss": 2.6672,
+      "step": 597
+    },
+    {
+      "epoch": 0.17005041558795475,
+      "grad_norm": 2.171520471572876,
+      "learning_rate": 3.355357655846022e-05,
+      "loss": 2.2089,
+      "step": 598
+    },
+    {
+      "epoch": 0.17033478083141287,
+      "grad_norm": 2.1892077922821045,
+      "learning_rate": 3.354208560758403e-05,
+      "loss": 2.2999,
+      "step": 599
+    },
+    {
+      "epoch": 0.170619146074871,
+      "grad_norm": 1.8929511308670044,
+      "learning_rate": 3.3530594656707847e-05,
+      "loss": 2.0265,
+      "step": 600
+    },
+    {
+      "epoch": 0.17090351131832912,
+      "grad_norm": 2.2485220432281494,
+      "learning_rate": 3.351910370583166e-05,
+      "loss": 3.4067,
+      "step": 601
+    },
+    {
+      "epoch": 0.17118787656178724,
+      "grad_norm": 2.0794124603271484,
+      "learning_rate": 3.3507612754955476e-05,
+      "loss": 3.0856,
+      "step": 602
+    },
+    {
+      "epoch": 0.17147224180524534,
+      "grad_norm": 1.9198224544525146,
+      "learning_rate": 3.349612180407929e-05,
+      "loss": 2.7309,
+      "step": 603
+    },
+    {
+      "epoch": 0.17175660704870346,
+      "grad_norm": 1.9311753511428833,
+      "learning_rate": 3.3484630853203105e-05,
+      "loss": 2.5093,
+      "step": 604
+    },
+    {
+      "epoch": 0.1720409722921616,
+      "grad_norm": 2.2155637741088867,
+      "learning_rate": 3.3473139902326916e-05,
+      "loss": 2.4131,
+      "step": 605
+    },
+    {
+      "epoch": 0.1723253375356197,
+      "grad_norm": 2.4310479164123535,
+      "learning_rate": 3.3461648951450733e-05,
+      "loss": 2.3546,
+      "step": 606
+    },
+    {
+      "epoch": 0.17260970277907783,
+      "grad_norm": 2.170459747314453,
+      "learning_rate": 3.345015800057455e-05,
+      "loss": 2.1606,
+      "step": 607
+    },
+    {
+      "epoch": 0.17289406802253596,
+      "grad_norm": 2.082836627960205,
+      "learning_rate": 3.343866704969836e-05,
+      "loss": 2.1429,
+      "step": 608
+    },
+    {
+      "epoch": 0.17317843326599405,
+      "grad_norm": 2.6757774353027344,
+      "learning_rate": 3.342717609882218e-05,
+      "loss": 3.3129,
+      "step": 609
+    },
+    {
+      "epoch": 0.17346279850945218,
+      "grad_norm": 1.9752081632614136,
+      "learning_rate": 3.341568514794599e-05,
+      "loss": 2.8585,
+      "step": 610
+    },
+    {
+      "epoch": 0.1737471637529103,
+      "grad_norm": 2.269960880279541,
+      "learning_rate": 3.340419419706981e-05,
+      "loss": 2.5409,
+      "step": 611
+    },
+    {
+      "epoch": 0.17403152899636842,
+      "grad_norm": 2.2884328365325928,
+      "learning_rate": 3.339270324619363e-05,
+      "loss": 2.6172,
+      "step": 612
+    },
+    {
+      "epoch": 0.17431589423982655,
+      "grad_norm": 2.2851369380950928,
+      "learning_rate": 3.3381212295317445e-05,
+      "loss": 2.6341,
+      "step": 613
+    },
+    {
+      "epoch": 0.17460025948328464,
+      "grad_norm": 1.8490828275680542,
+      "learning_rate": 3.3369721344441256e-05,
+      "loss": 2.2735,
+      "step": 614
+    },
+    {
+      "epoch": 0.17488462472674277,
+      "grad_norm": 1.9733513593673706,
+      "learning_rate": 3.3358230393565074e-05,
+      "loss": 2.2607,
+      "step": 615
+    },
+    {
+      "epoch": 0.1751689899702009,
+      "grad_norm": 1.9823896884918213,
+      "learning_rate": 3.3346739442688885e-05,
+      "loss": 2.0564,
+      "step": 616
+    },
+    {
+      "epoch": 0.175453355213659,
+      "grad_norm": 2.479914665222168,
+      "learning_rate": 3.33352484918127e-05,
+      "loss": 3.352,
+      "step": 617
+    },
+    {
+      "epoch": 0.17573772045711714,
+      "grad_norm": 2.3320770263671875,
+      "learning_rate": 3.3323757540936514e-05,
+      "loss": 2.8721,
+      "step": 618
+    },
+    {
+      "epoch": 0.17602208570057526,
+      "grad_norm": 2.451782703399658,
+      "learning_rate": 3.331226659006033e-05,
+      "loss": 2.8098,
+      "step": 619
+    },
+    {
+      "epoch": 0.17630645094403335,
+      "grad_norm": 2.2985215187072754,
+      "learning_rate": 3.330077563918414e-05,
+      "loss": 2.7563,
+      "step": 620
+    },
+    {
+      "epoch": 0.17659081618749148,
+      "grad_norm": 2.119804859161377,
+      "learning_rate": 3.328928468830796e-05,
+      "loss": 2.5972,
+      "step": 621
+    },
+    {
+      "epoch": 0.1768751814309496,
+      "grad_norm": 2.0667002201080322,
+      "learning_rate": 3.327779373743178e-05,
+      "loss": 2.2254,
+      "step": 622
+    },
+    {
+      "epoch": 0.17715954667440773,
+      "grad_norm": 1.9303321838378906,
+      "learning_rate": 3.326630278655559e-05,
+      "loss": 2.4747,
+      "step": 623
+    },
+    {
+      "epoch": 0.17744391191786585,
+      "grad_norm": 2.147676706314087,
+      "learning_rate": 3.325481183567941e-05,
+      "loss": 1.9882,
+      "step": 624
+    },
+    {
+      "epoch": 0.17772827716132394,
+      "grad_norm": 2.619396448135376,
+      "learning_rate": 3.324332088480322e-05,
+      "loss": 3.2464,
+      "step": 625
+    },
+    {
+      "epoch": 0.17801264240478207,
+      "grad_norm": 2.6688385009765625,
+      "learning_rate": 3.323182993392704e-05,
+      "loss": 3.0559,
+      "step": 626
+    },
+    {
+      "epoch": 0.1782970076482402,
+      "grad_norm": 1.9910519123077393,
+      "learning_rate": 3.322033898305085e-05,
+      "loss": 2.7364,
+      "step": 627
+    },
+    {
+      "epoch": 0.17858137289169831,
+      "grad_norm": 2.037667989730835,
+      "learning_rate": 3.3208848032174666e-05,
+      "loss": 2.3946,
+      "step": 628
+    },
+    {
+      "epoch": 0.17886573813515644,
+      "grad_norm": 2.0019378662109375,
+      "learning_rate": 3.319735708129848e-05,
+      "loss": 2.4888,
+      "step": 629
+    },
+    {
+      "epoch": 0.17915010337861456,
+      "grad_norm": 1.999667763710022,
+      "learning_rate": 3.3185866130422295e-05,
+      "loss": 2.5945,
+      "step": 630
+    },
+    {
+      "epoch": 0.17943446862207266,
+      "grad_norm": 2.3381664752960205,
+      "learning_rate": 3.317437517954611e-05,
+      "loss": 2.0536,
+      "step": 631
+    },
+    {
+      "epoch": 0.17971883386553078,
+      "grad_norm": 2.1102583408355713,
+      "learning_rate": 3.3162884228669924e-05,
+      "loss": 2.1843,
+      "step": 632
+    },
+    {
+      "epoch": 0.1800031991089889,
+      "grad_norm": 4.009433269500732,
+      "learning_rate": 3.315139327779374e-05,
+      "loss": 3.122,
+      "step": 633
+    },
+    {
+      "epoch": 0.18028756435244703,
+      "grad_norm": 2.4775569438934326,
+      "learning_rate": 3.313990232691755e-05,
+      "loss": 3.1036,
+      "step": 634
+    },
+    {
+      "epoch": 0.18057192959590515,
+      "grad_norm": 2.1339893341064453,
+      "learning_rate": 3.312841137604137e-05,
+      "loss": 2.8175,
+      "step": 635
+    },
+    {
+      "epoch": 0.18085629483936325,
+      "grad_norm": 1.9820502996444702,
+      "learning_rate": 3.311692042516518e-05,
+      "loss": 2.8805,
+      "step": 636
+    },
+    {
+      "epoch": 0.18114066008282137,
+      "grad_norm": 1.9298404455184937,
+      "learning_rate": 3.3105429474289e-05,
+      "loss": 2.5823,
+      "step": 637
+    },
+    {
+      "epoch": 0.1814250253262795,
+      "grad_norm": 1.915541172027588,
+      "learning_rate": 3.309393852341281e-05,
+      "loss": 2.4167,
+      "step": 638
+    },
+    {
+      "epoch": 0.18170939056973762,
+      "grad_norm": 2.0158815383911133,
+      "learning_rate": 3.308244757253663e-05,
+      "loss": 2.2251,
+      "step": 639
+    },
+    {
+      "epoch": 0.18199375581319574,
+      "grad_norm": 1.840295672416687,
+      "learning_rate": 3.307095662166045e-05,
+      "loss": 2.2033,
+      "step": 640
+    },
+    {
+      "epoch": 0.18227812105665386,
+      "grad_norm": 2.2416555881500244,
+      "learning_rate": 3.305946567078426e-05,
+      "loss": 3.302,
+      "step": 641
+    },
+    {
+      "epoch": 0.18256248630011196,
+      "grad_norm": 2.0106818675994873,
+      "learning_rate": 3.3047974719908076e-05,
+      "loss": 2.8823,
+      "step": 642
+    },
+    {
+      "epoch": 0.18284685154357008,
+      "grad_norm": 2.0697238445281982,
+      "learning_rate": 3.303648376903189e-05,
+      "loss": 2.8577,
+      "step": 643
+    },
+    {
+      "epoch": 0.1831312167870282,
+      "grad_norm": 1.8144038915634155,
+      "learning_rate": 3.3024992818155705e-05,
+      "loss": 2.5007,
+      "step": 644
+    },
+    {
+      "epoch": 0.18341558203048633,
+      "grad_norm": 1.9172033071517944,
+      "learning_rate": 3.3013501867279516e-05,
+      "loss": 2.5931,
+      "step": 645
+    },
+    {
+      "epoch": 0.18369994727394445,
+      "grad_norm": 2.0230233669281006,
+      "learning_rate": 3.3002010916403334e-05,
+      "loss": 2.3733,
+      "step": 646
+    },
+    {
+      "epoch": 0.18398431251740255,
+      "grad_norm": 1.8354929685592651,
+      "learning_rate": 3.2990519965527145e-05,
+      "loss": 2.1317,
+      "step": 647
+    },
+    {
+      "epoch": 0.18426867776086067,
+      "grad_norm": 2.032435894012451,
+      "learning_rate": 3.297902901465096e-05,
+      "loss": 1.9827,
+      "step": 648
+    },
+    {
+      "epoch": 0.1845530430043188,
+      "grad_norm": 2.6801607608795166,
+      "learning_rate": 3.296753806377478e-05,
+      "loss": 3.3537,
+      "step": 649
+    },
+    {
+      "epoch": 0.18483740824777692,
+      "grad_norm": 2.1917881965637207,
+      "learning_rate": 3.29560471128986e-05,
+      "loss": 3.0921,
+      "step": 650
+    },
+    {
+      "epoch": 0.18512177349123504,
+      "grad_norm": 1.906441330909729,
+      "learning_rate": 3.294455616202241e-05,
+      "loss": 2.8275,
+      "step": 651
+    },
+    {
+      "epoch": 0.18540613873469317,
+      "grad_norm": 1.8117749691009521,
+      "learning_rate": 3.293306521114623e-05,
+      "loss": 2.5992,
+      "step": 652
+    },
+    {
+      "epoch": 0.18569050397815126,
+      "grad_norm": 1.962047815322876,
+      "learning_rate": 3.292157426027004e-05,
+      "loss": 2.4936,
+      "step": 653
+    },
+    {
+      "epoch": 0.18597486922160938,
+      "grad_norm": 1.9997988939285278,
+      "learning_rate": 3.2910083309393856e-05,
+      "loss": 2.3512,
+      "step": 654
+    },
+    {
+      "epoch": 0.1862592344650675,
+      "grad_norm": 2.2575020790100098,
+      "learning_rate": 3.2898592358517674e-05,
+      "loss": 2.3052,
+      "step": 655
+    },
+    {
+      "epoch": 0.18654359970852563,
+      "grad_norm": 2.3353512287139893,
+      "learning_rate": 3.2887101407641485e-05,
+      "loss": 2.1795,
+      "step": 656
+    },
+    {
+      "epoch": 0.18682796495198375,
+      "grad_norm": 2.2830419540405273,
+      "learning_rate": 3.28756104567653e-05,
+      "loss": 3.2798,
+      "step": 657
+    },
+    {
+      "epoch": 0.18711233019544185,
+      "grad_norm": 2.1097829341888428,
+      "learning_rate": 3.2864119505889114e-05,
+      "loss": 3.1461,
+      "step": 658
+    },
+    {
+      "epoch": 0.18739669543889997,
+      "grad_norm": 2.0392520427703857,
+      "learning_rate": 3.285262855501293e-05,
+      "loss": 2.865,
+      "step": 659
+    },
+    {
+      "epoch": 0.1876810606823581,
+      "grad_norm": 1.9128637313842773,
+      "learning_rate": 3.2841137604136743e-05,
+      "loss": 2.4607,
+      "step": 660
+    },
+    {
+      "epoch": 0.18796542592581622,
+      "grad_norm": 2.3386354446411133,
+      "learning_rate": 3.282964665326056e-05,
+      "loss": 2.5321,
+      "step": 661
+    },
+    {
+      "epoch": 0.18824979116927434,
+      "grad_norm": 2.1804091930389404,
+      "learning_rate": 3.281815570238437e-05,
+      "loss": 2.372,
+      "step": 662
+    },
+    {
+      "epoch": 0.18853415641273247,
+      "grad_norm": 2.139617919921875,
+      "learning_rate": 3.280666475150819e-05,
+      "loss": 2.2559,
+      "step": 663
+    },
+    {
+      "epoch": 0.18881852165619056,
+      "grad_norm": 2.0755178928375244,
+      "learning_rate": 3.279517380063201e-05,
+      "loss": 2.3333,
+      "step": 664
+    },
+    {
+      "epoch": 0.1891028868996487,
+      "grad_norm": 2.1971609592437744,
+      "learning_rate": 3.278368284975582e-05,
+      "loss": 3.1885,
+      "step": 665
+    },
+    {
+      "epoch": 0.1893872521431068,
+      "grad_norm": 2.080216884613037,
+      "learning_rate": 3.277219189887964e-05,
+      "loss": 2.8508,
+      "step": 666
+    },
+    {
+      "epoch": 0.18967161738656493,
+      "grad_norm": 2.1171205043792725,
+      "learning_rate": 3.276070094800345e-05,
+      "loss": 2.5981,
+      "step": 667
+    },
+    {
+      "epoch": 0.18995598263002306,
+      "grad_norm": 2.2514328956604004,
+      "learning_rate": 3.2749209997127266e-05,
+      "loss": 2.8185,
+      "step": 668
+    },
+    {
+      "epoch": 0.19024034787348115,
+      "grad_norm": 2.328221082687378,
+      "learning_rate": 3.273771904625108e-05,
+      "loss": 2.6257,
+      "step": 669
+    },
+    {
+      "epoch": 0.19052471311693928,
+      "grad_norm": 2.0298240184783936,
+      "learning_rate": 3.2726228095374895e-05,
+      "loss": 2.2706,
+      "step": 670
+    },
+    {
+      "epoch": 0.1908090783603974,
+      "grad_norm": 1.9545693397521973,
+      "learning_rate": 3.2714737144498706e-05,
+      "loss": 2.1073,
+      "step": 671
+    },
+    {
+      "epoch": 0.19109344360385552,
+      "grad_norm": 2.0932371616363525,
+      "learning_rate": 3.2703246193622524e-05,
+      "loss": 2.1358,
+      "step": 672
+    },
+    {
+      "epoch": 0.19137780884731365,
+      "grad_norm": 2.68083119392395,
+      "learning_rate": 3.269175524274634e-05,
+      "loss": 3.3797,
+      "step": 673
+    },
+    {
+      "epoch": 0.19166217409077174,
+      "grad_norm": 2.340952157974243,
+      "learning_rate": 3.268026429187015e-05,
+      "loss": 2.7555,
+      "step": 674
+    },
+    {
+      "epoch": 0.19194653933422987,
+      "grad_norm": 2.199805736541748,
+      "learning_rate": 3.266877334099397e-05,
+      "loss": 2.6694,
+      "step": 675
+    },
+    {
+      "epoch": 0.192230904577688,
+      "grad_norm": 2.1734702587127686,
+      "learning_rate": 3.265728239011778e-05,
+      "loss": 2.8257,
+      "step": 676
+    },
+    {
+      "epoch": 0.1925152698211461,
+      "grad_norm": 2.420964479446411,
+      "learning_rate": 3.26457914392416e-05,
+      "loss": 2.4486,
+      "step": 677
+    },
+    {
+      "epoch": 0.19279963506460424,
+      "grad_norm": 1.9791522026062012,
+      "learning_rate": 3.263430048836541e-05,
+      "loss": 2.3201,
+      "step": 678
+    },
+    {
+      "epoch": 0.19308400030806236,
+      "grad_norm": 2.2870163917541504,
+      "learning_rate": 3.262280953748923e-05,
+      "loss": 2.1877,
+      "step": 679
+    },
+    {
+      "epoch": 0.19336836555152045,
+      "grad_norm": 2.085327625274658,
+      "learning_rate": 3.261131858661304e-05,
+      "loss": 2.0054,
+      "step": 680
+    },
+    {
+      "epoch": 0.19365273079497858,
+      "grad_norm": 3.2444210052490234,
+      "learning_rate": 3.259982763573686e-05,
+      "loss": 3.2208,
+      "step": 681
+    },
+    {
+      "epoch": 0.1939370960384367,
+      "grad_norm": 2.47835111618042,
+      "learning_rate": 3.2588336684860676e-05,
+      "loss": 3.1121,
+      "step": 682
+    },
+    {
+      "epoch": 0.19422146128189482,
+      "grad_norm": 2.056704521179199,
+      "learning_rate": 3.257684573398449e-05,
+      "loss": 2.8531,
+      "step": 683
+    },
+    {
+      "epoch": 0.19450582652535295,
+      "grad_norm": 1.9477993249893188,
+      "learning_rate": 3.2565354783108305e-05,
+      "loss": 2.4773,
+      "step": 684
+    },
+    {
+      "epoch": 0.19479019176881104,
+      "grad_norm": 2.019652843475342,
+      "learning_rate": 3.2553863832232116e-05,
+      "loss": 2.5318,
+      "step": 685
+    },
+    {
+      "epoch": 0.19507455701226917,
+      "grad_norm": 2.164877414703369,
+      "learning_rate": 3.2542372881355934e-05,
+      "loss": 2.2676,
+      "step": 686
+    },
+    {
+      "epoch": 0.1953589222557273,
+      "grad_norm": 1.9417436122894287,
+      "learning_rate": 3.253088193047975e-05,
+      "loss": 2.1097,
+      "step": 687
+    },
+    {
+      "epoch": 0.19564328749918541,
+      "grad_norm": 2.0217719078063965,
+      "learning_rate": 3.251939097960357e-05,
+      "loss": 1.9464,
+      "step": 688
+    },
+    {
+      "epoch": 0.19592765274264354,
+      "grad_norm": 2.3088855743408203,
+      "learning_rate": 3.250790002872738e-05,
+      "loss": 3.1466,
+      "step": 689
+    },
+    {
+      "epoch": 0.19621201798610166,
+      "grad_norm": 2.2107951641082764,
+      "learning_rate": 3.24964090778512e-05,
+      "loss": 2.898,
+      "step": 690
+    },
+    {
+      "epoch": 0.19649638322955976,
+      "grad_norm": 1.8729802370071411,
+      "learning_rate": 3.248491812697501e-05,
+      "loss": 2.5574,
+      "step": 691
+    },
+    {
+      "epoch": 0.19678074847301788,
+      "grad_norm": 1.948519229888916,
+      "learning_rate": 3.247342717609883e-05,
+      "loss": 2.644,
+      "step": 692
+    },
+    {
+      "epoch": 0.197065113716476,
+      "grad_norm": 2.479282855987549,
+      "learning_rate": 3.246193622522264e-05,
+      "loss": 2.6382,
+      "step": 693
+    },
+    {
+      "epoch": 0.19734947895993413,
+      "grad_norm": 2.2494328022003174,
+      "learning_rate": 3.245044527434646e-05,
+      "loss": 2.0895,
+      "step": 694
+    },
+    {
+      "epoch": 0.19763384420339225,
+      "grad_norm": 1.9501408338546753,
+      "learning_rate": 3.243895432347027e-05,
+      "loss": 2.2874,
+      "step": 695
+    },
+    {
+      "epoch": 0.19791820944685035,
+      "grad_norm": 1.865615963935852,
+      "learning_rate": 3.2427463372594086e-05,
+      "loss": 2.0848,
+      "step": 696
+    },
+    {
+      "epoch": 0.19820257469030847,
+      "grad_norm": 2.376199722290039,
+      "learning_rate": 3.2415972421717904e-05,
+      "loss": 3.3114,
+      "step": 697
+    },
+    {
+      "epoch": 0.1984869399337666,
+      "grad_norm": 1.978826880455017,
+      "learning_rate": 3.2404481470841715e-05,
+      "loss": 2.6896,
+      "step": 698
+    },
+    {
+      "epoch": 0.19877130517722472,
+      "grad_norm": 1.8862276077270508,
+      "learning_rate": 3.239299051996553e-05,
+      "loss": 2.4757,
+      "step": 699
+    },
+    {
+      "epoch": 0.19905567042068284,
+      "grad_norm": 2.253225564956665,
+      "learning_rate": 3.2381499569089344e-05,
+      "loss": 2.7394,
+      "step": 700
+    },
+    {
+      "epoch": 0.19934003566414096,
+      "grad_norm": 2.258186101913452,
+      "learning_rate": 3.237000861821316e-05,
+      "loss": 2.6104,
+      "step": 701
+    },
+    {
+      "epoch": 0.19962440090759906,
+      "grad_norm": 1.78932785987854,
+      "learning_rate": 3.235851766733697e-05,
+      "loss": 2.2367,
+      "step": 702
+    },
+    {
+      "epoch": 0.19990876615105718,
+      "grad_norm": 1.7839140892028809,
+      "learning_rate": 3.234702671646079e-05,
+      "loss": 2.0307,
+      "step": 703
+    },
+    {
+      "epoch": 0.2001931313945153,
+      "grad_norm": 1.9078693389892578,
+      "learning_rate": 3.23355357655846e-05,
+      "loss": 1.8974,
+      "step": 704
+    },
+    {
+      "epoch": 0.20047749663797343,
+      "grad_norm": 2.489147901535034,
+      "learning_rate": 3.232404481470842e-05,
+      "loss": 3.0721,
+      "step": 705
+    },
+    {
+      "epoch": 0.20076186188143155,
+      "grad_norm": 2.5214760303497314,
+      "learning_rate": 3.231255386383224e-05,
+      "loss": 2.9676,
+      "step": 706
+    },
+    {
+      "epoch": 0.20104622712488965,
+      "grad_norm": 2.182615280151367,
+      "learning_rate": 3.230106291295605e-05,
+      "loss": 2.9474,
+      "step": 707
+    },
+    {
+      "epoch": 0.20133059236834777,
+      "grad_norm": 1.98862624168396,
+      "learning_rate": 3.2289571962079866e-05,
+      "loss": 2.7995,
+      "step": 708
+    },
+    {
+      "epoch": 0.2016149576118059,
+      "grad_norm": 1.9838632345199585,
+      "learning_rate": 3.227808101120368e-05,
+      "loss": 2.4476,
+      "step": 709
+    },
+    {
+      "epoch": 0.20189932285526402,
+      "grad_norm": 1.9730520248413086,
+      "learning_rate": 3.2266590060327495e-05,
+      "loss": 2.3272,
+      "step": 710
+    },
+    {
+      "epoch": 0.20218368809872214,
+      "grad_norm": 1.880563497543335,
+      "learning_rate": 3.2255099109451307e-05,
+      "loss": 2.0332,
+      "step": 711
+    },
+    {
+      "epoch": 0.20246805334218027,
+      "grad_norm": 1.99998140335083,
+      "learning_rate": 3.2243608158575124e-05,
+      "loss": 2.2419,
+      "step": 712
+    },
+    {
+      "epoch": 0.20275241858563836,
+      "grad_norm": 2.910835027694702,
+      "learning_rate": 3.2232117207698936e-05,
+      "loss": 3.2948,
+      "step": 713
+    },
+    {
+      "epoch": 0.20303678382909648,
+      "grad_norm": 2.386836528778076,
+      "learning_rate": 3.2220626256822753e-05,
+      "loss": 2.7419,
+      "step": 714
+    },
+    {
+      "epoch": 0.2033211490725546,
+      "grad_norm": 1.9169079065322876,
+      "learning_rate": 3.220913530594657e-05,
+      "loss": 2.6171,
+      "step": 715
+    },
+    {
+      "epoch": 0.20360551431601273,
+      "grad_norm": 1.7654900550842285,
+      "learning_rate": 3.219764435507038e-05,
+      "loss": 2.7043,
+      "step": 716
+    },
+    {
+      "epoch": 0.20388987955947085,
+      "grad_norm": 1.8636305332183838,
+      "learning_rate": 3.21861534041942e-05,
+      "loss": 2.4612,
+      "step": 717
+    },
+    {
+      "epoch": 0.20417424480292895,
+      "grad_norm": 2.11191463470459,
+      "learning_rate": 3.217466245331801e-05,
+      "loss": 2.3006,
+      "step": 718
+    },
+    {
+      "epoch": 0.20445861004638707,
+      "grad_norm": 2.05869460105896,
+      "learning_rate": 3.216317150244183e-05,
+      "loss": 2.2266,
+      "step": 719
+    },
+    {
+      "epoch": 0.2047429752898452,
+      "grad_norm": 1.921918511390686,
+      "learning_rate": 3.215168055156564e-05,
+      "loss": 1.9656,
+      "step": 720
+    },
+    {
+      "epoch": 0.20502734053330332,
+      "grad_norm": 3.3702752590179443,
+      "learning_rate": 3.214018960068946e-05,
+      "loss": 3.208,
+      "step": 721
+    },
+    {
+      "epoch": 0.20531170577676144,
+      "grad_norm": 2.2111897468566895,
+      "learning_rate": 3.212869864981327e-05,
+      "loss": 2.8252,
+      "step": 722
+    },
+    {
+      "epoch": 0.20559607102021957,
+      "grad_norm": 1.9135637283325195,
+      "learning_rate": 3.2117207698937094e-05,
+      "loss": 2.3325,
+      "step": 723
+    },
+    {
+      "epoch": 0.20588043626367766,
+      "grad_norm": 1.9720232486724854,
+      "learning_rate": 3.2105716748060905e-05,
+      "loss": 2.5609,
+      "step": 724
+    },
+    {
+      "epoch": 0.2061648015071358,
+      "grad_norm": 2.031601667404175,
+      "learning_rate": 3.209422579718472e-05,
+      "loss": 2.6579,
+      "step": 725
+    },
+    {
+      "epoch": 0.2064491667505939,
+      "grad_norm": 1.992948055267334,
+      "learning_rate": 3.2082734846308534e-05,
+      "loss": 2.4377,
+      "step": 726
+    },
+    {
+      "epoch": 0.20673353199405203,
+      "grad_norm": 2.026014566421509,
+      "learning_rate": 3.207124389543235e-05,
+      "loss": 1.9672,
+      "step": 727
+    },
+    {
+      "epoch": 0.20701789723751016,
+      "grad_norm": 1.9836902618408203,
+      "learning_rate": 3.205975294455616e-05,
+      "loss": 2.2928,
+      "step": 728
+    },
+    {
+      "epoch": 0.20730226248096825,
+      "grad_norm": 2.541131019592285,
+      "learning_rate": 3.204826199367998e-05,
+      "loss": 3.3255,
+      "step": 729
+    },
+    {
+      "epoch": 0.20758662772442638,
+      "grad_norm": 2.0298244953155518,
+      "learning_rate": 3.20367710428038e-05,
+      "loss": 2.6907,
+      "step": 730
+    },
+    {
+      "epoch": 0.2078709929678845,
+      "grad_norm": 1.8890761137008667,
+      "learning_rate": 3.202528009192761e-05,
+      "loss": 2.5107,
+      "step": 731
+    },
+    {
+      "epoch": 0.20815535821134262,
+      "grad_norm": 1.8663610219955444,
+      "learning_rate": 3.201378914105143e-05,
+      "loss": 2.4509,
+      "step": 732
+    },
+    {
+      "epoch": 0.20843972345480075,
+      "grad_norm": 1.906890630722046,
+      "learning_rate": 3.200229819017524e-05,
+      "loss": 2.5638,
+      "step": 733
+    },
+    {
+      "epoch": 0.20872408869825884,
+      "grad_norm": 1.9237983226776123,
+      "learning_rate": 3.199080723929906e-05,
+      "loss": 2.4034,
+      "step": 734
+    },
+    {
+      "epoch": 0.20900845394171697,
+      "grad_norm": 2.0110905170440674,
+      "learning_rate": 3.197931628842287e-05,
+      "loss": 2.242,
+      "step": 735
+    },
+    {
+      "epoch": 0.2092928191851751,
+      "grad_norm": 2.0842487812042236,
+      "learning_rate": 3.1967825337546686e-05,
+      "loss": 2.1287,
+      "step": 736
+    },
+    {
+      "epoch": 0.2095771844286332,
+      "grad_norm": 2.772614002227783,
+      "learning_rate": 3.19563343866705e-05,
+      "loss": 3.0786,
+      "step": 737
+    },
+    {
+      "epoch": 0.20986154967209134,
+      "grad_norm": 2.3587253093719482,
+      "learning_rate": 3.1944843435794315e-05,
+      "loss": 2.7391,
+      "step": 738
+    },
+    {
+      "epoch": 0.21014591491554946,
+      "grad_norm": 1.897799015045166,
+      "learning_rate": 3.193335248491813e-05,
+      "loss": 2.5878,
+      "step": 739
+    },
+    {
+      "epoch": 0.21043028015900755,
+      "grad_norm": 1.919022560119629,
+      "learning_rate": 3.1921861534041944e-05,
+      "loss": 2.6763,
+      "step": 740
+    },
+    {
+      "epoch": 0.21071464540246568,
+      "grad_norm": 2.1270596981048584,
+      "learning_rate": 3.191037058316576e-05,
+      "loss": 2.3583,
+      "step": 741
+    },
+    {
+      "epoch": 0.2109990106459238,
+      "grad_norm": 1.8899205923080444,
+      "learning_rate": 3.189887963228957e-05,
+      "loss": 2.4652,
+      "step": 742
+    },
+    {
+      "epoch": 0.21128337588938192,
+      "grad_norm": 1.9961659908294678,
+      "learning_rate": 3.188738868141339e-05,
+      "loss": 2.275,
+      "step": 743
+    },
+    {
+      "epoch": 0.21156774113284005,
+      "grad_norm": 2.287783622741699,
+      "learning_rate": 3.18758977305372e-05,
+      "loss": 2.0559,
+      "step": 744
+    },
+    {
+      "epoch": 0.21185210637629814,
+      "grad_norm": 2.3738584518432617,
+      "learning_rate": 3.186440677966102e-05,
+      "loss": 3.1947,
+      "step": 745
+    },
+    {
+      "epoch": 0.21213647161975627,
+      "grad_norm": 2.055588960647583,
+      "learning_rate": 3.185291582878483e-05,
+      "loss": 2.881,
+      "step": 746
+    },
+    {
+      "epoch": 0.2124208368632144,
+      "grad_norm": 2.0361130237579346,
+      "learning_rate": 3.184142487790865e-05,
+      "loss": 2.6499,
+      "step": 747
+    },
+    {
+      "epoch": 0.21270520210667251,
+      "grad_norm": 1.918145775794983,
+      "learning_rate": 3.182993392703247e-05,
+      "loss": 2.3748,
+      "step": 748
+    },
+    {
+      "epoch": 0.21298956735013064,
+      "grad_norm": 2.2371819019317627,
+      "learning_rate": 3.181844297615628e-05,
+      "loss": 2.5316,
+      "step": 749
+    },
+    {
+      "epoch": 0.21327393259358876,
+      "grad_norm": 2.0490219593048096,
+      "learning_rate": 3.1806952025280096e-05,
+      "loss": 2.3496,
+      "step": 750
+    },
+    {
+      "epoch": 0.21355829783704686,
+      "grad_norm": 1.884697437286377,
+      "learning_rate": 3.179546107440391e-05,
+      "loss": 2.168,
+      "step": 751
+    },
+    {
+      "epoch": 0.21384266308050498,
+      "grad_norm": 1.9147650003433228,
+      "learning_rate": 3.1783970123527725e-05,
+      "loss": 2.1791,
+      "step": 752
+    },
+    {
+      "epoch": 0.2141270283239631,
+      "grad_norm": 2.1511213779449463,
+      "learning_rate": 3.1772479172651536e-05,
+      "loss": 3.073,
+      "step": 753
+    },
+    {
+      "epoch": 0.21441139356742123,
+      "grad_norm": 1.8281089067459106,
+      "learning_rate": 3.1760988221775354e-05,
+      "loss": 2.7645,
+      "step": 754
+    },
+    {
+      "epoch": 0.21469575881087935,
+      "grad_norm": 1.7798327207565308,
+      "learning_rate": 3.1749497270899165e-05,
+      "loss": 2.6769,
+      "step": 755
+    },
+    {
+      "epoch": 0.21498012405433745,
+      "grad_norm": 1.9437909126281738,
+      "learning_rate": 3.173800632002298e-05,
+      "loss": 2.7385,
+      "step": 756
+    },
+    {
+      "epoch": 0.21526448929779557,
+      "grad_norm": 2.2421510219573975,
+      "learning_rate": 3.17265153691468e-05,
+      "loss": 2.3904,
+      "step": 757
+    },
+    {
+      "epoch": 0.2155488545412537,
+      "grad_norm": 2.002697467803955,
+      "learning_rate": 3.171502441827061e-05,
+      "loss": 2.1965,
+      "step": 758
+    },
+    {
+      "epoch": 0.21583321978471182,
+      "grad_norm": 2.018341541290283,
+      "learning_rate": 3.170353346739443e-05,
+      "loss": 2.1783,
+      "step": 759
+    },
+    {
+      "epoch": 0.21611758502816994,
+      "grad_norm": 1.9171587228775024,
+      "learning_rate": 3.169204251651825e-05,
+      "loss": 1.9632,
+      "step": 760
+    },
+    {
+      "epoch": 0.21640195027162806,
+      "grad_norm": 2.128800868988037,
+      "learning_rate": 3.168055156564206e-05,
+      "loss": 3.1176,
+      "step": 761
+    },
+    {
+      "epoch": 0.21668631551508616,
+      "grad_norm": 2.2135539054870605,
+      "learning_rate": 3.1669060614765876e-05,
+      "loss": 2.9083,
+      "step": 762
+    },
+    {
+      "epoch": 0.21697068075854428,
+      "grad_norm": 2.01832914352417,
+      "learning_rate": 3.1657569663889694e-05,
+      "loss": 2.5822,
+      "step": 763
+    },
+    {
+      "epoch": 0.2172550460020024,
+      "grad_norm": 1.827118992805481,
+      "learning_rate": 3.1646078713013505e-05,
+      "loss": 2.5527,
+      "step": 764
+    },
+    {
+      "epoch": 0.21753941124546053,
+      "grad_norm": 1.9703270196914673,
+      "learning_rate": 3.163458776213732e-05,
+      "loss": 2.61,
+      "step": 765
+    },
+    {
+      "epoch": 0.21782377648891865,
+      "grad_norm": 2.0175693035125732,
+      "learning_rate": 3.1623096811261134e-05,
+      "loss": 2.1853,
+      "step": 766
+    },
+    {
+      "epoch": 0.21810814173237675,
+      "grad_norm": 2.11944580078125,
+      "learning_rate": 3.161160586038495e-05,
+      "loss": 1.7576,
+      "step": 767
+    },
+    {
+      "epoch": 0.21839250697583487,
+      "grad_norm": 2.100249767303467,
+      "learning_rate": 3.1600114909508763e-05,
+      "loss": 1.9592,
+      "step": 768
+    },
+    {
+      "epoch": 0.218676872219293,
+      "grad_norm": 2.8305320739746094,
+      "learning_rate": 3.158862395863258e-05,
+      "loss": 3.3594,
+      "step": 769
+    },
+    {
+      "epoch": 0.21896123746275112,
+      "grad_norm": 2.1420650482177734,
+      "learning_rate": 3.157713300775639e-05,
+      "loss": 2.847,
+      "step": 770
+    },
+    {
+      "epoch": 0.21924560270620924,
+      "grad_norm": 1.9816900491714478,
+      "learning_rate": 3.156564205688021e-05,
+      "loss": 2.5702,
+      "step": 771
+    },
+    {
+      "epoch": 0.21952996794966737,
+      "grad_norm": 1.9326121807098389,
+      "learning_rate": 3.155415110600403e-05,
+      "loss": 2.6882,
+      "step": 772
+    },
+    {
+      "epoch": 0.21981433319312546,
+      "grad_norm": 2.1733944416046143,
+      "learning_rate": 3.154266015512784e-05,
+      "loss": 2.7475,
+      "step": 773
+    },
+    {
+      "epoch": 0.22009869843658358,
+      "grad_norm": 1.7192537784576416,
+      "learning_rate": 3.153116920425166e-05,
+      "loss": 2.3554,
+      "step": 774
+    },
+    {
+      "epoch": 0.2203830636800417,
+      "grad_norm": 1.9640390872955322,
+      "learning_rate": 3.151967825337547e-05,
+      "loss": 2.1575,
+      "step": 775
+    },
+    {
+      "epoch": 0.22066742892349983,
+      "grad_norm": 2.0537362098693848,
+      "learning_rate": 3.1508187302499286e-05,
+      "loss": 2.2478,
+      "step": 776
+    },
+    {
+      "epoch": 0.22095179416695795,
+      "grad_norm": 2.381430149078369,
+      "learning_rate": 3.14966963516231e-05,
+      "loss": 2.936,
+      "step": 777
+    },
+    {
+      "epoch": 0.22123615941041605,
+      "grad_norm": 2.202801465988159,
+      "learning_rate": 3.1485205400746915e-05,
+      "loss": 2.9745,
+      "step": 778
+    },
+    {
+      "epoch": 0.22152052465387417,
+      "grad_norm": 1.809591293334961,
+      "learning_rate": 3.1473714449870726e-05,
+      "loss": 2.6893,
+      "step": 779
+    },
+    {
+      "epoch": 0.2218048898973323,
+      "grad_norm": 1.792960524559021,
+      "learning_rate": 3.1462223498994544e-05,
+      "loss": 2.4085,
+      "step": 780
+    },
+    {
+      "epoch": 0.22208925514079042,
+      "grad_norm": 1.961672067642212,
+      "learning_rate": 3.145073254811836e-05,
+      "loss": 2.5024,
+      "step": 781
+    },
+    {
+      "epoch": 0.22237362038424854,
+      "grad_norm": 1.8171442747116089,
+      "learning_rate": 3.143924159724217e-05,
+      "loss": 2.3142,
+      "step": 782
+    },
+    {
+      "epoch": 0.22265798562770667,
+      "grad_norm": 1.9773929119110107,
+      "learning_rate": 3.142775064636599e-05,
+      "loss": 2.1619,
+      "step": 783
+    },
+    {
+      "epoch": 0.22294235087116476,
+      "grad_norm": 1.815152645111084,
+      "learning_rate": 3.14162596954898e-05,
+      "loss": 2.0866,
+      "step": 784
+    },
+    {
+      "epoch": 0.2232267161146229,
+      "grad_norm": 2.7723894119262695,
+      "learning_rate": 3.140476874461362e-05,
+      "loss": 3.3017,
+      "step": 785
+    },
+    {
+      "epoch": 0.223511081358081,
+      "grad_norm": 1.9993256330490112,
+      "learning_rate": 3.139327779373743e-05,
+      "loss": 2.6683,
+      "step": 786
+    },
+    {
+      "epoch": 0.22379544660153913,
+      "grad_norm": 1.7609732151031494,
+      "learning_rate": 3.138178684286125e-05,
+      "loss": 2.5893,
+      "step": 787
+    },
+    {
+      "epoch": 0.22407981184499726,
+      "grad_norm": 1.7786259651184082,
+      "learning_rate": 3.137029589198506e-05,
+      "loss": 2.4328,
+      "step": 788
+    },
+    {
+      "epoch": 0.22436417708845535,
+      "grad_norm": 1.887364149093628,
+      "learning_rate": 3.135880494110888e-05,
+      "loss": 2.5221,
+      "step": 789
+    },
+    {
+      "epoch": 0.22464854233191348,
+      "grad_norm": 1.8501337766647339,
+      "learning_rate": 3.1347313990232696e-05,
+      "loss": 2.2003,
+      "step": 790
+    },
+    {
+      "epoch": 0.2249329075753716,
+      "grad_norm": 1.7945717573165894,
+      "learning_rate": 3.133582303935651e-05,
+      "loss": 2.1568,
+      "step": 791
+    },
+    {
+      "epoch": 0.22521727281882972,
+      "grad_norm": 1.8720512390136719,
+      "learning_rate": 3.1324332088480325e-05,
+      "loss": 2.104,
+      "step": 792
+    },
+    {
+      "epoch": 0.22550163806228785,
+      "grad_norm": 2.1769962310791016,
+      "learning_rate": 3.1312841137604136e-05,
+      "loss": 3.2416,
+      "step": 793
+    },
+    {
+      "epoch": 0.22578600330574594,
+      "grad_norm": 2.1001853942871094,
+      "learning_rate": 3.1301350186727954e-05,
+      "loss": 2.6237,
+      "step": 794
+    },
+    {
+      "epoch": 0.22607036854920406,
+      "grad_norm": 1.819381594657898,
+      "learning_rate": 3.1289859235851765e-05,
+      "loss": 2.5273,
+      "step": 795
+    },
+    {
+      "epoch": 0.2263547337926622,
+      "grad_norm": 1.706929326057434,
+      "learning_rate": 3.127836828497558e-05,
+      "loss": 2.5795,
+      "step": 796
+    },
+    {
+      "epoch": 0.2266390990361203,
+      "grad_norm": 2.0382096767425537,
+      "learning_rate": 3.12668773340994e-05,
+      "loss": 2.5978,
+      "step": 797
+    },
+    {
+      "epoch": 0.22692346427957844,
+      "grad_norm": 1.7794004678726196,
+      "learning_rate": 3.125538638322322e-05,
+      "loss": 2.2474,
+      "step": 798
+    },
+    {
+      "epoch": 0.22720782952303656,
+      "grad_norm": 1.8751369714736938,
+      "learning_rate": 3.124389543234703e-05,
+      "loss": 2.0745,
+      "step": 799
+    },
+    {
+      "epoch": 0.22749219476649465,
+      "grad_norm": 1.9768788814544678,
+      "learning_rate": 3.123240448147085e-05,
+      "loss": 2.2523,
+      "step": 800
+    },
+    {
+      "epoch": 0.22777656000995278,
+      "grad_norm": 2.430659532546997,
+      "learning_rate": 3.122091353059466e-05,
+      "loss": 3.2516,
+      "step": 801
+    },
+    {
+      "epoch": 0.2280609252534109,
+      "grad_norm": 1.9983506202697754,
+      "learning_rate": 3.120942257971848e-05,
+      "loss": 2.7283,
+      "step": 802
+    },
+    {
+      "epoch": 0.22834529049686902,
+      "grad_norm": 1.8245633840560913,
+      "learning_rate": 3.119793162884229e-05,
+      "loss": 2.4896,
+      "step": 803
+    },
+    {
+      "epoch": 0.22862965574032715,
+      "grad_norm": 1.724178433418274,
+      "learning_rate": 3.1186440677966106e-05,
+      "loss": 2.5445,
+      "step": 804
+    },
+    {
+      "epoch": 0.22891402098378524,
+      "grad_norm": 1.851968765258789,
+      "learning_rate": 3.1174949727089924e-05,
+      "loss": 2.3454,
+      "step": 805
+    },
+    {
+      "epoch": 0.22919838622724337,
+      "grad_norm": 1.7947938442230225,
+      "learning_rate": 3.1163458776213735e-05,
+      "loss": 2.2357,
+      "step": 806
+    },
+    {
+      "epoch": 0.2294827514707015,
+      "grad_norm": 2.108400821685791,
+      "learning_rate": 3.115196782533755e-05,
+      "loss": 2.2438,
+      "step": 807
+    },
+    {
+      "epoch": 0.2297671167141596,
+      "grad_norm": 2.0238804817199707,
+      "learning_rate": 3.1140476874461364e-05,
+      "loss": 1.9718,
+      "step": 808
+    },
+    {
+      "epoch": 0.23005148195761774,
+      "grad_norm": 2.4201109409332275,
+      "learning_rate": 3.112898592358518e-05,
+      "loss": 3.2521,
+      "step": 809
+    },
+    {
+      "epoch": 0.23033584720107586,
+      "grad_norm": 1.9365209341049194,
+      "learning_rate": 3.111749497270899e-05,
+      "loss": 2.9715,
+      "step": 810
+    },
+    {
+      "epoch": 0.23062021244453396,
+      "grad_norm": 1.670932650566101,
+      "learning_rate": 3.110600402183281e-05,
+      "loss": 2.6555,
+      "step": 811
+    },
+    {
+      "epoch": 0.23090457768799208,
+      "grad_norm": 1.8441979885101318,
+      "learning_rate": 3.109451307095662e-05,
+      "loss": 2.6097,
+      "step": 812
+    },
+    {
+      "epoch": 0.2311889429314502,
+      "grad_norm": 2.0533034801483154,
+      "learning_rate": 3.108302212008044e-05,
+      "loss": 2.4527,
+      "step": 813
+    },
+    {
+      "epoch": 0.23147330817490833,
+      "grad_norm": 1.9839503765106201,
+      "learning_rate": 3.107153116920426e-05,
+      "loss": 2.3322,
+      "step": 814
+    },
+    {
+      "epoch": 0.23175767341836645,
+      "grad_norm": 1.8730031251907349,
+      "learning_rate": 3.106004021832807e-05,
+      "loss": 2.0321,
+      "step": 815
+    },
+    {
+      "epoch": 0.23204203866182455,
+      "grad_norm": 2.0080573558807373,
+      "learning_rate": 3.1048549267451886e-05,
+      "loss": 2.295,
+      "step": 816
+    },
+    {
+      "epoch": 0.23232640390528267,
+      "grad_norm": 2.2020275592803955,
+      "learning_rate": 3.10370583165757e-05,
+      "loss": 3.0295,
+      "step": 817
+    },
+    {
+      "epoch": 0.2326107691487408,
+      "grad_norm": 1.7750929594039917,
+      "learning_rate": 3.1025567365699515e-05,
+      "loss": 2.8698,
+      "step": 818
+    },
+    {
+      "epoch": 0.23289513439219892,
+      "grad_norm": 1.6771564483642578,
+      "learning_rate": 3.1014076414823326e-05,
+      "loss": 2.417,
+      "step": 819
+    },
+    {
+      "epoch": 0.23317949963565704,
+      "grad_norm": 1.8319196701049805,
+      "learning_rate": 3.1002585463947144e-05,
+      "loss": 2.3378,
+      "step": 820
+    },
+    {
+      "epoch": 0.23346386487911516,
+      "grad_norm": 2.1067543029785156,
+      "learning_rate": 3.0991094513070955e-05,
+      "loss": 2.434,
+      "step": 821
+    },
+    {
+      "epoch": 0.23374823012257326,
+      "grad_norm": 2.0174291133880615,
+      "learning_rate": 3.097960356219477e-05,
+      "loss": 2.2196,
+      "step": 822
+    },
+    {
+      "epoch": 0.23403259536603138,
+      "grad_norm": 1.8617424964904785,
+      "learning_rate": 3.096811261131859e-05,
+      "loss": 2.0332,
+      "step": 823
+    },
+    {
+      "epoch": 0.2343169606094895,
+      "grad_norm": 1.8981695175170898,
+      "learning_rate": 3.09566216604424e-05,
+      "loss": 2.0747,
+      "step": 824
+    },
+    {
+      "epoch": 0.23460132585294763,
+      "grad_norm": 2.3870060443878174,
+      "learning_rate": 3.094513070956622e-05,
+      "loss": 3.1643,
+      "step": 825
+    },
+    {
+      "epoch": 0.23488569109640575,
+      "grad_norm": 2.0108344554901123,
+      "learning_rate": 3.093363975869003e-05,
+      "loss": 2.8178,
+      "step": 826
+    },
+    {
+      "epoch": 0.23517005633986385,
+      "grad_norm": 2.198622941970825,
+      "learning_rate": 3.092214880781385e-05,
+      "loss": 2.7069,
+      "step": 827
+    },
+    {
+      "epoch": 0.23545442158332197,
+      "grad_norm": 2.032430648803711,
+      "learning_rate": 3.091065785693766e-05,
+      "loss": 2.4352,
+      "step": 828
+    },
+    {
+      "epoch": 0.2357387868267801,
+      "grad_norm": 1.968604564666748,
+      "learning_rate": 3.089916690606148e-05,
+      "loss": 2.2391,
+      "step": 829
+    },
+    {
+      "epoch": 0.23602315207023822,
+      "grad_norm": 1.6530208587646484,
+      "learning_rate": 3.088767595518529e-05,
+      "loss": 2.223,
+      "step": 830
+    },
+    {
+      "epoch": 0.23630751731369634,
+      "grad_norm": 2.004181385040283,
+      "learning_rate": 3.087618500430911e-05,
+      "loss": 2.0482,
+      "step": 831
+    },
+    {
+      "epoch": 0.23659188255715446,
+      "grad_norm": 2.0865046977996826,
+      "learning_rate": 3.0864694053432925e-05,
+      "loss": 2.1199,
+      "step": 832
+    },
+    {
+      "epoch": 0.23687624780061256,
+      "grad_norm": 2.4281649589538574,
+      "learning_rate": 3.0853203102556736e-05,
+      "loss": 3.1147,
+      "step": 833
+    },
+    {
+      "epoch": 0.23716061304407068,
+      "grad_norm": 2.327183246612549,
+      "learning_rate": 3.0841712151680554e-05,
+      "loss": 2.7999,
+      "step": 834
+    },
+    {
+      "epoch": 0.2374449782875288,
+      "grad_norm": 1.8897318840026855,
+      "learning_rate": 3.083022120080437e-05,
+      "loss": 2.6127,
+      "step": 835
+    },
+    {
+      "epoch": 0.23772934353098693,
+      "grad_norm": 1.755764126777649,
+      "learning_rate": 3.081873024992818e-05,
+      "loss": 2.5684,
+      "step": 836
+    },
+    {
+      "epoch": 0.23801370877444505,
+      "grad_norm": 1.9734392166137695,
+      "learning_rate": 3.0807239299052e-05,
+      "loss": 2.3049,
+      "step": 837
+    },
+    {
+      "epoch": 0.23829807401790315,
+      "grad_norm": 1.9231913089752197,
+      "learning_rate": 3.079574834817582e-05,
+      "loss": 2.3946,
+      "step": 838
+    },
+    {
+      "epoch": 0.23858243926136127,
+      "grad_norm": 2.0235962867736816,
+      "learning_rate": 3.078425739729963e-05,
+      "loss": 2.2068,
+      "step": 839
+    },
+    {
+      "epoch": 0.2388668045048194,
+      "grad_norm": 2.155137538909912,
+      "learning_rate": 3.077276644642345e-05,
+      "loss": 2.0929,
+      "step": 840
+    },
+    {
+      "epoch": 0.23915116974827752,
+      "grad_norm": 2.36690092086792,
+      "learning_rate": 3.076127549554726e-05,
+      "loss": 3.1167,
+      "step": 841
+    },
+    {
+      "epoch": 0.23943553499173564,
+      "grad_norm": 2.195228099822998,
+      "learning_rate": 3.074978454467108e-05,
+      "loss": 3.0185,
+      "step": 842
+    },
+    {
+      "epoch": 0.23971990023519377,
+      "grad_norm": 1.9150617122650146,
+      "learning_rate": 3.073829359379489e-05,
+      "loss": 2.6971,
+      "step": 843
+    },
+    {
+      "epoch": 0.24000426547865186,
+      "grad_norm": 1.8287773132324219,
+      "learning_rate": 3.0726802642918706e-05,
+      "loss": 2.6606,
+      "step": 844
+    },
+    {
+      "epoch": 0.24028863072210999,
+      "grad_norm": 1.9123808145523071,
+      "learning_rate": 3.071531169204252e-05,
+      "loss": 2.3256,
+      "step": 845
+    },
+    {
+      "epoch": 0.2405729959655681,
+      "grad_norm": 2.077585458755493,
+      "learning_rate": 3.0703820741166335e-05,
+      "loss": 2.1924,
+      "step": 846
+    },
+    {
+      "epoch": 0.24085736120902623,
+      "grad_norm": 2.0052852630615234,
+      "learning_rate": 3.069232979029015e-05,
+      "loss": 1.979,
+      "step": 847
+    },
+    {
+      "epoch": 0.24114172645248436,
+      "grad_norm": 2.1259374618530273,
+      "learning_rate": 3.0680838839413964e-05,
+      "loss": 2.0226,
+      "step": 848
+    },
+    {
+      "epoch": 0.24142609169594245,
+      "grad_norm": 2.7498273849487305,
+      "learning_rate": 3.066934788853778e-05,
+      "loss": 3.2752,
+      "step": 849
+    },
+    {
+      "epoch": 0.24171045693940058,
+      "grad_norm": 1.9522055387496948,
+      "learning_rate": 3.065785693766159e-05,
+      "loss": 2.7302,
+      "step": 850
+    },
+    {
+      "epoch": 0.2419948221828587,
+      "grad_norm": 1.9767372608184814,
+      "learning_rate": 3.064636598678541e-05,
+      "loss": 2.7091,
+      "step": 851
+    },
+    {
+      "epoch": 0.24227918742631682,
+      "grad_norm": 1.8098385334014893,
+      "learning_rate": 3.063487503590922e-05,
+      "loss": 2.5372,
+      "step": 852
+    },
+    {
+      "epoch": 0.24256355266977495,
+      "grad_norm": 1.995309829711914,
+      "learning_rate": 3.062338408503304e-05,
+      "loss": 2.3491,
+      "step": 853
+    },
+    {
+      "epoch": 0.24284791791323304,
+      "grad_norm": 1.838243842124939,
+      "learning_rate": 3.061189313415685e-05,
+      "loss": 2.3286,
+      "step": 854
+    },
+    {
+      "epoch": 0.24313228315669116,
+      "grad_norm": 1.8721081018447876,
+      "learning_rate": 3.060040218328067e-05,
+      "loss": 2.1154,
+      "step": 855
+    },
+    {
+      "epoch": 0.2434166484001493,
+      "grad_norm": 2.050607919692993,
+      "learning_rate": 3.0588911232404487e-05,
+      "loss": 1.938,
+      "step": 856
+    },
+    {
+      "epoch": 0.2437010136436074,
+      "grad_norm": 2.1715247631073,
+      "learning_rate": 3.05774202815283e-05,
+      "loss": 2.9436,
+      "step": 857
+    },
+    {
+      "epoch": 0.24398537888706553,
+      "grad_norm": 2.1600615978240967,
+      "learning_rate": 3.0565929330652116e-05,
+      "loss": 2.6224,
+      "step": 858
+    },
+    {
+      "epoch": 0.24426974413052366,
+      "grad_norm": 1.890982747077942,
+      "learning_rate": 3.055443837977593e-05,
+      "loss": 2.6999,
+      "step": 859
+    },
+    {
+      "epoch": 0.24455410937398175,
+      "grad_norm": 1.9753085374832153,
+      "learning_rate": 3.0542947428899745e-05,
+      "loss": 2.4645,
+      "step": 860
+    },
+    {
+      "epoch": 0.24483847461743988,
+      "grad_norm": 2.096250295639038,
+      "learning_rate": 3.0531456478023556e-05,
+      "loss": 2.534,
+      "step": 861
+    },
+    {
+      "epoch": 0.245122839860898,
+      "grad_norm": 1.9475681781768799,
+      "learning_rate": 3.0519965527147374e-05,
+      "loss": 2.1234,
+      "step": 862
+    },
+    {
+      "epoch": 0.24540720510435612,
+      "grad_norm": 1.8136141300201416,
+      "learning_rate": 3.0508474576271188e-05,
+      "loss": 1.917,
+      "step": 863
+    },
+    {
+      "epoch": 0.24569157034781425,
+      "grad_norm": 2.0293166637420654,
+      "learning_rate": 3.0496983625395003e-05,
+      "loss": 2.2178,
+      "step": 864
+    },
+    {
+      "epoch": 0.24597593559127234,
+      "grad_norm": 2.072183847427368,
+      "learning_rate": 3.0485492674518817e-05,
+      "loss": 2.9106,
+      "step": 865
+    },
+    {
+      "epoch": 0.24626030083473047,
+      "grad_norm": 1.7826175689697266,
+      "learning_rate": 3.047400172364263e-05,
+      "loss": 2.9296,
+      "step": 866
+    },
+    {
+      "epoch": 0.2465446660781886,
+      "grad_norm": 1.876105546951294,
+      "learning_rate": 3.0462510772766446e-05,
+      "loss": 2.6027,
+      "step": 867
+    },
+    {
+      "epoch": 0.2468290313216467,
+      "grad_norm": 1.8935751914978027,
+      "learning_rate": 3.045101982189026e-05,
+      "loss": 2.4412,
+      "step": 868
+    },
+    {
+      "epoch": 0.24711339656510484,
+      "grad_norm": 2.0665159225463867,
+      "learning_rate": 3.043952887101408e-05,
+      "loss": 2.271,
+      "step": 869
+    },
+    {
+      "epoch": 0.24739776180856296,
+      "grad_norm": 1.7189526557922363,
+      "learning_rate": 3.0428037920137893e-05,
+      "loss": 2.2023,
+      "step": 870
+    },
+    {
+      "epoch": 0.24768212705202106,
+      "grad_norm": 1.9814810752868652,
+      "learning_rate": 3.041654696926171e-05,
+      "loss": 2.1293,
+      "step": 871
+    },
+    {
+      "epoch": 0.24796649229547918,
+      "grad_norm": 1.9453988075256348,
+      "learning_rate": 3.0405056018385525e-05,
+      "loss": 2.0083,
+      "step": 872
+    },
+    {
+      "epoch": 0.2482508575389373,
+      "grad_norm": 2.7860076427459717,
+      "learning_rate": 3.039356506750934e-05,
+      "loss": 3.0749,
+      "step": 873
+    },
+    {
+      "epoch": 0.24853522278239543,
+      "grad_norm": 1.9733988046646118,
+      "learning_rate": 3.0382074116633154e-05,
+      "loss": 2.6634,
+      "step": 874
+    },
+    {
+      "epoch": 0.24881958802585355,
+      "grad_norm": 1.8551710844039917,
+      "learning_rate": 3.0370583165756972e-05,
+      "loss": 2.4293,
+      "step": 875
+    },
+    {
+      "epoch": 0.24910395326931165,
+      "grad_norm": 1.7632685899734497,
+      "learning_rate": 3.0359092214880787e-05,
+      "loss": 2.2907,
+      "step": 876
+    },
+    {
+      "epoch": 0.24938831851276977,
+      "grad_norm": 1.9409010410308838,
+      "learning_rate": 3.03476012640046e-05,
+      "loss": 2.4988,
+      "step": 877
+    },
+    {
+      "epoch": 0.2496726837562279,
+      "grad_norm": 1.7946302890777588,
+      "learning_rate": 3.0336110313128416e-05,
+      "loss": 2.0384,
+      "step": 878
+    },
+    {
+      "epoch": 0.24995704899968602,
+      "grad_norm": 2.018123149871826,
+      "learning_rate": 3.032461936225223e-05,
+      "loss": 1.9721,
+      "step": 879
+    },
+    {
+      "epoch": 0.2502414142431441,
+      "grad_norm": 1.960157036781311,
+      "learning_rate": 3.0313128411376045e-05,
+      "loss": 2.194,
+      "step": 880
+    },
+    {
+      "epoch": 0.25052577948660226,
+      "grad_norm": 2.0886595249176025,
+      "learning_rate": 3.030163746049986e-05,
+      "loss": 2.9749,
+      "step": 881
+    },
+    {
+      "epoch": 0.25081014473006036,
+      "grad_norm": 1.888434886932373,
+      "learning_rate": 3.0290146509623674e-05,
+      "loss": 2.8315,
+      "step": 882
+    },
+    {
+      "epoch": 0.2510945099735185,
+      "grad_norm": 1.9592381715774536,
+      "learning_rate": 3.0278655558747488e-05,
+      "loss": 2.4951,
+      "step": 883
+    },
+    {
+      "epoch": 0.2513788752169766,
+      "grad_norm": 1.7165327072143555,
+      "learning_rate": 3.0267164607871306e-05,
+      "loss": 2.5657,
+      "step": 884
+    },
+    {
+      "epoch": 0.2516632404604347,
+      "grad_norm": 2.0365653038024902,
+      "learning_rate": 3.025567365699512e-05,
+      "loss": 2.5552,
+      "step": 885
+    },
+    {
+      "epoch": 0.25194760570389285,
+      "grad_norm": 1.9661747217178345,
+      "learning_rate": 3.0244182706118935e-05,
+      "loss": 2.2448,
+      "step": 886
+    },
+    {
+      "epoch": 0.25223197094735095,
+      "grad_norm": 1.8233864307403564,
+      "learning_rate": 3.023269175524275e-05,
+      "loss": 2.0185,
+      "step": 887
+    },
+    {
+      "epoch": 0.2525163361908091,
+      "grad_norm": 2.1719167232513428,
+      "learning_rate": 3.0221200804366564e-05,
+      "loss": 2.224,
+      "step": 888
+    },
+    {
+      "epoch": 0.2528007014342672,
+      "grad_norm": 2.498225450515747,
+      "learning_rate": 3.020970985349038e-05,
+      "loss": 3.2098,
+      "step": 889
+    },
+    {
+      "epoch": 0.2530850666777253,
+      "grad_norm": 1.9225410223007202,
+      "learning_rate": 3.0198218902614193e-05,
+      "loss": 2.5611,
+      "step": 890
+    },
+    {
+      "epoch": 0.25336943192118344,
+      "grad_norm": 1.8655143976211548,
+      "learning_rate": 3.0186727951738008e-05,
+      "loss": 2.6871,
+      "step": 891
+    },
+    {
+      "epoch": 0.25365379716464154,
+      "grad_norm": 1.7953119277954102,
+      "learning_rate": 3.0175237000861822e-05,
+      "loss": 2.6092,
+      "step": 892
+    },
+    {
+      "epoch": 0.2539381624080997,
+      "grad_norm": 2.0147297382354736,
+      "learning_rate": 3.016374604998564e-05,
+      "loss": 2.45,
+      "step": 893
+    },
+    {
+      "epoch": 0.2542225276515578,
+      "grad_norm": 1.827158808708191,
+      "learning_rate": 3.0152255099109454e-05,
+      "loss": 2.0963,
+      "step": 894
+    },
+    {
+      "epoch": 0.2545068928950159,
+      "grad_norm": 1.9638267755508423,
+      "learning_rate": 3.014076414823327e-05,
+      "loss": 1.9395,
+      "step": 895
+    },
+    {
+      "epoch": 0.25479125813847403,
+      "grad_norm": 1.9481165409088135,
+      "learning_rate": 3.0129273197357083e-05,
+      "loss": 2.171,
+      "step": 896
+    },
+    {
+      "epoch": 0.2550756233819321,
+      "grad_norm": 2.1484596729278564,
+      "learning_rate": 3.0117782246480898e-05,
+      "loss": 3.081,
+      "step": 897
+    },
+    {
+      "epoch": 0.2553599886253903,
+      "grad_norm": 1.8102747201919556,
+      "learning_rate": 3.0106291295604712e-05,
+      "loss": 2.6534,
+      "step": 898
+    },
+    {
+      "epoch": 0.2556443538688484,
+      "grad_norm": 1.7093149423599243,
+      "learning_rate": 3.0094800344728527e-05,
+      "loss": 2.5904,
+      "step": 899
+    },
+    {
+      "epoch": 0.2559287191123065,
+      "grad_norm": 1.7110258340835571,
+      "learning_rate": 3.008330939385234e-05,
+      "loss": 2.6272,
+      "step": 900
+    },
+    {
+      "epoch": 0.2562130843557646,
+      "grad_norm": 1.8184210062026978,
+      "learning_rate": 3.0071818442976156e-05,
+      "loss": 2.3048,
+      "step": 901
+    },
+    {
+      "epoch": 0.2564974495992227,
+      "grad_norm": 1.7951569557189941,
+      "learning_rate": 3.0060327492099974e-05,
+      "loss": 2.1765,
+      "step": 902
+    },
+    {
+      "epoch": 0.25678181484268087,
+      "grad_norm": 1.8752012252807617,
+      "learning_rate": 3.0048836541223788e-05,
+      "loss": 2.065,
+      "step": 903
+    },
+    {
+      "epoch": 0.25706618008613896,
+      "grad_norm": 1.838429570198059,
+      "learning_rate": 3.0037345590347603e-05,
+      "loss": 1.9967,
+      "step": 904
+    },
+    {
+      "epoch": 0.2573505453295971,
+      "grad_norm": 2.226714849472046,
+      "learning_rate": 3.0025854639471417e-05,
+      "loss": 3.2629,
+      "step": 905
+    },
+    {
+      "epoch": 0.2576349105730552,
+      "grad_norm": 1.8551560640335083,
+      "learning_rate": 3.0014363688595232e-05,
+      "loss": 2.7787,
+      "step": 906
+    },
+    {
+      "epoch": 0.2579192758165133,
+      "grad_norm": 1.9678069353103638,
+      "learning_rate": 3.0002872737719046e-05,
+      "loss": 2.7359,
+      "step": 907
+    },
+    {
+      "epoch": 0.25820364105997146,
+      "grad_norm": 1.7606141567230225,
+      "learning_rate": 2.9991381786842864e-05,
+      "loss": 2.4614,
+      "step": 908
+    },
+    {
+      "epoch": 0.25848800630342955,
+      "grad_norm": 1.9030214548110962,
+      "learning_rate": 2.9979890835966682e-05,
+      "loss": 2.6742,
+      "step": 909
+    },
+    {
+      "epoch": 0.2587723715468877,
+      "grad_norm": 1.7596023082733154,
+      "learning_rate": 2.9968399885090497e-05,
+      "loss": 2.4495,
+      "step": 910
+    },
+    {
+      "epoch": 0.2590567367903458,
+      "grad_norm": 1.7681137323379517,
+      "learning_rate": 2.995690893421431e-05,
+      "loss": 1.8863,
+      "step": 911
+    },
+    {
+      "epoch": 0.2593411020338039,
+      "grad_norm": 1.8547093868255615,
+      "learning_rate": 2.9945417983338126e-05,
+      "loss": 1.8526,
+      "step": 912
+    },
+    {
+      "epoch": 0.25962546727726205,
+      "grad_norm": 2.3479535579681396,
+      "learning_rate": 2.993392703246194e-05,
+      "loss": 3.0925,
+      "step": 913
+    },
+    {
+      "epoch": 0.25990983252072014,
+      "grad_norm": 2.0483782291412354,
+      "learning_rate": 2.9922436081585755e-05,
+      "loss": 2.8808,
+      "step": 914
+    },
+    {
+      "epoch": 0.2601941977641783,
+      "grad_norm": 1.8194479942321777,
+      "learning_rate": 2.991094513070957e-05,
+      "loss": 2.7473,
+      "step": 915
+    },
+    {
+      "epoch": 0.2604785630076364,
+      "grad_norm": 1.7477372884750366,
+      "learning_rate": 2.9899454179833384e-05,
+      "loss": 2.5109,
+      "step": 916
+    },
+    {
+      "epoch": 0.2607629282510945,
+      "grad_norm": 1.8002467155456543,
+      "learning_rate": 2.9887963228957198e-05,
+      "loss": 2.3975,
+      "step": 917
+    },
+    {
+      "epoch": 0.26104729349455263,
+      "grad_norm": 1.7082115411758423,
+      "learning_rate": 2.9876472278081016e-05,
+      "loss": 2.0252,
+      "step": 918
+    },
+    {
+      "epoch": 0.26133165873801073,
+      "grad_norm": 1.8353215456008911,
+      "learning_rate": 2.986498132720483e-05,
+      "loss": 1.9635,
+      "step": 919
+    },
+    {
+      "epoch": 0.2616160239814689,
+      "grad_norm": 1.9129843711853027,
+      "learning_rate": 2.9853490376328645e-05,
+      "loss": 1.8942,
+      "step": 920
+    },
+    {
+      "epoch": 0.261900389224927,
+      "grad_norm": 2.272021770477295,
+      "learning_rate": 2.984199942545246e-05,
+      "loss": 3.1266,
+      "step": 921
+    },
+    {
+      "epoch": 0.26218475446838513,
+      "grad_norm": 1.8906854391098022,
+      "learning_rate": 2.9830508474576274e-05,
+      "loss": 2.7551,
+      "step": 922
+    },
+    {
+      "epoch": 0.2624691197118432,
+      "grad_norm": 1.786133050918579,
+      "learning_rate": 2.981901752370009e-05,
+      "loss": 2.529,
+      "step": 923
+    },
+    {
+      "epoch": 0.2627534849553013,
+      "grad_norm": 1.7091472148895264,
+      "learning_rate": 2.9807526572823903e-05,
+      "loss": 2.4381,
+      "step": 924
+    },
+    {
+      "epoch": 0.26303785019875947,
+      "grad_norm": 2.077052354812622,
+      "learning_rate": 2.9796035621947717e-05,
+      "loss": 2.4781,
+      "step": 925
+    },
+    {
+      "epoch": 0.26332221544221757,
+      "grad_norm": 1.774610996246338,
+      "learning_rate": 2.9784544671071532e-05,
+      "loss": 2.156,
+      "step": 926
+    },
+    {
+      "epoch": 0.2636065806856757,
+      "grad_norm": 1.7777446508407593,
+      "learning_rate": 2.977305372019535e-05,
+      "loss": 2.1573,
+      "step": 927
+    },
+    {
+      "epoch": 0.2638909459291338,
+      "grad_norm": 1.9106909036636353,
+      "learning_rate": 2.9761562769319164e-05,
+      "loss": 1.9771,
+      "step": 928
+    },
+    {
+      "epoch": 0.2641753111725919,
+      "grad_norm": 2.073526620864868,
+      "learning_rate": 2.975007181844298e-05,
+      "loss": 3.0332,
+      "step": 929
+    },
+    {
+      "epoch": 0.26445967641605006,
+      "grad_norm": 1.7309014797210693,
+      "learning_rate": 2.9738580867566793e-05,
+      "loss": 2.7578,
+      "step": 930
+    },
+    {
+      "epoch": 0.26474404165950816,
+      "grad_norm": 1.7269665002822876,
+      "learning_rate": 2.9727089916690608e-05,
+      "loss": 2.5512,
+      "step": 931
+    },
+    {
+      "epoch": 0.2650284069029663,
+      "grad_norm": 1.8180932998657227,
+      "learning_rate": 2.9715598965814422e-05,
+      "loss": 2.6815,
+      "step": 932
+    },
+    {
+      "epoch": 0.2653127721464244,
+      "grad_norm": 1.8024221658706665,
+      "learning_rate": 2.9704108014938237e-05,
+      "loss": 2.306,
+      "step": 933
+    },
+    {
+      "epoch": 0.2655971373898825,
+      "grad_norm": 1.8379640579223633,
+      "learning_rate": 2.969261706406205e-05,
+      "loss": 2.1043,
+      "step": 934
+    },
+    {
+      "epoch": 0.26588150263334065,
+      "grad_norm": 1.738338828086853,
+      "learning_rate": 2.9681126113185866e-05,
+      "loss": 1.9482,
+      "step": 935
+    },
+    {
+      "epoch": 0.26616586787679875,
+      "grad_norm": 1.7984474897384644,
+      "learning_rate": 2.9669635162309684e-05,
+      "loss": 2.0764,
+      "step": 936
+    },
+    {
+      "epoch": 0.2664502331202569,
+      "grad_norm": 2.280787229537964,
+      "learning_rate": 2.9658144211433498e-05,
+      "loss": 2.9034,
+      "step": 937
+    },
+    {
+      "epoch": 0.266734598363715,
+      "grad_norm": 1.7945760488510132,
+      "learning_rate": 2.9646653260557313e-05,
+      "loss": 2.5243,
+      "step": 938
+    },
+    {
+      "epoch": 0.2670189636071731,
+      "grad_norm": 1.812246322631836,
+      "learning_rate": 2.9635162309681127e-05,
+      "loss": 2.4984,
+      "step": 939
+    },
+    {
+      "epoch": 0.26730332885063124,
+      "grad_norm": 1.9260566234588623,
+      "learning_rate": 2.962367135880494e-05,
+      "loss": 2.5072,
+      "step": 940
+    },
+    {
+      "epoch": 0.26758769409408933,
+      "grad_norm": 2.1809868812561035,
+      "learning_rate": 2.9612180407928756e-05,
+      "loss": 2.2971,
+      "step": 941
+    },
+    {
+      "epoch": 0.2678720593375475,
+      "grad_norm": 2.036618232727051,
+      "learning_rate": 2.960068945705257e-05,
+      "loss": 2.3523,
+      "step": 942
+    },
+    {
+      "epoch": 0.2681564245810056,
+      "grad_norm": 2.0626561641693115,
+      "learning_rate": 2.9589198506176385e-05,
+      "loss": 2.1757,
+      "step": 943
+    },
+    {
+      "epoch": 0.26844078982446373,
+      "grad_norm": 1.8505395650863647,
+      "learning_rate": 2.95777075553002e-05,
+      "loss": 2.0444,
+      "step": 944
+    },
+    {
+      "epoch": 0.26872515506792183,
+      "grad_norm": 2.1572868824005127,
+      "learning_rate": 2.956621660442402e-05,
+      "loss": 3.1699,
+      "step": 945
+    },
+    {
+      "epoch": 0.2690095203113799,
+      "grad_norm": 2.3714983463287354,
+      "learning_rate": 2.9554725653547835e-05,
+      "loss": 2.6934,
+      "step": 946
+    },
+    {
+      "epoch": 0.2692938855548381,
+      "grad_norm": 2.13543701171875,
+      "learning_rate": 2.954323470267165e-05,
+      "loss": 2.4942,
+      "step": 947
+    },
+    {
+      "epoch": 0.26957825079829617,
+      "grad_norm": 2.133861541748047,
+      "learning_rate": 2.9531743751795464e-05,
+      "loss": 2.5744,
+      "step": 948
+    },
+    {
+      "epoch": 0.2698626160417543,
+      "grad_norm": 2.4103362560272217,
+      "learning_rate": 2.952025280091928e-05,
+      "loss": 2.4989,
+      "step": 949
+    },
+    {
+      "epoch": 0.2701469812852124,
+      "grad_norm": 1.9884144067764282,
+      "learning_rate": 2.9508761850043093e-05,
+      "loss": 2.3324,
+      "step": 950
+    },
+    {
+      "epoch": 0.2704313465286705,
+      "grad_norm": 1.9855095148086548,
+      "learning_rate": 2.949727089916691e-05,
+      "loss": 2.2511,
+      "step": 951
+    },
+    {
+      "epoch": 0.27071571177212866,
+      "grad_norm": 1.8973934650421143,
+      "learning_rate": 2.9485779948290726e-05,
+      "loss": 2.0523,
+      "step": 952
+    },
+    {
+      "epoch": 0.27100007701558676,
+      "grad_norm": 2.413694143295288,
+      "learning_rate": 2.947428899741454e-05,
+      "loss": 3.0049,
+      "step": 953
+    },
+    {
+      "epoch": 0.2712844422590449,
+      "grad_norm": 2.2881429195404053,
+      "learning_rate": 2.9462798046538355e-05,
+      "loss": 2.5529,
+      "step": 954
+    },
+    {
+      "epoch": 0.271568807502503,
+      "grad_norm": 1.9285370111465454,
+      "learning_rate": 2.945130709566217e-05,
+      "loss": 2.3945,
+      "step": 955
+    },
+    {
+      "epoch": 0.2718531727459611,
+      "grad_norm": 1.9740569591522217,
+      "learning_rate": 2.9439816144785984e-05,
+      "loss": 2.4983,
+      "step": 956
+    },
+    {
+      "epoch": 0.27213753798941925,
+      "grad_norm": 1.8763717412948608,
+      "learning_rate": 2.9428325193909798e-05,
+      "loss": 2.4825,
+      "step": 957
+    },
+    {
+      "epoch": 0.27242190323287735,
+      "grad_norm": 1.8496977090835571,
+      "learning_rate": 2.9416834243033613e-05,
+      "loss": 2.1357,
+      "step": 958
+    },
+    {
+      "epoch": 0.2727062684763355,
+      "grad_norm": 1.9461162090301514,
+      "learning_rate": 2.9405343292157427e-05,
+      "loss": 2.1733,
+      "step": 959
+    },
+    {
+      "epoch": 0.2729906337197936,
+      "grad_norm": 2.133432388305664,
+      "learning_rate": 2.9393852341281245e-05,
+      "loss": 2.145,
+      "step": 960
+    },
+    {
+      "epoch": 0.2732749989632517,
+      "grad_norm": 2.7687673568725586,
+      "learning_rate": 2.938236139040506e-05,
+      "loss": 3.0316,
+      "step": 961
+    },
+    {
+      "epoch": 0.27355936420670984,
+      "grad_norm": 1.9343767166137695,
+      "learning_rate": 2.9370870439528874e-05,
+      "loss": 2.6818,
+      "step": 962
+    },
+    {
+      "epoch": 0.27384372945016794,
+      "grad_norm": 1.7821241617202759,
+      "learning_rate": 2.935937948865269e-05,
+      "loss": 2.5806,
+      "step": 963
+    },
+    {
+      "epoch": 0.2741280946936261,
+      "grad_norm": 1.7667675018310547,
+      "learning_rate": 2.9347888537776503e-05,
+      "loss": 2.6895,
+      "step": 964
+    },
+    {
+      "epoch": 0.2744124599370842,
+      "grad_norm": 1.863686203956604,
+      "learning_rate": 2.9336397586900318e-05,
+      "loss": 2.3492,
+      "step": 965
+    },
+    {
+      "epoch": 0.2746968251805423,
+      "grad_norm": 1.9857769012451172,
+      "learning_rate": 2.9324906636024132e-05,
+      "loss": 2.3327,
+      "step": 966
+    },
+    {
+      "epoch": 0.27498119042400043,
+      "grad_norm": 1.7995944023132324,
+      "learning_rate": 2.9313415685147947e-05,
+      "loss": 2.0556,
+      "step": 967
+    },
+    {
+      "epoch": 0.27526555566745853,
+      "grad_norm": 1.8642942905426025,
+      "learning_rate": 2.930192473427176e-05,
+      "loss": 1.875,
+      "step": 968
+    },
+    {
+      "epoch": 0.2755499209109167,
+      "grad_norm": 1.9679466485977173,
+      "learning_rate": 2.929043378339558e-05,
+      "loss": 3.0327,
+      "step": 969
+    },
+    {
+      "epoch": 0.2758342861543748,
+      "grad_norm": 1.868360161781311,
+      "learning_rate": 2.9278942832519394e-05,
+      "loss": 2.4991,
+      "step": 970
+    },
+    {
+      "epoch": 0.2761186513978329,
+      "grad_norm": 1.9371778964996338,
+      "learning_rate": 2.9267451881643208e-05,
+      "loss": 2.651,
+      "step": 971
+    },
+    {
+      "epoch": 0.276403016641291,
+      "grad_norm": 1.7306244373321533,
+      "learning_rate": 2.9255960930767022e-05,
+      "loss": 2.4744,
+      "step": 972
+    },
+    {
+      "epoch": 0.2766873818847491,
+      "grad_norm": 1.9607441425323486,
+      "learning_rate": 2.9244469979890837e-05,
+      "loss": 2.3596,
+      "step": 973
+    },
+    {
+      "epoch": 0.27697174712820727,
+      "grad_norm": 1.9397883415222168,
+      "learning_rate": 2.923297902901465e-05,
+      "loss": 2.1509,
+      "step": 974
+    },
+    {
+      "epoch": 0.27725611237166536,
+      "grad_norm": 1.9228720664978027,
+      "learning_rate": 2.9221488078138466e-05,
+      "loss": 1.9686,
+      "step": 975
+    },
+    {
+      "epoch": 0.2775404776151235,
+      "grad_norm": 1.9440864324569702,
+      "learning_rate": 2.920999712726228e-05,
+      "loss": 2.0284,
+      "step": 976
+    },
+    {
+      "epoch": 0.2778248428585816,
+      "grad_norm": 2.285125970840454,
+      "learning_rate": 2.9198506176386095e-05,
+      "loss": 2.8307,
+      "step": 977
+    },
+    {
+      "epoch": 0.2781092081020397,
+      "grad_norm": 1.86907160282135,
+      "learning_rate": 2.9187015225509913e-05,
+      "loss": 2.6328,
+      "step": 978
+    },
+    {
+      "epoch": 0.27839357334549786,
+      "grad_norm": 1.7151695489883423,
+      "learning_rate": 2.9175524274633727e-05,
+      "loss": 2.7284,
+      "step": 979
+    },
+    {
+      "epoch": 0.27867793858895595,
+      "grad_norm": 1.6392958164215088,
+      "learning_rate": 2.9164033323757542e-05,
+      "loss": 2.6941,
+      "step": 980
+    },
+    {
+      "epoch": 0.2789623038324141,
+      "grad_norm": 1.9084131717681885,
+      "learning_rate": 2.9152542372881356e-05,
+      "loss": 2.2339,
+      "step": 981
+    },
+    {
+      "epoch": 0.2792466690758722,
+      "grad_norm": 1.7988026142120361,
+      "learning_rate": 2.9141051422005174e-05,
+      "loss": 2.1509,
+      "step": 982
+    },
+    {
+      "epoch": 0.2795310343193303,
+      "grad_norm": 1.8712517023086548,
+      "learning_rate": 2.912956047112899e-05,
+      "loss": 1.9558,
+      "step": 983
+    },
+    {
+      "epoch": 0.27981539956278845,
+      "grad_norm": 2.0422396659851074,
+      "learning_rate": 2.9118069520252807e-05,
+      "loss": 2.1704,
+      "step": 984
+    },
+    {
+      "epoch": 0.28009976480624654,
+      "grad_norm": 2.0253920555114746,
+      "learning_rate": 2.910657856937662e-05,
+      "loss": 3.0636,
+      "step": 985
+    },
+    {
+      "epoch": 0.2803841300497047,
+      "grad_norm": 2.0420773029327393,
+      "learning_rate": 2.9095087618500436e-05,
+      "loss": 2.556,
+      "step": 986
+    },
+    {
+      "epoch": 0.2806684952931628,
+      "grad_norm": 1.864678978919983,
+      "learning_rate": 2.908359666762425e-05,
+      "loss": 2.4779,
+      "step": 987
+    },
+    {
+      "epoch": 0.2809528605366209,
+      "grad_norm": 1.7852369546890259,
+      "learning_rate": 2.9072105716748065e-05,
+      "loss": 2.409,
+      "step": 988
+    },
+    {
+      "epoch": 0.28123722578007904,
+      "grad_norm": 1.8626235723495483,
+      "learning_rate": 2.906061476587188e-05,
+      "loss": 2.2681,
+      "step": 989
+    },
+    {
+      "epoch": 0.28152159102353713,
+      "grad_norm": 2.052464485168457,
+      "learning_rate": 2.9049123814995694e-05,
+      "loss": 2.3266,
+      "step": 990
+    },
+    {
+      "epoch": 0.2818059562669953,
+      "grad_norm": 1.991090178489685,
+      "learning_rate": 2.9037632864119508e-05,
+      "loss": 2.1501,
+      "step": 991
+    },
+    {
+      "epoch": 0.2820903215104534,
+      "grad_norm": 1.8865472078323364,
+      "learning_rate": 2.9026141913243323e-05,
+      "loss": 2.0682,
+      "step": 992
+    },
+    {
+      "epoch": 0.28237468675391153,
+      "grad_norm": 2.1116745471954346,
+      "learning_rate": 2.901465096236714e-05,
+      "loss": 2.9731,
+      "step": 993
+    },
+    {
+      "epoch": 0.2826590519973696,
+      "grad_norm": 1.8774727582931519,
+      "learning_rate": 2.9003160011490955e-05,
+      "loss": 2.7532,
+      "step": 994
+    },
+    {
+      "epoch": 0.2829434172408277,
+      "grad_norm": 1.7747597694396973,
+      "learning_rate": 2.899166906061477e-05,
+      "loss": 2.6523,
+      "step": 995
+    },
+    {
+      "epoch": 0.2832277824842859,
+      "grad_norm": 1.8037197589874268,
+      "learning_rate": 2.8980178109738584e-05,
+      "loss": 2.5828,
+      "step": 996
+    },
+    {
+      "epoch": 0.28351214772774397,
+      "grad_norm": 2.067134141921997,
+      "learning_rate": 2.89686871588624e-05,
+      "loss": 2.5429,
+      "step": 997
+    },
+    {
+      "epoch": 0.2837965129712021,
+      "grad_norm": 2.0115020275115967,
+      "learning_rate": 2.8957196207986213e-05,
+      "loss": 2.1842,
+      "step": 998
+    },
+    {
+      "epoch": 0.2840808782146602,
+      "grad_norm": 1.8412772417068481,
+      "learning_rate": 2.8945705257110027e-05,
+      "loss": 2.1457,
+      "step": 999
+    },
+    {
+      "epoch": 0.2843652434581183,
+      "grad_norm": 1.7536171674728394,
+      "learning_rate": 2.8934214306233842e-05,
+      "loss": 1.9325,
+      "step": 1000
+    },
+    {
+      "epoch": 0.28464960870157646,
+      "grad_norm": 2.1407527923583984,
+      "learning_rate": 2.8922723355357656e-05,
+      "loss": 3.1672,
+      "step": 1001
+    },
+    {
+      "epoch": 0.28493397394503456,
+      "grad_norm": 1.750623345375061,
+      "learning_rate": 2.8911232404481474e-05,
+      "loss": 2.5675,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2852183391884927,
+      "grad_norm": 1.8202438354492188,
+      "learning_rate": 2.889974145360529e-05,
+      "loss": 2.5414,
+      "step": 1003
+    },
+    {
+      "epoch": 0.2855027044319508,
+      "grad_norm": 1.8455865383148193,
+      "learning_rate": 2.8888250502729103e-05,
+      "loss": 2.4263,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2857870696754089,
+      "grad_norm": 1.9973098039627075,
+      "learning_rate": 2.8876759551852918e-05,
+      "loss": 2.2655,
+      "step": 1005
+    },
+    {
+      "epoch": 0.28607143491886705,
+      "grad_norm": 1.8605165481567383,
+      "learning_rate": 2.8865268600976732e-05,
+      "loss": 2.0906,
+      "step": 1006
+    },
+    {
+      "epoch": 0.28635580016232515,
+      "grad_norm": 1.9424145221710205,
+      "learning_rate": 2.8853777650100547e-05,
+      "loss": 2.1295,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2866401654057833,
+      "grad_norm": 1.8717995882034302,
+      "learning_rate": 2.884228669922436e-05,
+      "loss": 2.0872,
+      "step": 1008
+    },
+    {
+      "epoch": 0.2869245306492414,
+      "grad_norm": 1.9393534660339355,
+      "learning_rate": 2.8830795748348176e-05,
+      "loss": 3.0324,
+      "step": 1009
+    },
+    {
+      "epoch": 0.2872088958926995,
+      "grad_norm": 2.0312883853912354,
+      "learning_rate": 2.881930479747199e-05,
+      "loss": 2.7265,
+      "step": 1010
+    },
+    {
+      "epoch": 0.28749326113615764,
+      "grad_norm": 2.081376791000366,
+      "learning_rate": 2.8807813846595808e-05,
+      "loss": 2.5069,
+      "step": 1011
+    },
+    {
+      "epoch": 0.28777762637961574,
+      "grad_norm": 2.1162948608398438,
+      "learning_rate": 2.8796322895719623e-05,
+      "loss": 2.4216,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2880619916230739,
+      "grad_norm": 1.9769059419631958,
+      "learning_rate": 2.8784831944843437e-05,
+      "loss": 2.3386,
+      "step": 1013
+    },
+    {
+      "epoch": 0.288346356866532,
+      "grad_norm": 1.8281911611557007,
+      "learning_rate": 2.8773340993967252e-05,
+      "loss": 2.152,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2886307221099901,
+      "grad_norm": 1.8548706769943237,
+      "learning_rate": 2.8761850043091066e-05,
+      "loss": 1.9768,
+      "step": 1015
+    },
+    {
+      "epoch": 0.28891508735344823,
+      "grad_norm": 1.9629120826721191,
+      "learning_rate": 2.875035909221488e-05,
+      "loss": 2.0301,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2891994525969063,
+      "grad_norm": 2.21577525138855,
+      "learning_rate": 2.8738868141338695e-05,
+      "loss": 2.7658,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2894838178403645,
+      "grad_norm": 1.955876111984253,
+      "learning_rate": 2.872737719046251e-05,
+      "loss": 2.5863,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2897681830838226,
+      "grad_norm": 1.8049885034561157,
+      "learning_rate": 2.871588623958633e-05,
+      "loss": 2.7557,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2900525483272807,
+      "grad_norm": 1.851650357246399,
+      "learning_rate": 2.8704395288710145e-05,
+      "loss": 2.3583,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2903369135707388,
+      "grad_norm": 1.8646175861358643,
+      "learning_rate": 2.869290433783396e-05,
+      "loss": 2.3485,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2906212788141969,
+      "grad_norm": 1.6461886167526245,
+      "learning_rate": 2.8681413386957774e-05,
+      "loss": 2.2649,
+      "step": 1022
+    },
+    {
+      "epoch": 0.29090564405765507,
+      "grad_norm": 1.8911197185516357,
+      "learning_rate": 2.866992243608159e-05,
+      "loss": 2.0709,
+      "step": 1023
+    },
+    {
+      "epoch": 0.29119000930111316,
+      "grad_norm": 1.8676636219024658,
+      "learning_rate": 2.8658431485205403e-05,
+      "loss": 2.1007,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2914743745445713,
+      "grad_norm": 2.4858603477478027,
+      "learning_rate": 2.8646940534329218e-05,
+      "loss": 3.0081,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2917587397880294,
+      "grad_norm": 1.9424095153808594,
+      "learning_rate": 2.8635449583453036e-05,
+      "loss": 2.6527,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2920431050314875,
+      "grad_norm": 1.7152304649353027,
+      "learning_rate": 2.862395863257685e-05,
+      "loss": 2.6888,
+      "step": 1027
+    },
+    {
+      "epoch": 0.29232747027494566,
+      "grad_norm": 1.6697314977645874,
+      "learning_rate": 2.8612467681700665e-05,
+      "loss": 2.5852,
+      "step": 1028
+    },
+    {
+      "epoch": 0.29261183551840375,
+      "grad_norm": 1.8688958883285522,
+      "learning_rate": 2.860097673082448e-05,
+      "loss": 2.4024,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2928962007618619,
+      "grad_norm": 1.8187808990478516,
+      "learning_rate": 2.8589485779948294e-05,
+      "loss": 2.1133,
+      "step": 1030
+    },
+    {
+      "epoch": 0.29318056600532,
+      "grad_norm": 1.6755000352859497,
+      "learning_rate": 2.857799482907211e-05,
+      "loss": 1.9874,
+      "step": 1031
+    },
+    {
+      "epoch": 0.2934649312487781,
+      "grad_norm": 1.9520643949508667,
+      "learning_rate": 2.8566503878195923e-05,
+      "loss": 2.0306,
+      "step": 1032
+    },
+    {
+      "epoch": 0.29374929649223624,
+      "grad_norm": 2.2317845821380615,
+      "learning_rate": 2.8555012927319737e-05,
+      "loss": 2.956,
+      "step": 1033
+    },
+    {
+      "epoch": 0.29403366173569434,
+      "grad_norm": 1.9723296165466309,
+      "learning_rate": 2.8543521976443552e-05,
+      "loss": 2.6379,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2943180269791525,
+      "grad_norm": 1.8846988677978516,
+      "learning_rate": 2.853203102556737e-05,
+      "loss": 2.7342,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2946023922226106,
+      "grad_norm": 1.7443634271621704,
+      "learning_rate": 2.8520540074691184e-05,
+      "loss": 2.437,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2948867574660687,
+      "grad_norm": 1.8220562934875488,
+      "learning_rate": 2.8509049123815e-05,
+      "loss": 2.1899,
+      "step": 1037
+    },
+    {
+      "epoch": 0.29517112270952683,
+      "grad_norm": 1.8991121053695679,
+      "learning_rate": 2.8497558172938813e-05,
+      "loss": 2.087,
+      "step": 1038
+    },
+    {
+      "epoch": 0.29545548795298493,
+      "grad_norm": 1.7543880939483643,
+      "learning_rate": 2.8486067222062628e-05,
+      "loss": 1.7512,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2957398531964431,
+      "grad_norm": 1.95134699344635,
+      "learning_rate": 2.8474576271186442e-05,
+      "loss": 2.0684,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2960242184399012,
+      "grad_norm": 2.655040979385376,
+      "learning_rate": 2.8463085320310257e-05,
+      "loss": 2.9242,
+      "step": 1041
+    },
+    {
+      "epoch": 0.29630858368335933,
+      "grad_norm": 1.837572693824768,
+      "learning_rate": 2.845159436943407e-05,
+      "loss": 2.4314,
+      "step": 1042
+    },
+    {
+      "epoch": 0.2965929489268174,
+      "grad_norm": 1.903134822845459,
+      "learning_rate": 2.8440103418557886e-05,
+      "loss": 2.6958,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2968773141702755,
+      "grad_norm": 1.6535474061965942,
+      "learning_rate": 2.8428612467681704e-05,
+      "loss": 2.301,
+      "step": 1044
+    },
+    {
+      "epoch": 0.29716167941373367,
+      "grad_norm": 2.0542876720428467,
+      "learning_rate": 2.8417121516805518e-05,
+      "loss": 2.4347,
+      "step": 1045
+    },
+    {
+      "epoch": 0.29744604465719177,
+      "grad_norm": 1.8558207750320435,
+      "learning_rate": 2.8405630565929333e-05,
+      "loss": 2.2441,
+      "step": 1046
+    },
+    {
+      "epoch": 0.2977304099006499,
+      "grad_norm": 1.7878432273864746,
+      "learning_rate": 2.8394139615053147e-05,
+      "loss": 2.0426,
+      "step": 1047
+    },
+    {
+      "epoch": 0.298014775144108,
+      "grad_norm": 1.9130539894104004,
+      "learning_rate": 2.838264866417696e-05,
+      "loss": 2.1265,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2982991403875661,
+      "grad_norm": 2.4959192276000977,
+      "learning_rate": 2.8371157713300776e-05,
+      "loss": 2.8223,
+      "step": 1049
+    },
+    {
+      "epoch": 0.29858350563102426,
+      "grad_norm": 1.7817506790161133,
+      "learning_rate": 2.835966676242459e-05,
+      "loss": 2.6125,
+      "step": 1050
+    },
+    {
+      "epoch": 0.29886787087448236,
+      "grad_norm": 1.8295172452926636,
+      "learning_rate": 2.8348175811548405e-05,
+      "loss": 2.5712,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2991522361179405,
+      "grad_norm": 1.658846139907837,
+      "learning_rate": 2.833668486067222e-05,
+      "loss": 2.4154,
+      "step": 1052
+    },
+    {
+      "epoch": 0.2994366013613986,
+      "grad_norm": 1.8340063095092773,
+      "learning_rate": 2.8325193909796037e-05,
+      "loss": 2.3476,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2997209666048567,
+      "grad_norm": 1.722588062286377,
+      "learning_rate": 2.8313702958919852e-05,
+      "loss": 2.0986,
+      "step": 1054
+    },
+    {
+      "epoch": 0.30000533184831485,
+      "grad_norm": 1.6516780853271484,
+      "learning_rate": 2.8302212008043666e-05,
+      "loss": 2.1311,
+      "step": 1055
+    },
+    {
+      "epoch": 0.30028969709177294,
+      "grad_norm": 2.02481746673584,
+      "learning_rate": 2.8290721057167484e-05,
+      "loss": 1.7839,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3005740623352311,
+      "grad_norm": 2.038386821746826,
+      "learning_rate": 2.82792301062913e-05,
+      "loss": 2.9095,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3008584275786892,
+      "grad_norm": 1.7752598524093628,
+      "learning_rate": 2.8267739155415113e-05,
+      "loss": 2.9031,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3011427928221473,
+      "grad_norm": 1.6934603452682495,
+      "learning_rate": 2.825624820453893e-05,
+      "loss": 2.4481,
+      "step": 1059
+    },
+    {
+      "epoch": 0.30142715806560544,
+      "grad_norm": 1.6980534791946411,
+      "learning_rate": 2.8244757253662746e-05,
+      "loss": 2.4492,
+      "step": 1060
+    },
+    {
+      "epoch": 0.30171152330906353,
+      "grad_norm": 1.7166688442230225,
+      "learning_rate": 2.823326630278656e-05,
+      "loss": 2.2513,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3019958885525217,
+      "grad_norm": 1.8027641773223877,
+      "learning_rate": 2.8221775351910375e-05,
+      "loss": 2.2973,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3022802537959798,
+      "grad_norm": 2.2114596366882324,
+      "learning_rate": 2.821028440103419e-05,
+      "loss": 1.9347,
+      "step": 1063
+    },
+    {
+      "epoch": 0.30256461903943793,
+      "grad_norm": 1.8748873472213745,
+      "learning_rate": 2.8198793450158004e-05,
+      "loss": 1.8972,
+      "step": 1064
+    },
+    {
+      "epoch": 0.302848984282896,
+      "grad_norm": 1.924533486366272,
+      "learning_rate": 2.8187302499281818e-05,
+      "loss": 3.2114,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3031333495263541,
+      "grad_norm": 1.7401739358901978,
+      "learning_rate": 2.8175811548405633e-05,
+      "loss": 2.7883,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3034177147698123,
+      "grad_norm": 1.6388260126113892,
+      "learning_rate": 2.8164320597529447e-05,
+      "loss": 2.5342,
+      "step": 1067
+    },
+    {
+      "epoch": 0.30370208001327037,
+      "grad_norm": 1.66873037815094,
+      "learning_rate": 2.8152829646653265e-05,
+      "loss": 2.348,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3039864452567285,
+      "grad_norm": 1.798418402671814,
+      "learning_rate": 2.814133869577708e-05,
+      "loss": 2.3285,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3042708105001866,
+      "grad_norm": 1.6459908485412598,
+      "learning_rate": 2.8129847744900894e-05,
+      "loss": 1.9097,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3045551757436447,
+      "grad_norm": 1.6589969396591187,
+      "learning_rate": 2.811835679402471e-05,
+      "loss": 1.9083,
+      "step": 1071
+    },
+    {
+      "epoch": 0.30483954098710286,
+      "grad_norm": 1.7330818176269531,
+      "learning_rate": 2.8106865843148523e-05,
+      "loss": 2.0022,
+      "step": 1072
+    },
+    {
+      "epoch": 0.30512390623056096,
+      "grad_norm": 2.162515878677368,
+      "learning_rate": 2.8095374892272338e-05,
+      "loss": 3.2133,
+      "step": 1073
+    },
+    {
+      "epoch": 0.3054082714740191,
+      "grad_norm": 1.890426516532898,
+      "learning_rate": 2.8083883941396152e-05,
+      "loss": 2.5098,
+      "step": 1074
+    },
+    {
+      "epoch": 0.3056926367174772,
+      "grad_norm": 1.9428486824035645,
+      "learning_rate": 2.8072392990519967e-05,
+      "loss": 2.8393,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3059770019609353,
+      "grad_norm": 1.7772703170776367,
+      "learning_rate": 2.806090203964378e-05,
+      "loss": 2.5238,
+      "step": 1076
+    },
+    {
+      "epoch": 0.30626136720439345,
+      "grad_norm": 2.0038602352142334,
+      "learning_rate": 2.80494110887676e-05,
+      "loss": 2.5622,
+      "step": 1077
+    },
+    {
+      "epoch": 0.30654573244785155,
+      "grad_norm": 1.7946076393127441,
+      "learning_rate": 2.8037920137891413e-05,
+      "loss": 2.1265,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3068300976913097,
+      "grad_norm": 1.694143533706665,
+      "learning_rate": 2.8026429187015228e-05,
+      "loss": 2.1898,
+      "step": 1079
+    },
+    {
+      "epoch": 0.3071144629347678,
+      "grad_norm": 1.8604297637939453,
+      "learning_rate": 2.8014938236139042e-05,
+      "loss": 1.8618,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3073988281782259,
+      "grad_norm": 2.2564175128936768,
+      "learning_rate": 2.8003447285262857e-05,
+      "loss": 2.8106,
+      "step": 1081
+    },
+    {
+      "epoch": 0.30768319342168404,
+      "grad_norm": 2.0401079654693604,
+      "learning_rate": 2.799195633438667e-05,
+      "loss": 2.7885,
+      "step": 1082
+    },
+    {
+      "epoch": 0.30796755866514214,
+      "grad_norm": 1.7820488214492798,
+      "learning_rate": 2.7980465383510486e-05,
+      "loss": 2.4126,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3082519239086003,
+      "grad_norm": 1.7269959449768066,
+      "learning_rate": 2.79689744326343e-05,
+      "loss": 2.3303,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3085362891520584,
+      "grad_norm": 1.8052058219909668,
+      "learning_rate": 2.7957483481758115e-05,
+      "loss": 2.3413,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3088206543955165,
+      "grad_norm": 1.7034716606140137,
+      "learning_rate": 2.7945992530881933e-05,
+      "loss": 2.2142,
+      "step": 1086
+    },
+    {
+      "epoch": 0.30910501963897463,
+      "grad_norm": 1.924057960510254,
+      "learning_rate": 2.7934501580005747e-05,
+      "loss": 2.0875,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3093893848824327,
+      "grad_norm": 1.9003182649612427,
+      "learning_rate": 2.7923010629129562e-05,
+      "loss": 1.9661,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3096737501258909,
+      "grad_norm": 2.4958977699279785,
+      "learning_rate": 2.7911519678253376e-05,
+      "loss": 3.0476,
+      "step": 1089
+    },
+    {
+      "epoch": 0.309958115369349,
+      "grad_norm": 2.0897626876831055,
+      "learning_rate": 2.790002872737719e-05,
+      "loss": 2.5666,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3102424806128071,
+      "grad_norm": 1.7425907850265503,
+      "learning_rate": 2.7888537776501005e-05,
+      "loss": 2.3865,
+      "step": 1091
+    },
+    {
+      "epoch": 0.3105268458562652,
+      "grad_norm": 1.892893671989441,
+      "learning_rate": 2.787704682562482e-05,
+      "loss": 2.4982,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3108112110997233,
+      "grad_norm": 2.0055623054504395,
+      "learning_rate": 2.786555587474864e-05,
+      "loss": 2.3453,
+      "step": 1093
+    },
+    {
+      "epoch": 0.31109557634318147,
+      "grad_norm": 1.7691258192062378,
+      "learning_rate": 2.7854064923872456e-05,
+      "loss": 2.1304,
+      "step": 1094
+    },
+    {
+      "epoch": 0.31137994158663956,
+      "grad_norm": 1.7920550107955933,
+      "learning_rate": 2.784257397299627e-05,
+      "loss": 2.0597,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3116643068300977,
+      "grad_norm": 1.657732367515564,
+      "learning_rate": 2.7831083022120085e-05,
+      "loss": 1.8743,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3119486720735558,
+      "grad_norm": 2.2476625442504883,
+      "learning_rate": 2.78195920712439e-05,
+      "loss": 3.1486,
+      "step": 1097
+    },
+    {
+      "epoch": 0.3122330373170139,
+      "grad_norm": 1.9806352853775024,
+      "learning_rate": 2.7808101120367714e-05,
+      "loss": 2.6852,
+      "step": 1098
+    },
+    {
+      "epoch": 0.31251740256047206,
+      "grad_norm": 1.7334058284759521,
+      "learning_rate": 2.7796610169491528e-05,
+      "loss": 2.5363,
+      "step": 1099
+    },
+    {
+      "epoch": 0.31280176780393015,
+      "grad_norm": 1.769322395324707,
+      "learning_rate": 2.7785119218615343e-05,
+      "loss": 2.3318,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3130861330473883,
+      "grad_norm": 1.7031893730163574,
+      "learning_rate": 2.777362826773916e-05,
+      "loss": 2.4197,
+      "step": 1101
+    },
+    {
+      "epoch": 0.3133704982908464,
+      "grad_norm": 1.877233862876892,
+      "learning_rate": 2.7762137316862975e-05,
+      "loss": 2.049,
+      "step": 1102
+    },
+    {
+      "epoch": 0.3136548635343045,
+      "grad_norm": 1.7407324314117432,
+      "learning_rate": 2.775064636598679e-05,
+      "loss": 1.8772,
+      "step": 1103
+    },
+    {
+      "epoch": 0.31393922877776265,
+      "grad_norm": 1.7369098663330078,
+      "learning_rate": 2.7739155415110604e-05,
+      "loss": 1.9947,
+      "step": 1104
+    },
+    {
+      "epoch": 0.31422359402122074,
+      "grad_norm": 2.291818141937256,
+      "learning_rate": 2.772766446423442e-05,
+      "loss": 2.9378,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3145079592646789,
+      "grad_norm": 1.7521897554397583,
+      "learning_rate": 2.7716173513358233e-05,
+      "loss": 2.6269,
+      "step": 1106
+    },
+    {
+      "epoch": 0.314792324508137,
+      "grad_norm": 1.6815153360366821,
+      "learning_rate": 2.7704682562482047e-05,
+      "loss": 2.3316,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3150766897515951,
+      "grad_norm": 1.6132457256317139,
+      "learning_rate": 2.7693191611605862e-05,
+      "loss": 2.1689,
+      "step": 1108
+    },
+    {
+      "epoch": 0.31536105499505324,
+      "grad_norm": 1.8604917526245117,
+      "learning_rate": 2.7681700660729676e-05,
+      "loss": 2.3652,
+      "step": 1109
+    },
+    {
+      "epoch": 0.31564542023851133,
+      "grad_norm": 1.6801247596740723,
+      "learning_rate": 2.7670209709853494e-05,
+      "loss": 2.0579,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3159297854819695,
+      "grad_norm": 1.8979250192642212,
+      "learning_rate": 2.765871875897731e-05,
+      "loss": 2.0514,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3162141507254276,
+      "grad_norm": 1.8273488283157349,
+      "learning_rate": 2.7647227808101123e-05,
+      "loss": 2.0111,
+      "step": 1112
+    },
+    {
+      "epoch": 0.31649851596888573,
+      "grad_norm": 2.283964157104492,
+      "learning_rate": 2.7635736857224938e-05,
+      "loss": 2.9822,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3167828812123438,
+      "grad_norm": 1.8019022941589355,
+      "learning_rate": 2.7624245906348752e-05,
+      "loss": 2.7109,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3170672464558019,
+      "grad_norm": 1.7521454095840454,
+      "learning_rate": 2.7612754955472567e-05,
+      "loss": 2.5455,
+      "step": 1115
+    },
+    {
+      "epoch": 0.31735161169926007,
+      "grad_norm": 1.7717288732528687,
+      "learning_rate": 2.760126400459638e-05,
+      "loss": 2.2537,
+      "step": 1116
+    },
+    {
+      "epoch": 0.31763597694271817,
+      "grad_norm": 1.9506962299346924,
+      "learning_rate": 2.7589773053720196e-05,
+      "loss": 2.2884,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3179203421861763,
+      "grad_norm": 1.6748727560043335,
+      "learning_rate": 2.757828210284401e-05,
+      "loss": 2.1113,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3182047074296344,
+      "grad_norm": 1.8400766849517822,
+      "learning_rate": 2.7566791151967828e-05,
+      "loss": 2.0633,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3184890726730925,
+      "grad_norm": 1.9170554876327515,
+      "learning_rate": 2.7555300201091643e-05,
+      "loss": 1.9511,
+      "step": 1120
+    },
+    {
+      "epoch": 0.31877343791655066,
+      "grad_norm": 2.18106746673584,
+      "learning_rate": 2.7543809250215457e-05,
+      "loss": 3.0597,
+      "step": 1121
+    },
+    {
+      "epoch": 0.31905780316000876,
+      "grad_norm": 1.7323399782180786,
+      "learning_rate": 2.753231829933927e-05,
+      "loss": 2.3877,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3193421684034669,
+      "grad_norm": 1.7470673322677612,
+      "learning_rate": 2.7520827348463086e-05,
+      "loss": 2.5492,
+      "step": 1123
+    },
+    {
+      "epoch": 0.319626533646925,
+      "grad_norm": 1.627826452255249,
+      "learning_rate": 2.75093363975869e-05,
+      "loss": 2.3832,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3199108988903831,
+      "grad_norm": 1.8971792459487915,
+      "learning_rate": 2.7497845446710715e-05,
+      "loss": 2.2562,
+      "step": 1125
+    },
+    {
+      "epoch": 0.32019526413384125,
+      "grad_norm": 1.8873558044433594,
+      "learning_rate": 2.748635449583453e-05,
+      "loss": 2.1615,
+      "step": 1126
+    },
+    {
+      "epoch": 0.32047962937729935,
+      "grad_norm": 1.8183141946792603,
+      "learning_rate": 2.7474863544958344e-05,
+      "loss": 2.0914,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3207639946207575,
+      "grad_norm": 1.8885418176651,
+      "learning_rate": 2.7463372594082162e-05,
+      "loss": 1.9792,
+      "step": 1128
+    },
+    {
+      "epoch": 0.3210483598642156,
+      "grad_norm": 2.204162359237671,
+      "learning_rate": 2.7451881643205977e-05,
+      "loss": 2.9695,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3213327251076737,
+      "grad_norm": 1.727662444114685,
+      "learning_rate": 2.7440390692329794e-05,
+      "loss": 2.5204,
+      "step": 1130
+    },
+    {
+      "epoch": 0.32161709035113184,
+      "grad_norm": 1.721366286277771,
+      "learning_rate": 2.742889974145361e-05,
+      "loss": 2.5624,
+      "step": 1131
+    },
+    {
+      "epoch": 0.32190145559458994,
+      "grad_norm": 1.598515510559082,
+      "learning_rate": 2.7417408790577423e-05,
+      "loss": 2.3068,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3221858208380481,
+      "grad_norm": 1.7761754989624023,
+      "learning_rate": 2.7405917839701238e-05,
+      "loss": 2.3844,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3224701860815062,
+      "grad_norm": 1.6555432081222534,
+      "learning_rate": 2.7394426888825056e-05,
+      "loss": 2.0163,
+      "step": 1134
+    },
+    {
+      "epoch": 0.32275455132496433,
+      "grad_norm": 1.6981431245803833,
+      "learning_rate": 2.738293593794887e-05,
+      "loss": 1.9993,
+      "step": 1135
+    },
+    {
+      "epoch": 0.32303891656842243,
+      "grad_norm": 1.702057957649231,
+      "learning_rate": 2.7371444987072685e-05,
+      "loss": 1.8252,
+      "step": 1136
+    },
+    {
+      "epoch": 0.3233232818118805,
+      "grad_norm": 1.9638831615447998,
+      "learning_rate": 2.73599540361965e-05,
+      "loss": 3.1115,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3236076470553387,
+      "grad_norm": 1.7202662229537964,
+      "learning_rate": 2.7348463085320314e-05,
+      "loss": 2.5926,
+      "step": 1138
+    },
+    {
+      "epoch": 0.32389201229879677,
+      "grad_norm": 1.6885807514190674,
+      "learning_rate": 2.7336972134444128e-05,
+      "loss": 2.5377,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3241763775422549,
+      "grad_norm": 1.6235450506210327,
+      "learning_rate": 2.7325481183567943e-05,
+      "loss": 2.5253,
+      "step": 1140
+    },
+    {
+      "epoch": 0.324460742785713,
+      "grad_norm": 1.7628118991851807,
+      "learning_rate": 2.7313990232691757e-05,
+      "loss": 2.4275,
+      "step": 1141
+    },
+    {
+      "epoch": 0.3247451080291711,
+      "grad_norm": 1.8225659132003784,
+      "learning_rate": 2.7302499281815572e-05,
+      "loss": 2.0278,
+      "step": 1142
+    },
+    {
+      "epoch": 0.32502947327262927,
+      "grad_norm": 1.8690272569656372,
+      "learning_rate": 2.729100833093939e-05,
+      "loss": 2.0703,
+      "step": 1143
+    },
+    {
+      "epoch": 0.32531383851608736,
+      "grad_norm": 1.951603889465332,
+      "learning_rate": 2.7279517380063204e-05,
+      "loss": 1.7402,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3255982037595455,
+      "grad_norm": 2.100116491317749,
+      "learning_rate": 2.726802642918702e-05,
+      "loss": 3.0271,
+      "step": 1145
+    },
+    {
+      "epoch": 0.3258825690030036,
+      "grad_norm": 1.7133632898330688,
+      "learning_rate": 2.7256535478310833e-05,
+      "loss": 2.5711,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3261669342464617,
+      "grad_norm": 1.6197803020477295,
+      "learning_rate": 2.7245044527434648e-05,
+      "loss": 2.4896,
+      "step": 1147
+    },
+    {
+      "epoch": 0.32645129948991986,
+      "grad_norm": 1.7283875942230225,
+      "learning_rate": 2.7233553576558462e-05,
+      "loss": 2.3986,
+      "step": 1148
+    },
+    {
+      "epoch": 0.32673566473337795,
+      "grad_norm": 1.7507110834121704,
+      "learning_rate": 2.7222062625682277e-05,
+      "loss": 2.1887,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3270200299768361,
+      "grad_norm": 1.7281140089035034,
+      "learning_rate": 2.721057167480609e-05,
+      "loss": 2.2903,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3273043952202942,
+      "grad_norm": 1.7815526723861694,
+      "learning_rate": 2.7199080723929906e-05,
+      "loss": 2.0782,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3275887604637523,
+      "grad_norm": 1.8577396869659424,
+      "learning_rate": 2.7187589773053724e-05,
+      "loss": 1.7293,
+      "step": 1152
+    },
+    {
+      "epoch": 0.32787312570721044,
+      "grad_norm": 2.047935962677002,
+      "learning_rate": 2.7176098822177538e-05,
+      "loss": 2.8512,
+      "step": 1153
+    },
+    {
+      "epoch": 0.32815749095066854,
+      "grad_norm": 1.804479718208313,
+      "learning_rate": 2.7164607871301353e-05,
+      "loss": 2.7431,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3284418561941267,
+      "grad_norm": 1.7864207029342651,
+      "learning_rate": 2.7153116920425167e-05,
+      "loss": 2.4874,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3287262214375848,
+      "grad_norm": 1.6858755350112915,
+      "learning_rate": 2.714162596954898e-05,
+      "loss": 2.3767,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3290105866810429,
+      "grad_norm": 1.7890626192092896,
+      "learning_rate": 2.7130135018672796e-05,
+      "loss": 2.2339,
+      "step": 1157
+    },
+    {
+      "epoch": 0.32929495192450103,
+      "grad_norm": 1.7149113416671753,
+      "learning_rate": 2.711864406779661e-05,
+      "loss": 2.2292,
+      "step": 1158
+    },
+    {
+      "epoch": 0.32957931716795913,
+      "grad_norm": 1.7810120582580566,
+      "learning_rate": 2.7107153116920425e-05,
+      "loss": 2.0133,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3298636824114173,
+      "grad_norm": 1.7565312385559082,
+      "learning_rate": 2.709566216604424e-05,
+      "loss": 2.0087,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3301480476548754,
+      "grad_norm": 2.1705124378204346,
+      "learning_rate": 2.7084171215168057e-05,
+      "loss": 3.0282,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3304324128983335,
+      "grad_norm": 1.7916361093521118,
+      "learning_rate": 2.7072680264291872e-05,
+      "loss": 2.7158,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3307167781417916,
+      "grad_norm": 1.6537439823150635,
+      "learning_rate": 2.7061189313415686e-05,
+      "loss": 2.4969,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3310011433852497,
+      "grad_norm": 1.7181519269943237,
+      "learning_rate": 2.70496983625395e-05,
+      "loss": 2.3481,
+      "step": 1164
+    },
+    {
+      "epoch": 0.33128550862870787,
+      "grad_norm": 1.7014647722244263,
+      "learning_rate": 2.7038207411663315e-05,
+      "loss": 2.2512,
+      "step": 1165
+    },
+    {
+      "epoch": 0.33156987387216597,
+      "grad_norm": 1.7426505088806152,
+      "learning_rate": 2.702671646078713e-05,
+      "loss": 2.1619,
+      "step": 1166
+    },
+    {
+      "epoch": 0.3318542391156241,
+      "grad_norm": 1.7927415370941162,
+      "learning_rate": 2.701522550991095e-05,
+      "loss": 1.9217,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3321386043590822,
+      "grad_norm": 1.9813625812530518,
+      "learning_rate": 2.7003734559034766e-05,
+      "loss": 1.89,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3324229696025403,
+      "grad_norm": 2.2803986072540283,
+      "learning_rate": 2.699224360815858e-05,
+      "loss": 3.1151,
+      "step": 1169
+    },
+    {
+      "epoch": 0.33270733484599846,
+      "grad_norm": 1.8023364543914795,
+      "learning_rate": 2.6980752657282395e-05,
+      "loss": 2.5497,
+      "step": 1170
+    },
+    {
+      "epoch": 0.33299170008945655,
+      "grad_norm": 1.5802347660064697,
+      "learning_rate": 2.696926170640621e-05,
+      "loss": 2.3567,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3332760653329147,
+      "grad_norm": 1.7448047399520874,
+      "learning_rate": 2.6957770755530024e-05,
+      "loss": 2.4281,
+      "step": 1172
+    },
+    {
+      "epoch": 0.3335604305763728,
+      "grad_norm": 1.9900225400924683,
+      "learning_rate": 2.6946279804653838e-05,
+      "loss": 2.4395,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3338447958198309,
+      "grad_norm": 1.7050385475158691,
+      "learning_rate": 2.6934788853777653e-05,
+      "loss": 2.0546,
+      "step": 1174
+    },
+    {
+      "epoch": 0.33412916106328905,
+      "grad_norm": 1.9430030584335327,
+      "learning_rate": 2.6923297902901467e-05,
+      "loss": 2.0789,
+      "step": 1175
+    },
+    {
+      "epoch": 0.33441352630674714,
+      "grad_norm": 1.910507321357727,
+      "learning_rate": 2.6911806952025285e-05,
+      "loss": 1.9512,
+      "step": 1176
+    },
+    {
+      "epoch": 0.3346978915502053,
+      "grad_norm": 2.156999349594116,
+      "learning_rate": 2.69003160011491e-05,
+      "loss": 2.7457,
+      "step": 1177
+    },
+    {
+      "epoch": 0.3349822567936634,
+      "grad_norm": 1.7482166290283203,
+      "learning_rate": 2.6888825050272914e-05,
+      "loss": 2.7449,
+      "step": 1178
+    },
+    {
+      "epoch": 0.3352666220371215,
+      "grad_norm": 1.692886233329773,
+      "learning_rate": 2.687733409939673e-05,
+      "loss": 2.4689,
+      "step": 1179
+    },
+    {
+      "epoch": 0.33555098728057964,
+      "grad_norm": 1.6204313039779663,
+      "learning_rate": 2.6865843148520543e-05,
+      "loss": 2.3802,
+      "step": 1180
+    },
+    {
+      "epoch": 0.33583535252403773,
+      "grad_norm": 1.7569348812103271,
+      "learning_rate": 2.6854352197644357e-05,
+      "loss": 2.2957,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3361197177674959,
+      "grad_norm": 1.6739596128463745,
+      "learning_rate": 2.6842861246768172e-05,
+      "loss": 1.9977,
+      "step": 1182
+    },
+    {
+      "epoch": 0.336404083010954,
+      "grad_norm": 1.858077883720398,
+      "learning_rate": 2.6831370295891986e-05,
+      "loss": 2.0055,
+      "step": 1183
+    },
+    {
+      "epoch": 0.33668844825441213,
+      "grad_norm": 1.8774821758270264,
+      "learning_rate": 2.68198793450158e-05,
+      "loss": 1.8794,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3369728134978702,
+      "grad_norm": 2.0827038288116455,
+      "learning_rate": 2.680838839413962e-05,
+      "loss": 2.8639,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3372571787413283,
+      "grad_norm": 1.8128411769866943,
+      "learning_rate": 2.6796897443263433e-05,
+      "loss": 2.4063,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3375415439847865,
+      "grad_norm": 1.9702905416488647,
+      "learning_rate": 2.6785406492387248e-05,
+      "loss": 2.4151,
+      "step": 1187
+    },
+    {
+      "epoch": 0.33782590922824457,
+      "grad_norm": 1.90627121925354,
+      "learning_rate": 2.6773915541511062e-05,
+      "loss": 2.4171,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3381102744717027,
+      "grad_norm": 2.10098934173584,
+      "learning_rate": 2.6762424590634877e-05,
+      "loss": 2.3142,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3383946397151608,
+      "grad_norm": 1.7690738439559937,
+      "learning_rate": 2.675093363975869e-05,
+      "loss": 1.9259,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3386790049586189,
+      "grad_norm": 1.794833779335022,
+      "learning_rate": 2.6739442688882506e-05,
+      "loss": 1.9992,
+      "step": 1191
+    },
+    {
+      "epoch": 0.33896337020207706,
+      "grad_norm": 1.848533034324646,
+      "learning_rate": 2.672795173800632e-05,
+      "loss": 1.9893,
+      "step": 1192
+    },
+    {
+      "epoch": 0.33924773544553516,
+      "grad_norm": 2.1067445278167725,
+      "learning_rate": 2.6716460787130135e-05,
+      "loss": 3.1749,
+      "step": 1193
+    },
+    {
+      "epoch": 0.3395321006889933,
+      "grad_norm": 1.9302208423614502,
+      "learning_rate": 2.6704969836253953e-05,
+      "loss": 2.6356,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3398164659324514,
+      "grad_norm": 1.7722642421722412,
+      "learning_rate": 2.6693478885377767e-05,
+      "loss": 2.4753,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3401008311759095,
+      "grad_norm": 1.7334580421447754,
+      "learning_rate": 2.6681987934501582e-05,
+      "loss": 2.2498,
+      "step": 1196
+    },
+    {
+      "epoch": 0.34038519641936765,
+      "grad_norm": 1.7633260488510132,
+      "learning_rate": 2.6670496983625396e-05,
+      "loss": 2.2229,
+      "step": 1197
+    },
+    {
+      "epoch": 0.34066956166282575,
+      "grad_norm": 1.721509337425232,
+      "learning_rate": 2.665900603274921e-05,
+      "loss": 2.2436,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3409539269062839,
+      "grad_norm": 1.93727445602417,
+      "learning_rate": 2.6647515081873025e-05,
+      "loss": 1.9182,
+      "step": 1199
+    },
+    {
+      "epoch": 0.341238292149742,
+      "grad_norm": 1.9112025499343872,
+      "learning_rate": 2.663602413099684e-05,
+      "loss": 1.8985,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3415226573932001,
+      "grad_norm": 2.4720232486724854,
+      "learning_rate": 2.6624533180120654e-05,
+      "loss": 2.9957,
+      "step": 1201
+    },
+    {
+      "epoch": 0.34180702263665824,
+      "grad_norm": 2.0485682487487793,
+      "learning_rate": 2.661304222924447e-05,
+      "loss": 2.5158,
+      "step": 1202
+    },
+    {
+      "epoch": 0.34209138788011634,
+      "grad_norm": 1.8902740478515625,
+      "learning_rate": 2.6601551278368287e-05,
+      "loss": 2.3761,
+      "step": 1203
+    },
+    {
+      "epoch": 0.3423757531235745,
+      "grad_norm": 1.7648299932479858,
+      "learning_rate": 2.6590060327492104e-05,
+      "loss": 2.0841,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3426601183670326,
+      "grad_norm": 1.8326659202575684,
+      "learning_rate": 2.657856937661592e-05,
+      "loss": 2.3462,
+      "step": 1205
+    },
+    {
+      "epoch": 0.3429444836104907,
+      "grad_norm": 1.8549422025680542,
+      "learning_rate": 2.6567078425739733e-05,
+      "loss": 2.1884,
+      "step": 1206
+    },
+    {
+      "epoch": 0.34322884885394883,
+      "grad_norm": 1.930391550064087,
+      "learning_rate": 2.6555587474863548e-05,
+      "loss": 1.9793,
+      "step": 1207
+    },
+    {
+      "epoch": 0.3435132140974069,
+      "grad_norm": 1.8481113910675049,
+      "learning_rate": 2.6544096523987362e-05,
+      "loss": 2.0828,
+      "step": 1208
+    },
+    {
+      "epoch": 0.3437975793408651,
+      "grad_norm": 2.0146400928497314,
+      "learning_rate": 2.653260557311118e-05,
+      "loss": 2.8335,
+      "step": 1209
+    },
+    {
+      "epoch": 0.3440819445843232,
+      "grad_norm": 1.820291519165039,
+      "learning_rate": 2.6521114622234995e-05,
+      "loss": 2.7538,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3443663098277813,
+      "grad_norm": 1.721548080444336,
+      "learning_rate": 2.650962367135881e-05,
+      "loss": 2.4637,
+      "step": 1211
+    },
+    {
+      "epoch": 0.3446506750712394,
+      "grad_norm": 1.7482106685638428,
+      "learning_rate": 2.6498132720482624e-05,
+      "loss": 2.2598,
+      "step": 1212
+    },
+    {
+      "epoch": 0.3449350403146975,
+      "grad_norm": 1.8595374822616577,
+      "learning_rate": 2.648664176960644e-05,
+      "loss": 2.4933,
+      "step": 1213
+    },
+    {
+      "epoch": 0.34521940555815567,
+      "grad_norm": 1.690114974975586,
+      "learning_rate": 2.6475150818730253e-05,
+      "loss": 2.1184,
+      "step": 1214
+    },
+    {
+      "epoch": 0.34550377080161376,
+      "grad_norm": 1.6892638206481934,
+      "learning_rate": 2.6463659867854067e-05,
+      "loss": 1.9942,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3457881360450719,
+      "grad_norm": 1.8123127222061157,
+      "learning_rate": 2.6452168916977882e-05,
+      "loss": 2.1055,
+      "step": 1216
+    },
+    {
+      "epoch": 0.34607250128853,
+      "grad_norm": 2.2426178455352783,
+      "learning_rate": 2.6440677966101696e-05,
+      "loss": 3.0519,
+      "step": 1217
+    },
+    {
+      "epoch": 0.3463568665319881,
+      "grad_norm": 1.7458500862121582,
+      "learning_rate": 2.6429187015225514e-05,
+      "loss": 2.6327,
+      "step": 1218
+    },
+    {
+      "epoch": 0.34664123177544626,
+      "grad_norm": 1.7254029512405396,
+      "learning_rate": 2.641769606434933e-05,
+      "loss": 2.4534,
+      "step": 1219
+    },
+    {
+      "epoch": 0.34692559701890435,
+      "grad_norm": 1.6477024555206299,
+      "learning_rate": 2.6406205113473143e-05,
+      "loss": 2.3594,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3472099622623625,
+      "grad_norm": 1.6863490343093872,
+      "learning_rate": 2.6394714162596958e-05,
+      "loss": 2.2598,
+      "step": 1221
+    },
+    {
+      "epoch": 0.3474943275058206,
+      "grad_norm": 1.8328365087509155,
+      "learning_rate": 2.6383223211720772e-05,
+      "loss": 2.1452,
+      "step": 1222
+    },
+    {
+      "epoch": 0.3477786927492787,
+      "grad_norm": 1.7846546173095703,
+      "learning_rate": 2.6371732260844587e-05,
+      "loss": 2.0761,
+      "step": 1223
+    },
+    {
+      "epoch": 0.34806305799273685,
+      "grad_norm": 2.078010082244873,
+      "learning_rate": 2.63602413099684e-05,
+      "loss": 1.9126,
+      "step": 1224
+    },
+    {
+      "epoch": 0.34834742323619494,
+      "grad_norm": 2.1503591537475586,
+      "learning_rate": 2.6348750359092216e-05,
+      "loss": 2.875,
+      "step": 1225
+    },
+    {
+      "epoch": 0.3486317884796531,
+      "grad_norm": 1.839574933052063,
+      "learning_rate": 2.633725940821603e-05,
+      "loss": 2.4719,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3489161537231112,
+      "grad_norm": 1.8120521306991577,
+      "learning_rate": 2.6325768457339848e-05,
+      "loss": 2.2545,
+      "step": 1227
+    },
+    {
+      "epoch": 0.3492005189665693,
+      "grad_norm": 1.7394510507583618,
+      "learning_rate": 2.6314277506463663e-05,
+      "loss": 2.2625,
+      "step": 1228
+    },
+    {
+      "epoch": 0.34948488421002744,
+      "grad_norm": 1.7700285911560059,
+      "learning_rate": 2.6302786555587477e-05,
+      "loss": 2.2649,
+      "step": 1229
+    },
+    {
+      "epoch": 0.34976924945348553,
+      "grad_norm": 1.6092114448547363,
+      "learning_rate": 2.629129560471129e-05,
+      "loss": 1.9294,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3500536146969437,
+      "grad_norm": 1.8207448720932007,
+      "learning_rate": 2.6279804653835106e-05,
+      "loss": 1.9099,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3503379799404018,
+      "grad_norm": 1.8957747220993042,
+      "learning_rate": 2.626831370295892e-05,
+      "loss": 1.8944,
+      "step": 1232
+    },
+    {
+      "epoch": 0.35062234518385993,
+      "grad_norm": 2.1414735317230225,
+      "learning_rate": 2.6256822752082735e-05,
+      "loss": 2.878,
+      "step": 1233
+    },
+    {
+      "epoch": 0.350906710427318,
+      "grad_norm": 1.7470128536224365,
+      "learning_rate": 2.624533180120655e-05,
+      "loss": 2.6301,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3511910756707761,
+      "grad_norm": 1.71774423122406,
+      "learning_rate": 2.6233840850330364e-05,
+      "loss": 2.5042,
+      "step": 1235
+    },
+    {
+      "epoch": 0.35147544091423427,
+      "grad_norm": 1.7391661405563354,
+      "learning_rate": 2.6222349899454182e-05,
+      "loss": 2.5183,
+      "step": 1236
+    },
+    {
+      "epoch": 0.35175980615769237,
+      "grad_norm": 1.9395679235458374,
+      "learning_rate": 2.6210858948577996e-05,
+      "loss": 2.2172,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3520441714011505,
+      "grad_norm": 1.9249560832977295,
+      "learning_rate": 2.619936799770181e-05,
+      "loss": 2.0831,
+      "step": 1238
+    },
+    {
+      "epoch": 0.3523285366446086,
+      "grad_norm": 1.7382254600524902,
+      "learning_rate": 2.6187877046825625e-05,
+      "loss": 2.1057,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3526129018880667,
+      "grad_norm": 1.7688852548599243,
+      "learning_rate": 2.617638609594944e-05,
+      "loss": 1.9215,
+      "step": 1240
+    },
+    {
+      "epoch": 0.35289726713152486,
+      "grad_norm": 2.368236541748047,
+      "learning_rate": 2.6164895145073258e-05,
+      "loss": 2.7489,
+      "step": 1241
+    },
+    {
+      "epoch": 0.35318163237498296,
+      "grad_norm": 1.797239065170288,
+      "learning_rate": 2.6153404194197072e-05,
+      "loss": 2.568,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3534659976184411,
+      "grad_norm": 1.586428165435791,
+      "learning_rate": 2.614191324332089e-05,
+      "loss": 2.2321,
+      "step": 1243
+    },
+    {
+      "epoch": 0.3537503628618992,
+      "grad_norm": 1.7970319986343384,
+      "learning_rate": 2.6130422292444705e-05,
+      "loss": 2.405,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3540347281053573,
+      "grad_norm": 2.1091277599334717,
+      "learning_rate": 2.611893134156852e-05,
+      "loss": 2.5763,
+      "step": 1245
+    },
+    {
+      "epoch": 0.35431909334881545,
+      "grad_norm": 1.7261741161346436,
+      "learning_rate": 2.6107440390692334e-05,
+      "loss": 2.3919,
+      "step": 1246
+    },
+    {
+      "epoch": 0.35460345859227355,
+      "grad_norm": 1.808210015296936,
+      "learning_rate": 2.6095949439816148e-05,
+      "loss": 2.0507,
+      "step": 1247
+    },
+    {
+      "epoch": 0.3548878238357317,
+      "grad_norm": 2.0044949054718018,
+      "learning_rate": 2.6084458488939963e-05,
+      "loss": 1.9141,
+      "step": 1248
+    },
+    {
+      "epoch": 0.3551721890791898,
+      "grad_norm": 2.0011098384857178,
+      "learning_rate": 2.6072967538063777e-05,
+      "loss": 2.7051,
+      "step": 1249
+    },
+    {
+      "epoch": 0.3554565543226479,
+      "grad_norm": 1.7383923530578613,
+      "learning_rate": 2.606147658718759e-05,
+      "loss": 2.5117,
+      "step": 1250
+    },
+    {
+      "epoch": 0.35574091956610604,
+      "grad_norm": 1.8515959978103638,
+      "learning_rate": 2.6049985636311406e-05,
+      "loss": 2.5653,
+      "step": 1251
+    },
+    {
+      "epoch": 0.35602528480956414,
+      "grad_norm": 1.7738157510757446,
+      "learning_rate": 2.6038494685435224e-05,
+      "loss": 2.4249,
+      "step": 1252
+    },
+    {
+      "epoch": 0.3563096500530223,
+      "grad_norm": 1.834094524383545,
+      "learning_rate": 2.602700373455904e-05,
+      "loss": 2.3869,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3565940152964804,
+      "grad_norm": 1.6823465824127197,
+      "learning_rate": 2.6015512783682853e-05,
+      "loss": 1.9128,
+      "step": 1254
+    },
+    {
+      "epoch": 0.35687838053993853,
+      "grad_norm": 1.706059217453003,
+      "learning_rate": 2.6004021832806668e-05,
+      "loss": 1.8869,
+      "step": 1255
+    },
+    {
+      "epoch": 0.35716274578339663,
+      "grad_norm": 1.8522998094558716,
+      "learning_rate": 2.5992530881930482e-05,
+      "loss": 1.9591,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3574471110268547,
+      "grad_norm": 2.1196982860565186,
+      "learning_rate": 2.5981039931054297e-05,
+      "loss": 2.9377,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3577314762703129,
+      "grad_norm": 1.8649749755859375,
+      "learning_rate": 2.596954898017811e-05,
+      "loss": 2.5061,
+      "step": 1258
+    },
+    {
+      "epoch": 0.35801584151377097,
+      "grad_norm": 1.6296436786651611,
+      "learning_rate": 2.5958058029301926e-05,
+      "loss": 2.5032,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3583002067572291,
+      "grad_norm": 1.5869815349578857,
+      "learning_rate": 2.594656707842574e-05,
+      "loss": 2.4272,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3585845720006872,
+      "grad_norm": 1.7618229389190674,
+      "learning_rate": 2.5935076127549558e-05,
+      "loss": 2.3934,
+      "step": 1261
+    },
+    {
+      "epoch": 0.3588689372441453,
+      "grad_norm": 1.716230034828186,
+      "learning_rate": 2.5923585176673372e-05,
+      "loss": 2.0693,
+      "step": 1262
+    },
+    {
+      "epoch": 0.35915330248760347,
+      "grad_norm": 1.8889720439910889,
+      "learning_rate": 2.5912094225797187e-05,
+      "loss": 1.9207,
+      "step": 1263
+    },
+    {
+      "epoch": 0.35943766773106156,
+      "grad_norm": 1.825950264930725,
+      "learning_rate": 2.5900603274921e-05,
+      "loss": 1.867,
+      "step": 1264
+    },
+    {
+      "epoch": 0.3597220329745197,
+      "grad_norm": 2.1356546878814697,
+      "learning_rate": 2.5889112324044816e-05,
+      "loss": 3.0785,
+      "step": 1265
+    },
+    {
+      "epoch": 0.3600063982179778,
+      "grad_norm": 1.845746636390686,
+      "learning_rate": 2.587762137316863e-05,
+      "loss": 2.608,
+      "step": 1266
+    },
+    {
+      "epoch": 0.3602907634614359,
+      "grad_norm": 1.7854721546173096,
+      "learning_rate": 2.5866130422292445e-05,
+      "loss": 2.3848,
+      "step": 1267
+    },
+    {
+      "epoch": 0.36057512870489405,
+      "grad_norm": 1.6726619005203247,
+      "learning_rate": 2.585463947141626e-05,
+      "loss": 2.5518,
+      "step": 1268
+    },
+    {
+      "epoch": 0.36085949394835215,
+      "grad_norm": 1.716640830039978,
+      "learning_rate": 2.5843148520540077e-05,
+      "loss": 2.3346,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3611438591918103,
+      "grad_norm": 1.5924497842788696,
+      "learning_rate": 2.5831657569663892e-05,
+      "loss": 1.9842,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3614282244352684,
+      "grad_norm": 1.7810031175613403,
+      "learning_rate": 2.5820166618787706e-05,
+      "loss": 1.8548,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3617125896787265,
+      "grad_norm": 1.8115458488464355,
+      "learning_rate": 2.580867566791152e-05,
+      "loss": 2.0252,
+      "step": 1272
+    },
+    {
+      "epoch": 0.36199695492218464,
+      "grad_norm": 2.0684995651245117,
+      "learning_rate": 2.5797184717035335e-05,
+      "loss": 2.9993,
+      "step": 1273
+    },
+    {
+      "epoch": 0.36228132016564274,
+      "grad_norm": 1.792303204536438,
+      "learning_rate": 2.578569376615915e-05,
+      "loss": 2.5244,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3625656854091009,
+      "grad_norm": 1.5862210988998413,
+      "learning_rate": 2.5774202815282964e-05,
+      "loss": 2.4175,
+      "step": 1275
+    },
+    {
+      "epoch": 0.362850050652559,
+      "grad_norm": 1.706787109375,
+      "learning_rate": 2.576271186440678e-05,
+      "loss": 2.4619,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3631344158960171,
+      "grad_norm": 1.7666358947753906,
+      "learning_rate": 2.5751220913530593e-05,
+      "loss": 2.4914,
+      "step": 1277
+    },
+    {
+      "epoch": 0.36341878113947523,
+      "grad_norm": 1.6655855178833008,
+      "learning_rate": 2.5739729962654415e-05,
+      "loss": 2.0598,
+      "step": 1278
+    },
+    {
+      "epoch": 0.36370314638293333,
+      "grad_norm": 1.6447235345840454,
+      "learning_rate": 2.572823901177823e-05,
+      "loss": 1.734,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3639875116263915,
+      "grad_norm": 1.7836151123046875,
+      "learning_rate": 2.5716748060902044e-05,
+      "loss": 1.9249,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3642718768698496,
+      "grad_norm": 1.9049650430679321,
+      "learning_rate": 2.5705257110025858e-05,
+      "loss": 2.6863,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3645562421133077,
+      "grad_norm": 1.6191275119781494,
+      "learning_rate": 2.5693766159149673e-05,
+      "loss": 2.4752,
+      "step": 1282
+    },
+    {
+      "epoch": 0.3648406073567658,
+      "grad_norm": 1.6854596138000488,
+      "learning_rate": 2.5682275208273487e-05,
+      "loss": 2.4672,
+      "step": 1283
+    },
+    {
+      "epoch": 0.3651249726002239,
+      "grad_norm": 1.725872278213501,
+      "learning_rate": 2.56707842573973e-05,
+      "loss": 2.4114,
+      "step": 1284
+    },
+    {
+      "epoch": 0.36540933784368207,
+      "grad_norm": 1.7640068531036377,
+      "learning_rate": 2.565929330652112e-05,
+      "loss": 2.1844,
+      "step": 1285
+    },
+    {
+      "epoch": 0.36569370308714017,
+      "grad_norm": 1.6957398653030396,
+      "learning_rate": 2.5647802355644934e-05,
+      "loss": 2.1487,
+      "step": 1286
+    },
+    {
+      "epoch": 0.3659780683305983,
+      "grad_norm": 1.8824515342712402,
+      "learning_rate": 2.563631140476875e-05,
+      "loss": 2.0372,
+      "step": 1287
+    },
+    {
+      "epoch": 0.3662624335740564,
+      "grad_norm": 1.8746623992919922,
+      "learning_rate": 2.5624820453892563e-05,
+      "loss": 1.9418,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3665467988175145,
+      "grad_norm": 2.1677398681640625,
+      "learning_rate": 2.5613329503016377e-05,
+      "loss": 3.0094,
+      "step": 1289
+    },
+    {
+      "epoch": 0.36683116406097266,
+      "grad_norm": 1.7507233619689941,
+      "learning_rate": 2.5601838552140192e-05,
+      "loss": 2.5547,
+      "step": 1290
+    },
+    {
+      "epoch": 0.36711552930443075,
+      "grad_norm": 1.6727383136749268,
+      "learning_rate": 2.5590347601264006e-05,
+      "loss": 2.49,
+      "step": 1291
+    },
+    {
+      "epoch": 0.3673998945478889,
+      "grad_norm": 1.6071473360061646,
+      "learning_rate": 2.557885665038782e-05,
+      "loss": 2.5057,
+      "step": 1292
+    },
+    {
+      "epoch": 0.367684259791347,
+      "grad_norm": 1.853559970855713,
+      "learning_rate": 2.5567365699511635e-05,
+      "loss": 2.135,
+      "step": 1293
+    },
+    {
+      "epoch": 0.3679686250348051,
+      "grad_norm": 1.656827449798584,
+      "learning_rate": 2.5555874748635453e-05,
+      "loss": 1.9428,
+      "step": 1294
+    },
+    {
+      "epoch": 0.36825299027826325,
+      "grad_norm": 1.7858695983886719,
+      "learning_rate": 2.5544383797759268e-05,
+      "loss": 2.0289,
+      "step": 1295
+    },
+    {
+      "epoch": 0.36853735552172134,
+      "grad_norm": 1.7706724405288696,
+      "learning_rate": 2.5532892846883082e-05,
+      "loss": 1.7794,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3688217207651795,
+      "grad_norm": 1.8838144540786743,
+      "learning_rate": 2.5521401896006897e-05,
+      "loss": 2.7027,
+      "step": 1297
+    },
+    {
+      "epoch": 0.3691060860086376,
+      "grad_norm": 1.6255005598068237,
+      "learning_rate": 2.550991094513071e-05,
+      "loss": 2.7461,
+      "step": 1298
+    },
+    {
+      "epoch": 0.3693904512520957,
+      "grad_norm": 1.5868300199508667,
+      "learning_rate": 2.5498419994254526e-05,
+      "loss": 2.4882,
+      "step": 1299
+    },
+    {
+      "epoch": 0.36967481649555384,
+      "grad_norm": 1.6790016889572144,
+      "learning_rate": 2.548692904337834e-05,
+      "loss": 2.1273,
+      "step": 1300
+    },
+    {
+      "epoch": 0.36995918173901193,
+      "grad_norm": 1.8113175630569458,
+      "learning_rate": 2.5475438092502155e-05,
+      "loss": 2.2047,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3702435469824701,
+      "grad_norm": 1.5671573877334595,
+      "learning_rate": 2.546394714162597e-05,
+      "loss": 1.9255,
+      "step": 1302
+    },
+    {
+      "epoch": 0.3705279122259282,
+      "grad_norm": 1.6399245262145996,
+      "learning_rate": 2.5452456190749787e-05,
+      "loss": 1.8351,
+      "step": 1303
+    },
+    {
+      "epoch": 0.37081227746938633,
+      "grad_norm": 1.7190747261047363,
+      "learning_rate": 2.54409652398736e-05,
+      "loss": 2.0176,
+      "step": 1304
+    },
+    {
+      "epoch": 0.3710966427128444,
+      "grad_norm": 2.0205039978027344,
+      "learning_rate": 2.5429474288997416e-05,
+      "loss": 2.9932,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3713810079563025,
+      "grad_norm": 1.7755295038223267,
+      "learning_rate": 2.541798333812123e-05,
+      "loss": 2.5928,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3716653731997607,
+      "grad_norm": 1.6791408061981201,
+      "learning_rate": 2.5406492387245045e-05,
+      "loss": 2.4596,
+      "step": 1307
+    },
+    {
+      "epoch": 0.37194973844321877,
+      "grad_norm": 1.682665467262268,
+      "learning_rate": 2.539500143636886e-05,
+      "loss": 2.284,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3722341036866769,
+      "grad_norm": 1.661361813545227,
+      "learning_rate": 2.5383510485492674e-05,
+      "loss": 2.4567,
+      "step": 1309
+    },
+    {
+      "epoch": 0.372518468930135,
+      "grad_norm": 1.7297680377960205,
+      "learning_rate": 2.537201953461649e-05,
+      "loss": 1.993,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3728028341735931,
+      "grad_norm": 1.7310127019882202,
+      "learning_rate": 2.5360528583740303e-05,
+      "loss": 2.0113,
+      "step": 1311
+    },
+    {
+      "epoch": 0.37308719941705126,
+      "grad_norm": 1.7791327238082886,
+      "learning_rate": 2.534903763286412e-05,
+      "loss": 1.8896,
+      "step": 1312
+    },
+    {
+      "epoch": 0.37337156466050936,
+      "grad_norm": 2.11208438873291,
+      "learning_rate": 2.5337546681987936e-05,
+      "loss": 3.0813,
+      "step": 1313
+    },
+    {
+      "epoch": 0.3736559299039675,
+      "grad_norm": 1.8183382749557495,
+      "learning_rate": 2.532605573111175e-05,
+      "loss": 2.7019,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3739402951474256,
+      "grad_norm": 1.6248207092285156,
+      "learning_rate": 2.5314564780235568e-05,
+      "loss": 2.422,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3742246603908837,
+      "grad_norm": 1.640504002571106,
+      "learning_rate": 2.5303073829359382e-05,
+      "loss": 2.3868,
+      "step": 1316
+    },
+    {
+      "epoch": 0.37450902563434185,
+      "grad_norm": 1.7609598636627197,
+      "learning_rate": 2.5291582878483197e-05,
+      "loss": 2.263,
+      "step": 1317
+    },
+    {
+      "epoch": 0.37479339087779995,
+      "grad_norm": 1.7517386674880981,
+      "learning_rate": 2.5280091927607015e-05,
+      "loss": 2.024,
+      "step": 1318
+    },
+    {
+      "epoch": 0.3750777561212581,
+      "grad_norm": 1.6678249835968018,
+      "learning_rate": 2.526860097673083e-05,
+      "loss": 1.8084,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3753621213647162,
+      "grad_norm": 1.7585467100143433,
+      "learning_rate": 2.5257110025854644e-05,
+      "loss": 1.7552,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3756464866081743,
+      "grad_norm": 2.2945332527160645,
+      "learning_rate": 2.5245619074978458e-05,
+      "loss": 3.0485,
+      "step": 1321
+    },
+    {
+      "epoch": 0.37593085185163244,
+      "grad_norm": 1.6256428956985474,
+      "learning_rate": 2.5234128124102273e-05,
+      "loss": 2.3769,
+      "step": 1322
+    },
+    {
+      "epoch": 0.37621521709509054,
+      "grad_norm": 1.6193193197250366,
+      "learning_rate": 2.5222637173226087e-05,
+      "loss": 2.3007,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3764995823385487,
+      "grad_norm": 1.6534637212753296,
+      "learning_rate": 2.5211146222349902e-05,
+      "loss": 2.5241,
+      "step": 1324
+    },
+    {
+      "epoch": 0.3767839475820068,
+      "grad_norm": 1.7041131258010864,
+      "learning_rate": 2.5199655271473716e-05,
+      "loss": 2.3242,
+      "step": 1325
+    },
+    {
+      "epoch": 0.37706831282546494,
+      "grad_norm": 1.697442889213562,
+      "learning_rate": 2.518816432059753e-05,
+      "loss": 1.9187,
+      "step": 1326
+    },
+    {
+      "epoch": 0.37735267806892303,
+      "grad_norm": 1.7257381677627563,
+      "learning_rate": 2.517667336972135e-05,
+      "loss": 1.7997,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3776370433123811,
+      "grad_norm": 1.8041011095046997,
+      "learning_rate": 2.5165182418845163e-05,
+      "loss": 1.9322,
+      "step": 1328
+    },
+    {
+      "epoch": 0.3779214085558393,
+      "grad_norm": 2.060734748840332,
+      "learning_rate": 2.5153691467968978e-05,
+      "loss": 2.7627,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3782057737992974,
+      "grad_norm": 1.7511411905288696,
+      "learning_rate": 2.5142200517092792e-05,
+      "loss": 2.2934,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3784901390427555,
+      "grad_norm": 1.6538666486740112,
+      "learning_rate": 2.5130709566216607e-05,
+      "loss": 2.2186,
+      "step": 1331
+    },
+    {
+      "epoch": 0.3787745042862136,
+      "grad_norm": 1.672150731086731,
+      "learning_rate": 2.511921861534042e-05,
+      "loss": 2.3287,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3790588695296717,
+      "grad_norm": 1.8601032495498657,
+      "learning_rate": 2.5107727664464236e-05,
+      "loss": 2.1725,
+      "step": 1333
+    },
+    {
+      "epoch": 0.37934323477312987,
+      "grad_norm": 1.6767491102218628,
+      "learning_rate": 2.509623671358805e-05,
+      "loss": 2.1525,
+      "step": 1334
+    },
+    {
+      "epoch": 0.37962760001658796,
+      "grad_norm": 1.72711980342865,
+      "learning_rate": 2.5084745762711865e-05,
+      "loss": 2.1467,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3799119652600461,
+      "grad_norm": 1.7311689853668213,
+      "learning_rate": 2.5073254811835683e-05,
+      "loss": 1.8984,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3801963305035042,
+      "grad_norm": 2.0441644191741943,
+      "learning_rate": 2.5061763860959497e-05,
+      "loss": 3.0045,
+      "step": 1337
+    },
+    {
+      "epoch": 0.3804806957469623,
+      "grad_norm": 1.9976049661636353,
+      "learning_rate": 2.505027291008331e-05,
+      "loss": 2.6461,
+      "step": 1338
+    },
+    {
+      "epoch": 0.38076506099042046,
+      "grad_norm": 1.6794681549072266,
+      "learning_rate": 2.5038781959207126e-05,
+      "loss": 2.4184,
+      "step": 1339
+    },
+    {
+      "epoch": 0.38104942623387855,
+      "grad_norm": 1.573089599609375,
+      "learning_rate": 2.502729100833094e-05,
+      "loss": 2.3171,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3813337914773367,
+      "grad_norm": 1.8047806024551392,
+      "learning_rate": 2.5015800057454755e-05,
+      "loss": 2.4302,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3816181567207948,
+      "grad_norm": 1.679032564163208,
+      "learning_rate": 2.500430910657857e-05,
+      "loss": 2.0206,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3819025219642529,
+      "grad_norm": 1.6827946901321411,
+      "learning_rate": 2.4992818155702384e-05,
+      "loss": 1.8739,
+      "step": 1343
+    },
+    {
+      "epoch": 0.38218688720771105,
+      "grad_norm": 2.0795507431030273,
+      "learning_rate": 2.49813272048262e-05,
+      "loss": 1.9664,
+      "step": 1344
+    },
+    {
+      "epoch": 0.38247125245116914,
+      "grad_norm": 2.012922763824463,
+      "learning_rate": 2.4969836253950016e-05,
+      "loss": 2.757,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3827556176946273,
+      "grad_norm": 1.62966787815094,
+      "learning_rate": 2.495834530307383e-05,
+      "loss": 2.5295,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3830399829380854,
+      "grad_norm": 1.6679270267486572,
+      "learning_rate": 2.4946854352197645e-05,
+      "loss": 2.459,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3833243481815435,
+      "grad_norm": 1.615654468536377,
+      "learning_rate": 2.493536340132146e-05,
+      "loss": 2.1912,
+      "step": 1348
+    },
+    {
+      "epoch": 0.38360871342500164,
+      "grad_norm": 1.8527601957321167,
+      "learning_rate": 2.4923872450445274e-05,
+      "loss": 2.2239,
+      "step": 1349
+    },
+    {
+      "epoch": 0.38389307866845973,
+      "grad_norm": 1.7967239618301392,
+      "learning_rate": 2.491238149956909e-05,
+      "loss": 1.9359,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3841774439119179,
+      "grad_norm": 1.7077158689498901,
+      "learning_rate": 2.4900890548692903e-05,
+      "loss": 1.737,
+      "step": 1351
+    },
+    {
+      "epoch": 0.384461809155376,
+      "grad_norm": 1.7377227544784546,
+      "learning_rate": 2.4889399597816725e-05,
+      "loss": 1.934,
+      "step": 1352
+    },
+    {
+      "epoch": 0.38474617439883413,
+      "grad_norm": 1.9111387729644775,
+      "learning_rate": 2.487790864694054e-05,
+      "loss": 2.776,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3850305396422922,
+      "grad_norm": 1.6393592357635498,
+      "learning_rate": 2.4866417696064354e-05,
+      "loss": 2.4337,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3853149048857503,
+      "grad_norm": 1.7476547956466675,
+      "learning_rate": 2.4854926745188168e-05,
+      "loss": 2.377,
+      "step": 1355
+    },
+    {
+      "epoch": 0.38559927012920847,
+      "grad_norm": 1.8083293437957764,
+      "learning_rate": 2.4843435794311983e-05,
+      "loss": 2.3762,
+      "step": 1356
+    },
+    {
+      "epoch": 0.38588363537266657,
+      "grad_norm": 1.9728158712387085,
+      "learning_rate": 2.4831944843435797e-05,
+      "loss": 2.242,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3861680006161247,
+      "grad_norm": 1.5531713962554932,
+      "learning_rate": 2.482045389255961e-05,
+      "loss": 2.0673,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3864523658595828,
+      "grad_norm": 1.5641489028930664,
+      "learning_rate": 2.4808962941683426e-05,
+      "loss": 1.7698,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3867367311030409,
+      "grad_norm": 1.7081369161605835,
+      "learning_rate": 2.4797471990807244e-05,
+      "loss": 1.7401,
+      "step": 1360
+    },
+    {
+      "epoch": 0.38702109634649906,
+      "grad_norm": 1.9915111064910889,
+      "learning_rate": 2.478598103993106e-05,
+      "loss": 2.9381,
+      "step": 1361
+    },
+    {
+      "epoch": 0.38730546158995716,
+      "grad_norm": 1.6150346994400024,
+      "learning_rate": 2.4774490089054873e-05,
+      "loss": 2.5681,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3875898268334153,
+      "grad_norm": 1.6368330717086792,
+      "learning_rate": 2.4762999138178687e-05,
+      "loss": 2.5464,
+      "step": 1363
+    },
+    {
+      "epoch": 0.3878741920768734,
+      "grad_norm": 1.6457427740097046,
+      "learning_rate": 2.4751508187302502e-05,
+      "loss": 2.1136,
+      "step": 1364
+    },
+    {
+      "epoch": 0.3881585573203315,
+      "grad_norm": 1.6669577360153198,
+      "learning_rate": 2.4740017236426316e-05,
+      "loss": 2.3689,
+      "step": 1365
+    },
+    {
+      "epoch": 0.38844292256378965,
+      "grad_norm": 1.6207609176635742,
+      "learning_rate": 2.472852628555013e-05,
+      "loss": 2.0466,
+      "step": 1366
+    },
+    {
+      "epoch": 0.38872728780724775,
+      "grad_norm": 1.6581090688705444,
+      "learning_rate": 2.4717035334673945e-05,
+      "loss": 2.1255,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3890116530507059,
+      "grad_norm": 1.7444984912872314,
+      "learning_rate": 2.470554438379776e-05,
+      "loss": 1.9976,
+      "step": 1368
+    },
+    {
+      "epoch": 0.389296018294164,
+      "grad_norm": 2.0723159313201904,
+      "learning_rate": 2.4694053432921578e-05,
+      "loss": 3.1322,
+      "step": 1369
+    },
+    {
+      "epoch": 0.3895803835376221,
+      "grad_norm": 1.8740094900131226,
+      "learning_rate": 2.4682562482045392e-05,
+      "loss": 2.346,
+      "step": 1370
+    },
+    {
+      "epoch": 0.38986474878108024,
+      "grad_norm": 1.7882744073867798,
+      "learning_rate": 2.4671071531169207e-05,
+      "loss": 2.3347,
+      "step": 1371
+    },
+    {
+      "epoch": 0.39014911402453833,
+      "grad_norm": 1.7288506031036377,
+      "learning_rate": 2.465958058029302e-05,
+      "loss": 2.2182,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3904334792679965,
+      "grad_norm": 1.6869946718215942,
+      "learning_rate": 2.4648089629416836e-05,
+      "loss": 2.203,
+      "step": 1373
+    },
+    {
+      "epoch": 0.3907178445114546,
+      "grad_norm": 1.6515893936157227,
+      "learning_rate": 2.463659867854065e-05,
+      "loss": 2.1165,
+      "step": 1374
+    },
+    {
+      "epoch": 0.39100220975491273,
+      "grad_norm": 1.7650010585784912,
+      "learning_rate": 2.4625107727664465e-05,
+      "loss": 1.8461,
+      "step": 1375
+    },
+    {
+      "epoch": 0.39128657499837083,
+      "grad_norm": 1.602225661277771,
+      "learning_rate": 2.461361677678828e-05,
+      "loss": 1.8419,
+      "step": 1376
+    },
+    {
+      "epoch": 0.3915709402418289,
+      "grad_norm": 1.930778980255127,
+      "learning_rate": 2.4602125825912094e-05,
+      "loss": 2.7933,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3918553054852871,
+      "grad_norm": 1.921008586883545,
+      "learning_rate": 2.4590634875035912e-05,
+      "loss": 2.6368,
+      "step": 1378
+    },
+    {
+      "epoch": 0.39213967072874517,
+      "grad_norm": 1.7024277448654175,
+      "learning_rate": 2.4579143924159726e-05,
+      "loss": 2.3727,
+      "step": 1379
+    },
+    {
+      "epoch": 0.3924240359722033,
+      "grad_norm": 1.522916316986084,
+      "learning_rate": 2.456765297328354e-05,
+      "loss": 2.2389,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3927084012156614,
+      "grad_norm": 1.8298519849777222,
+      "learning_rate": 2.4556162022407355e-05,
+      "loss": 2.3436,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3929927664591195,
+      "grad_norm": 1.6771541833877563,
+      "learning_rate": 2.454467107153117e-05,
+      "loss": 1.8835,
+      "step": 1382
+    },
+    {
+      "epoch": 0.39327713170257766,
+      "grad_norm": 1.8293532133102417,
+      "learning_rate": 2.4533180120654984e-05,
+      "loss": 2.0321,
+      "step": 1383
+    },
+    {
+      "epoch": 0.39356149694603576,
+      "grad_norm": 1.8741817474365234,
+      "learning_rate": 2.45216891697788e-05,
+      "loss": 1.9599,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3938458621894939,
+      "grad_norm": 2.148267984390259,
+      "learning_rate": 2.4510198218902613e-05,
+      "loss": 2.7614,
+      "step": 1385
+    },
+    {
+      "epoch": 0.394130227432952,
+      "grad_norm": 1.771562933921814,
+      "learning_rate": 2.4498707268026428e-05,
+      "loss": 2.6609,
+      "step": 1386
+    },
+    {
+      "epoch": 0.3944145926764101,
+      "grad_norm": 1.6484909057617188,
+      "learning_rate": 2.4487216317150246e-05,
+      "loss": 2.3871,
+      "step": 1387
+    },
+    {
+      "epoch": 0.39469895791986825,
+      "grad_norm": 1.6296532154083252,
+      "learning_rate": 2.447572536627406e-05,
+      "loss": 2.3645,
+      "step": 1388
+    },
+    {
+      "epoch": 0.39498332316332635,
+      "grad_norm": 1.7516010999679565,
+      "learning_rate": 2.4464234415397878e-05,
+      "loss": 2.1963,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3952676884067845,
+      "grad_norm": 1.6003243923187256,
+      "learning_rate": 2.4452743464521692e-05,
+      "loss": 1.9164,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3955520536502426,
+      "grad_norm": 1.8290302753448486,
+      "learning_rate": 2.4441252513645507e-05,
+      "loss": 1.9689,
+      "step": 1391
+    },
+    {
+      "epoch": 0.3958364188937007,
+      "grad_norm": 1.694347858428955,
+      "learning_rate": 2.442976156276932e-05,
+      "loss": 1.7561,
+      "step": 1392
+    },
+    {
+      "epoch": 0.39612078413715884,
+      "grad_norm": 1.91927969455719,
+      "learning_rate": 2.441827061189314e-05,
+      "loss": 2.8435,
+      "step": 1393
+    },
+    {
+      "epoch": 0.39640514938061694,
+      "grad_norm": 1.766818642616272,
+      "learning_rate": 2.4406779661016954e-05,
+      "loss": 2.6195,
+      "step": 1394
+    },
+    {
+      "epoch": 0.3966895146240751,
+      "grad_norm": 1.7529215812683105,
+      "learning_rate": 2.439528871014077e-05,
+      "loss": 2.455,
+      "step": 1395
+    },
+    {
+      "epoch": 0.3969738798675332,
+      "grad_norm": 1.6795532703399658,
+      "learning_rate": 2.4383797759264583e-05,
+      "loss": 2.4504,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3972582451109913,
+      "grad_norm": 1.8063428401947021,
+      "learning_rate": 2.4372306808388397e-05,
+      "loss": 2.3136,
+      "step": 1397
+    },
+    {
+      "epoch": 0.39754261035444943,
+      "grad_norm": 1.8748013973236084,
+      "learning_rate": 2.4360815857512212e-05,
+      "loss": 1.9549,
+      "step": 1398
+    },
+    {
+      "epoch": 0.39782697559790753,
+      "grad_norm": 1.702768087387085,
+      "learning_rate": 2.4349324906636026e-05,
+      "loss": 1.9597,
+      "step": 1399
+    },
+    {
+      "epoch": 0.3981113408413657,
+      "grad_norm": 1.8588896989822388,
+      "learning_rate": 2.433783395575984e-05,
+      "loss": 1.9016,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3983957060848238,
+      "grad_norm": 1.9412899017333984,
+      "learning_rate": 2.4326343004883655e-05,
+      "loss": 2.6964,
+      "step": 1401
+    },
+    {
+      "epoch": 0.3986800713282819,
+      "grad_norm": 1.590203046798706,
+      "learning_rate": 2.4314852054007473e-05,
+      "loss": 2.5482,
+      "step": 1402
+    },
+    {
+      "epoch": 0.39896443657174,
+      "grad_norm": 1.613073468208313,
+      "learning_rate": 2.4303361103131288e-05,
+      "loss": 2.3643,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3992488018151981,
+      "grad_norm": 1.7016730308532715,
+      "learning_rate": 2.4291870152255102e-05,
+      "loss": 2.179,
+      "step": 1404
+    },
+    {
+      "epoch": 0.39953316705865627,
+      "grad_norm": 1.7055073976516724,
+      "learning_rate": 2.4280379201378917e-05,
+      "loss": 2.3988,
+      "step": 1405
+    },
+    {
+      "epoch": 0.39981753230211436,
+      "grad_norm": 1.6752086877822876,
+      "learning_rate": 2.426888825050273e-05,
+      "loss": 2.3758,
+      "step": 1406
+    },
+    {
+      "epoch": 0.4001018975455725,
+      "grad_norm": 1.6053355932235718,
+      "learning_rate": 2.4257397299626546e-05,
+      "loss": 1.9306,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4003862627890306,
+      "grad_norm": 1.7850687503814697,
+      "learning_rate": 2.424590634875036e-05,
+      "loss": 1.8416,
+      "step": 1408
+    },
+    {
+      "epoch": 0.4006706280324887,
+      "grad_norm": 2.1916213035583496,
+      "learning_rate": 2.4234415397874175e-05,
+      "loss": 2.8873,
+      "step": 1409
+    },
+    {
+      "epoch": 0.40095499327594686,
+      "grad_norm": 1.7736766338348389,
+      "learning_rate": 2.422292444699799e-05,
+      "loss": 2.5443,
+      "step": 1410
+    },
+    {
+      "epoch": 0.40123935851940495,
+      "grad_norm": 1.630659818649292,
+      "learning_rate": 2.4211433496121807e-05,
+      "loss": 2.4075,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4015237237628631,
+      "grad_norm": 1.5184587240219116,
+      "learning_rate": 2.419994254524562e-05,
+      "loss": 2.2805,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4018080890063212,
+      "grad_norm": 1.721583366394043,
+      "learning_rate": 2.4188451594369436e-05,
+      "loss": 2.3186,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4020924542497793,
+      "grad_norm": 1.7377300262451172,
+      "learning_rate": 2.417696064349325e-05,
+      "loss": 1.9795,
+      "step": 1414
+    },
+    {
+      "epoch": 0.40237681949323745,
+      "grad_norm": 1.7117558717727661,
+      "learning_rate": 2.4165469692617065e-05,
+      "loss": 2.2045,
+      "step": 1415
+    },
+    {
+      "epoch": 0.40266118473669554,
+      "grad_norm": 1.7114864587783813,
+      "learning_rate": 2.415397874174088e-05,
+      "loss": 1.799,
+      "step": 1416
+    },
+    {
+      "epoch": 0.4029455499801537,
+      "grad_norm": 1.7981421947479248,
+      "learning_rate": 2.4142487790864694e-05,
+      "loss": 3.078,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4032299152236118,
+      "grad_norm": 1.600596308708191,
+      "learning_rate": 2.413099683998851e-05,
+      "loss": 2.3393,
+      "step": 1418
+    },
+    {
+      "epoch": 0.4035142804670699,
+      "grad_norm": 1.4729454517364502,
+      "learning_rate": 2.4119505889112323e-05,
+      "loss": 2.3691,
+      "step": 1419
+    },
+    {
+      "epoch": 0.40379864571052804,
+      "grad_norm": 1.6638671159744263,
+      "learning_rate": 2.410801493823614e-05,
+      "loss": 2.2835,
+      "step": 1420
+    },
+    {
+      "epoch": 0.40408301095398613,
+      "grad_norm": 1.79701566696167,
+      "learning_rate": 2.4096523987359955e-05,
+      "loss": 2.2159,
+      "step": 1421
+    },
+    {
+      "epoch": 0.4043673761974443,
+      "grad_norm": 1.6714825630187988,
+      "learning_rate": 2.408503303648377e-05,
+      "loss": 1.8111,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4046517414409024,
+      "grad_norm": 1.736799716949463,
+      "learning_rate": 2.4073542085607584e-05,
+      "loss": 1.9677,
+      "step": 1423
+    },
+    {
+      "epoch": 0.40493610668436053,
+      "grad_norm": 1.7546263933181763,
+      "learning_rate": 2.40620511347314e-05,
+      "loss": 1.8836,
+      "step": 1424
+    },
+    {
+      "epoch": 0.4052204719278186,
+      "grad_norm": 1.9755287170410156,
+      "learning_rate": 2.4050560183855213e-05,
+      "loss": 2.9325,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4055048371712767,
+      "grad_norm": 1.6950325965881348,
+      "learning_rate": 2.4039069232979035e-05,
+      "loss": 2.7484,
+      "step": 1426
+    },
+    {
+      "epoch": 0.4057892024147349,
+      "grad_norm": 1.787856936454773,
+      "learning_rate": 2.402757828210285e-05,
+      "loss": 2.5154,
+      "step": 1427
+    },
+    {
+      "epoch": 0.40607356765819297,
+      "grad_norm": 1.7529258728027344,
+      "learning_rate": 2.4016087331226664e-05,
+      "loss": 2.3995,
+      "step": 1428
+    },
+    {
+      "epoch": 0.4063579329016511,
+      "grad_norm": 1.8270007371902466,
+      "learning_rate": 2.4004596380350478e-05,
+      "loss": 2.2466,
+      "step": 1429
+    },
+    {
+      "epoch": 0.4066422981451092,
+      "grad_norm": 1.6671589612960815,
+      "learning_rate": 2.3993105429474293e-05,
+      "loss": 2.0727,
+      "step": 1430
+    },
+    {
+      "epoch": 0.4069266633885673,
+      "grad_norm": 1.8417056798934937,
+      "learning_rate": 2.3981614478598107e-05,
+      "loss": 1.9906,
+      "step": 1431
+    },
+    {
+      "epoch": 0.40721102863202546,
+      "grad_norm": 1.794356346130371,
+      "learning_rate": 2.397012352772192e-05,
+      "loss": 1.9631,
+      "step": 1432
+    },
+    {
+      "epoch": 0.40749539387548356,
+      "grad_norm": 1.9737939834594727,
+      "learning_rate": 2.3958632576845736e-05,
+      "loss": 2.8793,
+      "step": 1433
+    },
+    {
+      "epoch": 0.4077797591189417,
+      "grad_norm": 1.7797672748565674,
+      "learning_rate": 2.394714162596955e-05,
+      "loss": 2.6237,
+      "step": 1434
+    },
+    {
+      "epoch": 0.4080641243623998,
+      "grad_norm": 1.782353401184082,
+      "learning_rate": 2.393565067509337e-05,
+      "loss": 2.3761,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4083484896058579,
+      "grad_norm": 1.662298560142517,
+      "learning_rate": 2.3924159724217183e-05,
+      "loss": 2.3679,
+      "step": 1436
+    },
+    {
+      "epoch": 0.40863285484931605,
+      "grad_norm": 1.7160699367523193,
+      "learning_rate": 2.3912668773340998e-05,
+      "loss": 2.0721,
+      "step": 1437
+    },
+    {
+      "epoch": 0.40891722009277415,
+      "grad_norm": 1.5304116010665894,
+      "learning_rate": 2.3901177822464812e-05,
+      "loss": 1.9679,
+      "step": 1438
+    },
+    {
+      "epoch": 0.4092015853362323,
+      "grad_norm": 1.6749613285064697,
+      "learning_rate": 2.3889686871588627e-05,
+      "loss": 1.9677,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4094859505796904,
+      "grad_norm": 1.8116540908813477,
+      "learning_rate": 2.387819592071244e-05,
+      "loss": 1.7566,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4097703158231485,
+      "grad_norm": 2.061837911605835,
+      "learning_rate": 2.3866704969836256e-05,
+      "loss": 2.768,
+      "step": 1441
+    },
+    {
+      "epoch": 0.41005468106660664,
+      "grad_norm": 1.6977015733718872,
+      "learning_rate": 2.385521401896007e-05,
+      "loss": 2.4958,
+      "step": 1442
+    },
+    {
+      "epoch": 0.41033904631006474,
+      "grad_norm": 1.6543117761611938,
+      "learning_rate": 2.3843723068083885e-05,
+      "loss": 2.4396,
+      "step": 1443
+    },
+    {
+      "epoch": 0.4106234115535229,
+      "grad_norm": 1.5568379163742065,
+      "learning_rate": 2.3832232117207702e-05,
+      "loss": 2.2158,
+      "step": 1444
+    },
+    {
+      "epoch": 0.410907776796981,
+      "grad_norm": 1.6220964193344116,
+      "learning_rate": 2.3820741166331517e-05,
+      "loss": 2.3038,
+      "step": 1445
+    },
+    {
+      "epoch": 0.41119214204043913,
+      "grad_norm": 1.6945890188217163,
+      "learning_rate": 2.380925021545533e-05,
+      "loss": 1.9376,
+      "step": 1446
+    },
+    {
+      "epoch": 0.41147650728389723,
+      "grad_norm": 1.5313851833343506,
+      "learning_rate": 2.3797759264579146e-05,
+      "loss": 1.888,
+      "step": 1447
+    },
+    {
+      "epoch": 0.4117608725273553,
+      "grad_norm": 1.7938929796218872,
+      "learning_rate": 2.378626831370296e-05,
+      "loss": 1.9618,
+      "step": 1448
+    },
+    {
+      "epoch": 0.4120452377708135,
+      "grad_norm": 2.1637260913848877,
+      "learning_rate": 2.3774777362826775e-05,
+      "loss": 2.864,
+      "step": 1449
+    },
+    {
+      "epoch": 0.4123296030142716,
+      "grad_norm": 1.8949286937713623,
+      "learning_rate": 2.376328641195059e-05,
+      "loss": 2.675,
+      "step": 1450
+    },
+    {
+      "epoch": 0.4126139682577297,
+      "grad_norm": 1.711017370223999,
+      "learning_rate": 2.3751795461074404e-05,
+      "loss": 2.4241,
+      "step": 1451
+    },
+    {
+      "epoch": 0.4128983335011878,
+      "grad_norm": 1.6025915145874023,
+      "learning_rate": 2.374030451019822e-05,
+      "loss": 2.3166,
+      "step": 1452
+    },
+    {
+      "epoch": 0.4131826987446459,
+      "grad_norm": 1.7555315494537354,
+      "learning_rate": 2.3728813559322036e-05,
+      "loss": 2.0809,
+      "step": 1453
+    },
+    {
+      "epoch": 0.41346706398810407,
+      "grad_norm": 1.5702033042907715,
+      "learning_rate": 2.371732260844585e-05,
+      "loss": 1.981,
+      "step": 1454
+    },
+    {
+      "epoch": 0.41375142923156216,
+      "grad_norm": 1.6825945377349854,
+      "learning_rate": 2.3705831657569665e-05,
+      "loss": 1.766,
+      "step": 1455
+    },
+    {
+      "epoch": 0.4140357944750203,
+      "grad_norm": 1.711393117904663,
+      "learning_rate": 2.369434070669348e-05,
+      "loss": 2.0005,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4143201597184784,
+      "grad_norm": 1.917224407196045,
+      "learning_rate": 2.3682849755817294e-05,
+      "loss": 2.8846,
+      "step": 1457
+    },
+    {
+      "epoch": 0.4146045249619365,
+      "grad_norm": 1.728300929069519,
+      "learning_rate": 2.367135880494111e-05,
+      "loss": 2.63,
+      "step": 1458
+    },
+    {
+      "epoch": 0.41488889020539466,
+      "grad_norm": 1.6932635307312012,
+      "learning_rate": 2.3659867854064923e-05,
+      "loss": 2.3097,
+      "step": 1459
+    },
+    {
+      "epoch": 0.41517325544885275,
+      "grad_norm": 1.6644999980926514,
+      "learning_rate": 2.3648376903188738e-05,
+      "loss": 2.3912,
+      "step": 1460
+    },
+    {
+      "epoch": 0.4154576206923109,
+      "grad_norm": 1.7875523567199707,
+      "learning_rate": 2.3636885952312552e-05,
+      "loss": 2.3286,
+      "step": 1461
+    },
+    {
+      "epoch": 0.415741985935769,
+      "grad_norm": 1.7267717123031616,
+      "learning_rate": 2.362539500143637e-05,
+      "loss": 2.021,
+      "step": 1462
+    },
+    {
+      "epoch": 0.4160263511792271,
+      "grad_norm": 1.854394555091858,
+      "learning_rate": 2.3613904050560188e-05,
+      "loss": 2.0849,
+      "step": 1463
+    },
+    {
+      "epoch": 0.41631071642268525,
+      "grad_norm": 1.7931166887283325,
+      "learning_rate": 2.3602413099684003e-05,
+      "loss": 1.9079,
+      "step": 1464
+    },
+    {
+      "epoch": 0.41659508166614334,
+      "grad_norm": 1.9061031341552734,
+      "learning_rate": 2.3590922148807817e-05,
+      "loss": 2.7233,
+      "step": 1465
+    },
+    {
+      "epoch": 0.4168794469096015,
+      "grad_norm": 1.6702359914779663,
+      "learning_rate": 2.357943119793163e-05,
+      "loss": 2.6487,
+      "step": 1466
+    },
+    {
+      "epoch": 0.4171638121530596,
+      "grad_norm": 1.5894525051116943,
+      "learning_rate": 2.3567940247055446e-05,
+      "loss": 2.3472,
+      "step": 1467
+    },
+    {
+      "epoch": 0.4174481773965177,
+      "grad_norm": 1.6212420463562012,
+      "learning_rate": 2.3556449296179264e-05,
+      "loss": 2.0305,
+      "step": 1468
+    },
+    {
+      "epoch": 0.41773254263997583,
+      "grad_norm": 1.8889355659484863,
+      "learning_rate": 2.354495834530308e-05,
+      "loss": 2.5149,
+      "step": 1469
+    },
+    {
+      "epoch": 0.41801690788343393,
+      "grad_norm": 1.4953631162643433,
+      "learning_rate": 2.3533467394426893e-05,
+      "loss": 1.877,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4183012731268921,
+      "grad_norm": 1.6321781873703003,
+      "learning_rate": 2.3521976443550707e-05,
+      "loss": 1.9186,
+      "step": 1471
+    },
+    {
+      "epoch": 0.4185856383703502,
+      "grad_norm": 1.6362193822860718,
+      "learning_rate": 2.3510485492674522e-05,
+      "loss": 1.7191,
+      "step": 1472
+    },
+    {
+      "epoch": 0.41887000361380833,
+      "grad_norm": 2.1711504459381104,
+      "learning_rate": 2.3498994541798336e-05,
+      "loss": 2.9604,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4191543688572664,
+      "grad_norm": 1.5688064098358154,
+      "learning_rate": 2.348750359092215e-05,
+      "loss": 2.4455,
+      "step": 1474
+    },
+    {
+      "epoch": 0.4194387341007245,
+      "grad_norm": 1.5791515111923218,
+      "learning_rate": 2.3476012640045965e-05,
+      "loss": 2.3827,
+      "step": 1475
+    },
+    {
+      "epoch": 0.41972309934418267,
+      "grad_norm": 1.5574970245361328,
+      "learning_rate": 2.346452168916978e-05,
+      "loss": 2.2648,
+      "step": 1476
+    },
+    {
+      "epoch": 0.42000746458764077,
+      "grad_norm": 1.607318639755249,
+      "learning_rate": 2.3453030738293598e-05,
+      "loss": 2.1887,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4202918298310989,
+      "grad_norm": 1.6023167371749878,
+      "learning_rate": 2.3441539787417412e-05,
+      "loss": 1.8484,
+      "step": 1478
+    },
+    {
+      "epoch": 0.420576195074557,
+      "grad_norm": 1.614344835281372,
+      "learning_rate": 2.3430048836541227e-05,
+      "loss": 1.7518,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4208605603180151,
+      "grad_norm": 1.7779884338378906,
+      "learning_rate": 2.341855788566504e-05,
+      "loss": 1.9386,
+      "step": 1480
+    },
+    {
+      "epoch": 0.42114492556147326,
+      "grad_norm": 1.913327932357788,
+      "learning_rate": 2.3407066934788856e-05,
+      "loss": 2.8805,
+      "step": 1481
+    },
+    {
+      "epoch": 0.42142929080493136,
+      "grad_norm": 1.888013482093811,
+      "learning_rate": 2.339557598391267e-05,
+      "loss": 2.4296,
+      "step": 1482
+    },
+    {
+      "epoch": 0.4217136560483895,
+      "grad_norm": 1.841597080230713,
+      "learning_rate": 2.3384085033036485e-05,
+      "loss": 2.4056,
+      "step": 1483
+    },
+    {
+      "epoch": 0.4219980212918476,
+      "grad_norm": 1.6443198919296265,
+      "learning_rate": 2.33725940821603e-05,
+      "loss": 2.2189,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4222823865353057,
+      "grad_norm": 1.7513799667358398,
+      "learning_rate": 2.3361103131284114e-05,
+      "loss": 2.3345,
+      "step": 1485
+    },
+    {
+      "epoch": 0.42256675177876385,
+      "grad_norm": 1.6593313217163086,
+      "learning_rate": 2.334961218040793e-05,
+      "loss": 2.0602,
+      "step": 1486
+    },
+    {
+      "epoch": 0.42285111702222195,
+      "grad_norm": 1.7148724794387817,
+      "learning_rate": 2.3338121229531746e-05,
+      "loss": 1.6755,
+      "step": 1487
+    },
+    {
+      "epoch": 0.4231354822656801,
+      "grad_norm": 1.6917473077774048,
+      "learning_rate": 2.332663027865556e-05,
+      "loss": 1.8464,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4234198475091382,
+      "grad_norm": 2.1567392349243164,
+      "learning_rate": 2.3315139327779375e-05,
+      "loss": 2.9124,
+      "step": 1489
+    },
+    {
+      "epoch": 0.4237042127525963,
+      "grad_norm": 1.7503013610839844,
+      "learning_rate": 2.330364837690319e-05,
+      "loss": 2.5927,
+      "step": 1490
+    },
+    {
+      "epoch": 0.42398857799605444,
+      "grad_norm": 1.6987637281417847,
+      "learning_rate": 2.3292157426027004e-05,
+      "loss": 2.4439,
+      "step": 1491
+    },
+    {
+      "epoch": 0.42427294323951253,
+      "grad_norm": 1.6804696321487427,
+      "learning_rate": 2.328066647515082e-05,
+      "loss": 2.3747,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4245573084829707,
+      "grad_norm": 1.7630577087402344,
+      "learning_rate": 2.3269175524274633e-05,
+      "loss": 2.4236,
+      "step": 1493
+    },
+    {
+      "epoch": 0.4248416737264288,
+      "grad_norm": 1.5913448333740234,
+      "learning_rate": 2.3257684573398448e-05,
+      "loss": 2.0831,
+      "step": 1494
+    },
+    {
+      "epoch": 0.42512603896988693,
+      "grad_norm": 1.6789977550506592,
+      "learning_rate": 2.3246193622522266e-05,
+      "loss": 1.9894,
+      "step": 1495
+    },
+    {
+      "epoch": 0.42541040421334503,
+      "grad_norm": 1.732582688331604,
+      "learning_rate": 2.323470267164608e-05,
+      "loss": 1.7747,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4256947694568031,
+      "grad_norm": 1.926916241645813,
+      "learning_rate": 2.3223211720769895e-05,
+      "loss": 2.9434,
+      "step": 1497
+    },
+    {
+      "epoch": 0.4259791347002613,
+      "grad_norm": 1.6622836589813232,
+      "learning_rate": 2.321172076989371e-05,
+      "loss": 2.4302,
+      "step": 1498
+    },
+    {
+      "epoch": 0.42626349994371937,
+      "grad_norm": 1.64722740650177,
+      "learning_rate": 2.3200229819017524e-05,
+      "loss": 2.1876,
+      "step": 1499
+    },
+    {
+      "epoch": 0.4265478651871775,
+      "grad_norm": 1.7458999156951904,
+      "learning_rate": 2.318873886814134e-05,
+      "loss": 2.4106,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4268322304306356,
+      "grad_norm": 1.6324485540390015,
+      "learning_rate": 2.317724791726516e-05,
+      "loss": 2.3124,
+      "step": 1501
+    },
+    {
+      "epoch": 0.4271165956740937,
+      "grad_norm": 1.6437077522277832,
+      "learning_rate": 2.3165756966388974e-05,
+      "loss": 1.9083,
+      "step": 1502
+    },
+    {
+      "epoch": 0.42740096091755186,
+      "grad_norm": 1.6657003164291382,
+      "learning_rate": 2.3154266015512788e-05,
+      "loss": 1.7993,
+      "step": 1503
+    },
+    {
+      "epoch": 0.42768532616100996,
+      "grad_norm": 1.6173607110977173,
+      "learning_rate": 2.3142775064636603e-05,
+      "loss": 1.7787,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4279696914044681,
+      "grad_norm": 2.0600807666778564,
+      "learning_rate": 2.3131284113760417e-05,
+      "loss": 2.8575,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4282540566479262,
+      "grad_norm": 1.759519338607788,
+      "learning_rate": 2.3119793162884232e-05,
+      "loss": 2.4221,
+      "step": 1506
+    },
+    {
+      "epoch": 0.4285384218913843,
+      "grad_norm": 1.6098270416259766,
+      "learning_rate": 2.3108302212008046e-05,
+      "loss": 2.4158,
+      "step": 1507
+    },
+    {
+      "epoch": 0.42882278713484245,
+      "grad_norm": 1.5840624570846558,
+      "learning_rate": 2.309681126113186e-05,
+      "loss": 2.1282,
+      "step": 1508
+    },
+    {
+      "epoch": 0.42910715237830055,
+      "grad_norm": 1.7449572086334229,
+      "learning_rate": 2.3085320310255675e-05,
+      "loss": 2.289,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4293915176217587,
+      "grad_norm": 1.5223720073699951,
+      "learning_rate": 2.3073829359379493e-05,
+      "loss": 1.7867,
+      "step": 1510
+    },
+    {
+      "epoch": 0.4296758828652168,
+      "grad_norm": 1.6522331237792969,
+      "learning_rate": 2.3062338408503308e-05,
+      "loss": 2.1099,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4299602481086749,
+      "grad_norm": 1.652082085609436,
+      "learning_rate": 2.3050847457627122e-05,
+      "loss": 1.9209,
+      "step": 1512
+    },
+    {
+      "epoch": 0.43024461335213304,
+      "grad_norm": 1.941834568977356,
+      "learning_rate": 2.3039356506750937e-05,
+      "loss": 2.6694,
+      "step": 1513
+    },
+    {
+      "epoch": 0.43052897859559114,
+      "grad_norm": 1.7142715454101562,
+      "learning_rate": 2.302786555587475e-05,
+      "loss": 2.4722,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4308133438390493,
+      "grad_norm": 1.7237780094146729,
+      "learning_rate": 2.3016374604998566e-05,
+      "loss": 2.2415,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4310977090825074,
+      "grad_norm": 1.696518063545227,
+      "learning_rate": 2.300488365412238e-05,
+      "loss": 2.1774,
+      "step": 1516
+    },
+    {
+      "epoch": 0.43138207432596554,
+      "grad_norm": 1.6398987770080566,
+      "learning_rate": 2.2993392703246195e-05,
+      "loss": 2.2672,
+      "step": 1517
+    },
+    {
+      "epoch": 0.43166643956942363,
+      "grad_norm": 1.7152602672576904,
+      "learning_rate": 2.298190175237001e-05,
+      "loss": 2.1277,
+      "step": 1518
+    },
+    {
+      "epoch": 0.43195080481288173,
+      "grad_norm": 1.612593173980713,
+      "learning_rate": 2.2970410801493827e-05,
+      "loss": 2.1015,
+      "step": 1519
+    },
+    {
+      "epoch": 0.4322351700563399,
+      "grad_norm": 1.6528387069702148,
+      "learning_rate": 2.295891985061764e-05,
+      "loss": 1.7369,
+      "step": 1520
+    },
+    {
+      "epoch": 0.432519535299798,
+      "grad_norm": 2.0139312744140625,
+      "learning_rate": 2.2947428899741456e-05,
+      "loss": 2.8552,
+      "step": 1521
+    },
+    {
+      "epoch": 0.4328039005432561,
+      "grad_norm": 1.6151010990142822,
+      "learning_rate": 2.293593794886527e-05,
+      "loss": 2.1954,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4330882657867142,
+      "grad_norm": 1.5266865491867065,
+      "learning_rate": 2.2924446997989085e-05,
+      "loss": 2.2809,
+      "step": 1523
+    },
+    {
+      "epoch": 0.4333726310301723,
+      "grad_norm": 1.5820162296295166,
+      "learning_rate": 2.29129560471129e-05,
+      "loss": 2.2162,
+      "step": 1524
+    },
+    {
+      "epoch": 0.43365699627363047,
+      "grad_norm": 1.7275466918945312,
+      "learning_rate": 2.2901465096236714e-05,
+      "loss": 2.2792,
+      "step": 1525
+    },
+    {
+      "epoch": 0.43394136151708856,
+      "grad_norm": 1.6316412687301636,
+      "learning_rate": 2.288997414536053e-05,
+      "loss": 1.9186,
+      "step": 1526
+    },
+    {
+      "epoch": 0.4342257267605467,
+      "grad_norm": 1.7064950466156006,
+      "learning_rate": 2.2878483194484343e-05,
+      "loss": 1.7606,
+      "step": 1527
+    },
+    {
+      "epoch": 0.4345100920040048,
+      "grad_norm": 1.7663518190383911,
+      "learning_rate": 2.286699224360816e-05,
+      "loss": 1.7999,
+      "step": 1528
+    },
+    {
+      "epoch": 0.4347944572474629,
+      "grad_norm": 1.942549705505371,
+      "learning_rate": 2.2855501292731975e-05,
+      "loss": 3.1272,
+      "step": 1529
+    },
+    {
+      "epoch": 0.43507882249092106,
+      "grad_norm": 1.546832799911499,
+      "learning_rate": 2.284401034185579e-05,
+      "loss": 2.4118,
+      "step": 1530
+    },
+    {
+      "epoch": 0.43536318773437915,
+      "grad_norm": 1.5846697092056274,
+      "learning_rate": 2.2832519390979604e-05,
+      "loss": 2.3054,
+      "step": 1531
+    },
+    {
+      "epoch": 0.4356475529778373,
+      "grad_norm": 1.6705207824707031,
+      "learning_rate": 2.282102844010342e-05,
+      "loss": 2.465,
+      "step": 1532
+    },
+    {
+      "epoch": 0.4359319182212954,
+      "grad_norm": 1.7256457805633545,
+      "learning_rate": 2.2809537489227233e-05,
+      "loss": 2.2058,
+      "step": 1533
+    },
+    {
+      "epoch": 0.4362162834647535,
+      "grad_norm": 1.6743327379226685,
+      "learning_rate": 2.2798046538351048e-05,
+      "loss": 1.999,
+      "step": 1534
+    },
+    {
+      "epoch": 0.43650064870821165,
+      "grad_norm": 1.8469429016113281,
+      "learning_rate": 2.2786555587474862e-05,
+      "loss": 2.0113,
+      "step": 1535
+    },
+    {
+      "epoch": 0.43678501395166974,
+      "grad_norm": 1.6663111448287964,
+      "learning_rate": 2.2775064636598677e-05,
+      "loss": 1.9054,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4370693791951279,
+      "grad_norm": 1.980746865272522,
+      "learning_rate": 2.2763573685722498e-05,
+      "loss": 2.5987,
+      "step": 1537
+    },
+    {
+      "epoch": 0.437353744438586,
+      "grad_norm": 1.5249035358428955,
+      "learning_rate": 2.2752082734846313e-05,
+      "loss": 2.5415,
+      "step": 1538
+    },
+    {
+      "epoch": 0.4376381096820441,
+      "grad_norm": 1.5359803438186646,
+      "learning_rate": 2.2740591783970127e-05,
+      "loss": 2.2511,
+      "step": 1539
+    },
+    {
+      "epoch": 0.43792247492550224,
+      "grad_norm": 1.671136498451233,
+      "learning_rate": 2.272910083309394e-05,
+      "loss": 2.264,
+      "step": 1540
+    },
+    {
+      "epoch": 0.43820684016896033,
+      "grad_norm": 1.6809831857681274,
+      "learning_rate": 2.2717609882217756e-05,
+      "loss": 2.2934,
+      "step": 1541
+    },
+    {
+      "epoch": 0.4384912054124185,
+      "grad_norm": 1.785244107246399,
+      "learning_rate": 2.270611893134157e-05,
+      "loss": 1.8482,
+      "step": 1542
+    },
+    {
+      "epoch": 0.4387755706558766,
+      "grad_norm": 1.749173641204834,
+      "learning_rate": 2.269462798046539e-05,
+      "loss": 2.0663,
+      "step": 1543
+    },
+    {
+      "epoch": 0.43905993589933473,
+      "grad_norm": 1.7575947046279907,
+      "learning_rate": 2.2683137029589203e-05,
+      "loss": 1.9117,
+      "step": 1544
+    },
+    {
+      "epoch": 0.4393443011427928,
+      "grad_norm": 1.795770287513733,
+      "learning_rate": 2.2671646078713017e-05,
+      "loss": 2.9625,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4396286663862509,
+      "grad_norm": 1.6358546018600464,
+      "learning_rate": 2.2660155127836832e-05,
+      "loss": 2.4055,
+      "step": 1546
+    },
+    {
+      "epoch": 0.4399130316297091,
+      "grad_norm": 1.5694764852523804,
+      "learning_rate": 2.2648664176960646e-05,
+      "loss": 2.5108,
+      "step": 1547
+    },
+    {
+      "epoch": 0.44019739687316717,
+      "grad_norm": 1.6293818950653076,
+      "learning_rate": 2.263717322608446e-05,
+      "loss": 2.3389,
+      "step": 1548
+    },
+    {
+      "epoch": 0.4404817621166253,
+      "grad_norm": 1.683681845664978,
+      "learning_rate": 2.2625682275208275e-05,
+      "loss": 2.1871,
+      "step": 1549
+    },
+    {
+      "epoch": 0.4407661273600834,
+      "grad_norm": 1.668007254600525,
+      "learning_rate": 2.261419132433209e-05,
+      "loss": 1.8618,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4410504926035415,
+      "grad_norm": 1.5370334386825562,
+      "learning_rate": 2.2602700373455904e-05,
+      "loss": 1.6422,
+      "step": 1551
+    },
+    {
+      "epoch": 0.44133485784699966,
+      "grad_norm": 1.6079347133636475,
+      "learning_rate": 2.2591209422579722e-05,
+      "loss": 1.7268,
+      "step": 1552
+    },
+    {
+      "epoch": 0.44161922309045776,
+      "grad_norm": 1.7942856550216675,
+      "learning_rate": 2.2579718471703537e-05,
+      "loss": 2.8017,
+      "step": 1553
+    },
+    {
+      "epoch": 0.4419035883339159,
+      "grad_norm": 1.694395899772644,
+      "learning_rate": 2.256822752082735e-05,
+      "loss": 2.4659,
+      "step": 1554
+    },
+    {
+      "epoch": 0.442187953577374,
+      "grad_norm": 1.6033554077148438,
+      "learning_rate": 2.2556736569951166e-05,
+      "loss": 2.321,
+      "step": 1555
+    },
+    {
+      "epoch": 0.4424723188208321,
+      "grad_norm": 1.4896838665008545,
+      "learning_rate": 2.254524561907498e-05,
+      "loss": 2.3603,
+      "step": 1556
+    },
+    {
+      "epoch": 0.44275668406429025,
+      "grad_norm": 1.5675745010375977,
+      "learning_rate": 2.2533754668198795e-05,
+      "loss": 2.199,
+      "step": 1557
+    },
+    {
+      "epoch": 0.44304104930774835,
+      "grad_norm": 1.5907286405563354,
+      "learning_rate": 2.252226371732261e-05,
+      "loss": 2.0474,
+      "step": 1558
+    },
+    {
+      "epoch": 0.4433254145512065,
+      "grad_norm": 1.67164146900177,
+      "learning_rate": 2.2510772766446424e-05,
+      "loss": 1.9384,
+      "step": 1559
+    },
+    {
+      "epoch": 0.4436097797946646,
+      "grad_norm": 1.6073744297027588,
+      "learning_rate": 2.249928181557024e-05,
+      "loss": 1.5771,
+      "step": 1560
+    },
+    {
+      "epoch": 0.4438941450381227,
+      "grad_norm": 1.9481167793273926,
+      "learning_rate": 2.2487790864694056e-05,
+      "loss": 2.8547,
+      "step": 1561
+    },
+    {
+      "epoch": 0.44417851028158084,
+      "grad_norm": 1.627387523651123,
+      "learning_rate": 2.247629991381787e-05,
+      "loss": 2.661,
+      "step": 1562
+    },
+    {
+      "epoch": 0.44446287552503894,
+      "grad_norm": 1.665747046470642,
+      "learning_rate": 2.2464808962941685e-05,
+      "loss": 2.2633,
+      "step": 1563
+    },
+    {
+      "epoch": 0.4447472407684971,
+      "grad_norm": 1.6608874797821045,
+      "learning_rate": 2.24533180120655e-05,
+      "loss": 2.3801,
+      "step": 1564
+    },
+    {
+      "epoch": 0.4450316060119552,
+      "grad_norm": 1.778861165046692,
+      "learning_rate": 2.2441827061189314e-05,
+      "loss": 2.1286,
+      "step": 1565
+    },
+    {
+      "epoch": 0.44531597125541333,
+      "grad_norm": 1.678720474243164,
+      "learning_rate": 2.243033611031313e-05,
+      "loss": 2.0441,
+      "step": 1566
+    },
+    {
+      "epoch": 0.44560033649887143,
+      "grad_norm": 1.6838959455490112,
+      "learning_rate": 2.2418845159436943e-05,
+      "loss": 1.8895,
+      "step": 1567
+    },
+    {
+      "epoch": 0.4458847017423295,
+      "grad_norm": 1.7138351202011108,
+      "learning_rate": 2.2407354208560758e-05,
+      "loss": 1.8344,
+      "step": 1568
+    },
+    {
+      "epoch": 0.4461690669857877,
+      "grad_norm": 1.9342299699783325,
+      "learning_rate": 2.2395863257684572e-05,
+      "loss": 2.7738,
+      "step": 1569
+    },
+    {
+      "epoch": 0.4464534322292458,
+      "grad_norm": 1.6085205078125,
+      "learning_rate": 2.238437230680839e-05,
+      "loss": 2.4156,
+      "step": 1570
+    },
+    {
+      "epoch": 0.4467377974727039,
+      "grad_norm": 1.598381757736206,
+      "learning_rate": 2.2372881355932205e-05,
+      "loss": 2.323,
+      "step": 1571
+    },
+    {
+      "epoch": 0.447022162716162,
+      "grad_norm": 1.5285613536834717,
+      "learning_rate": 2.236139040505602e-05,
+      "loss": 2.3501,
+      "step": 1572
+    },
+    {
+      "epoch": 0.4473065279596201,
+      "grad_norm": 1.6888141632080078,
+      "learning_rate": 2.2349899454179834e-05,
+      "loss": 2.3968,
+      "step": 1573
+    },
+    {
+      "epoch": 0.44759089320307827,
+      "grad_norm": 1.6421995162963867,
+      "learning_rate": 2.233840850330365e-05,
+      "loss": 1.7312,
+      "step": 1574
+    },
+    {
+      "epoch": 0.44787525844653636,
+      "grad_norm": 1.7016944885253906,
+      "learning_rate": 2.2326917552427466e-05,
+      "loss": 1.8975,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4481596236899945,
+      "grad_norm": 1.8153733015060425,
+      "learning_rate": 2.231542660155128e-05,
+      "loss": 2.0029,
+      "step": 1576
+    },
+    {
+      "epoch": 0.4484439889334526,
+      "grad_norm": 1.941930890083313,
+      "learning_rate": 2.23039356506751e-05,
+      "loss": 3.027,
+      "step": 1577
+    },
+    {
+      "epoch": 0.4487283541769107,
+      "grad_norm": 1.643111228942871,
+      "learning_rate": 2.2292444699798913e-05,
+      "loss": 2.5447,
+      "step": 1578
+    },
+    {
+      "epoch": 0.44901271942036886,
+      "grad_norm": 1.5642459392547607,
+      "learning_rate": 2.2280953748922727e-05,
+      "loss": 2.3714,
+      "step": 1579
+    },
+    {
+      "epoch": 0.44929708466382695,
+      "grad_norm": 1.6498128175735474,
+      "learning_rate": 2.2269462798046542e-05,
+      "loss": 2.3655,
+      "step": 1580
+    },
+    {
+      "epoch": 0.4495814499072851,
+      "grad_norm": 1.7358462810516357,
+      "learning_rate": 2.2257971847170356e-05,
+      "loss": 2.2202,
+      "step": 1581
+    },
+    {
+      "epoch": 0.4498658151507432,
+      "grad_norm": 1.718059778213501,
+      "learning_rate": 2.224648089629417e-05,
+      "loss": 1.8404,
+      "step": 1582
+    },
+    {
+      "epoch": 0.4501501803942013,
+      "grad_norm": 1.6849524974822998,
+      "learning_rate": 2.2234989945417985e-05,
+      "loss": 1.857,
+      "step": 1583
+    },
+    {
+      "epoch": 0.45043454563765944,
+      "grad_norm": 1.6961820125579834,
+      "learning_rate": 2.22234989945418e-05,
+      "loss": 1.8101,
+      "step": 1584
+    },
+    {
+      "epoch": 0.45071891088111754,
+      "grad_norm": 1.9128514528274536,
+      "learning_rate": 2.2212008043665618e-05,
+      "loss": 2.9717,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4510032761245757,
+      "grad_norm": 1.4966126680374146,
+      "learning_rate": 2.2200517092789432e-05,
+      "loss": 2.6604,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4512876413680338,
+      "grad_norm": 1.5449092388153076,
+      "learning_rate": 2.2189026141913247e-05,
+      "loss": 2.1829,
+      "step": 1587
+    },
+    {
+      "epoch": 0.4515720066114919,
+      "grad_norm": 1.6908026933670044,
+      "learning_rate": 2.217753519103706e-05,
+      "loss": 2.2731,
+      "step": 1588
+    },
+    {
+      "epoch": 0.45185637185495003,
+      "grad_norm": 1.561723232269287,
+      "learning_rate": 2.2166044240160876e-05,
+      "loss": 2.1934,
+      "step": 1589
+    },
+    {
+      "epoch": 0.45214073709840813,
+      "grad_norm": 1.6748071908950806,
+      "learning_rate": 2.215455328928469e-05,
+      "loss": 2.112,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4524251023418663,
+      "grad_norm": 1.5931808948516846,
+      "learning_rate": 2.2143062338408505e-05,
+      "loss": 1.853,
+      "step": 1591
+    },
+    {
+      "epoch": 0.4527094675853244,
+      "grad_norm": 1.7409586906433105,
+      "learning_rate": 2.213157138753232e-05,
+      "loss": 1.7871,
+      "step": 1592
+    },
+    {
+      "epoch": 0.45299383282878253,
+      "grad_norm": 1.953231930732727,
+      "learning_rate": 2.2120080436656134e-05,
+      "loss": 2.8715,
+      "step": 1593
+    },
+    {
+      "epoch": 0.4532781980722406,
+      "grad_norm": 1.6238800287246704,
+      "learning_rate": 2.210858948577995e-05,
+      "loss": 2.5115,
+      "step": 1594
+    },
+    {
+      "epoch": 0.4535625633156987,
+      "grad_norm": 1.643713116645813,
+      "learning_rate": 2.2097098534903766e-05,
+      "loss": 2.2356,
+      "step": 1595
+    },
+    {
+      "epoch": 0.45384692855915687,
+      "grad_norm": 1.5235264301300049,
+      "learning_rate": 2.208560758402758e-05,
+      "loss": 2.348,
+      "step": 1596
+    },
+    {
+      "epoch": 0.45413129380261497,
+      "grad_norm": 1.6755728721618652,
+      "learning_rate": 2.2074116633151395e-05,
+      "loss": 2.2908,
+      "step": 1597
+    },
+    {
+      "epoch": 0.4544156590460731,
+      "grad_norm": 1.6341913938522339,
+      "learning_rate": 2.206262568227521e-05,
+      "loss": 1.9315,
+      "step": 1598
+    },
+    {
+      "epoch": 0.4547000242895312,
+      "grad_norm": 1.8000038862228394,
+      "learning_rate": 2.2051134731399024e-05,
+      "loss": 1.8148,
+      "step": 1599
+    },
+    {
+      "epoch": 0.4549843895329893,
+      "grad_norm": 1.7715755701065063,
+      "learning_rate": 2.203964378052284e-05,
+      "loss": 1.9843,
+      "step": 1600
+    },
+    {
+      "epoch": 0.45526875477644746,
+      "grad_norm": 1.9607915878295898,
+      "learning_rate": 2.2028152829646653e-05,
+      "loss": 3.0164,
+      "step": 1601
+    },
+    {
+      "epoch": 0.45555312001990556,
+      "grad_norm": 1.5718859434127808,
+      "learning_rate": 2.2016661878770468e-05,
+      "loss": 2.5762,
+      "step": 1602
+    },
+    {
+      "epoch": 0.4558374852633637,
+      "grad_norm": 1.6624130010604858,
+      "learning_rate": 2.2005170927894285e-05,
+      "loss": 2.3932,
+      "step": 1603
+    },
+    {
+      "epoch": 0.4561218505068218,
+      "grad_norm": 1.6082323789596558,
+      "learning_rate": 2.19936799770181e-05,
+      "loss": 2.4261,
+      "step": 1604
+    },
+    {
+      "epoch": 0.4564062157502799,
+      "grad_norm": 1.7497981786727905,
+      "learning_rate": 2.1982189026141914e-05,
+      "loss": 2.3493,
+      "step": 1605
+    },
+    {
+      "epoch": 0.45669058099373805,
+      "grad_norm": 1.559987187385559,
+      "learning_rate": 2.197069807526573e-05,
+      "loss": 1.8154,
+      "step": 1606
+    },
+    {
+      "epoch": 0.45697494623719614,
+      "grad_norm": 1.724812388420105,
+      "learning_rate": 2.1959207124389543e-05,
+      "loss": 1.6993,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4572593114806543,
+      "grad_norm": 1.6445400714874268,
+      "learning_rate": 2.1947716173513358e-05,
+      "loss": 1.7773,
+      "step": 1608
+    },
+    {
+      "epoch": 0.4575436767241124,
+      "grad_norm": 1.8340013027191162,
+      "learning_rate": 2.1936225222637172e-05,
+      "loss": 2.8774,
+      "step": 1609
+    },
+    {
+      "epoch": 0.4578280419675705,
+      "grad_norm": 1.6000295877456665,
+      "learning_rate": 2.1924734271760987e-05,
+      "loss": 2.6376,
+      "step": 1610
+    },
+    {
+      "epoch": 0.45811240721102864,
+      "grad_norm": 1.6122288703918457,
+      "learning_rate": 2.1913243320884808e-05,
+      "loss": 2.4566,
+      "step": 1611
+    },
+    {
+      "epoch": 0.45839677245448673,
+      "grad_norm": 1.62675940990448,
+      "learning_rate": 2.1901752370008623e-05,
+      "loss": 2.4088,
+      "step": 1612
+    },
+    {
+      "epoch": 0.4586811376979449,
+      "grad_norm": 1.7197331190109253,
+      "learning_rate": 2.1890261419132437e-05,
+      "loss": 2.2432,
+      "step": 1613
+    },
+    {
+      "epoch": 0.458965502941403,
+      "grad_norm": 1.5948972702026367,
+      "learning_rate": 2.187877046825625e-05,
+      "loss": 1.9617,
+      "step": 1614
+    },
+    {
+      "epoch": 0.45924986818486113,
+      "grad_norm": 1.689054012298584,
+      "learning_rate": 2.1867279517380066e-05,
+      "loss": 1.9137,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4595342334283192,
+      "grad_norm": 1.829690933227539,
+      "learning_rate": 2.185578856650388e-05,
+      "loss": 1.8653,
+      "step": 1616
+    },
+    {
+      "epoch": 0.4598185986717773,
+      "grad_norm": 1.973512887954712,
+      "learning_rate": 2.1844297615627695e-05,
+      "loss": 2.8129,
+      "step": 1617
+    },
+    {
+      "epoch": 0.4601029639152355,
+      "grad_norm": 1.612326979637146,
+      "learning_rate": 2.183280666475151e-05,
+      "loss": 2.6233,
+      "step": 1618
+    },
+    {
+      "epoch": 0.46038732915869357,
+      "grad_norm": 1.473063588142395,
+      "learning_rate": 2.1821315713875328e-05,
+      "loss": 2.3022,
+      "step": 1619
+    },
+    {
+      "epoch": 0.4606716944021517,
+      "grad_norm": 1.552682876586914,
+      "learning_rate": 2.1809824762999142e-05,
+      "loss": 2.1985,
+      "step": 1620
+    },
+    {
+      "epoch": 0.4609560596456098,
+      "grad_norm": 1.7268232107162476,
+      "learning_rate": 2.1798333812122957e-05,
+      "loss": 2.2878,
+      "step": 1621
+    },
+    {
+      "epoch": 0.4612404248890679,
+      "grad_norm": 1.5425605773925781,
+      "learning_rate": 2.178684286124677e-05,
+      "loss": 1.8743,
+      "step": 1622
+    },
+    {
+      "epoch": 0.46152479013252606,
+      "grad_norm": 1.7569302320480347,
+      "learning_rate": 2.1775351910370586e-05,
+      "loss": 1.7704,
+      "step": 1623
+    },
+    {
+      "epoch": 0.46180915537598416,
+      "grad_norm": 1.6393110752105713,
+      "learning_rate": 2.17638609594944e-05,
+      "loss": 1.8674,
+      "step": 1624
+    },
+    {
+      "epoch": 0.4620935206194423,
+      "grad_norm": 1.9522907733917236,
+      "learning_rate": 2.1752370008618215e-05,
+      "loss": 2.9953,
+      "step": 1625
+    },
+    {
+      "epoch": 0.4623778858629004,
+      "grad_norm": 1.7476742267608643,
+      "learning_rate": 2.174087905774203e-05,
+      "loss": 2.4966,
+      "step": 1626
+    },
+    {
+      "epoch": 0.4626622511063585,
+      "grad_norm": 1.642991065979004,
+      "learning_rate": 2.1729388106865844e-05,
+      "loss": 2.4291,
+      "step": 1627
+    },
+    {
+      "epoch": 0.46294661634981665,
+      "grad_norm": 1.6043994426727295,
+      "learning_rate": 2.171789715598966e-05,
+      "loss": 2.1552,
+      "step": 1628
+    },
+    {
+      "epoch": 0.46323098159327475,
+      "grad_norm": 1.703909158706665,
+      "learning_rate": 2.1706406205113476e-05,
+      "loss": 2.2067,
+      "step": 1629
+    },
+    {
+      "epoch": 0.4635153468367329,
+      "grad_norm": 1.5352064371109009,
+      "learning_rate": 2.169491525423729e-05,
+      "loss": 2.0239,
+      "step": 1630
+    },
+    {
+      "epoch": 0.463799712080191,
+      "grad_norm": 1.6908349990844727,
+      "learning_rate": 2.1683424303361105e-05,
+      "loss": 1.9427,
+      "step": 1631
+    },
+    {
+      "epoch": 0.4640840773236491,
+      "grad_norm": 1.7095677852630615,
+      "learning_rate": 2.167193335248492e-05,
+      "loss": 1.7988,
+      "step": 1632
+    },
+    {
+      "epoch": 0.46436844256710724,
+      "grad_norm": 2.046193838119507,
+      "learning_rate": 2.1660442401608734e-05,
+      "loss": 3.045,
+      "step": 1633
+    },
+    {
+      "epoch": 0.46465280781056534,
+      "grad_norm": 1.6458848714828491,
+      "learning_rate": 2.164895145073255e-05,
+      "loss": 2.6204,
+      "step": 1634
+    },
+    {
+      "epoch": 0.4649371730540235,
+      "grad_norm": 1.6441450119018555,
+      "learning_rate": 2.1637460499856363e-05,
+      "loss": 2.482,
+      "step": 1635
+    },
+    {
+      "epoch": 0.4652215382974816,
+      "grad_norm": 1.6394556760787964,
+      "learning_rate": 2.1625969548980177e-05,
+      "loss": 2.2559,
+      "step": 1636
+    },
+    {
+      "epoch": 0.46550590354093974,
+      "grad_norm": 1.8199498653411865,
+      "learning_rate": 2.1614478598103995e-05,
+      "loss": 2.2969,
+      "step": 1637
+    },
+    {
+      "epoch": 0.46579026878439783,
+      "grad_norm": 1.5875297784805298,
+      "learning_rate": 2.160298764722781e-05,
+      "loss": 1.9558,
+      "step": 1638
+    },
+    {
+      "epoch": 0.4660746340278559,
+      "grad_norm": 1.6678173542022705,
+      "learning_rate": 2.1591496696351624e-05,
+      "loss": 1.9319,
+      "step": 1639
+    },
+    {
+      "epoch": 0.4663589992713141,
+      "grad_norm": 1.7809865474700928,
+      "learning_rate": 2.158000574547544e-05,
+      "loss": 1.8821,
+      "step": 1640
+    },
+    {
+      "epoch": 0.4666433645147722,
+      "grad_norm": 2.026603937149048,
+      "learning_rate": 2.1568514794599253e-05,
+      "loss": 2.7958,
+      "step": 1641
+    },
+    {
+      "epoch": 0.4669277297582303,
+      "grad_norm": 1.7228740453720093,
+      "learning_rate": 2.1557023843723068e-05,
+      "loss": 2.591,
+      "step": 1642
+    },
+    {
+      "epoch": 0.4672120950016884,
+      "grad_norm": 1.6805092096328735,
+      "learning_rate": 2.1545532892846882e-05,
+      "loss": 2.4039,
+      "step": 1643
+    },
+    {
+      "epoch": 0.4674964602451465,
+      "grad_norm": 1.605502724647522,
+      "learning_rate": 2.1534041941970697e-05,
+      "loss": 2.2221,
+      "step": 1644
+    },
+    {
+      "epoch": 0.46778082548860467,
+      "grad_norm": 1.750071406364441,
+      "learning_rate": 2.152255099109451e-05,
+      "loss": 2.2253,
+      "step": 1645
+    },
+    {
+      "epoch": 0.46806519073206276,
+      "grad_norm": 1.907105565071106,
+      "learning_rate": 2.151106004021833e-05,
+      "loss": 1.9662,
+      "step": 1646
+    },
+    {
+      "epoch": 0.4683495559755209,
+      "grad_norm": 1.6502389907836914,
+      "learning_rate": 2.1499569089342144e-05,
+      "loss": 1.8562,
+      "step": 1647
+    },
+    {
+      "epoch": 0.468633921218979,
+      "grad_norm": 1.72910737991333,
+      "learning_rate": 2.148807813846596e-05,
+      "loss": 1.8048,
+      "step": 1648
+    },
+    {
+      "epoch": 0.4689182864624371,
+      "grad_norm": 1.7769256830215454,
+      "learning_rate": 2.1476587187589776e-05,
+      "loss": 2.9131,
+      "step": 1649
+    },
+    {
+      "epoch": 0.46920265170589526,
+      "grad_norm": 1.667056918144226,
+      "learning_rate": 2.146509623671359e-05,
+      "loss": 2.3479,
+      "step": 1650
+    },
+    {
+      "epoch": 0.46948701694935335,
+      "grad_norm": 1.6939767599105835,
+      "learning_rate": 2.1453605285837405e-05,
+      "loss": 2.4278,
+      "step": 1651
+    },
+    {
+      "epoch": 0.4697713821928115,
+      "grad_norm": 1.723417043685913,
+      "learning_rate": 2.1442114334961223e-05,
+      "loss": 2.2906,
+      "step": 1652
+    },
+    {
+      "epoch": 0.4700557474362696,
+      "grad_norm": 1.7457711696624756,
+      "learning_rate": 2.1430623384085037e-05,
+      "loss": 2.1299,
+      "step": 1653
+    },
+    {
+      "epoch": 0.4703401126797277,
+      "grad_norm": 1.643261432647705,
+      "learning_rate": 2.1419132433208852e-05,
+      "loss": 1.99,
+      "step": 1654
+    },
+    {
+      "epoch": 0.47062447792318585,
+      "grad_norm": 1.7526366710662842,
+      "learning_rate": 2.1407641482332666e-05,
+      "loss": 1.8781,
+      "step": 1655
+    },
+    {
+      "epoch": 0.47090884316664394,
+      "grad_norm": 1.6480202674865723,
+      "learning_rate": 2.139615053145648e-05,
+      "loss": 1.7935,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4711932084101021,
+      "grad_norm": 1.997900128364563,
+      "learning_rate": 2.1384659580580295e-05,
+      "loss": 3.0299,
+      "step": 1657
+    },
+    {
+      "epoch": 0.4714775736535602,
+      "grad_norm": 1.7013146877288818,
+      "learning_rate": 2.137316862970411e-05,
+      "loss": 2.548,
+      "step": 1658
+    },
+    {
+      "epoch": 0.4717619388970183,
+      "grad_norm": 1.5773371458053589,
+      "learning_rate": 2.1361677678827924e-05,
+      "loss": 2.2042,
+      "step": 1659
+    },
+    {
+      "epoch": 0.47204630414047644,
+      "grad_norm": 1.5602200031280518,
+      "learning_rate": 2.135018672795174e-05,
+      "loss": 2.1936,
+      "step": 1660
+    },
+    {
+      "epoch": 0.47233066938393453,
+      "grad_norm": 1.7694247961044312,
+      "learning_rate": 2.1338695777075557e-05,
+      "loss": 2.2457,
+      "step": 1661
+    },
+    {
+      "epoch": 0.4726150346273927,
+      "grad_norm": 1.5969568490982056,
+      "learning_rate": 2.132720482619937e-05,
+      "loss": 2.0625,
+      "step": 1662
+    },
+    {
+      "epoch": 0.4728993998708508,
+      "grad_norm": 1.6549208164215088,
+      "learning_rate": 2.1315713875323186e-05,
+      "loss": 1.9321,
+      "step": 1663
+    },
+    {
+      "epoch": 0.47318376511430893,
+      "grad_norm": 1.806251049041748,
+      "learning_rate": 2.1304222924447e-05,
+      "loss": 2.003,
+      "step": 1664
+    },
+    {
+      "epoch": 0.473468130357767,
+      "grad_norm": 2.0554563999176025,
+      "learning_rate": 2.1292731973570815e-05,
+      "loss": 2.967,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4737524956012251,
+      "grad_norm": 1.6516426801681519,
+      "learning_rate": 2.128124102269463e-05,
+      "loss": 2.5175,
+      "step": 1666
+    },
+    {
+      "epoch": 0.47403686084468327,
+      "grad_norm": 1.5493863821029663,
+      "learning_rate": 2.1269750071818444e-05,
+      "loss": 2.385,
+      "step": 1667
+    },
+    {
+      "epoch": 0.47432122608814137,
+      "grad_norm": 1.759292483329773,
+      "learning_rate": 2.1258259120942258e-05,
+      "loss": 2.3617,
+      "step": 1668
+    },
+    {
+      "epoch": 0.4746055913315995,
+      "grad_norm": 1.6405413150787354,
+      "learning_rate": 2.1246768170066073e-05,
+      "loss": 2.2015,
+      "step": 1669
+    },
+    {
+      "epoch": 0.4748899565750576,
+      "grad_norm": 1.686164140701294,
+      "learning_rate": 2.123527721918989e-05,
+      "loss": 2.0404,
+      "step": 1670
+    },
+    {
+      "epoch": 0.4751743218185157,
+      "grad_norm": 1.6657214164733887,
+      "learning_rate": 2.1223786268313705e-05,
+      "loss": 1.705,
+      "step": 1671
+    },
+    {
+      "epoch": 0.47545868706197386,
+      "grad_norm": 1.5510362386703491,
+      "learning_rate": 2.121229531743752e-05,
+      "loss": 1.8379,
+      "step": 1672
+    },
+    {
+      "epoch": 0.47574305230543196,
+      "grad_norm": 1.9575543403625488,
+      "learning_rate": 2.1200804366561334e-05,
+      "loss": 2.9591,
+      "step": 1673
+    },
+    {
+      "epoch": 0.4760274175488901,
+      "grad_norm": 1.660929799079895,
+      "learning_rate": 2.118931341568515e-05,
+      "loss": 2.4613,
+      "step": 1674
+    },
+    {
+      "epoch": 0.4763117827923482,
+      "grad_norm": 1.6097886562347412,
+      "learning_rate": 2.1177822464808963e-05,
+      "loss": 2.3457,
+      "step": 1675
+    },
+    {
+      "epoch": 0.4765961480358063,
+      "grad_norm": 1.6404685974121094,
+      "learning_rate": 2.1166331513932778e-05,
+      "loss": 2.2377,
+      "step": 1676
+    },
+    {
+      "epoch": 0.47688051327926445,
+      "grad_norm": 1.6690386533737183,
+      "learning_rate": 2.1154840563056592e-05,
+      "loss": 2.088,
+      "step": 1677
+    },
+    {
+      "epoch": 0.47716487852272255,
+      "grad_norm": 1.7172131538391113,
+      "learning_rate": 2.1143349612180407e-05,
+      "loss": 2.1428,
+      "step": 1678
+    },
+    {
+      "epoch": 0.4774492437661807,
+      "grad_norm": 1.645805835723877,
+      "learning_rate": 2.1131858661304225e-05,
+      "loss": 1.7262,
+      "step": 1679
+    },
+    {
+      "epoch": 0.4777336090096388,
+      "grad_norm": 1.6956250667572021,
+      "learning_rate": 2.112036771042804e-05,
+      "loss": 1.7153,
+      "step": 1680
+    },
+    {
+      "epoch": 0.4780179742530969,
+      "grad_norm": 1.8658703565597534,
+      "learning_rate": 2.1108876759551854e-05,
+      "loss": 2.9007,
+      "step": 1681
+    },
+    {
+      "epoch": 0.47830233949655504,
+      "grad_norm": 1.5524179935455322,
+      "learning_rate": 2.1097385808675668e-05,
+      "loss": 2.5446,
+      "step": 1682
+    },
+    {
+      "epoch": 0.47858670474001314,
+      "grad_norm": 1.4524405002593994,
+      "learning_rate": 2.1085894857799483e-05,
+      "loss": 2.2187,
+      "step": 1683
+    },
+    {
+      "epoch": 0.4788710699834713,
+      "grad_norm": 1.5633457899093628,
+      "learning_rate": 2.1074403906923297e-05,
+      "loss": 2.3211,
+      "step": 1684
+    },
+    {
+      "epoch": 0.4791554352269294,
+      "grad_norm": 1.6988346576690674,
+      "learning_rate": 2.1062912956047118e-05,
+      "loss": 2.3174,
+      "step": 1685
+    },
+    {
+      "epoch": 0.47943980047038753,
+      "grad_norm": 1.747249722480774,
+      "learning_rate": 2.1051422005170933e-05,
+      "loss": 1.9955,
+      "step": 1686
+    },
+    {
+      "epoch": 0.47972416571384563,
+      "grad_norm": 1.5953621864318848,
+      "learning_rate": 2.1039931054294747e-05,
+      "loss": 1.939,
+      "step": 1687
+    },
+    {
+      "epoch": 0.4800085309573037,
+      "grad_norm": 1.6366169452667236,
+      "learning_rate": 2.1028440103418562e-05,
+      "loss": 1.7721,
+      "step": 1688
+    },
+    {
+      "epoch": 0.4802928962007619,
+      "grad_norm": 1.9023154973983765,
+      "learning_rate": 2.1016949152542376e-05,
+      "loss": 2.8568,
+      "step": 1689
+    },
+    {
+      "epoch": 0.48057726144421997,
+      "grad_norm": 1.4789276123046875,
+      "learning_rate": 2.100545820166619e-05,
+      "loss": 2.4112,
+      "step": 1690
+    },
+    {
+      "epoch": 0.4808616266876781,
+      "grad_norm": 1.5606416463851929,
+      "learning_rate": 2.0993967250790005e-05,
+      "loss": 2.2638,
+      "step": 1691
+    },
+    {
+      "epoch": 0.4811459919311362,
+      "grad_norm": 1.5675312280654907,
+      "learning_rate": 2.098247629991382e-05,
+      "loss": 2.3888,
+      "step": 1692
+    },
+    {
+      "epoch": 0.4814303571745943,
+      "grad_norm": 1.6891237497329712,
+      "learning_rate": 2.0970985349037634e-05,
+      "loss": 2.1829,
+      "step": 1693
+    },
+    {
+      "epoch": 0.48171472241805247,
+      "grad_norm": 1.6167731285095215,
+      "learning_rate": 2.0959494398161452e-05,
+      "loss": 1.9001,
+      "step": 1694
+    },
+    {
+      "epoch": 0.48199908766151056,
+      "grad_norm": 1.6935869455337524,
+      "learning_rate": 2.0948003447285267e-05,
+      "loss": 1.8639,
+      "step": 1695
+    },
+    {
+      "epoch": 0.4822834529049687,
+      "grad_norm": 1.6257230043411255,
+      "learning_rate": 2.093651249640908e-05,
+      "loss": 1.78,
+      "step": 1696
+    },
+    {
+      "epoch": 0.4825678181484268,
+      "grad_norm": 2.0507266521453857,
+      "learning_rate": 2.0925021545532896e-05,
+      "loss": 2.8274,
+      "step": 1697
+    },
+    {
+      "epoch": 0.4828521833918849,
+      "grad_norm": 1.5421440601348877,
+      "learning_rate": 2.091353059465671e-05,
+      "loss": 2.5799,
+      "step": 1698
+    },
+    {
+      "epoch": 0.48313654863534305,
+      "grad_norm": 1.5370677709579468,
+      "learning_rate": 2.0902039643780525e-05,
+      "loss": 2.2449,
+      "step": 1699
+    },
+    {
+      "epoch": 0.48342091387880115,
+      "grad_norm": 1.6264358758926392,
+      "learning_rate": 2.089054869290434e-05,
+      "loss": 2.3207,
+      "step": 1700
+    },
+    {
+      "epoch": 0.4837052791222593,
+      "grad_norm": 1.8033897876739502,
+      "learning_rate": 2.0879057742028154e-05,
+      "loss": 2.2928,
+      "step": 1701
+    },
+    {
+      "epoch": 0.4839896443657174,
+      "grad_norm": 1.610359787940979,
+      "learning_rate": 2.0867566791151968e-05,
+      "loss": 1.896,
+      "step": 1702
+    },
+    {
+      "epoch": 0.4842740096091755,
+      "grad_norm": 1.8859881162643433,
+      "learning_rate": 2.0856075840275786e-05,
+      "loss": 1.648,
+      "step": 1703
+    },
+    {
+      "epoch": 0.48455837485263364,
+      "grad_norm": 1.7698290348052979,
+      "learning_rate": 2.08445848893996e-05,
+      "loss": 1.6407,
+      "step": 1704
+    },
+    {
+      "epoch": 0.48484274009609174,
+      "grad_norm": 1.8778108358383179,
+      "learning_rate": 2.0833093938523415e-05,
+      "loss": 2.8316,
+      "step": 1705
+    },
+    {
+      "epoch": 0.4851271053395499,
+      "grad_norm": 1.6026455163955688,
+      "learning_rate": 2.082160298764723e-05,
+      "loss": 2.4976,
+      "step": 1706
+    },
+    {
+      "epoch": 0.485411470583008,
+      "grad_norm": 1.5125036239624023,
+      "learning_rate": 2.0810112036771044e-05,
+      "loss": 2.2848,
+      "step": 1707
+    },
+    {
+      "epoch": 0.4856958358264661,
+      "grad_norm": 1.6800127029418945,
+      "learning_rate": 2.079862108589486e-05,
+      "loss": 2.1576,
+      "step": 1708
+    },
+    {
+      "epoch": 0.48598020106992423,
+      "grad_norm": 1.6857738494873047,
+      "learning_rate": 2.0787130135018673e-05,
+      "loss": 2.1469,
+      "step": 1709
+    },
+    {
+      "epoch": 0.48626456631338233,
+      "grad_norm": 1.5208419561386108,
+      "learning_rate": 2.0775639184142487e-05,
+      "loss": 1.9286,
+      "step": 1710
+    },
+    {
+      "epoch": 0.4865489315568405,
+      "grad_norm": 1.5637043714523315,
+      "learning_rate": 2.0764148233266302e-05,
+      "loss": 1.7151,
+      "step": 1711
+    },
+    {
+      "epoch": 0.4868332968002986,
+      "grad_norm": 1.6756120920181274,
+      "learning_rate": 2.075265728239012e-05,
+      "loss": 1.8036,
+      "step": 1712
+    },
+    {
+      "epoch": 0.4871176620437567,
+      "grad_norm": 1.8630014657974243,
+      "learning_rate": 2.0741166331513934e-05,
+      "loss": 2.9317,
+      "step": 1713
+    },
+    {
+      "epoch": 0.4874020272872148,
+      "grad_norm": 1.732492208480835,
+      "learning_rate": 2.072967538063775e-05,
+      "loss": 2.4387,
+      "step": 1714
+    },
+    {
+      "epoch": 0.4876863925306729,
+      "grad_norm": 1.6385549306869507,
+      "learning_rate": 2.0718184429761563e-05,
+      "loss": 2.1674,
+      "step": 1715
+    },
+    {
+      "epoch": 0.48797075777413107,
+      "grad_norm": 1.5367069244384766,
+      "learning_rate": 2.0706693478885378e-05,
+      "loss": 2.2549,
+      "step": 1716
+    },
+    {
+      "epoch": 0.48825512301758917,
+      "grad_norm": 1.765271782875061,
+      "learning_rate": 2.0695202528009192e-05,
+      "loss": 2.2325,
+      "step": 1717
+    },
+    {
+      "epoch": 0.4885394882610473,
+      "grad_norm": 1.6616159677505493,
+      "learning_rate": 2.0683711577133007e-05,
+      "loss": 1.806,
+      "step": 1718
+    },
+    {
+      "epoch": 0.4888238535045054,
+      "grad_norm": 1.7250800132751465,
+      "learning_rate": 2.067222062625682e-05,
+      "loss": 1.8495,
+      "step": 1719
+    },
+    {
+      "epoch": 0.4891082187479635,
+      "grad_norm": 1.7794257402420044,
+      "learning_rate": 2.0660729675380636e-05,
+      "loss": 1.7158,
+      "step": 1720
+    },
+    {
+      "epoch": 0.48939258399142166,
+      "grad_norm": 1.9646224975585938,
+      "learning_rate": 2.0649238724504454e-05,
+      "loss": 2.9298,
+      "step": 1721
+    },
+    {
+      "epoch": 0.48967694923487975,
+      "grad_norm": 1.6093361377716064,
+      "learning_rate": 2.063774777362827e-05,
+      "loss": 2.4871,
+      "step": 1722
+    },
+    {
+      "epoch": 0.4899613144783379,
+      "grad_norm": 1.5131632089614868,
+      "learning_rate": 2.0626256822752086e-05,
+      "loss": 2.2751,
+      "step": 1723
+    },
+    {
+      "epoch": 0.490245679721796,
+      "grad_norm": 1.5674437284469604,
+      "learning_rate": 2.06147658718759e-05,
+      "loss": 2.2159,
+      "step": 1724
+    },
+    {
+      "epoch": 0.4905300449652541,
+      "grad_norm": 1.5765401124954224,
+      "learning_rate": 2.0603274920999715e-05,
+      "loss": 2.2252,
+      "step": 1725
+    },
+    {
+      "epoch": 0.49081441020871225,
+      "grad_norm": 1.6332764625549316,
+      "learning_rate": 2.059178397012353e-05,
+      "loss": 1.9848,
+      "step": 1726
+    },
+    {
+      "epoch": 0.49109877545217034,
+      "grad_norm": 1.555993914604187,
+      "learning_rate": 2.0580293019247347e-05,
+      "loss": 1.8058,
+      "step": 1727
+    },
+    {
+      "epoch": 0.4913831406956285,
+      "grad_norm": 1.792938470840454,
+      "learning_rate": 2.0568802068371162e-05,
+      "loss": 1.7801,
+      "step": 1728
+    },
+    {
+      "epoch": 0.4916675059390866,
+      "grad_norm": 1.881766676902771,
+      "learning_rate": 2.0557311117494976e-05,
+      "loss": 2.734,
+      "step": 1729
+    },
+    {
+      "epoch": 0.4919518711825447,
+      "grad_norm": 1.5868914127349854,
+      "learning_rate": 2.054582016661879e-05,
+      "loss": 2.5361,
+      "step": 1730
+    },
+    {
+      "epoch": 0.49223623642600284,
+      "grad_norm": 1.59299898147583,
+      "learning_rate": 2.0534329215742605e-05,
+      "loss": 2.3963,
+      "step": 1731
+    },
+    {
+      "epoch": 0.49252060166946093,
+      "grad_norm": 1.7048550844192505,
+      "learning_rate": 2.052283826486642e-05,
+      "loss": 2.1933,
+      "step": 1732
+    },
+    {
+      "epoch": 0.4928049669129191,
+      "grad_norm": 1.5875484943389893,
+      "learning_rate": 2.0511347313990234e-05,
+      "loss": 2.2065,
+      "step": 1733
+    },
+    {
+      "epoch": 0.4930893321563772,
+      "grad_norm": 1.7140761613845825,
+      "learning_rate": 2.049985636311405e-05,
+      "loss": 1.912,
+      "step": 1734
+    },
+    {
+      "epoch": 0.49337369739983533,
+      "grad_norm": 1.6499665975570679,
+      "learning_rate": 2.0488365412237863e-05,
+      "loss": 1.9668,
+      "step": 1735
+    },
+    {
+      "epoch": 0.4936580626432934,
+      "grad_norm": 1.7199233770370483,
+      "learning_rate": 2.047687446136168e-05,
+      "loss": 1.7098,
+      "step": 1736
+    },
+    {
+      "epoch": 0.4939424278867515,
+      "grad_norm": 1.7340093851089478,
+      "learning_rate": 2.0465383510485496e-05,
+      "loss": 2.9317,
+      "step": 1737
+    },
+    {
+      "epoch": 0.4942267931302097,
+      "grad_norm": 1.4972059726715088,
+      "learning_rate": 2.045389255960931e-05,
+      "loss": 2.4092,
+      "step": 1738
+    },
+    {
+      "epoch": 0.49451115837366777,
+      "grad_norm": 1.5227432250976562,
+      "learning_rate": 2.0442401608733125e-05,
+      "loss": 2.3194,
+      "step": 1739
+    },
+    {
+      "epoch": 0.4947955236171259,
+      "grad_norm": 1.5438145399093628,
+      "learning_rate": 2.043091065785694e-05,
+      "loss": 2.4278,
+      "step": 1740
+    },
+    {
+      "epoch": 0.495079888860584,
+      "grad_norm": 1.767443060874939,
+      "learning_rate": 2.0419419706980754e-05,
+      "loss": 2.2614,
+      "step": 1741
+    },
+    {
+      "epoch": 0.4953642541040421,
+      "grad_norm": 1.471556544303894,
+      "learning_rate": 2.040792875610457e-05,
+      "loss": 1.8055,
+      "step": 1742
+    },
+    {
+      "epoch": 0.49564861934750026,
+      "grad_norm": 1.977705478668213,
+      "learning_rate": 2.0396437805228383e-05,
+      "loss": 1.8515,
+      "step": 1743
+    },
+    {
+      "epoch": 0.49593298459095836,
+      "grad_norm": 1.7835551500320435,
+      "learning_rate": 2.0384946854352197e-05,
+      "loss": 1.8965,
+      "step": 1744
+    },
+    {
+      "epoch": 0.4962173498344165,
+      "grad_norm": 1.7893133163452148,
+      "learning_rate": 2.0373455903476015e-05,
+      "loss": 2.6412,
+      "step": 1745
+    },
+    {
+      "epoch": 0.4965017150778746,
+      "grad_norm": 1.5060399770736694,
+      "learning_rate": 2.036196495259983e-05,
+      "loss": 2.176,
+      "step": 1746
+    },
+    {
+      "epoch": 0.4967860803213327,
+      "grad_norm": 1.6172282695770264,
+      "learning_rate": 2.0350474001723644e-05,
+      "loss": 2.2377,
+      "step": 1747
+    },
+    {
+      "epoch": 0.49707044556479085,
+      "grad_norm": 1.5684025287628174,
+      "learning_rate": 2.033898305084746e-05,
+      "loss": 2.0984,
+      "step": 1748
+    },
+    {
+      "epoch": 0.49735481080824895,
+      "grad_norm": 1.6685534715652466,
+      "learning_rate": 2.0327492099971273e-05,
+      "loss": 2.2496,
+      "step": 1749
+    },
+    {
+      "epoch": 0.4976391760517071,
+      "grad_norm": 1.6324350833892822,
+      "learning_rate": 2.0316001149095088e-05,
+      "loss": 1.854,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4979235412951652,
+      "grad_norm": 1.642563819885254,
+      "learning_rate": 2.0304510198218902e-05,
+      "loss": 1.8248,
+      "step": 1751
+    },
+    {
+      "epoch": 0.4982079065386233,
+      "grad_norm": 1.6866960525512695,
+      "learning_rate": 2.0293019247342717e-05,
+      "loss": 2.0292,
+      "step": 1752
+    },
+    {
+      "epoch": 0.49849227178208144,
+      "grad_norm": 1.9261857271194458,
+      "learning_rate": 2.028152829646653e-05,
+      "loss": 2.8267,
+      "step": 1753
+    },
+    {
+      "epoch": 0.49877663702553954,
+      "grad_norm": 1.5823674201965332,
+      "learning_rate": 2.027003734559035e-05,
+      "loss": 2.3225,
+      "step": 1754
+    },
+    {
+      "epoch": 0.4990610022689977,
+      "grad_norm": 1.5094738006591797,
+      "learning_rate": 2.0258546394714164e-05,
+      "loss": 2.2377,
+      "step": 1755
+    },
+    {
+      "epoch": 0.4993453675124558,
+      "grad_norm": 1.5388329029083252,
+      "learning_rate": 2.0247055443837978e-05,
+      "loss": 2.4092,
+      "step": 1756
+    },
+    {
+      "epoch": 0.49962973275591394,
+      "grad_norm": 1.5663154125213623,
+      "learning_rate": 2.0235564492961793e-05,
+      "loss": 2.1407,
+      "step": 1757
+    },
+    {
+      "epoch": 0.49991409799937203,
+      "grad_norm": 1.633455753326416,
+      "learning_rate": 2.0224073542085607e-05,
+      "loss": 2.0007,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5001984632428301,
+      "grad_norm": 1.7223169803619385,
+      "learning_rate": 2.0212582591209425e-05,
+      "loss": 2.0154,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5004828284862882,
+      "grad_norm": 1.8982787132263184,
+      "learning_rate": 2.0201091640333243e-05,
+      "loss": 1.9042,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5007671937297464,
+      "grad_norm": 1.855883240699768,
+      "learning_rate": 2.0189600689457057e-05,
+      "loss": 2.6409,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5010515589732045,
+      "grad_norm": 1.5711190700531006,
+      "learning_rate": 2.0178109738580872e-05,
+      "loss": 2.5989,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5013359242166626,
+      "grad_norm": 1.565077781677246,
+      "learning_rate": 2.0166618787704686e-05,
+      "loss": 2.5103,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5016202894601207,
+      "grad_norm": 1.5790541172027588,
+      "learning_rate": 2.01551278368285e-05,
+      "loss": 2.2145,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5019046547035788,
+      "grad_norm": 1.712709665298462,
+      "learning_rate": 2.0143636885952315e-05,
+      "loss": 2.2995,
+      "step": 1765
+    },
+    {
+      "epoch": 0.502189019947037,
+      "grad_norm": 1.6883087158203125,
+      "learning_rate": 2.013214593507613e-05,
+      "loss": 1.8596,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5024733851904951,
+      "grad_norm": 1.6610372066497803,
+      "learning_rate": 2.0120654984199944e-05,
+      "loss": 1.9617,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5027577504339532,
+      "grad_norm": 1.5222752094268799,
+      "learning_rate": 2.010916403332376e-05,
+      "loss": 1.8691,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5030421156774113,
+      "grad_norm": 1.8320504426956177,
+      "learning_rate": 2.0097673082447577e-05,
+      "loss": 2.722,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5033264809208694,
+      "grad_norm": 1.5912829637527466,
+      "learning_rate": 2.008618213157139e-05,
+      "loss": 2.4662,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5036108461643276,
+      "grad_norm": 1.522308588027954,
+      "learning_rate": 2.0074691180695206e-05,
+      "loss": 2.3028,
+      "step": 1771
+    },
+    {
+      "epoch": 0.5038952114077857,
+      "grad_norm": 1.5681263208389282,
+      "learning_rate": 2.006320022981902e-05,
+      "loss": 2.2298,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5041795766512438,
+      "grad_norm": 1.6516104936599731,
+      "learning_rate": 2.0051709278942835e-05,
+      "loss": 2.2985,
+      "step": 1773
+    },
+    {
+      "epoch": 0.5044639418947019,
+      "grad_norm": 1.5645509958267212,
+      "learning_rate": 2.004021832806665e-05,
+      "loss": 1.942,
+      "step": 1774
+    },
+    {
+      "epoch": 0.50474830713816,
+      "grad_norm": 1.656414270401001,
+      "learning_rate": 2.0028727377190464e-05,
+      "loss": 1.7982,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5050326723816182,
+      "grad_norm": 1.6322808265686035,
+      "learning_rate": 2.0017236426314278e-05,
+      "loss": 1.9504,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5053170376250763,
+      "grad_norm": 1.7643145322799683,
+      "learning_rate": 2.0005745475438093e-05,
+      "loss": 2.765,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5056014028685344,
+      "grad_norm": 1.471491813659668,
+      "learning_rate": 1.999425452456191e-05,
+      "loss": 2.2441,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5058857681119925,
+      "grad_norm": 1.5488718748092651,
+      "learning_rate": 1.9982763573685725e-05,
+      "loss": 2.3885,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5061701333554506,
+      "grad_norm": 1.5598701238632202,
+      "learning_rate": 1.997127262280954e-05,
+      "loss": 2.3161,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5064544985989088,
+      "grad_norm": 1.5675969123840332,
+      "learning_rate": 1.9959781671933354e-05,
+      "loss": 2.3845,
+      "step": 1781
+    },
+    {
+      "epoch": 0.5067388638423669,
+      "grad_norm": 1.645579218864441,
+      "learning_rate": 1.994829072105717e-05,
+      "loss": 1.949,
+      "step": 1782
+    },
+    {
+      "epoch": 0.507023229085825,
+      "grad_norm": 1.514928936958313,
+      "learning_rate": 1.9936799770180983e-05,
+      "loss": 1.8749,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5073075943292831,
+      "grad_norm": 1.7080934047698975,
+      "learning_rate": 1.9925308819304798e-05,
+      "loss": 1.6213,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5075919595727412,
+      "grad_norm": 2.008892297744751,
+      "learning_rate": 1.9913817868428612e-05,
+      "loss": 2.8028,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5078763248161994,
+      "grad_norm": 1.6706644296646118,
+      "learning_rate": 1.9902326917552427e-05,
+      "loss": 2.1027,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5081606900596575,
+      "grad_norm": 1.5543371438980103,
+      "learning_rate": 1.9890835966676244e-05,
+      "loss": 2.4243,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5084450553031156,
+      "grad_norm": 1.555815577507019,
+      "learning_rate": 1.987934501580006e-05,
+      "loss": 2.3077,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5087294205465737,
+      "grad_norm": 1.6928585767745972,
+      "learning_rate": 1.9867854064923873e-05,
+      "loss": 2.0664,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5090137857900318,
+      "grad_norm": 1.5728840827941895,
+      "learning_rate": 1.985636311404769e-05,
+      "loss": 2.0858,
+      "step": 1790
+    },
+    {
+      "epoch": 0.50929815103349,
+      "grad_norm": 1.7079205513000488,
+      "learning_rate": 1.9844872163171506e-05,
+      "loss": 1.8084,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5095825162769481,
+      "grad_norm": 1.6424134969711304,
+      "learning_rate": 1.983338121229532e-05,
+      "loss": 1.7926,
+      "step": 1792
+    },
+    {
+      "epoch": 0.5098668815204062,
+      "grad_norm": 1.9183549880981445,
+      "learning_rate": 1.9821890261419135e-05,
+      "loss": 2.7518,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5101512467638643,
+      "grad_norm": 1.5652717351913452,
+      "learning_rate": 1.981039931054295e-05,
+      "loss": 2.2392,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5104356120073225,
+      "grad_norm": 1.5693566799163818,
+      "learning_rate": 1.9798908359666764e-05,
+      "loss": 2.5368,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5107199772507806,
+      "grad_norm": 1.57747220993042,
+      "learning_rate": 1.978741740879058e-05,
+      "loss": 2.2448,
+      "step": 1796
+    },
+    {
+      "epoch": 0.5110043424942387,
+      "grad_norm": 1.6220982074737549,
+      "learning_rate": 1.9775926457914393e-05,
+      "loss": 1.9995,
+      "step": 1797
+    },
+    {
+      "epoch": 0.5112887077376967,
+      "grad_norm": 1.5088688135147095,
+      "learning_rate": 1.9764435507038207e-05,
+      "loss": 1.9295,
+      "step": 1798
+    },
+    {
+      "epoch": 0.5115730729811548,
+      "grad_norm": 1.5756409168243408,
+      "learning_rate": 1.9752944556162025e-05,
+      "loss": 1.8758,
+      "step": 1799
+    },
+    {
+      "epoch": 0.511857438224613,
+      "grad_norm": 1.6049522161483765,
+      "learning_rate": 1.974145360528584e-05,
+      "loss": 1.6545,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5121418034680711,
+      "grad_norm": 1.9730896949768066,
+      "learning_rate": 1.9729962654409654e-05,
+      "loss": 2.7769,
+      "step": 1801
+    },
+    {
+      "epoch": 0.5124261687115292,
+      "grad_norm": 1.4819504022598267,
+      "learning_rate": 1.971847170353347e-05,
+      "loss": 2.2833,
+      "step": 1802
+    },
+    {
+      "epoch": 0.5127105339549873,
+      "grad_norm": 1.6456480026245117,
+      "learning_rate": 1.9706980752657283e-05,
+      "loss": 2.3185,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5129948991984454,
+      "grad_norm": 1.5290210247039795,
+      "learning_rate": 1.9695489801781098e-05,
+      "loss": 2.4907,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5132792644419036,
+      "grad_norm": 1.5902469158172607,
+      "learning_rate": 1.9683998850904912e-05,
+      "loss": 2.2612,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5135636296853617,
+      "grad_norm": 1.622529149055481,
+      "learning_rate": 1.967250790002873e-05,
+      "loss": 1.9592,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5138479949288198,
+      "grad_norm": 1.5335959196090698,
+      "learning_rate": 1.9661016949152545e-05,
+      "loss": 1.7958,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5141323601722779,
+      "grad_norm": 1.5710722208023071,
+      "learning_rate": 1.964952599827636e-05,
+      "loss": 1.7861,
+      "step": 1808
+    },
+    {
+      "epoch": 0.514416725415736,
+      "grad_norm": 1.9415019750595093,
+      "learning_rate": 1.9638035047400174e-05,
+      "loss": 2.8067,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5147010906591942,
+      "grad_norm": 1.686766266822815,
+      "learning_rate": 1.9626544096523988e-05,
+      "loss": 2.4224,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5149854559026523,
+      "grad_norm": 1.6373934745788574,
+      "learning_rate": 1.9615053145647806e-05,
+      "loss": 2.3926,
+      "step": 1811
+    },
+    {
+      "epoch": 0.5152698211461104,
+      "grad_norm": 1.5860379934310913,
+      "learning_rate": 1.960356219477162e-05,
+      "loss": 2.213,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5155541863895685,
+      "grad_norm": 1.5629751682281494,
+      "learning_rate": 1.9592071243895435e-05,
+      "loss": 2.2401,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5158385516330266,
+      "grad_norm": 1.6226354837417603,
+      "learning_rate": 1.958058029301925e-05,
+      "loss": 1.8855,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5161229168764848,
+      "grad_norm": 1.6399784088134766,
+      "learning_rate": 1.9569089342143064e-05,
+      "loss": 1.9354,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5164072821199429,
+      "grad_norm": 1.8511910438537598,
+      "learning_rate": 1.955759839126688e-05,
+      "loss": 1.7383,
+      "step": 1816
+    },
+    {
+      "epoch": 0.516691647363401,
+      "grad_norm": 1.8263779878616333,
+      "learning_rate": 1.9546107440390693e-05,
+      "loss": 2.7295,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5169760126068591,
+      "grad_norm": 1.721635341644287,
+      "learning_rate": 1.9534616489514507e-05,
+      "loss": 2.4588,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5172603778503172,
+      "grad_norm": 1.6186027526855469,
+      "learning_rate": 1.9523125538638322e-05,
+      "loss": 2.4666,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5175447430937754,
+      "grad_norm": 1.6999125480651855,
+      "learning_rate": 1.951163458776214e-05,
+      "loss": 2.21,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5178291083372335,
+      "grad_norm": 1.6916403770446777,
+      "learning_rate": 1.9500143636885954e-05,
+      "loss": 2.037,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5181134735806916,
+      "grad_norm": 1.541050672531128,
+      "learning_rate": 1.948865268600977e-05,
+      "loss": 1.8888,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5183978388241497,
+      "grad_norm": 1.5618538856506348,
+      "learning_rate": 1.9477161735133583e-05,
+      "loss": 1.8478,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5186822040676078,
+      "grad_norm": 1.6209863424301147,
+      "learning_rate": 1.94656707842574e-05,
+      "loss": 1.6114,
+      "step": 1824
+    },
+    {
+      "epoch": 0.518966569311066,
+      "grad_norm": 1.7759110927581787,
+      "learning_rate": 1.9454179833381216e-05,
+      "loss": 2.6206,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5192509345545241,
+      "grad_norm": 1.5494519472122192,
+      "learning_rate": 1.944268888250503e-05,
+      "loss": 2.3457,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5195352997979822,
+      "grad_norm": 1.6379374265670776,
+      "learning_rate": 1.9431197931628845e-05,
+      "loss": 2.581,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5198196650414403,
+      "grad_norm": 1.5918887853622437,
+      "learning_rate": 1.941970698075266e-05,
+      "loss": 2.1037,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5201040302848984,
+      "grad_norm": 1.5805288553237915,
+      "learning_rate": 1.9408216029876474e-05,
+      "loss": 2.2555,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5203883955283566,
+      "grad_norm": 1.653382420539856,
+      "learning_rate": 1.9396725079000288e-05,
+      "loss": 1.7606,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5206727607718147,
+      "grad_norm": 1.5923447608947754,
+      "learning_rate": 1.9385234128124103e-05,
+      "loss": 1.6076,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5209571260152728,
+      "grad_norm": 1.7855467796325684,
+      "learning_rate": 1.937374317724792e-05,
+      "loss": 1.8589,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5212414912587309,
+      "grad_norm": 1.8567280769348145,
+      "learning_rate": 1.9362252226371735e-05,
+      "loss": 2.7604,
+      "step": 1833
+    },
+    {
+      "epoch": 0.521525856502189,
+      "grad_norm": 1.6221263408660889,
+      "learning_rate": 1.935076127549555e-05,
+      "loss": 2.4281,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5218102217456472,
+      "grad_norm": 1.6806731224060059,
+      "learning_rate": 1.9339270324619364e-05,
+      "loss": 2.4162,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5220945869891053,
+      "grad_norm": 1.55227792263031,
+      "learning_rate": 1.932777937374318e-05,
+      "loss": 2.0863,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5223789522325634,
+      "grad_norm": 1.5924606323242188,
+      "learning_rate": 1.9316288422866993e-05,
+      "loss": 2.004,
+      "step": 1837
+    },
+    {
+      "epoch": 0.5226633174760215,
+      "grad_norm": 1.6249966621398926,
+      "learning_rate": 1.9304797471990808e-05,
+      "loss": 2.0434,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5229476827194796,
+      "grad_norm": 1.702721357345581,
+      "learning_rate": 1.9293306521114622e-05,
+      "loss": 1.8394,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5232320479629378,
+      "grad_norm": 1.7455133199691772,
+      "learning_rate": 1.9281815570238437e-05,
+      "loss": 1.5562,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5235164132063959,
+      "grad_norm": 1.8886913061141968,
+      "learning_rate": 1.9270324619362254e-05,
+      "loss": 2.6912,
+      "step": 1841
+    },
+    {
+      "epoch": 0.523800778449854,
+      "grad_norm": 1.5527037382125854,
+      "learning_rate": 1.925883366848607e-05,
+      "loss": 2.4175,
+      "step": 1842
+    },
+    {
+      "epoch": 0.524085143693312,
+      "grad_norm": 1.5094817876815796,
+      "learning_rate": 1.9247342717609883e-05,
+      "loss": 2.299,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5243695089367703,
+      "grad_norm": 1.5052844285964966,
+      "learning_rate": 1.92358517667337e-05,
+      "loss": 2.3148,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5246538741802284,
+      "grad_norm": 1.5606803894042969,
+      "learning_rate": 1.9224360815857516e-05,
+      "loss": 2.2779,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5249382394236864,
+      "grad_norm": 1.6374552249908447,
+      "learning_rate": 1.921286986498133e-05,
+      "loss": 1.9273,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5252226046671445,
+      "grad_norm": 1.6898576021194458,
+      "learning_rate": 1.9201378914105145e-05,
+      "loss": 1.8248,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5255069699106026,
+      "grad_norm": 1.6066820621490479,
+      "learning_rate": 1.918988796322896e-05,
+      "loss": 1.8573,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5257913351540608,
+      "grad_norm": 1.8908296823501587,
+      "learning_rate": 1.9178397012352774e-05,
+      "loss": 2.9586,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5260757003975189,
+      "grad_norm": 1.7448999881744385,
+      "learning_rate": 1.9166906061476588e-05,
+      "loss": 2.5243,
+      "step": 1850
+    },
+    {
+      "epoch": 0.526360065640977,
+      "grad_norm": 1.5253217220306396,
+      "learning_rate": 1.9155415110600403e-05,
+      "loss": 2.3295,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5266444308844351,
+      "grad_norm": 1.5445467233657837,
+      "learning_rate": 1.9143924159724217e-05,
+      "loss": 2.3908,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5269287961278932,
+      "grad_norm": 1.622870922088623,
+      "learning_rate": 1.9132433208848035e-05,
+      "loss": 2.0477,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5272131613713514,
+      "grad_norm": 1.7335789203643799,
+      "learning_rate": 1.912094225797185e-05,
+      "loss": 1.821,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5274975266148095,
+      "grad_norm": 1.7417482137680054,
+      "learning_rate": 1.9109451307095664e-05,
+      "loss": 1.7534,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5277818918582676,
+      "grad_norm": 1.7831518650054932,
+      "learning_rate": 1.909796035621948e-05,
+      "loss": 2.0581,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5280662571017257,
+      "grad_norm": 1.9567564725875854,
+      "learning_rate": 1.9086469405343293e-05,
+      "loss": 2.7799,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5283506223451838,
+      "grad_norm": 1.6343289613723755,
+      "learning_rate": 1.9074978454467108e-05,
+      "loss": 2.5412,
+      "step": 1858
+    },
+    {
+      "epoch": 0.528634987588642,
+      "grad_norm": 1.4926238059997559,
+      "learning_rate": 1.9063487503590922e-05,
+      "loss": 2.3056,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5289193528321001,
+      "grad_norm": 1.5979171991348267,
+      "learning_rate": 1.9051996552714737e-05,
+      "loss": 2.4987,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5292037180755582,
+      "grad_norm": 1.6570512056350708,
+      "learning_rate": 1.9040505601838555e-05,
+      "loss": 2.0305,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5294880833190163,
+      "grad_norm": 1.517991065979004,
+      "learning_rate": 1.902901465096237e-05,
+      "loss": 1.672,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5297724485624744,
+      "grad_norm": 1.5992649793624878,
+      "learning_rate": 1.9017523700086184e-05,
+      "loss": 1.8708,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5300568138059326,
+      "grad_norm": 1.6488360166549683,
+      "learning_rate": 1.9006032749209998e-05,
+      "loss": 1.7175,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5303411790493907,
+      "grad_norm": 1.810335397720337,
+      "learning_rate": 1.8994541798333816e-05,
+      "loss": 2.8107,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5306255442928488,
+      "grad_norm": 1.5411492586135864,
+      "learning_rate": 1.898305084745763e-05,
+      "loss": 2.4042,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5309099095363069,
+      "grad_norm": 1.6033616065979004,
+      "learning_rate": 1.8971559896581445e-05,
+      "loss": 2.5207,
+      "step": 1867
+    },
+    {
+      "epoch": 0.531194274779765,
+      "grad_norm": 1.451252818107605,
+      "learning_rate": 1.896006894570526e-05,
+      "loss": 2.2705,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5314786400232232,
+      "grad_norm": 1.614965558052063,
+      "learning_rate": 1.8948577994829074e-05,
+      "loss": 2.3188,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5317630052666813,
+      "grad_norm": 1.570432186126709,
+      "learning_rate": 1.893708704395289e-05,
+      "loss": 1.901,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5320473705101394,
+      "grad_norm": 1.603621244430542,
+      "learning_rate": 1.8925596093076703e-05,
+      "loss": 1.811,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5323317357535975,
+      "grad_norm": 1.7074507474899292,
+      "learning_rate": 1.8914105142200517e-05,
+      "loss": 1.7383,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5326161009970556,
+      "grad_norm": 1.7993121147155762,
+      "learning_rate": 1.8902614191324332e-05,
+      "loss": 2.8412,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5329004662405138,
+      "grad_norm": 1.4372435808181763,
+      "learning_rate": 1.889112324044815e-05,
+      "loss": 2.4849,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5331848314839719,
+      "grad_norm": 1.4756051301956177,
+      "learning_rate": 1.8879632289571964e-05,
+      "loss": 2.3527,
+      "step": 1875
+    },
+    {
+      "epoch": 0.53346919672743,
+      "grad_norm": 1.54763925075531,
+      "learning_rate": 1.886814133869578e-05,
+      "loss": 2.1188,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5337535619708881,
+      "grad_norm": 1.6387181282043457,
+      "learning_rate": 1.8856650387819593e-05,
+      "loss": 2.2232,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5340379272143462,
+      "grad_norm": 1.5445959568023682,
+      "learning_rate": 1.8845159436943408e-05,
+      "loss": 1.96,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5343222924578044,
+      "grad_norm": 1.6059736013412476,
+      "learning_rate": 1.8833668486067222e-05,
+      "loss": 1.9384,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5346066577012625,
+      "grad_norm": 1.5326387882232666,
+      "learning_rate": 1.882217753519104e-05,
+      "loss": 1.8572,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5348910229447206,
+      "grad_norm": 1.9614207744598389,
+      "learning_rate": 1.8810686584314855e-05,
+      "loss": 2.8123,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5351753881881787,
+      "grad_norm": 1.532706618309021,
+      "learning_rate": 1.879919563343867e-05,
+      "loss": 2.3232,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5354597534316368,
+      "grad_norm": 1.612025499343872,
+      "learning_rate": 1.8787704682562484e-05,
+      "loss": 2.4058,
+      "step": 1883
+    },
+    {
+      "epoch": 0.535744118675095,
+      "grad_norm": 1.535540223121643,
+      "learning_rate": 1.8776213731686298e-05,
+      "loss": 2.2764,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5360284839185531,
+      "grad_norm": 1.7752254009246826,
+      "learning_rate": 1.8764722780810113e-05,
+      "loss": 2.1374,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5363128491620112,
+      "grad_norm": 1.6558598279953003,
+      "learning_rate": 1.875323182993393e-05,
+      "loss": 1.9204,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5365972144054693,
+      "grad_norm": 1.5669152736663818,
+      "learning_rate": 1.8741740879057745e-05,
+      "loss": 1.9425,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5368815796489275,
+      "grad_norm": 1.711891531944275,
+      "learning_rate": 1.873024992818156e-05,
+      "loss": 1.8079,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5371659448923856,
+      "grad_norm": 1.939569115638733,
+      "learning_rate": 1.8718758977305374e-05,
+      "loss": 2.7149,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5374503101358437,
+      "grad_norm": 1.6168116331100464,
+      "learning_rate": 1.870726802642919e-05,
+      "loss": 2.6034,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5377346753793018,
+      "grad_norm": 1.5140300989151,
+      "learning_rate": 1.8695777075553003e-05,
+      "loss": 2.2716,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5380190406227598,
+      "grad_norm": 1.440793752670288,
+      "learning_rate": 1.8684286124676817e-05,
+      "loss": 2.2802,
+      "step": 1892
+    },
+    {
+      "epoch": 0.538303405866218,
+      "grad_norm": 1.5787031650543213,
+      "learning_rate": 1.8672795173800632e-05,
+      "loss": 2.3291,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5385877711096761,
+      "grad_norm": 1.5322247743606567,
+      "learning_rate": 1.8661304222924446e-05,
+      "loss": 2.0667,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5388721363531342,
+      "grad_norm": 1.6390894651412964,
+      "learning_rate": 1.8649813272048264e-05,
+      "loss": 1.8755,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5391565015965923,
+      "grad_norm": 1.7358583211898804,
+      "learning_rate": 1.863832232117208e-05,
+      "loss": 1.8302,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5394408668400504,
+      "grad_norm": 1.8744136095046997,
+      "learning_rate": 1.8626831370295893e-05,
+      "loss": 2.8263,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5397252320835086,
+      "grad_norm": 1.5263346433639526,
+      "learning_rate": 1.861534041941971e-05,
+      "loss": 2.1965,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5400095973269667,
+      "grad_norm": 1.5258394479751587,
+      "learning_rate": 1.8603849468543526e-05,
+      "loss": 2.2005,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5402939625704248,
+      "grad_norm": 1.5305837392807007,
+      "learning_rate": 1.859235851766734e-05,
+      "loss": 2.2732,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5405783278138829,
+      "grad_norm": 1.7153434753417969,
+      "learning_rate": 1.8580867566791155e-05,
+      "loss": 2.2237,
+      "step": 1901
+    },
+    {
+      "epoch": 0.540862693057341,
+      "grad_norm": 1.5634775161743164,
+      "learning_rate": 1.856937661591497e-05,
+      "loss": 1.9519,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5411470583007992,
+      "grad_norm": 1.6455775499343872,
+      "learning_rate": 1.8557885665038784e-05,
+      "loss": 1.8184,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5414314235442573,
+      "grad_norm": 1.5984224081039429,
+      "learning_rate": 1.8546394714162598e-05,
+      "loss": 1.7093,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5417157887877154,
+      "grad_norm": 2.2204861640930176,
+      "learning_rate": 1.8534903763286413e-05,
+      "loss": 2.7596,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5420001540311735,
+      "grad_norm": 1.8560435771942139,
+      "learning_rate": 1.8523412812410227e-05,
+      "loss": 2.5364,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5422845192746316,
+      "grad_norm": 1.5567165613174438,
+      "learning_rate": 1.8511921861534045e-05,
+      "loss": 2.2509,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5425688845180898,
+      "grad_norm": 1.4999059438705444,
+      "learning_rate": 1.850043091065786e-05,
+      "loss": 2.153,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5428532497615479,
+      "grad_norm": 1.6572489738464355,
+      "learning_rate": 1.8488939959781674e-05,
+      "loss": 1.9657,
+      "step": 1909
+    },
+    {
+      "epoch": 0.543137615005006,
+      "grad_norm": 1.5145512819290161,
+      "learning_rate": 1.847744900890549e-05,
+      "loss": 1.7628,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5434219802484641,
+      "grad_norm": 1.5486762523651123,
+      "learning_rate": 1.8465958058029303e-05,
+      "loss": 1.8377,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5437063454919222,
+      "grad_norm": 1.6072850227355957,
+      "learning_rate": 1.8454467107153118e-05,
+      "loss": 1.9599,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5439907107353804,
+      "grad_norm": 1.9221049547195435,
+      "learning_rate": 1.8442976156276932e-05,
+      "loss": 2.8359,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5442750759788385,
+      "grad_norm": 1.6474133729934692,
+      "learning_rate": 1.8431485205400747e-05,
+      "loss": 2.495,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5445594412222966,
+      "grad_norm": 1.5209523439407349,
+      "learning_rate": 1.841999425452456e-05,
+      "loss": 2.2543,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5448438064657547,
+      "grad_norm": 1.592429280281067,
+      "learning_rate": 1.840850330364838e-05,
+      "loss": 2.2111,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5451281717092128,
+      "grad_norm": 1.5777971744537354,
+      "learning_rate": 1.8397012352772193e-05,
+      "loss": 2.1635,
+      "step": 1917
+    },
+    {
+      "epoch": 0.545412536952671,
+      "grad_norm": 1.5663307905197144,
+      "learning_rate": 1.8385521401896008e-05,
+      "loss": 1.8126,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5456969021961291,
+      "grad_norm": 1.643549919128418,
+      "learning_rate": 1.8374030451019826e-05,
+      "loss": 1.9639,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5459812674395872,
+      "grad_norm": 1.7581907510757446,
+      "learning_rate": 1.836253950014364e-05,
+      "loss": 1.5631,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5462656326830453,
+      "grad_norm": 1.8245182037353516,
+      "learning_rate": 1.8351048549267455e-05,
+      "loss": 2.8506,
+      "step": 1921
+    },
+    {
+      "epoch": 0.5465499979265034,
+      "grad_norm": 1.53452730178833,
+      "learning_rate": 1.833955759839127e-05,
+      "loss": 2.3347,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5468343631699616,
+      "grad_norm": 1.4857192039489746,
+      "learning_rate": 1.8328066647515084e-05,
+      "loss": 2.3888,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5471187284134197,
+      "grad_norm": 1.5653821229934692,
+      "learning_rate": 1.83165756966389e-05,
+      "loss": 2.2097,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5474030936568778,
+      "grad_norm": 1.7162553071975708,
+      "learning_rate": 1.8305084745762713e-05,
+      "loss": 1.9651,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5476874589003359,
+      "grad_norm": 1.5265828371047974,
+      "learning_rate": 1.8293593794886527e-05,
+      "loss": 1.8044,
+      "step": 1926
+    },
+    {
+      "epoch": 0.547971824143794,
+      "grad_norm": 1.5719175338745117,
+      "learning_rate": 1.8282102844010342e-05,
+      "loss": 1.8203,
+      "step": 1927
+    },
+    {
+      "epoch": 0.5482561893872522,
+      "grad_norm": 1.6208674907684326,
+      "learning_rate": 1.827061189313416e-05,
+      "loss": 1.777,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5485405546307103,
+      "grad_norm": 1.8437464237213135,
+      "learning_rate": 1.8259120942257974e-05,
+      "loss": 2.8108,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5488249198741684,
+      "grad_norm": 1.4935005903244019,
+      "learning_rate": 1.824762999138179e-05,
+      "loss": 2.4553,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5491092851176265,
+      "grad_norm": 1.4502531290054321,
+      "learning_rate": 1.8236139040505603e-05,
+      "loss": 2.1391,
+      "step": 1931
+    },
+    {
+      "epoch": 0.5493936503610846,
+      "grad_norm": 1.467806100845337,
+      "learning_rate": 1.8224648089629418e-05,
+      "loss": 2.0988,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5496780156045428,
+      "grad_norm": 1.5642067193984985,
+      "learning_rate": 1.8213157138753232e-05,
+      "loss": 2.1998,
+      "step": 1933
+    },
+    {
+      "epoch": 0.5499623808480009,
+      "grad_norm": 1.6454665660858154,
+      "learning_rate": 1.8201666187877047e-05,
+      "loss": 1.8338,
+      "step": 1934
+    },
+    {
+      "epoch": 0.550246746091459,
+      "grad_norm": 1.665605068206787,
+      "learning_rate": 1.8190175237000865e-05,
+      "loss": 1.8698,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5505311113349171,
+      "grad_norm": 1.661318063735962,
+      "learning_rate": 1.817868428612468e-05,
+      "loss": 1.7389,
+      "step": 1936
+    },
+    {
+      "epoch": 0.5508154765783753,
+      "grad_norm": 1.778904676437378,
+      "learning_rate": 1.8167193335248494e-05,
+      "loss": 2.6794,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5510998418218334,
+      "grad_norm": 1.7070825099945068,
+      "learning_rate": 1.8155702384372308e-05,
+      "loss": 2.4678,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5513842070652915,
+      "grad_norm": 1.5696340799331665,
+      "learning_rate": 1.8144211433496123e-05,
+      "loss": 2.2423,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5516685723087495,
+      "grad_norm": 1.664480209350586,
+      "learning_rate": 1.813272048261994e-05,
+      "loss": 2.2692,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5519529375522076,
+      "grad_norm": 1.7250727415084839,
+      "learning_rate": 1.8121229531743755e-05,
+      "loss": 1.8935,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5522373027956659,
+      "grad_norm": 1.5921610593795776,
+      "learning_rate": 1.810973858086757e-05,
+      "loss": 1.9107,
+      "step": 1942
+    },
+    {
+      "epoch": 0.552521668039124,
+      "grad_norm": 1.5735629796981812,
+      "learning_rate": 1.8098247629991384e-05,
+      "loss": 1.8625,
+      "step": 1943
+    },
+    {
+      "epoch": 0.552806033282582,
+      "grad_norm": 1.7604540586471558,
+      "learning_rate": 1.80867566791152e-05,
+      "loss": 1.9191,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5530903985260401,
+      "grad_norm": 1.894787311553955,
+      "learning_rate": 1.8075265728239013e-05,
+      "loss": 2.8069,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5533747637694982,
+      "grad_norm": 1.5562604665756226,
+      "learning_rate": 1.8063774777362827e-05,
+      "loss": 2.5382,
+      "step": 1946
+    },
+    {
+      "epoch": 0.5536591290129564,
+      "grad_norm": 1.5096544027328491,
+      "learning_rate": 1.8052283826486642e-05,
+      "loss": 2.174,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5539434942564145,
+      "grad_norm": 1.5471742153167725,
+      "learning_rate": 1.8040792875610456e-05,
+      "loss": 2.3403,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5542278594998726,
+      "grad_norm": 1.6458371877670288,
+      "learning_rate": 1.8029301924734274e-05,
+      "loss": 2.0733,
+      "step": 1949
+    },
+    {
+      "epoch": 0.5545122247433307,
+      "grad_norm": 1.7109332084655762,
+      "learning_rate": 1.801781097385809e-05,
+      "loss": 1.8824,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5547965899867888,
+      "grad_norm": 1.7361643314361572,
+      "learning_rate": 1.8006320022981903e-05,
+      "loss": 1.8698,
+      "step": 1951
+    },
+    {
+      "epoch": 0.555080955230247,
+      "grad_norm": 1.6732168197631836,
+      "learning_rate": 1.7994829072105718e-05,
+      "loss": 1.774,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5553653204737051,
+      "grad_norm": 1.7529971599578857,
+      "learning_rate": 1.7983338121229532e-05,
+      "loss": 2.8649,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5556496857171632,
+      "grad_norm": 1.5565195083618164,
+      "learning_rate": 1.797184717035335e-05,
+      "loss": 2.4457,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5559340509606213,
+      "grad_norm": 1.5663349628448486,
+      "learning_rate": 1.7960356219477165e-05,
+      "loss": 2.2654,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5562184162040794,
+      "grad_norm": 1.509903907775879,
+      "learning_rate": 1.794886526860098e-05,
+      "loss": 2.1682,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5565027814475376,
+      "grad_norm": 1.7437077760696411,
+      "learning_rate": 1.7937374317724794e-05,
+      "loss": 2.1682,
+      "step": 1957
+    },
+    {
+      "epoch": 0.5567871466909957,
+      "grad_norm": 1.532533049583435,
+      "learning_rate": 1.7925883366848608e-05,
+      "loss": 1.7354,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5570715119344538,
+      "grad_norm": 1.6096084117889404,
+      "learning_rate": 1.7914392415972423e-05,
+      "loss": 1.8484,
+      "step": 1959
+    },
+    {
+      "epoch": 0.5573558771779119,
+      "grad_norm": 1.8258168697357178,
+      "learning_rate": 1.7902901465096237e-05,
+      "loss": 1.807,
+      "step": 1960
+    },
+    {
+      "epoch": 0.55764024242137,
+      "grad_norm": 2.004221200942993,
+      "learning_rate": 1.789141051422005e-05,
+      "loss": 2.7515,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5579246076648282,
+      "grad_norm": 1.593041181564331,
+      "learning_rate": 1.787991956334387e-05,
+      "loss": 2.3786,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5582089729082863,
+      "grad_norm": 1.5780787467956543,
+      "learning_rate": 1.7868428612467684e-05,
+      "loss": 2.3528,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5584933381517444,
+      "grad_norm": 1.5928815603256226,
+      "learning_rate": 1.78569376615915e-05,
+      "loss": 2.2974,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5587777033952025,
+      "grad_norm": 1.7977780103683472,
+      "learning_rate": 1.7845446710715313e-05,
+      "loss": 2.2807,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5590620686386606,
+      "grad_norm": 1.5917166471481323,
+      "learning_rate": 1.7833955759839128e-05,
+      "loss": 2.0174,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5593464338821188,
+      "grad_norm": 1.55897057056427,
+      "learning_rate": 1.7822464808962942e-05,
+      "loss": 1.8759,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5596307991255769,
+      "grad_norm": 1.6804914474487305,
+      "learning_rate": 1.7810973858086757e-05,
+      "loss": 1.7898,
+      "step": 1968
+    },
+    {
+      "epoch": 0.559915164369035,
+      "grad_norm": 1.7221648693084717,
+      "learning_rate": 1.779948290721057e-05,
+      "loss": 2.6948,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5601995296124931,
+      "grad_norm": 1.5615862607955933,
+      "learning_rate": 1.7787991956334386e-05,
+      "loss": 2.2914,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5604838948559512,
+      "grad_norm": 1.5630532503128052,
+      "learning_rate": 1.7776501005458203e-05,
+      "loss": 2.4075,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5607682600994094,
+      "grad_norm": 1.6465175151824951,
+      "learning_rate": 1.7765010054582018e-05,
+      "loss": 2.2431,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5610526253428675,
+      "grad_norm": 1.6053401231765747,
+      "learning_rate": 1.7753519103705832e-05,
+      "loss": 1.9444,
+      "step": 1973
+    },
+    {
+      "epoch": 0.5613369905863256,
+      "grad_norm": 1.5108147859573364,
+      "learning_rate": 1.774202815282965e-05,
+      "loss": 1.8403,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5616213558297837,
+      "grad_norm": 1.814316987991333,
+      "learning_rate": 1.7730537201953465e-05,
+      "loss": 1.7395,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5619057210732418,
+      "grad_norm": 1.6270887851715088,
+      "learning_rate": 1.771904625107728e-05,
+      "loss": 1.6755,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5621900863167,
+      "grad_norm": 1.9045475721359253,
+      "learning_rate": 1.7707555300201094e-05,
+      "loss": 2.8776,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5624744515601581,
+      "grad_norm": 1.6713111400604248,
+      "learning_rate": 1.769606434932491e-05,
+      "loss": 2.3999,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5627588168036162,
+      "grad_norm": 1.4755371809005737,
+      "learning_rate": 1.7684573398448723e-05,
+      "loss": 2.344,
+      "step": 1979
+    },
+    {
+      "epoch": 0.5630431820470743,
+      "grad_norm": 1.6609379053115845,
+      "learning_rate": 1.7673082447572537e-05,
+      "loss": 2.2334,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5633275472905324,
+      "grad_norm": 1.750351071357727,
+      "learning_rate": 1.7661591496696352e-05,
+      "loss": 2.2961,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5636119125339906,
+      "grad_norm": 1.47465181350708,
+      "learning_rate": 1.7650100545820166e-05,
+      "loss": 1.7603,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5638962777774487,
+      "grad_norm": 1.627179503440857,
+      "learning_rate": 1.7638609594943984e-05,
+      "loss": 1.8822,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5641806430209068,
+      "grad_norm": 1.570872187614441,
+      "learning_rate": 1.76271186440678e-05,
+      "loss": 1.8734,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5644650082643649,
+      "grad_norm": 1.7837358713150024,
+      "learning_rate": 1.7615627693191613e-05,
+      "loss": 2.7609,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5647493735078231,
+      "grad_norm": 1.5438477993011475,
+      "learning_rate": 1.7604136742315428e-05,
+      "loss": 2.2791,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5650337387512812,
+      "grad_norm": 1.561362385749817,
+      "learning_rate": 1.7592645791439242e-05,
+      "loss": 2.1921,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5653181039947393,
+      "grad_norm": 1.4173671007156372,
+      "learning_rate": 1.7581154840563057e-05,
+      "loss": 2.0708,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5656024692381973,
+      "grad_norm": 1.5849429368972778,
+      "learning_rate": 1.756966388968687e-05,
+      "loss": 2.1582,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5658868344816554,
+      "grad_norm": 1.5969016551971436,
+      "learning_rate": 1.7558172938810686e-05,
+      "loss": 1.9153,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5661711997251136,
+      "grad_norm": 1.72384774684906,
+      "learning_rate": 1.7546681987934504e-05,
+      "loss": 2.1121,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5664555649685717,
+      "grad_norm": 1.6643520593643188,
+      "learning_rate": 1.7535191037058318e-05,
+      "loss": 1.7982,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5667399302120298,
+      "grad_norm": 1.7195711135864258,
+      "learning_rate": 1.7523700086182133e-05,
+      "loss": 2.8548,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5670242954554879,
+      "grad_norm": 1.5349563360214233,
+      "learning_rate": 1.7512209135305947e-05,
+      "loss": 2.5764,
+      "step": 1994
+    },
+    {
+      "epoch": 0.567308660698946,
+      "grad_norm": 1.562687873840332,
+      "learning_rate": 1.7500718184429765e-05,
+      "loss": 2.1619,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5675930259424042,
+      "grad_norm": 1.4912785291671753,
+      "learning_rate": 1.748922723355358e-05,
+      "loss": 2.2925,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5678773911858623,
+      "grad_norm": 1.5238652229309082,
+      "learning_rate": 1.7477736282677394e-05,
+      "loss": 2.0571,
+      "step": 1997
+    },
+    {
+      "epoch": 0.5681617564293204,
+      "grad_norm": 1.570028305053711,
+      "learning_rate": 1.746624533180121e-05,
+      "loss": 2.0634,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5684461216727785,
+      "grad_norm": 1.6997617483139038,
+      "learning_rate": 1.7454754380925023e-05,
+      "loss": 1.7017,
+      "step": 1999
+    },
+    {
+      "epoch": 0.5687304869162366,
+      "grad_norm": 1.747353434562683,
+      "learning_rate": 1.7443263430048837e-05,
+      "loss": 1.8186,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5690148521596948,
+      "grad_norm": 1.8693517446517944,
+      "learning_rate": 1.7431772479172652e-05,
+      "loss": 2.8259,
+      "step": 2001
+    },
+    {
+      "epoch": 0.5692992174031529,
+      "grad_norm": 1.6010723114013672,
+      "learning_rate": 1.7420281528296466e-05,
+      "loss": 2.4006,
+      "step": 2002
+    },
+    {
+      "epoch": 0.569583582646611,
+      "grad_norm": 1.4531267881393433,
+      "learning_rate": 1.740879057742028e-05,
+      "loss": 2.1439,
+      "step": 2003
+    },
+    {
+      "epoch": 0.5698679478900691,
+      "grad_norm": 1.6961573362350464,
+      "learning_rate": 1.73972996265441e-05,
+      "loss": 2.2823,
+      "step": 2004
+    },
+    {
+      "epoch": 0.5701523131335272,
+      "grad_norm": 1.5460631847381592,
+      "learning_rate": 1.7385808675667913e-05,
+      "loss": 2.1329,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5704366783769854,
+      "grad_norm": 1.644081950187683,
+      "learning_rate": 1.7374317724791728e-05,
+      "loss": 2.0236,
+      "step": 2006
+    },
+    {
+      "epoch": 0.5707210436204435,
+      "grad_norm": 1.6294313669204712,
+      "learning_rate": 1.7362826773915542e-05,
+      "loss": 2.004,
+      "step": 2007
+    },
+    {
+      "epoch": 0.5710054088639016,
+      "grad_norm": 1.7416433095932007,
+      "learning_rate": 1.7351335823039357e-05,
+      "loss": 1.9497,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5712897741073597,
+      "grad_norm": 1.8446623086929321,
+      "learning_rate": 1.7339844872163175e-05,
+      "loss": 2.8675,
+      "step": 2009
+    },
+    {
+      "epoch": 0.5715741393508178,
+      "grad_norm": 1.5892642736434937,
+      "learning_rate": 1.732835392128699e-05,
+      "loss": 2.3443,
+      "step": 2010
+    },
+    {
+      "epoch": 0.571858504594276,
+      "grad_norm": 1.5367027521133423,
+      "learning_rate": 1.7316862970410804e-05,
+      "loss": 2.2238,
+      "step": 2011
+    },
+    {
+      "epoch": 0.5721428698377341,
+      "grad_norm": 1.5910507440567017,
+      "learning_rate": 1.7305372019534618e-05,
+      "loss": 2.2285,
+      "step": 2012
+    },
+    {
+      "epoch": 0.5724272350811922,
+      "grad_norm": 1.6503397226333618,
+      "learning_rate": 1.7293881068658433e-05,
+      "loss": 2.1693,
+      "step": 2013
+    },
+    {
+      "epoch": 0.5727116003246503,
+      "grad_norm": 1.5087186098098755,
+      "learning_rate": 1.7282390117782247e-05,
+      "loss": 1.896,
+      "step": 2014
+    },
+    {
+      "epoch": 0.5729959655681084,
+      "grad_norm": 1.6095649003982544,
+      "learning_rate": 1.727089916690606e-05,
+      "loss": 1.773,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5732803308115666,
+      "grad_norm": 1.5037708282470703,
+      "learning_rate": 1.725940821602988e-05,
+      "loss": 1.7249,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5735646960550247,
+      "grad_norm": 1.788989543914795,
+      "learning_rate": 1.7247917265153694e-05,
+      "loss": 2.8071,
+      "step": 2017
+    },
+    {
+      "epoch": 0.5738490612984828,
+      "grad_norm": 1.5129423141479492,
+      "learning_rate": 1.723642631427751e-05,
+      "loss": 2.3458,
+      "step": 2018
+    },
+    {
+      "epoch": 0.5741334265419409,
+      "grad_norm": 1.4829000234603882,
+      "learning_rate": 1.7224935363401323e-05,
+      "loss": 2.2329,
+      "step": 2019
+    },
+    {
+      "epoch": 0.574417791785399,
+      "grad_norm": 1.5166312456130981,
+      "learning_rate": 1.7213444412525138e-05,
+      "loss": 2.1336,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5747021570288572,
+      "grad_norm": 1.6192970275878906,
+      "learning_rate": 1.7201953461648952e-05,
+      "loss": 2.2265,
+      "step": 2021
+    },
+    {
+      "epoch": 0.5749865222723153,
+      "grad_norm": 1.7159581184387207,
+      "learning_rate": 1.7190462510772767e-05,
+      "loss": 1.9326,
+      "step": 2022
+    },
+    {
+      "epoch": 0.5752708875157734,
+      "grad_norm": 1.5626370906829834,
+      "learning_rate": 1.717897155989658e-05,
+      "loss": 1.7684,
+      "step": 2023
+    },
+    {
+      "epoch": 0.5755552527592315,
+      "grad_norm": 1.6683768033981323,
+      "learning_rate": 1.7167480609020396e-05,
+      "loss": 1.9421,
+      "step": 2024
+    },
+    {
+      "epoch": 0.5758396180026896,
+      "grad_norm": 1.9297045469284058,
+      "learning_rate": 1.7155989658144213e-05,
+      "loss": 2.9594,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5761239832461478,
+      "grad_norm": 1.5986518859863281,
+      "learning_rate": 1.7144498707268028e-05,
+      "loss": 2.334,
+      "step": 2026
+    },
+    {
+      "epoch": 0.5764083484896059,
+      "grad_norm": 1.5038700103759766,
+      "learning_rate": 1.7133007756391842e-05,
+      "loss": 2.326,
+      "step": 2027
+    },
+    {
+      "epoch": 0.576692713733064,
+      "grad_norm": 1.4739896059036255,
+      "learning_rate": 1.712151680551566e-05,
+      "loss": 2.2516,
+      "step": 2028
+    },
+    {
+      "epoch": 0.5769770789765221,
+      "grad_norm": 1.5917458534240723,
+      "learning_rate": 1.7110025854639475e-05,
+      "loss": 2.0803,
+      "step": 2029
+    },
+    {
+      "epoch": 0.5772614442199802,
+      "grad_norm": 1.491292953491211,
+      "learning_rate": 1.709853490376329e-05,
+      "loss": 1.8907,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5775458094634384,
+      "grad_norm": 1.5379160642623901,
+      "learning_rate": 1.7087043952887104e-05,
+      "loss": 1.8301,
+      "step": 2031
+    },
+    {
+      "epoch": 0.5778301747068965,
+      "grad_norm": 1.6341428756713867,
+      "learning_rate": 1.7075553002010918e-05,
+      "loss": 1.7232,
+      "step": 2032
+    },
+    {
+      "epoch": 0.5781145399503546,
+      "grad_norm": 1.6281121969223022,
+      "learning_rate": 1.7064062051134733e-05,
+      "loss": 2.7037,
+      "step": 2033
+    },
+    {
+      "epoch": 0.5783989051938127,
+      "grad_norm": 1.5146305561065674,
+      "learning_rate": 1.7052571100258547e-05,
+      "loss": 2.4446,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5786832704372709,
+      "grad_norm": 1.5931079387664795,
+      "learning_rate": 1.7041080149382362e-05,
+      "loss": 2.2561,
+      "step": 2035
+    },
+    {
+      "epoch": 0.578967635680729,
+      "grad_norm": 1.4791065454483032,
+      "learning_rate": 1.7029589198506176e-05,
+      "loss": 2.3593,
+      "step": 2036
+    },
+    {
+      "epoch": 0.579252000924187,
+      "grad_norm": 1.6104813814163208,
+      "learning_rate": 1.7018098247629994e-05,
+      "loss": 2.1088,
+      "step": 2037
+    },
+    {
+      "epoch": 0.5795363661676451,
+      "grad_norm": 1.5709761381149292,
+      "learning_rate": 1.700660729675381e-05,
+      "loss": 1.7519,
+      "step": 2038
+    },
+    {
+      "epoch": 0.5798207314111032,
+      "grad_norm": 1.6428778171539307,
+      "learning_rate": 1.6995116345877623e-05,
+      "loss": 1.9199,
+      "step": 2039
+    },
+    {
+      "epoch": 0.5801050966545614,
+      "grad_norm": 1.7062665224075317,
+      "learning_rate": 1.6983625395001438e-05,
+      "loss": 1.9117,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5803894618980195,
+      "grad_norm": 1.9232088327407837,
+      "learning_rate": 1.6972134444125252e-05,
+      "loss": 2.822,
+      "step": 2041
+    },
+    {
+      "epoch": 0.5806738271414776,
+      "grad_norm": 1.44762122631073,
+      "learning_rate": 1.6960643493249067e-05,
+      "loss": 2.3914,
+      "step": 2042
+    },
+    {
+      "epoch": 0.5809581923849357,
+      "grad_norm": 1.5696653127670288,
+      "learning_rate": 1.694915254237288e-05,
+      "loss": 2.2531,
+      "step": 2043
+    },
+    {
+      "epoch": 0.5812425576283938,
+      "grad_norm": 1.4691784381866455,
+      "learning_rate": 1.6937661591496696e-05,
+      "loss": 2.225,
+      "step": 2044
+    },
+    {
+      "epoch": 0.581526922871852,
+      "grad_norm": 1.6073601245880127,
+      "learning_rate": 1.692617064062051e-05,
+      "loss": 1.9158,
+      "step": 2045
+    },
+    {
+      "epoch": 0.5818112881153101,
+      "grad_norm": 1.5055420398712158,
+      "learning_rate": 1.6914679689744328e-05,
+      "loss": 2.0171,
+      "step": 2046
+    },
+    {
+      "epoch": 0.5820956533587682,
+      "grad_norm": 1.6332578659057617,
+      "learning_rate": 1.6903188738868143e-05,
+      "loss": 1.7642,
+      "step": 2047
+    },
+    {
+      "epoch": 0.5823800186022263,
+      "grad_norm": 1.577101230621338,
+      "learning_rate": 1.6891697787991957e-05,
+      "loss": 1.8056,
+      "step": 2048
+    },
+    {
+      "epoch": 0.5826643838456844,
+      "grad_norm": 1.963454008102417,
+      "learning_rate": 1.6880206837115775e-05,
+      "loss": 2.8995,
+      "step": 2049
+    },
+    {
+      "epoch": 0.5829487490891426,
+      "grad_norm": 1.5220768451690674,
+      "learning_rate": 1.686871588623959e-05,
+      "loss": 2.2895,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5832331143326007,
+      "grad_norm": 1.5367721319198608,
+      "learning_rate": 1.6857224935363404e-05,
+      "loss": 2.2004,
+      "step": 2051
+    },
+    {
+      "epoch": 0.5835174795760588,
+      "grad_norm": 1.5274112224578857,
+      "learning_rate": 1.684573398448722e-05,
+      "loss": 2.2675,
+      "step": 2052
+    },
+    {
+      "epoch": 0.5838018448195169,
+      "grad_norm": 1.7082741260528564,
+      "learning_rate": 1.6834243033611033e-05,
+      "loss": 2.1209,
+      "step": 2053
+    },
+    {
+      "epoch": 0.584086210062975,
+      "grad_norm": 1.5683553218841553,
+      "learning_rate": 1.6822752082734847e-05,
+      "loss": 1.9012,
+      "step": 2054
+    },
+    {
+      "epoch": 0.5843705753064332,
+      "grad_norm": 1.6328856945037842,
+      "learning_rate": 1.6811261131858662e-05,
+      "loss": 1.9311,
+      "step": 2055
+    },
+    {
+      "epoch": 0.5846549405498913,
+      "grad_norm": 1.782463550567627,
+      "learning_rate": 1.6799770180982476e-05,
+      "loss": 1.8888,
+      "step": 2056
+    },
+    {
+      "epoch": 0.5849393057933494,
+      "grad_norm": 1.9230624437332153,
+      "learning_rate": 1.678827923010629e-05,
+      "loss": 3.0453,
+      "step": 2057
+    },
+    {
+      "epoch": 0.5852236710368075,
+      "grad_norm": 1.576603651046753,
+      "learning_rate": 1.677678827923011e-05,
+      "loss": 2.2623,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5855080362802656,
+      "grad_norm": 1.4937617778778076,
+      "learning_rate": 1.6765297328353923e-05,
+      "loss": 2.1262,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5857924015237238,
+      "grad_norm": 1.6465567350387573,
+      "learning_rate": 1.6753806377477738e-05,
+      "loss": 2.2345,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5860767667671819,
+      "grad_norm": 1.6568986177444458,
+      "learning_rate": 1.6742315426601552e-05,
+      "loss": 2.1957,
+      "step": 2061
+    },
+    {
+      "epoch": 0.58636113201064,
+      "grad_norm": 1.6795825958251953,
+      "learning_rate": 1.6730824475725367e-05,
+      "loss": 1.8565,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5866454972540981,
+      "grad_norm": 1.547038197517395,
+      "learning_rate": 1.671933352484918e-05,
+      "loss": 1.8675,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5869298624975562,
+      "grad_norm": 1.557686686515808,
+      "learning_rate": 1.6707842573972996e-05,
+      "loss": 1.7547,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5872142277410144,
+      "grad_norm": 1.8700066804885864,
+      "learning_rate": 1.6696351623096814e-05,
+      "loss": 2.7392,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5874985929844725,
+      "grad_norm": 1.6624739170074463,
+      "learning_rate": 1.6684860672220628e-05,
+      "loss": 2.343,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5877829582279306,
+      "grad_norm": 1.5265758037567139,
+      "learning_rate": 1.6673369721344443e-05,
+      "loss": 2.1868,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5880673234713887,
+      "grad_norm": 1.5248119831085205,
+      "learning_rate": 1.6661878770468257e-05,
+      "loss": 2.1564,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5883516887148468,
+      "grad_norm": 1.6401458978652954,
+      "learning_rate": 1.665038781959207e-05,
+      "loss": 2.2195,
+      "step": 2069
+    },
+    {
+      "epoch": 0.588636053958305,
+      "grad_norm": 1.5524535179138184,
+      "learning_rate": 1.663889686871589e-05,
+      "loss": 1.9903,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5889204192017631,
+      "grad_norm": 1.6761585474014282,
+      "learning_rate": 1.6627405917839704e-05,
+      "loss": 1.9134,
+      "step": 2071
+    },
+    {
+      "epoch": 0.5892047844452212,
+      "grad_norm": 1.6284050941467285,
+      "learning_rate": 1.661591496696352e-05,
+      "loss": 1.8622,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5894891496886793,
+      "grad_norm": 1.7775750160217285,
+      "learning_rate": 1.6604424016087333e-05,
+      "loss": 2.7387,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5897735149321374,
+      "grad_norm": 1.5672471523284912,
+      "learning_rate": 1.6592933065211147e-05,
+      "loss": 2.13,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5900578801755956,
+      "grad_norm": 2.415464401245117,
+      "learning_rate": 1.6581442114334962e-05,
+      "loss": 2.58,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5903422454190537,
+      "grad_norm": 1.550248622894287,
+      "learning_rate": 1.6569951163458776e-05,
+      "loss": 2.2516,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5906266106625118,
+      "grad_norm": 1.564470887184143,
+      "learning_rate": 1.655846021258259e-05,
+      "loss": 1.9016,
+      "step": 2077
+    },
+    {
+      "epoch": 0.5909109759059699,
+      "grad_norm": 1.5428235530853271,
+      "learning_rate": 1.6546969261706405e-05,
+      "loss": 1.9438,
+      "step": 2078
+    },
+    {
+      "epoch": 0.5911953411494281,
+      "grad_norm": 1.5634739398956299,
+      "learning_rate": 1.6535478310830223e-05,
+      "loss": 2.0286,
+      "step": 2079
+    },
+    {
+      "epoch": 0.5914797063928862,
+      "grad_norm": 1.6420611143112183,
+      "learning_rate": 1.6523987359954038e-05,
+      "loss": 1.7369,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5917640716363443,
+      "grad_norm": 1.8491171598434448,
+      "learning_rate": 1.6512496409077852e-05,
+      "loss": 2.7506,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5920484368798024,
+      "grad_norm": 1.6593526601791382,
+      "learning_rate": 1.6501005458201667e-05,
+      "loss": 2.4749,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5923328021232604,
+      "grad_norm": 1.6108300685882568,
+      "learning_rate": 1.648951450732548e-05,
+      "loss": 2.4418,
+      "step": 2083
+    },
+    {
+      "epoch": 0.5926171673667187,
+      "grad_norm": 1.489844799041748,
+      "learning_rate": 1.64780235564493e-05,
+      "loss": 2.2861,
+      "step": 2084
+    },
+    {
+      "epoch": 0.5929015326101768,
+      "grad_norm": 1.5152863264083862,
+      "learning_rate": 1.6466532605573114e-05,
+      "loss": 2.1108,
+      "step": 2085
+    },
+    {
+      "epoch": 0.5931858978536348,
+      "grad_norm": 1.5476469993591309,
+      "learning_rate": 1.6455041654696928e-05,
+      "loss": 1.9956,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5934702630970929,
+      "grad_norm": 1.568286418914795,
+      "learning_rate": 1.6443550703820743e-05,
+      "loss": 1.7746,
+      "step": 2087
+    },
+    {
+      "epoch": 0.593754628340551,
+      "grad_norm": 1.673620581626892,
+      "learning_rate": 1.6432059752944557e-05,
+      "loss": 1.8101,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5940389935840092,
+      "grad_norm": 1.8000257015228271,
+      "learning_rate": 1.6420568802068372e-05,
+      "loss": 2.8105,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5943233588274673,
+      "grad_norm": 1.4337655305862427,
+      "learning_rate": 1.6409077851192186e-05,
+      "loss": 2.507,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5946077240709254,
+      "grad_norm": 1.5374799966812134,
+      "learning_rate": 1.6397586900316004e-05,
+      "loss": 2.4551,
+      "step": 2091
+    },
+    {
+      "epoch": 0.5948920893143835,
+      "grad_norm": 1.586155891418457,
+      "learning_rate": 1.638609594943982e-05,
+      "loss": 2.0942,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5951764545578416,
+      "grad_norm": 1.6072989702224731,
+      "learning_rate": 1.6374604998563633e-05,
+      "loss": 1.9915,
+      "step": 2093
+    },
+    {
+      "epoch": 0.5954608198012998,
+      "grad_norm": 1.4546177387237549,
+      "learning_rate": 1.6363114047687448e-05,
+      "loss": 1.8326,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5957451850447579,
+      "grad_norm": 1.5891377925872803,
+      "learning_rate": 1.6351623096811262e-05,
+      "loss": 1.7843,
+      "step": 2095
+    },
+    {
+      "epoch": 0.596029550288216,
+      "grad_norm": 1.6567867994308472,
+      "learning_rate": 1.6340132145935077e-05,
+      "loss": 1.622,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5963139155316741,
+      "grad_norm": 1.8327213525772095,
+      "learning_rate": 1.632864119505889e-05,
+      "loss": 2.7612,
+      "step": 2097
+    },
+    {
+      "epoch": 0.5965982807751322,
+      "grad_norm": 1.5466746091842651,
+      "learning_rate": 1.6317150244182706e-05,
+      "loss": 2.2606,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5968826460185904,
+      "grad_norm": 1.4704080820083618,
+      "learning_rate": 1.630565929330652e-05,
+      "loss": 2.2202,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5971670112620485,
+      "grad_norm": 1.5352089405059814,
+      "learning_rate": 1.6294168342430338e-05,
+      "loss": 2.1893,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5974513765055066,
+      "grad_norm": 1.6276265382766724,
+      "learning_rate": 1.6282677391554152e-05,
+      "loss": 2.0548,
+      "step": 2101
+    },
+    {
+      "epoch": 0.5977357417489647,
+      "grad_norm": 1.4630635976791382,
+      "learning_rate": 1.6271186440677967e-05,
+      "loss": 1.799,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5980201069924228,
+      "grad_norm": 1.538522720336914,
+      "learning_rate": 1.6259695489801785e-05,
+      "loss": 1.6713,
+      "step": 2103
+    },
+    {
+      "epoch": 0.598304472235881,
+      "grad_norm": 1.6037315130233765,
+      "learning_rate": 1.62482045389256e-05,
+      "loss": 1.6671,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5985888374793391,
+      "grad_norm": 1.940868854522705,
+      "learning_rate": 1.6236713588049414e-05,
+      "loss": 2.8063,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5988732027227972,
+      "grad_norm": 1.5579652786254883,
+      "learning_rate": 1.622522263717323e-05,
+      "loss": 2.3295,
+      "step": 2106
+    },
+    {
+      "epoch": 0.5991575679662553,
+      "grad_norm": 1.6156672239303589,
+      "learning_rate": 1.6213731686297043e-05,
+      "loss": 2.2363,
+      "step": 2107
+    },
+    {
+      "epoch": 0.5994419332097134,
+      "grad_norm": 1.485302209854126,
+      "learning_rate": 1.6202240735420857e-05,
+      "loss": 2.0772,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5997262984531716,
+      "grad_norm": 1.7746816873550415,
+      "learning_rate": 1.6190749784544672e-05,
+      "loss": 2.1031,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6000106636966297,
+      "grad_norm": 1.6292321681976318,
+      "learning_rate": 1.6179258833668486e-05,
+      "loss": 1.9516,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6002950289400878,
+      "grad_norm": 1.6927971839904785,
+      "learning_rate": 1.61677678827923e-05,
+      "loss": 1.9776,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6005793941835459,
+      "grad_norm": 1.7361969947814941,
+      "learning_rate": 1.615627693191612e-05,
+      "loss": 1.6757,
+      "step": 2112
+    },
+    {
+      "epoch": 0.600863759427004,
+      "grad_norm": 1.7664377689361572,
+      "learning_rate": 1.6144785981039933e-05,
+      "loss": 2.7698,
+      "step": 2113
+    },
+    {
+      "epoch": 0.6011481246704622,
+      "grad_norm": 1.4889873266220093,
+      "learning_rate": 1.6133295030163748e-05,
+      "loss": 2.3977,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6014324899139203,
+      "grad_norm": 1.5173231363296509,
+      "learning_rate": 1.6121804079287562e-05,
+      "loss": 2.2699,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6017168551573784,
+      "grad_norm": 1.5664979219436646,
+      "learning_rate": 1.6110313128411377e-05,
+      "loss": 2.2895,
+      "step": 2116
+    },
+    {
+      "epoch": 0.6020012204008365,
+      "grad_norm": 1.7025697231292725,
+      "learning_rate": 1.609882217753519e-05,
+      "loss": 1.9624,
+      "step": 2117
+    },
+    {
+      "epoch": 0.6022855856442946,
+      "grad_norm": 1.7116023302078247,
+      "learning_rate": 1.6087331226659006e-05,
+      "loss": 2.1481,
+      "step": 2118
+    },
+    {
+      "epoch": 0.6025699508877528,
+      "grad_norm": 1.53733491897583,
+      "learning_rate": 1.607584027578282e-05,
+      "loss": 1.6645,
+      "step": 2119
+    },
+    {
+      "epoch": 0.6028543161312109,
+      "grad_norm": 1.602069616317749,
+      "learning_rate": 1.6064349324906635e-05,
+      "loss": 1.7552,
+      "step": 2120
+    },
+    {
+      "epoch": 0.603138681374669,
+      "grad_norm": 1.7021903991699219,
+      "learning_rate": 1.6052858374030453e-05,
+      "loss": 2.7208,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6034230466181271,
+      "grad_norm": 1.5884177684783936,
+      "learning_rate": 1.6041367423154267e-05,
+      "loss": 2.4558,
+      "step": 2122
+    },
+    {
+      "epoch": 0.6037074118615852,
+      "grad_norm": 1.600091814994812,
+      "learning_rate": 1.602987647227808e-05,
+      "loss": 2.3824,
+      "step": 2123
+    },
+    {
+      "epoch": 0.6039917771050434,
+      "grad_norm": 1.636725902557373,
+      "learning_rate": 1.60183855214019e-05,
+      "loss": 2.2555,
+      "step": 2124
+    },
+    {
+      "epoch": 0.6042761423485015,
+      "grad_norm": 1.5665066242218018,
+      "learning_rate": 1.6006894570525714e-05,
+      "loss": 2.2357,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6045605075919596,
+      "grad_norm": 1.6503568887710571,
+      "learning_rate": 1.599540361964953e-05,
+      "loss": 1.9319,
+      "step": 2126
+    },
+    {
+      "epoch": 0.6048448728354177,
+      "grad_norm": 1.774845838546753,
+      "learning_rate": 1.5983912668773343e-05,
+      "loss": 1.8753,
+      "step": 2127
+    },
+    {
+      "epoch": 0.6051292380788759,
+      "grad_norm": 1.626198649406433,
+      "learning_rate": 1.5972421717897157e-05,
+      "loss": 1.9712,
+      "step": 2128
+    },
+    {
+      "epoch": 0.605413603322334,
+      "grad_norm": 1.8184738159179688,
+      "learning_rate": 1.5960930767020972e-05,
+      "loss": 2.7941,
+      "step": 2129
+    },
+    {
+      "epoch": 0.605697968565792,
+      "grad_norm": 1.5862571001052856,
+      "learning_rate": 1.5949439816144786e-05,
+      "loss": 2.421,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6059823338092502,
+      "grad_norm": 1.5402880907058716,
+      "learning_rate": 1.59379488652686e-05,
+      "loss": 2.462,
+      "step": 2131
+    },
+    {
+      "epoch": 0.6062666990527082,
+      "grad_norm": 1.606608510017395,
+      "learning_rate": 1.5926457914392415e-05,
+      "loss": 2.1879,
+      "step": 2132
+    },
+    {
+      "epoch": 0.6065510642961665,
+      "grad_norm": 1.640724539756775,
+      "learning_rate": 1.5914966963516233e-05,
+      "loss": 1.9824,
+      "step": 2133
+    },
+    {
+      "epoch": 0.6068354295396245,
+      "grad_norm": 1.5415105819702148,
+      "learning_rate": 1.5903476012640048e-05,
+      "loss": 1.8615,
+      "step": 2134
+    },
+    {
+      "epoch": 0.6071197947830826,
+      "grad_norm": 1.5723168849945068,
+      "learning_rate": 1.5891985061763862e-05,
+      "loss": 1.8926,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6074041600265407,
+      "grad_norm": 1.752859115600586,
+      "learning_rate": 1.5880494110887677e-05,
+      "loss": 1.9364,
+      "step": 2136
+    },
+    {
+      "epoch": 0.6076885252699988,
+      "grad_norm": 1.7561215162277222,
+      "learning_rate": 1.586900316001149e-05,
+      "loss": 2.7265,
+      "step": 2137
+    },
+    {
+      "epoch": 0.607972890513457,
+      "grad_norm": 1.5687557458877563,
+      "learning_rate": 1.5857512209135306e-05,
+      "loss": 2.1813,
+      "step": 2138
+    },
+    {
+      "epoch": 0.6082572557569151,
+      "grad_norm": 1.5815242528915405,
+      "learning_rate": 1.5846021258259124e-05,
+      "loss": 2.4614,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6085416210003732,
+      "grad_norm": 1.4205738306045532,
+      "learning_rate": 1.5834530307382938e-05,
+      "loss": 2.0676,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6088259862438313,
+      "grad_norm": 1.630874752998352,
+      "learning_rate": 1.5823039356506753e-05,
+      "loss": 2.286,
+      "step": 2141
+    },
+    {
+      "epoch": 0.6091103514872894,
+      "grad_norm": 1.5306339263916016,
+      "learning_rate": 1.5811548405630567e-05,
+      "loss": 2.0203,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6093947167307476,
+      "grad_norm": 1.5277224779129028,
+      "learning_rate": 1.5800057454754382e-05,
+      "loss": 1.8091,
+      "step": 2143
+    },
+    {
+      "epoch": 0.6096790819742057,
+      "grad_norm": 1.5925323963165283,
+      "learning_rate": 1.5788566503878196e-05,
+      "loss": 1.7929,
+      "step": 2144
+    },
+    {
+      "epoch": 0.6099634472176638,
+      "grad_norm": 1.7087624073028564,
+      "learning_rate": 1.5777075553002014e-05,
+      "loss": 2.8013,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6102478124611219,
+      "grad_norm": 1.4798260927200317,
+      "learning_rate": 1.576558460212583e-05,
+      "loss": 2.6274,
+      "step": 2146
+    },
+    {
+      "epoch": 0.61053217770458,
+      "grad_norm": 1.5136479139328003,
+      "learning_rate": 1.5754093651249643e-05,
+      "loss": 2.2147,
+      "step": 2147
+    },
+    {
+      "epoch": 0.6108165429480382,
+      "grad_norm": 1.4526771306991577,
+      "learning_rate": 1.5742602700373458e-05,
+      "loss": 2.3316,
+      "step": 2148
+    },
+    {
+      "epoch": 0.6111009081914963,
+      "grad_norm": 1.671984076499939,
+      "learning_rate": 1.5731111749497272e-05,
+      "loss": 2.0726,
+      "step": 2149
+    },
+    {
+      "epoch": 0.6113852734349544,
+      "grad_norm": 1.5075078010559082,
+      "learning_rate": 1.5719620798621087e-05,
+      "loss": 2.0302,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6116696386784125,
+      "grad_norm": 1.594140648841858,
+      "learning_rate": 1.57081298477449e-05,
+      "loss": 1.9607,
+      "step": 2151
+    },
+    {
+      "epoch": 0.6119540039218706,
+      "grad_norm": 1.7576320171356201,
+      "learning_rate": 1.5696638896868716e-05,
+      "loss": 1.9199,
+      "step": 2152
+    },
+    {
+      "epoch": 0.6122383691653288,
+      "grad_norm": 1.8919973373413086,
+      "learning_rate": 1.568514794599253e-05,
+      "loss": 2.8498,
+      "step": 2153
+    },
+    {
+      "epoch": 0.6125227344087869,
+      "grad_norm": 1.564929723739624,
+      "learning_rate": 1.5673656995116348e-05,
+      "loss": 2.2753,
+      "step": 2154
+    },
+    {
+      "epoch": 0.612807099652245,
+      "grad_norm": 1.453860878944397,
+      "learning_rate": 1.5662166044240162e-05,
+      "loss": 2.2765,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6130914648957031,
+      "grad_norm": 1.5218002796173096,
+      "learning_rate": 1.5650675093363977e-05,
+      "loss": 2.1331,
+      "step": 2156
+    },
+    {
+      "epoch": 0.6133758301391612,
+      "grad_norm": 1.625298261642456,
+      "learning_rate": 1.563918414248779e-05,
+      "loss": 2.1687,
+      "step": 2157
+    },
+    {
+      "epoch": 0.6136601953826194,
+      "grad_norm": 1.6260716915130615,
+      "learning_rate": 1.562769319161161e-05,
+      "loss": 1.9181,
+      "step": 2158
+    },
+    {
+      "epoch": 0.6139445606260775,
+      "grad_norm": 1.7083520889282227,
+      "learning_rate": 1.5616202240735424e-05,
+      "loss": 1.9319,
+      "step": 2159
+    },
+    {
+      "epoch": 0.6142289258695356,
+      "grad_norm": 1.6045297384262085,
+      "learning_rate": 1.560471128985924e-05,
+      "loss": 1.6412,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6145132911129937,
+      "grad_norm": 1.6960153579711914,
+      "learning_rate": 1.5593220338983053e-05,
+      "loss": 2.6496,
+      "step": 2161
+    },
+    {
+      "epoch": 0.6147976563564518,
+      "grad_norm": 1.5479867458343506,
+      "learning_rate": 1.5581729388106867e-05,
+      "loss": 2.3291,
+      "step": 2162
+    },
+    {
+      "epoch": 0.61508202159991,
+      "grad_norm": 1.4145187139511108,
+      "learning_rate": 1.5570238437230682e-05,
+      "loss": 2.2698,
+      "step": 2163
+    },
+    {
+      "epoch": 0.6153663868433681,
+      "grad_norm": 1.6084132194519043,
+      "learning_rate": 1.5558747486354496e-05,
+      "loss": 2.3009,
+      "step": 2164
+    },
+    {
+      "epoch": 0.6156507520868262,
+      "grad_norm": 1.6727561950683594,
+      "learning_rate": 1.554725653547831e-05,
+      "loss": 2.2078,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6159351173302843,
+      "grad_norm": 1.6095800399780273,
+      "learning_rate": 1.553576558460213e-05,
+      "loss": 1.9408,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6162194825737424,
+      "grad_norm": 1.6844741106033325,
+      "learning_rate": 1.5524274633725943e-05,
+      "loss": 1.9492,
+      "step": 2167
+    },
+    {
+      "epoch": 0.6165038478172006,
+      "grad_norm": 1.600579857826233,
+      "learning_rate": 1.5512783682849758e-05,
+      "loss": 1.6956,
+      "step": 2168
+    },
+    {
+      "epoch": 0.6167882130606587,
+      "grad_norm": 1.7465009689331055,
+      "learning_rate": 1.5501292731973572e-05,
+      "loss": 2.7124,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6170725783041168,
+      "grad_norm": 1.466794729232788,
+      "learning_rate": 1.5489801781097387e-05,
+      "loss": 2.2909,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6173569435475749,
+      "grad_norm": 1.493714451789856,
+      "learning_rate": 1.54783108302212e-05,
+      "loss": 2.1008,
+      "step": 2171
+    },
+    {
+      "epoch": 0.617641308791033,
+      "grad_norm": 1.5686396360397339,
+      "learning_rate": 1.5466819879345016e-05,
+      "loss": 2.1567,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6179256740344912,
+      "grad_norm": 1.6050747632980347,
+      "learning_rate": 1.545532892846883e-05,
+      "loss": 2.2713,
+      "step": 2173
+    },
+    {
+      "epoch": 0.6182100392779493,
+      "grad_norm": 1.5931973457336426,
+      "learning_rate": 1.5443837977592645e-05,
+      "loss": 1.797,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6184944045214074,
+      "grad_norm": 1.662131667137146,
+      "learning_rate": 1.5432347026716463e-05,
+      "loss": 1.7863,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6187787697648655,
+      "grad_norm": 1.6877639293670654,
+      "learning_rate": 1.5420856075840277e-05,
+      "loss": 1.9392,
+      "step": 2176
+    },
+    {
+      "epoch": 0.6190631350083237,
+      "grad_norm": 1.8865476846694946,
+      "learning_rate": 1.540936512496409e-05,
+      "loss": 2.7452,
+      "step": 2177
+    },
+    {
+      "epoch": 0.6193475002517818,
+      "grad_norm": 1.5432252883911133,
+      "learning_rate": 1.539787417408791e-05,
+      "loss": 2.5342,
+      "step": 2178
+    },
+    {
+      "epoch": 0.6196318654952399,
+      "grad_norm": 1.513899564743042,
+      "learning_rate": 1.5386383223211724e-05,
+      "loss": 2.1821,
+      "step": 2179
+    },
+    {
+      "epoch": 0.619916230738698,
+      "grad_norm": 1.5674052238464355,
+      "learning_rate": 1.537489227233554e-05,
+      "loss": 2.2672,
+      "step": 2180
+    },
+    {
+      "epoch": 0.620200595982156,
+      "grad_norm": 1.5859272480010986,
+      "learning_rate": 1.5363401321459353e-05,
+      "loss": 2.1825,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6204849612256143,
+      "grad_norm": 1.5028396844863892,
+      "learning_rate": 1.5351910370583167e-05,
+      "loss": 1.7984,
+      "step": 2182
+    },
+    {
+      "epoch": 0.6207693264690723,
+      "grad_norm": 1.5256812572479248,
+      "learning_rate": 1.5340419419706982e-05,
+      "loss": 1.8812,
+      "step": 2183
+    },
+    {
+      "epoch": 0.6210536917125304,
+      "grad_norm": 1.6204417943954468,
+      "learning_rate": 1.5328928468830796e-05,
+      "loss": 1.8866,
+      "step": 2184
+    },
+    {
+      "epoch": 0.6213380569559885,
+      "grad_norm": 1.6233776807785034,
+      "learning_rate": 1.531743751795461e-05,
+      "loss": 2.7469,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6216224221994466,
+      "grad_norm": 1.5075547695159912,
+      "learning_rate": 1.5305946567078425e-05,
+      "loss": 2.4263,
+      "step": 2186
+    },
+    {
+      "epoch": 0.6219067874429048,
+      "grad_norm": 1.5008249282836914,
+      "learning_rate": 1.5294455616202243e-05,
+      "loss": 2.1231,
+      "step": 2187
+    },
+    {
+      "epoch": 0.6221911526863629,
+      "grad_norm": 1.5601019859313965,
+      "learning_rate": 1.5282964665326058e-05,
+      "loss": 2.3066,
+      "step": 2188
+    },
+    {
+      "epoch": 0.622475517929821,
+      "grad_norm": 1.6898422241210938,
+      "learning_rate": 1.5271473714449872e-05,
+      "loss": 2.2774,
+      "step": 2189
+    },
+    {
+      "epoch": 0.6227598831732791,
+      "grad_norm": 1.5225181579589844,
+      "learning_rate": 1.5259982763573687e-05,
+      "loss": 2.0043,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6230442484167372,
+      "grad_norm": 1.565322756767273,
+      "learning_rate": 1.5248491812697501e-05,
+      "loss": 1.7469,
+      "step": 2191
+    },
+    {
+      "epoch": 0.6233286136601954,
+      "grad_norm": 1.6591078042984009,
+      "learning_rate": 1.5237000861821316e-05,
+      "loss": 1.4503,
+      "step": 2192
+    },
+    {
+      "epoch": 0.6236129789036535,
+      "grad_norm": 1.8940781354904175,
+      "learning_rate": 1.522550991094513e-05,
+      "loss": 2.7695,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6238973441471116,
+      "grad_norm": 1.5857594013214111,
+      "learning_rate": 1.5214018960068946e-05,
+      "loss": 2.3266,
+      "step": 2194
+    },
+    {
+      "epoch": 0.6241817093905697,
+      "grad_norm": 1.531599998474121,
+      "learning_rate": 1.5202528009192763e-05,
+      "loss": 2.3562,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6244660746340278,
+      "grad_norm": 1.4814765453338623,
+      "learning_rate": 1.5191037058316577e-05,
+      "loss": 2.1034,
+      "step": 2196
+    },
+    {
+      "epoch": 0.624750439877486,
+      "grad_norm": 1.5699845552444458,
+      "learning_rate": 1.5179546107440393e-05,
+      "loss": 2.18,
+      "step": 2197
+    },
+    {
+      "epoch": 0.6250348051209441,
+      "grad_norm": 1.572761058807373,
+      "learning_rate": 1.5168055156564208e-05,
+      "loss": 1.9689,
+      "step": 2198
+    },
+    {
+      "epoch": 0.6253191703644022,
+      "grad_norm": 1.5882624387741089,
+      "learning_rate": 1.5156564205688022e-05,
+      "loss": 1.9178,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6256035356078603,
+      "grad_norm": 1.979588508605957,
+      "learning_rate": 1.5145073254811837e-05,
+      "loss": 1.8587,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6258879008513184,
+      "grad_norm": 1.7905445098876953,
+      "learning_rate": 1.5133582303935653e-05,
+      "loss": 2.6646,
+      "step": 2201
+    },
+    {
+      "epoch": 0.6261722660947766,
+      "grad_norm": 1.5020902156829834,
+      "learning_rate": 1.5122091353059468e-05,
+      "loss": 2.4876,
+      "step": 2202
+    },
+    {
+      "epoch": 0.6264566313382347,
+      "grad_norm": 1.551763653755188,
+      "learning_rate": 1.5110600402183282e-05,
+      "loss": 2.347,
+      "step": 2203
+    },
+    {
+      "epoch": 0.6267409965816928,
+      "grad_norm": 1.5747159719467163,
+      "learning_rate": 1.5099109451307097e-05,
+      "loss": 2.3735,
+      "step": 2204
+    },
+    {
+      "epoch": 0.6270253618251509,
+      "grad_norm": 1.567586064338684,
+      "learning_rate": 1.5087618500430911e-05,
+      "loss": 2.1383,
+      "step": 2205
+    },
+    {
+      "epoch": 0.627309727068609,
+      "grad_norm": 1.5105654001235962,
+      "learning_rate": 1.5076127549554727e-05,
+      "loss": 1.7974,
+      "step": 2206
+    },
+    {
+      "epoch": 0.6275940923120672,
+      "grad_norm": 1.600792646408081,
+      "learning_rate": 1.5064636598678542e-05,
+      "loss": 1.8991,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6278784575555253,
+      "grad_norm": 1.6282237768173218,
+      "learning_rate": 1.5053145647802356e-05,
+      "loss": 1.7965,
+      "step": 2208
+    },
+    {
+      "epoch": 0.6281628227989834,
+      "grad_norm": 1.7487695217132568,
+      "learning_rate": 1.504165469692617e-05,
+      "loss": 2.6253,
+      "step": 2209
+    },
+    {
+      "epoch": 0.6284471880424415,
+      "grad_norm": 1.627800703048706,
+      "learning_rate": 1.5030163746049987e-05,
+      "loss": 2.3377,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6287315532858996,
+      "grad_norm": 1.490409255027771,
+      "learning_rate": 1.5018672795173801e-05,
+      "loss": 2.1716,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6290159185293578,
+      "grad_norm": 1.5944234132766724,
+      "learning_rate": 1.5007181844297616e-05,
+      "loss": 2.2221,
+      "step": 2212
+    },
+    {
+      "epoch": 0.6293002837728159,
+      "grad_norm": 1.5915591716766357,
+      "learning_rate": 1.4995690893421432e-05,
+      "loss": 2.2536,
+      "step": 2213
+    },
+    {
+      "epoch": 0.629584649016274,
+      "grad_norm": 1.5468132495880127,
+      "learning_rate": 1.4984199942545248e-05,
+      "loss": 2.0741,
+      "step": 2214
+    },
+    {
+      "epoch": 0.6298690142597321,
+      "grad_norm": 1.5128260850906372,
+      "learning_rate": 1.4972708991669063e-05,
+      "loss": 1.7709,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6301533795031902,
+      "grad_norm": 1.6346701383590698,
+      "learning_rate": 1.4961218040792877e-05,
+      "loss": 1.7228,
+      "step": 2216
+    },
+    {
+      "epoch": 0.6304377447466484,
+      "grad_norm": 1.812436580657959,
+      "learning_rate": 1.4949727089916692e-05,
+      "loss": 2.7526,
+      "step": 2217
+    },
+    {
+      "epoch": 0.6307221099901065,
+      "grad_norm": 1.562059998512268,
+      "learning_rate": 1.4938236139040508e-05,
+      "loss": 2.4138,
+      "step": 2218
+    },
+    {
+      "epoch": 0.6310064752335646,
+      "grad_norm": 1.6043922901153564,
+      "learning_rate": 1.4926745188164322e-05,
+      "loss": 2.2918,
+      "step": 2219
+    },
+    {
+      "epoch": 0.6312908404770227,
+      "grad_norm": 1.5691092014312744,
+      "learning_rate": 1.4915254237288137e-05,
+      "loss": 2.2282,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6315752057204808,
+      "grad_norm": 1.5854427814483643,
+      "learning_rate": 1.4903763286411951e-05,
+      "loss": 2.2005,
+      "step": 2221
+    },
+    {
+      "epoch": 0.631859570963939,
+      "grad_norm": 1.552298903465271,
+      "learning_rate": 1.4892272335535766e-05,
+      "loss": 1.7499,
+      "step": 2222
+    },
+    {
+      "epoch": 0.6321439362073971,
+      "grad_norm": 1.7358125448226929,
+      "learning_rate": 1.4880781384659582e-05,
+      "loss": 1.886,
+      "step": 2223
+    },
+    {
+      "epoch": 0.6324283014508552,
+      "grad_norm": 1.6180943250656128,
+      "learning_rate": 1.4869290433783397e-05,
+      "loss": 1.7801,
+      "step": 2224
+    },
+    {
+      "epoch": 0.6327126666943133,
+      "grad_norm": 1.9027990102767944,
+      "learning_rate": 1.4857799482907211e-05,
+      "loss": 2.6241,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6329970319377715,
+      "grad_norm": 1.5672250986099243,
+      "learning_rate": 1.4846308532031026e-05,
+      "loss": 2.4785,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6332813971812296,
+      "grad_norm": 1.5182520151138306,
+      "learning_rate": 1.4834817581154842e-05,
+      "loss": 2.0982,
+      "step": 2227
+    },
+    {
+      "epoch": 0.6335657624246877,
+      "grad_norm": 1.4269053936004639,
+      "learning_rate": 1.4823326630278656e-05,
+      "loss": 2.3107,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6338501276681457,
+      "grad_norm": 1.6807674169540405,
+      "learning_rate": 1.481183567940247e-05,
+      "loss": 2.0994,
+      "step": 2229
+    },
+    {
+      "epoch": 0.6341344929116038,
+      "grad_norm": 1.5306531190872192,
+      "learning_rate": 1.4800344728526285e-05,
+      "loss": 1.8668,
+      "step": 2230
+    },
+    {
+      "epoch": 0.634418858155062,
+      "grad_norm": 1.683131456375122,
+      "learning_rate": 1.47888537776501e-05,
+      "loss": 1.7841,
+      "step": 2231
+    },
+    {
+      "epoch": 0.6347032233985201,
+      "grad_norm": 1.6257297992706299,
+      "learning_rate": 1.4777362826773918e-05,
+      "loss": 1.7767,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6349875886419782,
+      "grad_norm": 1.8746527433395386,
+      "learning_rate": 1.4765871875897732e-05,
+      "loss": 2.8278,
+      "step": 2233
+    },
+    {
+      "epoch": 0.6352719538854363,
+      "grad_norm": 1.5843089818954468,
+      "learning_rate": 1.4754380925021547e-05,
+      "loss": 2.3801,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6355563191288944,
+      "grad_norm": 1.465401291847229,
+      "learning_rate": 1.4742889974145363e-05,
+      "loss": 2.0188,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6358406843723526,
+      "grad_norm": 1.4717556238174438,
+      "learning_rate": 1.4731399023269177e-05,
+      "loss": 2.1638,
+      "step": 2236
+    },
+    {
+      "epoch": 0.6361250496158107,
+      "grad_norm": 1.733504056930542,
+      "learning_rate": 1.4719908072392992e-05,
+      "loss": 2.2512,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6364094148592688,
+      "grad_norm": 1.602497935295105,
+      "learning_rate": 1.4708417121516806e-05,
+      "loss": 1.9198,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6366937801027269,
+      "grad_norm": 1.6287175416946411,
+      "learning_rate": 1.4696926170640623e-05,
+      "loss": 1.8728,
+      "step": 2239
+    },
+    {
+      "epoch": 0.636978145346185,
+      "grad_norm": 1.640389084815979,
+      "learning_rate": 1.4685435219764437e-05,
+      "loss": 1.6453,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6372625105896432,
+      "grad_norm": 1.8420612812042236,
+      "learning_rate": 1.4673944268888252e-05,
+      "loss": 2.6964,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6375468758331013,
+      "grad_norm": 1.4961106777191162,
+      "learning_rate": 1.4662453318012066e-05,
+      "loss": 2.2855,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6378312410765594,
+      "grad_norm": 1.5498943328857422,
+      "learning_rate": 1.465096236713588e-05,
+      "loss": 2.3792,
+      "step": 2243
+    },
+    {
+      "epoch": 0.6381156063200175,
+      "grad_norm": 1.5839142799377441,
+      "learning_rate": 1.4639471416259697e-05,
+      "loss": 2.2614,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6383999715634756,
+      "grad_norm": 1.598946452140808,
+      "learning_rate": 1.4627980465383511e-05,
+      "loss": 2.2075,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6386843368069338,
+      "grad_norm": 1.521291732788086,
+      "learning_rate": 1.4616489514507326e-05,
+      "loss": 2.1571,
+      "step": 2246
+    },
+    {
+      "epoch": 0.6389687020503919,
+      "grad_norm": 1.5280981063842773,
+      "learning_rate": 1.460499856363114e-05,
+      "loss": 1.6476,
+      "step": 2247
+    },
+    {
+      "epoch": 0.63925306729385,
+      "grad_norm": 1.6235344409942627,
+      "learning_rate": 1.4593507612754956e-05,
+      "loss": 1.8018,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6395374325373081,
+      "grad_norm": 1.906637191772461,
+      "learning_rate": 1.4582016661878771e-05,
+      "loss": 2.8116,
+      "step": 2249
+    },
+    {
+      "epoch": 0.6398217977807662,
+      "grad_norm": 1.4746451377868652,
+      "learning_rate": 1.4570525711002587e-05,
+      "loss": 2.4155,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6401061630242244,
+      "grad_norm": 1.536787748336792,
+      "learning_rate": 1.4559034760126403e-05,
+      "loss": 2.2476,
+      "step": 2251
+    },
+    {
+      "epoch": 0.6403905282676825,
+      "grad_norm": 1.4829517602920532,
+      "learning_rate": 1.4547543809250218e-05,
+      "loss": 2.1656,
+      "step": 2252
+    },
+    {
+      "epoch": 0.6406748935111406,
+      "grad_norm": 1.6278111934661865,
+      "learning_rate": 1.4536052858374032e-05,
+      "loss": 2.025,
+      "step": 2253
+    },
+    {
+      "epoch": 0.6409592587545987,
+      "grad_norm": 1.5534253120422363,
+      "learning_rate": 1.4524561907497847e-05,
+      "loss": 1.9709,
+      "step": 2254
+    },
+    {
+      "epoch": 0.6412436239980568,
+      "grad_norm": 1.5600004196166992,
+      "learning_rate": 1.4513070956621661e-05,
+      "loss": 1.8011,
+      "step": 2255
+    },
+    {
+      "epoch": 0.641527989241515,
+      "grad_norm": 1.7009464502334595,
+      "learning_rate": 1.4501580005745477e-05,
+      "loss": 1.7419,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6418123544849731,
+      "grad_norm": 1.6545665264129639,
+      "learning_rate": 1.4490089054869292e-05,
+      "loss": 2.6367,
+      "step": 2257
+    },
+    {
+      "epoch": 0.6420967197284312,
+      "grad_norm": 1.542686939239502,
+      "learning_rate": 1.4478598103993106e-05,
+      "loss": 2.5222,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6423810849718893,
+      "grad_norm": 1.5785605907440186,
+      "learning_rate": 1.4467107153116921e-05,
+      "loss": 2.2131,
+      "step": 2259
+    },
+    {
+      "epoch": 0.6426654502153474,
+      "grad_norm": 1.6065642833709717,
+      "learning_rate": 1.4455616202240737e-05,
+      "loss": 2.1774,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6429498154588056,
+      "grad_norm": 1.5818967819213867,
+      "learning_rate": 1.4444125251364552e-05,
+      "loss": 2.0904,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6432341807022637,
+      "grad_norm": 1.6355520486831665,
+      "learning_rate": 1.4432634300488366e-05,
+      "loss": 1.8927,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6435185459457218,
+      "grad_norm": 1.528612732887268,
+      "learning_rate": 1.442114334961218e-05,
+      "loss": 1.8309,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6438029111891799,
+      "grad_norm": 1.5770155191421509,
+      "learning_rate": 1.4409652398735995e-05,
+      "loss": 1.6619,
+      "step": 2264
+    },
+    {
+      "epoch": 0.644087276432638,
+      "grad_norm": 1.7672600746154785,
+      "learning_rate": 1.4398161447859811e-05,
+      "loss": 2.5805,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6443716416760962,
+      "grad_norm": 1.5915486812591553,
+      "learning_rate": 1.4386670496983626e-05,
+      "loss": 2.2712,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6446560069195543,
+      "grad_norm": 1.3617949485778809,
+      "learning_rate": 1.437517954610744e-05,
+      "loss": 2.1554,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6449403721630124,
+      "grad_norm": 1.487271785736084,
+      "learning_rate": 1.4363688595231255e-05,
+      "loss": 2.3376,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6452247374064705,
+      "grad_norm": 1.9369791746139526,
+      "learning_rate": 1.4352197644355073e-05,
+      "loss": 2.1762,
+      "step": 2269
+    },
+    {
+      "epoch": 0.6455091026499287,
+      "grad_norm": 1.513728380203247,
+      "learning_rate": 1.4340706693478887e-05,
+      "loss": 1.8318,
+      "step": 2270
+    },
+    {
+      "epoch": 0.6457934678933868,
+      "grad_norm": 1.602360725402832,
+      "learning_rate": 1.4329215742602702e-05,
+      "loss": 1.8081,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6460778331368449,
+      "grad_norm": 1.706312656402588,
+      "learning_rate": 1.4317724791726518e-05,
+      "loss": 1.8961,
+      "step": 2272
+    },
+    {
+      "epoch": 0.646362198380303,
+      "grad_norm": 1.8679949045181274,
+      "learning_rate": 1.4306233840850332e-05,
+      "loss": 2.7414,
+      "step": 2273
+    },
+    {
+      "epoch": 0.646646563623761,
+      "grad_norm": 1.5807439088821411,
+      "learning_rate": 1.4294742889974147e-05,
+      "loss": 2.4749,
+      "step": 2274
+    },
+    {
+      "epoch": 0.6469309288672193,
+      "grad_norm": 1.6467163562774658,
+      "learning_rate": 1.4283251939097961e-05,
+      "loss": 2.4233,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6472152941106774,
+      "grad_norm": 1.611040711402893,
+      "learning_rate": 1.4271760988221776e-05,
+      "loss": 2.148,
+      "step": 2276
+    },
+    {
+      "epoch": 0.6474996593541354,
+      "grad_norm": 1.6380242109298706,
+      "learning_rate": 1.4260270037345592e-05,
+      "loss": 2.1739,
+      "step": 2277
+    },
+    {
+      "epoch": 0.6477840245975935,
+      "grad_norm": 1.6218241453170776,
+      "learning_rate": 1.4248779086469407e-05,
+      "loss": 2.0744,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6480683898410516,
+      "grad_norm": 1.4787720441818237,
+      "learning_rate": 1.4237288135593221e-05,
+      "loss": 1.6869,
+      "step": 2279
+    },
+    {
+      "epoch": 0.6483527550845098,
+      "grad_norm": 1.682263731956482,
+      "learning_rate": 1.4225797184717036e-05,
+      "loss": 1.7572,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6486371203279679,
+      "grad_norm": 1.9067274332046509,
+      "learning_rate": 1.4214306233840852e-05,
+      "loss": 2.7824,
+      "step": 2281
+    },
+    {
+      "epoch": 0.648921485571426,
+      "grad_norm": 1.4960787296295166,
+      "learning_rate": 1.4202815282964666e-05,
+      "loss": 2.3718,
+      "step": 2282
+    },
+    {
+      "epoch": 0.6492058508148841,
+      "grad_norm": 1.5059878826141357,
+      "learning_rate": 1.419132433208848e-05,
+      "loss": 2.1535,
+      "step": 2283
+    },
+    {
+      "epoch": 0.6494902160583422,
+      "grad_norm": 1.685180902481079,
+      "learning_rate": 1.4179833381212295e-05,
+      "loss": 2.2137,
+      "step": 2284
+    },
+    {
+      "epoch": 0.6497745813018004,
+      "grad_norm": 1.5441550016403198,
+      "learning_rate": 1.416834243033611e-05,
+      "loss": 2.3031,
+      "step": 2285
+    },
+    {
+      "epoch": 0.6500589465452585,
+      "grad_norm": 1.5095455646514893,
+      "learning_rate": 1.4156851479459926e-05,
+      "loss": 1.8482,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6503433117887166,
+      "grad_norm": 1.5256438255310059,
+      "learning_rate": 1.4145360528583742e-05,
+      "loss": 1.7711,
+      "step": 2287
+    },
+    {
+      "epoch": 0.6506276770321747,
+      "grad_norm": 1.6489698886871338,
+      "learning_rate": 1.4133869577707557e-05,
+      "loss": 1.64,
+      "step": 2288
+    },
+    {
+      "epoch": 0.6509120422756328,
+      "grad_norm": 1.7157350778579712,
+      "learning_rate": 1.4122378626831373e-05,
+      "loss": 2.8341,
+      "step": 2289
+    },
+    {
+      "epoch": 0.651196407519091,
+      "grad_norm": 1.5686113834381104,
+      "learning_rate": 1.4110887675955187e-05,
+      "loss": 2.6388,
+      "step": 2290
+    },
+    {
+      "epoch": 0.6514807727625491,
+      "grad_norm": 1.576197624206543,
+      "learning_rate": 1.4099396725079002e-05,
+      "loss": 2.2588,
+      "step": 2291
+    },
+    {
+      "epoch": 0.6517651380060072,
+      "grad_norm": 1.5347895622253418,
+      "learning_rate": 1.4087905774202816e-05,
+      "loss": 2.1164,
+      "step": 2292
+    },
+    {
+      "epoch": 0.6520495032494653,
+      "grad_norm": 1.6401748657226562,
+      "learning_rate": 1.4076414823326633e-05,
+      "loss": 2.1258,
+      "step": 2293
+    },
+    {
+      "epoch": 0.6523338684929234,
+      "grad_norm": 1.5072190761566162,
+      "learning_rate": 1.4064923872450447e-05,
+      "loss": 1.9321,
+      "step": 2294
+    },
+    {
+      "epoch": 0.6526182337363816,
+      "grad_norm": 1.5900171995162964,
+      "learning_rate": 1.4053432921574262e-05,
+      "loss": 1.8916,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6529025989798397,
+      "grad_norm": 1.6210306882858276,
+      "learning_rate": 1.4041941970698076e-05,
+      "loss": 1.6101,
+      "step": 2296
+    },
+    {
+      "epoch": 0.6531869642232978,
+      "grad_norm": 1.8477280139923096,
+      "learning_rate": 1.403045101982189e-05,
+      "loss": 2.6045,
+      "step": 2297
+    },
+    {
+      "epoch": 0.6534713294667559,
+      "grad_norm": 1.5514870882034302,
+      "learning_rate": 1.4018960068945707e-05,
+      "loss": 2.4303,
+      "step": 2298
+    },
+    {
+      "epoch": 0.653755694710214,
+      "grad_norm": 1.5255701541900635,
+      "learning_rate": 1.4007469118069521e-05,
+      "loss": 2.3312,
+      "step": 2299
+    },
+    {
+      "epoch": 0.6540400599536722,
+      "grad_norm": 1.5493617057800293,
+      "learning_rate": 1.3995978167193336e-05,
+      "loss": 2.3112,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6543244251971303,
+      "grad_norm": 1.477663278579712,
+      "learning_rate": 1.398448721631715e-05,
+      "loss": 2.0823,
+      "step": 2301
+    },
+    {
+      "epoch": 0.6546087904405884,
+      "grad_norm": 1.534895420074463,
+      "learning_rate": 1.3972996265440966e-05,
+      "loss": 1.9711,
+      "step": 2302
+    },
+    {
+      "epoch": 0.6548931556840465,
+      "grad_norm": 1.559821605682373,
+      "learning_rate": 1.3961505314564781e-05,
+      "loss": 1.9114,
+      "step": 2303
+    },
+    {
+      "epoch": 0.6551775209275046,
+      "grad_norm": 1.5802251100540161,
+      "learning_rate": 1.3950014363688595e-05,
+      "loss": 1.7692,
+      "step": 2304
+    },
+    {
+      "epoch": 0.6554618861709628,
+      "grad_norm": 1.8919142484664917,
+      "learning_rate": 1.393852341281241e-05,
+      "loss": 2.8514,
+      "step": 2305
+    },
+    {
+      "epoch": 0.6557462514144209,
+      "grad_norm": 1.5178377628326416,
+      "learning_rate": 1.3927032461936228e-05,
+      "loss": 2.2836,
+      "step": 2306
+    },
+    {
+      "epoch": 0.656030616657879,
+      "grad_norm": 1.4845335483551025,
+      "learning_rate": 1.3915541511060042e-05,
+      "loss": 2.4409,
+      "step": 2307
+    },
+    {
+      "epoch": 0.6563149819013371,
+      "grad_norm": 1.55623459815979,
+      "learning_rate": 1.3904050560183857e-05,
+      "loss": 2.245,
+      "step": 2308
+    },
+    {
+      "epoch": 0.6565993471447952,
+      "grad_norm": 1.5961564779281616,
+      "learning_rate": 1.3892559609307671e-05,
+      "loss": 2.1436,
+      "step": 2309
+    },
+    {
+      "epoch": 0.6568837123882534,
+      "grad_norm": 1.593917727470398,
+      "learning_rate": 1.3881068658431487e-05,
+      "loss": 2.1096,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6571680776317115,
+      "grad_norm": 1.4851388931274414,
+      "learning_rate": 1.3869577707555302e-05,
+      "loss": 1.7558,
+      "step": 2311
+    },
+    {
+      "epoch": 0.6574524428751696,
+      "grad_norm": 1.7430486679077148,
+      "learning_rate": 1.3858086756679116e-05,
+      "loss": 1.7607,
+      "step": 2312
+    },
+    {
+      "epoch": 0.6577368081186277,
+      "grad_norm": 1.8176724910736084,
+      "learning_rate": 1.3846595805802931e-05,
+      "loss": 2.7509,
+      "step": 2313
+    },
+    {
+      "epoch": 0.6580211733620858,
+      "grad_norm": 1.513588309288025,
+      "learning_rate": 1.3835104854926747e-05,
+      "loss": 2.405,
+      "step": 2314
+    },
+    {
+      "epoch": 0.658305538605544,
+      "grad_norm": 1.621985912322998,
+      "learning_rate": 1.3823613904050562e-05,
+      "loss": 2.3489,
+      "step": 2315
+    },
+    {
+      "epoch": 0.6585899038490021,
+      "grad_norm": 1.4367046356201172,
+      "learning_rate": 1.3812122953174376e-05,
+      "loss": 2.2808,
+      "step": 2316
+    },
+    {
+      "epoch": 0.6588742690924602,
+      "grad_norm": 1.577368974685669,
+      "learning_rate": 1.380063200229819e-05,
+      "loss": 2.1214,
+      "step": 2317
+    },
+    {
+      "epoch": 0.6591586343359183,
+      "grad_norm": 1.4581912755966187,
+      "learning_rate": 1.3789141051422005e-05,
+      "loss": 2.0159,
+      "step": 2318
+    },
+    {
+      "epoch": 0.6594429995793765,
+      "grad_norm": 1.500917911529541,
+      "learning_rate": 1.3777650100545821e-05,
+      "loss": 1.7401,
+      "step": 2319
+    },
+    {
+      "epoch": 0.6597273648228346,
+      "grad_norm": 1.6946115493774414,
+      "learning_rate": 1.3766159149669636e-05,
+      "loss": 1.9459,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6600117300662927,
+      "grad_norm": 1.7423439025878906,
+      "learning_rate": 1.375466819879345e-05,
+      "loss": 2.8295,
+      "step": 2321
+    },
+    {
+      "epoch": 0.6602960953097508,
+      "grad_norm": 1.5395139455795288,
+      "learning_rate": 1.3743177247917265e-05,
+      "loss": 2.4043,
+      "step": 2322
+    },
+    {
+      "epoch": 0.6605804605532088,
+      "grad_norm": 1.6088899374008179,
+      "learning_rate": 1.3731686297041081e-05,
+      "loss": 2.396,
+      "step": 2323
+    },
+    {
+      "epoch": 0.660864825796667,
+      "grad_norm": 1.542049765586853,
+      "learning_rate": 1.3720195346164897e-05,
+      "loss": 2.2012,
+      "step": 2324
+    },
+    {
+      "epoch": 0.6611491910401252,
+      "grad_norm": 1.5389901399612427,
+      "learning_rate": 1.3708704395288712e-05,
+      "loss": 2.02,
+      "step": 2325
+    },
+    {
+      "epoch": 0.6614335562835832,
+      "grad_norm": 1.6238247156143188,
+      "learning_rate": 1.3697213444412528e-05,
+      "loss": 1.9875,
+      "step": 2326
+    },
+    {
+      "epoch": 0.6617179215270413,
+      "grad_norm": 1.5157362222671509,
+      "learning_rate": 1.3685722493536342e-05,
+      "loss": 1.7811,
+      "step": 2327
+    },
+    {
+      "epoch": 0.6620022867704994,
+      "grad_norm": 1.6152019500732422,
+      "learning_rate": 1.3674231542660157e-05,
+      "loss": 1.7777,
+      "step": 2328
+    },
+    {
+      "epoch": 0.6622866520139576,
+      "grad_norm": 1.740449070930481,
+      "learning_rate": 1.3662740591783971e-05,
+      "loss": 2.7021,
+      "step": 2329
+    },
+    {
+      "epoch": 0.6625710172574157,
+      "grad_norm": 1.5307239294052124,
+      "learning_rate": 1.3651249640907786e-05,
+      "loss": 2.3465,
+      "step": 2330
+    },
+    {
+      "epoch": 0.6628553825008738,
+      "grad_norm": 1.5852240324020386,
+      "learning_rate": 1.3639758690031602e-05,
+      "loss": 2.477,
+      "step": 2331
+    },
+    {
+      "epoch": 0.6631397477443319,
+      "grad_norm": 1.6011126041412354,
+      "learning_rate": 1.3628267739155417e-05,
+      "loss": 2.2583,
+      "step": 2332
+    },
+    {
+      "epoch": 0.66342411298779,
+      "grad_norm": 1.549425482749939,
+      "learning_rate": 1.3616776788279231e-05,
+      "loss": 2.2773,
+      "step": 2333
+    },
+    {
+      "epoch": 0.6637084782312482,
+      "grad_norm": 1.449255347251892,
+      "learning_rate": 1.3605285837403046e-05,
+      "loss": 1.7077,
+      "step": 2334
+    },
+    {
+      "epoch": 0.6639928434747063,
+      "grad_norm": 1.624582290649414,
+      "learning_rate": 1.3593794886526862e-05,
+      "loss": 1.7985,
+      "step": 2335
+    },
+    {
+      "epoch": 0.6642772087181644,
+      "grad_norm": 1.6954373121261597,
+      "learning_rate": 1.3582303935650676e-05,
+      "loss": 1.7301,
+      "step": 2336
+    },
+    {
+      "epoch": 0.6645615739616225,
+      "grad_norm": 1.8019691705703735,
+      "learning_rate": 1.357081298477449e-05,
+      "loss": 2.6443,
+      "step": 2337
+    },
+    {
+      "epoch": 0.6648459392050806,
+      "grad_norm": 1.6157490015029907,
+      "learning_rate": 1.3559322033898305e-05,
+      "loss": 2.248,
+      "step": 2338
+    },
+    {
+      "epoch": 0.6651303044485388,
+      "grad_norm": 1.6532044410705566,
+      "learning_rate": 1.354783108302212e-05,
+      "loss": 2.4353,
+      "step": 2339
+    },
+    {
+      "epoch": 0.6654146696919969,
+      "grad_norm": 1.4396045207977295,
+      "learning_rate": 1.3536340132145936e-05,
+      "loss": 2.1747,
+      "step": 2340
+    },
+    {
+      "epoch": 0.665699034935455,
+      "grad_norm": 1.4797991514205933,
+      "learning_rate": 1.352484918126975e-05,
+      "loss": 2.0431,
+      "step": 2341
+    },
+    {
+      "epoch": 0.6659834001789131,
+      "grad_norm": 1.3988265991210938,
+      "learning_rate": 1.3513358230393565e-05,
+      "loss": 1.858,
+      "step": 2342
+    },
+    {
+      "epoch": 0.6662677654223712,
+      "grad_norm": 1.6098911762237549,
+      "learning_rate": 1.3501867279517383e-05,
+      "loss": 1.735,
+      "step": 2343
+    },
+    {
+      "epoch": 0.6665521306658294,
+      "grad_norm": 1.6491178274154663,
+      "learning_rate": 1.3490376328641197e-05,
+      "loss": 1.6631,
+      "step": 2344
+    },
+    {
+      "epoch": 0.6668364959092875,
+      "grad_norm": 1.7857693433761597,
+      "learning_rate": 1.3478885377765012e-05,
+      "loss": 2.6628,
+      "step": 2345
+    },
+    {
+      "epoch": 0.6671208611527456,
+      "grad_norm": 1.5431671142578125,
+      "learning_rate": 1.3467394426888826e-05,
+      "loss": 2.4461,
+      "step": 2346
+    },
+    {
+      "epoch": 0.6674052263962037,
+      "grad_norm": 1.5646437406539917,
+      "learning_rate": 1.3455903476012642e-05,
+      "loss": 2.3102,
+      "step": 2347
+    },
+    {
+      "epoch": 0.6676895916396618,
+      "grad_norm": 1.4588571786880493,
+      "learning_rate": 1.3444412525136457e-05,
+      "loss": 2.3354,
+      "step": 2348
+    },
+    {
+      "epoch": 0.66797395688312,
+      "grad_norm": 1.5933592319488525,
+      "learning_rate": 1.3432921574260271e-05,
+      "loss": 2.2924,
+      "step": 2349
+    },
+    {
+      "epoch": 0.6682583221265781,
+      "grad_norm": 1.7044545412063599,
+      "learning_rate": 1.3421430623384086e-05,
+      "loss": 1.9148,
+      "step": 2350
+    },
+    {
+      "epoch": 0.6685426873700362,
+      "grad_norm": 1.6366909742355347,
+      "learning_rate": 1.34099396725079e-05,
+      "loss": 1.6891,
+      "step": 2351
+    },
+    {
+      "epoch": 0.6688270526134943,
+      "grad_norm": 1.6043225526809692,
+      "learning_rate": 1.3398448721631717e-05,
+      "loss": 1.895,
+      "step": 2352
+    },
+    {
+      "epoch": 0.6691114178569524,
+      "grad_norm": 1.8012562990188599,
+      "learning_rate": 1.3386957770755531e-05,
+      "loss": 2.7992,
+      "step": 2353
+    },
+    {
+      "epoch": 0.6693957831004106,
+      "grad_norm": 1.4551502466201782,
+      "learning_rate": 1.3375466819879346e-05,
+      "loss": 2.3551,
+      "step": 2354
+    },
+    {
+      "epoch": 0.6696801483438687,
+      "grad_norm": 1.5812779664993286,
+      "learning_rate": 1.336397586900316e-05,
+      "loss": 2.2085,
+      "step": 2355
+    },
+    {
+      "epoch": 0.6699645135873268,
+      "grad_norm": 1.529133677482605,
+      "learning_rate": 1.3352484918126976e-05,
+      "loss": 2.4194,
+      "step": 2356
+    },
+    {
+      "epoch": 0.6702488788307849,
+      "grad_norm": 1.595837116241455,
+      "learning_rate": 1.3340993967250791e-05,
+      "loss": 2.1076,
+      "step": 2357
+    },
+    {
+      "epoch": 0.670533244074243,
+      "grad_norm": 1.6244161128997803,
+      "learning_rate": 1.3329503016374605e-05,
+      "loss": 1.8285,
+      "step": 2358
+    },
+    {
+      "epoch": 0.6708176093177012,
+      "grad_norm": 1.5791698694229126,
+      "learning_rate": 1.331801206549842e-05,
+      "loss": 1.8041,
+      "step": 2359
+    },
+    {
+      "epoch": 0.6711019745611593,
+      "grad_norm": 1.7883336544036865,
+      "learning_rate": 1.3306521114622234e-05,
+      "loss": 1.7612,
+      "step": 2360
+    },
+    {
+      "epoch": 0.6713863398046174,
+      "grad_norm": 1.7180142402648926,
+      "learning_rate": 1.3295030163746052e-05,
+      "loss": 2.9109,
+      "step": 2361
+    },
+    {
+      "epoch": 0.6716707050480755,
+      "grad_norm": 1.4666606187820435,
+      "learning_rate": 1.3283539212869867e-05,
+      "loss": 2.3537,
+      "step": 2362
+    },
+    {
+      "epoch": 0.6719550702915336,
+      "grad_norm": 1.450372338294983,
+      "learning_rate": 1.3272048261993681e-05,
+      "loss": 2.4139,
+      "step": 2363
+    },
+    {
+      "epoch": 0.6722394355349918,
+      "grad_norm": 1.5204261541366577,
+      "learning_rate": 1.3260557311117497e-05,
+      "loss": 2.2118,
+      "step": 2364
+    },
+    {
+      "epoch": 0.6725238007784499,
+      "grad_norm": 1.5298410654067993,
+      "learning_rate": 1.3249066360241312e-05,
+      "loss": 2.2147,
+      "step": 2365
+    },
+    {
+      "epoch": 0.672808166021908,
+      "grad_norm": 1.5501554012298584,
+      "learning_rate": 1.3237575409365126e-05,
+      "loss": 1.9778,
+      "step": 2366
+    },
+    {
+      "epoch": 0.6730925312653661,
+      "grad_norm": 1.5184673070907593,
+      "learning_rate": 1.3226084458488941e-05,
+      "loss": 1.7726,
+      "step": 2367
+    },
+    {
+      "epoch": 0.6733768965088243,
+      "grad_norm": 1.6755380630493164,
+      "learning_rate": 1.3214593507612757e-05,
+      "loss": 1.7409,
+      "step": 2368
+    },
+    {
+      "epoch": 0.6736612617522824,
+      "grad_norm": 1.799485683441162,
+      "learning_rate": 1.3203102556736572e-05,
+      "loss": 2.698,
+      "step": 2369
+    },
+    {
+      "epoch": 0.6739456269957405,
+      "grad_norm": 1.6598178148269653,
+      "learning_rate": 1.3191611605860386e-05,
+      "loss": 2.3039,
+      "step": 2370
+    },
+    {
+      "epoch": 0.6742299922391986,
+      "grad_norm": 1.4246594905853271,
+      "learning_rate": 1.31801206549842e-05,
+      "loss": 2.1442,
+      "step": 2371
+    },
+    {
+      "epoch": 0.6745143574826566,
+      "grad_norm": 1.5940912961959839,
+      "learning_rate": 1.3168629704108015e-05,
+      "loss": 2.3078,
+      "step": 2372
+    },
+    {
+      "epoch": 0.6747987227261149,
+      "grad_norm": 1.6379321813583374,
+      "learning_rate": 1.3157138753231831e-05,
+      "loss": 2.1717,
+      "step": 2373
+    },
+    {
+      "epoch": 0.675083087969573,
+      "grad_norm": 1.4846510887145996,
+      "learning_rate": 1.3145647802355646e-05,
+      "loss": 1.9535,
+      "step": 2374
+    },
+    {
+      "epoch": 0.675367453213031,
+      "grad_norm": 1.5312621593475342,
+      "learning_rate": 1.313415685147946e-05,
+      "loss": 1.7979,
+      "step": 2375
+    },
+    {
+      "epoch": 0.6756518184564891,
+      "grad_norm": 1.6713464260101318,
+      "learning_rate": 1.3122665900603275e-05,
+      "loss": 1.9716,
+      "step": 2376
+    },
+    {
+      "epoch": 0.6759361836999472,
+      "grad_norm": 1.6956110000610352,
+      "learning_rate": 1.3111174949727091e-05,
+      "loss": 2.5694,
+      "step": 2377
+    },
+    {
+      "epoch": 0.6762205489434054,
+      "grad_norm": 1.5432319641113281,
+      "learning_rate": 1.3099683998850905e-05,
+      "loss": 2.3722,
+      "step": 2378
+    },
+    {
+      "epoch": 0.6765049141868635,
+      "grad_norm": 1.5072269439697266,
+      "learning_rate": 1.308819304797472e-05,
+      "loss": 2.347,
+      "step": 2379
+    },
+    {
+      "epoch": 0.6767892794303216,
+      "grad_norm": 1.498325228691101,
+      "learning_rate": 1.3076702097098536e-05,
+      "loss": 2.2973,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6770736446737797,
+      "grad_norm": 1.5500154495239258,
+      "learning_rate": 1.3065211146222352e-05,
+      "loss": 2.1524,
+      "step": 2381
+    },
+    {
+      "epoch": 0.6773580099172378,
+      "grad_norm": 1.6011747121810913,
+      "learning_rate": 1.3053720195346167e-05,
+      "loss": 1.7051,
+      "step": 2382
+    },
+    {
+      "epoch": 0.677642375160696,
+      "grad_norm": 1.5794252157211304,
+      "learning_rate": 1.3042229244469981e-05,
+      "loss": 1.6992,
+      "step": 2383
+    },
+    {
+      "epoch": 0.6779267404041541,
+      "grad_norm": 1.5260065793991089,
+      "learning_rate": 1.3030738293593796e-05,
+      "loss": 1.7597,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6782111056476122,
+      "grad_norm": 1.744638204574585,
+      "learning_rate": 1.3019247342717612e-05,
+      "loss": 2.9247,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6784954708910703,
+      "grad_norm": 1.6044799089431763,
+      "learning_rate": 1.3007756391841427e-05,
+      "loss": 2.3649,
+      "step": 2386
+    },
+    {
+      "epoch": 0.6787798361345284,
+      "grad_norm": 1.4609475135803223,
+      "learning_rate": 1.2996265440965241e-05,
+      "loss": 2.2682,
+      "step": 2387
+    },
+    {
+      "epoch": 0.6790642013779866,
+      "grad_norm": 1.4124441146850586,
+      "learning_rate": 1.2984774490089056e-05,
+      "loss": 2.2928,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6793485666214447,
+      "grad_norm": 1.5999109745025635,
+      "learning_rate": 1.297328353921287e-05,
+      "loss": 2.0884,
+      "step": 2389
+    },
+    {
+      "epoch": 0.6796329318649028,
+      "grad_norm": 1.541373372077942,
+      "learning_rate": 1.2961792588336686e-05,
+      "loss": 1.8306,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6799172971083609,
+      "grad_norm": 1.6004866361618042,
+      "learning_rate": 1.29503016374605e-05,
+      "loss": 1.7566,
+      "step": 2391
+    },
+    {
+      "epoch": 0.680201662351819,
+      "grad_norm": 1.6891027688980103,
+      "learning_rate": 1.2938810686584315e-05,
+      "loss": 1.5669,
+      "step": 2392
+    },
+    {
+      "epoch": 0.6804860275952772,
+      "grad_norm": 1.9808518886566162,
+      "learning_rate": 1.292731973570813e-05,
+      "loss": 2.8488,
+      "step": 2393
+    },
+    {
+      "epoch": 0.6807703928387353,
+      "grad_norm": 1.5064984560012817,
+      "learning_rate": 1.2915828784831946e-05,
+      "loss": 2.449,
+      "step": 2394
+    },
+    {
+      "epoch": 0.6810547580821934,
+      "grad_norm": 1.4997645616531372,
+      "learning_rate": 1.290433783395576e-05,
+      "loss": 2.2506,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6813391233256515,
+      "grad_norm": 1.476811170578003,
+      "learning_rate": 1.2892846883079575e-05,
+      "loss": 2.0701,
+      "step": 2396
+    },
+    {
+      "epoch": 0.6816234885691096,
+      "grad_norm": 1.593111515045166,
+      "learning_rate": 1.288135593220339e-05,
+      "loss": 2.0792,
+      "step": 2397
+    },
+    {
+      "epoch": 0.6819078538125678,
+      "grad_norm": 1.5982271432876587,
+      "learning_rate": 1.2869864981327207e-05,
+      "loss": 1.815,
+      "step": 2398
+    },
+    {
+      "epoch": 0.6821922190560259,
+      "grad_norm": 1.615859866142273,
+      "learning_rate": 1.2858374030451022e-05,
+      "loss": 1.8881,
+      "step": 2399
+    },
+    {
+      "epoch": 0.682476584299484,
+      "grad_norm": 1.6525177955627441,
+      "learning_rate": 1.2846883079574836e-05,
+      "loss": 1.9245,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6827609495429421,
+      "grad_norm": 1.8938461542129517,
+      "learning_rate": 1.283539212869865e-05,
+      "loss": 2.9311,
+      "step": 2401
+    },
+    {
+      "epoch": 0.6830453147864002,
+      "grad_norm": 1.5215036869049072,
+      "learning_rate": 1.2823901177822467e-05,
+      "loss": 2.1695,
+      "step": 2402
+    },
+    {
+      "epoch": 0.6833296800298584,
+      "grad_norm": 1.4118622541427612,
+      "learning_rate": 1.2812410226946281e-05,
+      "loss": 2.3208,
+      "step": 2403
+    },
+    {
+      "epoch": 0.6836140452733165,
+      "grad_norm": 1.486366868019104,
+      "learning_rate": 1.2800919276070096e-05,
+      "loss": 2.3337,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6838984105167746,
+      "grad_norm": 1.5969897508621216,
+      "learning_rate": 1.278942832519391e-05,
+      "loss": 2.1099,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6841827757602327,
+      "grad_norm": 1.4879251718521118,
+      "learning_rate": 1.2777937374317727e-05,
+      "loss": 1.9579,
+      "step": 2406
+    },
+    {
+      "epoch": 0.6844671410036908,
+      "grad_norm": 1.6099344491958618,
+      "learning_rate": 1.2766446423441541e-05,
+      "loss": 2.0261,
+      "step": 2407
+    },
+    {
+      "epoch": 0.684751506247149,
+      "grad_norm": 1.8013091087341309,
+      "learning_rate": 1.2754955472565356e-05,
+      "loss": 1.7967,
+      "step": 2408
+    },
+    {
+      "epoch": 0.6850358714906071,
+      "grad_norm": 1.6463696956634521,
+      "learning_rate": 1.274346452168917e-05,
+      "loss": 2.7894,
+      "step": 2409
+    },
+    {
+      "epoch": 0.6853202367340652,
+      "grad_norm": 1.5260536670684814,
+      "learning_rate": 1.2731973570812985e-05,
+      "loss": 2.5281,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6856046019775233,
+      "grad_norm": 1.5363825559616089,
+      "learning_rate": 1.27204826199368e-05,
+      "loss": 1.9302,
+      "step": 2411
+    },
+    {
+      "epoch": 0.6858889672209814,
+      "grad_norm": 1.5252505540847778,
+      "learning_rate": 1.2708991669060615e-05,
+      "loss": 2.3634,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6861733324644396,
+      "grad_norm": 1.609386682510376,
+      "learning_rate": 1.269750071818443e-05,
+      "loss": 1.9163,
+      "step": 2413
+    },
+    {
+      "epoch": 0.6864576977078977,
+      "grad_norm": 1.467077612876892,
+      "learning_rate": 1.2686009767308244e-05,
+      "loss": 1.8251,
+      "step": 2414
+    },
+    {
+      "epoch": 0.6867420629513558,
+      "grad_norm": 1.5783828496932983,
+      "learning_rate": 1.267451881643206e-05,
+      "loss": 1.7815,
+      "step": 2415
+    },
+    {
+      "epoch": 0.6870264281948139,
+      "grad_norm": 1.699668049812317,
+      "learning_rate": 1.2663027865555875e-05,
+      "loss": 1.9008,
+      "step": 2416
+    },
+    {
+      "epoch": 0.6873107934382721,
+      "grad_norm": 1.7475955486297607,
+      "learning_rate": 1.2651536914679691e-05,
+      "loss": 2.4304,
+      "step": 2417
+    },
+    {
+      "epoch": 0.6875951586817302,
+      "grad_norm": 1.4119583368301392,
+      "learning_rate": 1.2640045963803507e-05,
+      "loss": 2.3749,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6878795239251883,
+      "grad_norm": 1.54697847366333,
+      "learning_rate": 1.2628555012927322e-05,
+      "loss": 2.1809,
+      "step": 2419
+    },
+    {
+      "epoch": 0.6881638891686463,
+      "grad_norm": 1.5237066745758057,
+      "learning_rate": 1.2617064062051136e-05,
+      "loss": 2.2084,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6884482544121044,
+      "grad_norm": 1.586302399635315,
+      "learning_rate": 1.2605573111174951e-05,
+      "loss": 1.9825,
+      "step": 2421
+    },
+    {
+      "epoch": 0.6887326196555627,
+      "grad_norm": 1.619920253753662,
+      "learning_rate": 1.2594082160298765e-05,
+      "loss": 1.9729,
+      "step": 2422
+    },
+    {
+      "epoch": 0.6890169848990207,
+      "grad_norm": 1.5657825469970703,
+      "learning_rate": 1.2582591209422582e-05,
+      "loss": 1.8286,
+      "step": 2423
+    },
+    {
+      "epoch": 0.6893013501424788,
+      "grad_norm": 1.5684531927108765,
+      "learning_rate": 1.2571100258546396e-05,
+      "loss": 1.7035,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6895857153859369,
+      "grad_norm": 1.7014023065567017,
+      "learning_rate": 1.255960930767021e-05,
+      "loss": 2.7972,
+      "step": 2425
+    },
+    {
+      "epoch": 0.689870080629395,
+      "grad_norm": 1.52834951877594,
+      "learning_rate": 1.2548118356794025e-05,
+      "loss": 2.3543,
+      "step": 2426
+    },
+    {
+      "epoch": 0.6901544458728532,
+      "grad_norm": 1.3747142553329468,
+      "learning_rate": 1.2536627405917841e-05,
+      "loss": 2.1109,
+      "step": 2427
+    },
+    {
+      "epoch": 0.6904388111163113,
+      "grad_norm": 1.534885287284851,
+      "learning_rate": 1.2525136455041656e-05,
+      "loss": 2.2004,
+      "step": 2428
+    },
+    {
+      "epoch": 0.6907231763597694,
+      "grad_norm": 1.6728806495666504,
+      "learning_rate": 1.251364550416547e-05,
+      "loss": 2.1012,
+      "step": 2429
+    },
+    {
+      "epoch": 0.6910075416032275,
+      "grad_norm": 1.5786833763122559,
+      "learning_rate": 1.2502154553289285e-05,
+      "loss": 2.0211,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6912919068466856,
+      "grad_norm": 1.4964499473571777,
+      "learning_rate": 1.24906636024131e-05,
+      "loss": 1.7223,
+      "step": 2431
+    },
+    {
+      "epoch": 0.6915762720901438,
+      "grad_norm": 1.5604283809661865,
+      "learning_rate": 1.2479172651536915e-05,
+      "loss": 1.73,
+      "step": 2432
+    },
+    {
+      "epoch": 0.6918606373336019,
+      "grad_norm": 2.041099786758423,
+      "learning_rate": 1.246768170066073e-05,
+      "loss": 2.7647,
+      "step": 2433
+    },
+    {
+      "epoch": 0.69214500257706,
+      "grad_norm": 1.5619685649871826,
+      "learning_rate": 1.2456190749784544e-05,
+      "loss": 2.2241,
+      "step": 2434
+    },
+    {
+      "epoch": 0.6924293678205181,
+      "grad_norm": 1.4527276754379272,
+      "learning_rate": 1.2444699798908362e-05,
+      "loss": 2.2705,
+      "step": 2435
+    },
+    {
+      "epoch": 0.6927137330639762,
+      "grad_norm": 1.5282810926437378,
+      "learning_rate": 1.2433208848032177e-05,
+      "loss": 2.0866,
+      "step": 2436
+    },
+    {
+      "epoch": 0.6929980983074344,
+      "grad_norm": 1.5058673620224,
+      "learning_rate": 1.2421717897155991e-05,
+      "loss": 1.8953,
+      "step": 2437
+    },
+    {
+      "epoch": 0.6932824635508925,
+      "grad_norm": 1.6031697988510132,
+      "learning_rate": 1.2410226946279806e-05,
+      "loss": 1.9201,
+      "step": 2438
+    },
+    {
+      "epoch": 0.6935668287943506,
+      "grad_norm": 1.544476866722107,
+      "learning_rate": 1.2398735995403622e-05,
+      "loss": 1.7643,
+      "step": 2439
+    },
+    {
+      "epoch": 0.6938511940378087,
+      "grad_norm": 1.6882764101028442,
+      "learning_rate": 1.2387245044527436e-05,
+      "loss": 1.7574,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6941355592812668,
+      "grad_norm": 1.8056814670562744,
+      "learning_rate": 1.2375754093651251e-05,
+      "loss": 2.5871,
+      "step": 2441
+    },
+    {
+      "epoch": 0.694419924524725,
+      "grad_norm": 1.553879976272583,
+      "learning_rate": 1.2364263142775065e-05,
+      "loss": 2.5545,
+      "step": 2442
+    },
+    {
+      "epoch": 0.6947042897681831,
+      "grad_norm": 1.5317288637161255,
+      "learning_rate": 1.235277219189888e-05,
+      "loss": 2.2901,
+      "step": 2443
+    },
+    {
+      "epoch": 0.6949886550116412,
+      "grad_norm": 1.6215896606445312,
+      "learning_rate": 1.2341281241022696e-05,
+      "loss": 2.2079,
+      "step": 2444
+    },
+    {
+      "epoch": 0.6952730202550993,
+      "grad_norm": 1.5378239154815674,
+      "learning_rate": 1.232979029014651e-05,
+      "loss": 2.0399,
+      "step": 2445
+    },
+    {
+      "epoch": 0.6955573854985574,
+      "grad_norm": 1.4927102327346802,
+      "learning_rate": 1.2318299339270325e-05,
+      "loss": 1.8721,
+      "step": 2446
+    },
+    {
+      "epoch": 0.6958417507420156,
+      "grad_norm": 1.48208749294281,
+      "learning_rate": 1.230680838839414e-05,
+      "loss": 1.8783,
+      "step": 2447
+    },
+    {
+      "epoch": 0.6961261159854737,
+      "grad_norm": 1.5241432189941406,
+      "learning_rate": 1.2295317437517956e-05,
+      "loss": 1.7324,
+      "step": 2448
+    },
+    {
+      "epoch": 0.6964104812289318,
+      "grad_norm": 1.8973292112350464,
+      "learning_rate": 1.228382648664177e-05,
+      "loss": 2.7123,
+      "step": 2449
+    },
+    {
+      "epoch": 0.6966948464723899,
+      "grad_norm": 1.539310097694397,
+      "learning_rate": 1.2272335535765585e-05,
+      "loss": 2.3683,
+      "step": 2450
+    },
+    {
+      "epoch": 0.696979211715848,
+      "grad_norm": 1.5558347702026367,
+      "learning_rate": 1.22608445848894e-05,
+      "loss": 2.0833,
+      "step": 2451
+    },
+    {
+      "epoch": 0.6972635769593062,
+      "grad_norm": 1.690713882446289,
+      "learning_rate": 1.2249353634013214e-05,
+      "loss": 2.1543,
+      "step": 2452
+    },
+    {
+      "epoch": 0.6975479422027643,
+      "grad_norm": 1.7084331512451172,
+      "learning_rate": 1.223786268313703e-05,
+      "loss": 2.2718,
+      "step": 2453
+    },
+    {
+      "epoch": 0.6978323074462224,
+      "grad_norm": 1.5845692157745361,
+      "learning_rate": 1.2226371732260846e-05,
+      "loss": 1.7258,
+      "step": 2454
+    },
+    {
+      "epoch": 0.6981166726896805,
+      "grad_norm": 1.7029811143875122,
+      "learning_rate": 1.221488078138466e-05,
+      "loss": 1.7476,
+      "step": 2455
+    },
+    {
+      "epoch": 0.6984010379331386,
+      "grad_norm": 1.6282069683074951,
+      "learning_rate": 1.2203389830508477e-05,
+      "loss": 1.7149,
+      "step": 2456
+    },
+    {
+      "epoch": 0.6986854031765968,
+      "grad_norm": 1.8195720911026,
+      "learning_rate": 1.2191898879632291e-05,
+      "loss": 2.6883,
+      "step": 2457
+    },
+    {
+      "epoch": 0.6989697684200549,
+      "grad_norm": 1.4874529838562012,
+      "learning_rate": 1.2180407928756106e-05,
+      "loss": 2.3012,
+      "step": 2458
+    },
+    {
+      "epoch": 0.699254133663513,
+      "grad_norm": 1.4426288604736328,
+      "learning_rate": 1.216891697787992e-05,
+      "loss": 2.2598,
+      "step": 2459
+    },
+    {
+      "epoch": 0.6995384989069711,
+      "grad_norm": 1.4771324396133423,
+      "learning_rate": 1.2157426027003737e-05,
+      "loss": 2.1432,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6998228641504293,
+      "grad_norm": 1.6694068908691406,
+      "learning_rate": 1.2145935076127551e-05,
+      "loss": 2.1884,
+      "step": 2461
+    },
+    {
+      "epoch": 0.7001072293938874,
+      "grad_norm": 1.5652145147323608,
+      "learning_rate": 1.2134444125251366e-05,
+      "loss": 1.8736,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7003915946373455,
+      "grad_norm": 1.6377602815628052,
+      "learning_rate": 1.212295317437518e-05,
+      "loss": 1.7648,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7006759598808036,
+      "grad_norm": 1.671975016593933,
+      "learning_rate": 1.2111462223498995e-05,
+      "loss": 1.8393,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7009603251242617,
+      "grad_norm": 1.6467212438583374,
+      "learning_rate": 1.209997127262281e-05,
+      "loss": 2.8314,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7012446903677199,
+      "grad_norm": 1.6222443580627441,
+      "learning_rate": 1.2088480321746625e-05,
+      "loss": 2.4483,
+      "step": 2466
+    },
+    {
+      "epoch": 0.701529055611178,
+      "grad_norm": 1.5044670104980469,
+      "learning_rate": 1.207698937087044e-05,
+      "loss": 2.4326,
+      "step": 2467
+    },
+    {
+      "epoch": 0.701813420854636,
+      "grad_norm": 1.5599663257598877,
+      "learning_rate": 1.2065498419994254e-05,
+      "loss": 2.0106,
+      "step": 2468
+    },
+    {
+      "epoch": 0.7020977860980941,
+      "grad_norm": 1.4764515161514282,
+      "learning_rate": 1.205400746911807e-05,
+      "loss": 2.0588,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7023821513415522,
+      "grad_norm": 1.4903563261032104,
+      "learning_rate": 1.2042516518241885e-05,
+      "loss": 1.8838,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7026665165850104,
+      "grad_norm": 1.5704025030136108,
+      "learning_rate": 1.20310255673657e-05,
+      "loss": 1.8783,
+      "step": 2471
+    },
+    {
+      "epoch": 0.7029508818284685,
+      "grad_norm": 1.671426773071289,
+      "learning_rate": 1.2019534616489517e-05,
+      "loss": 1.6243,
+      "step": 2472
+    },
+    {
+      "epoch": 0.7032352470719266,
+      "grad_norm": 1.8858857154846191,
+      "learning_rate": 1.2008043665613332e-05,
+      "loss": 2.6409,
+      "step": 2473
+    },
+    {
+      "epoch": 0.7035196123153847,
+      "grad_norm": 1.4502723217010498,
+      "learning_rate": 1.1996552714737146e-05,
+      "loss": 2.4368,
+      "step": 2474
+    },
+    {
+      "epoch": 0.7038039775588428,
+      "grad_norm": 1.4333308935165405,
+      "learning_rate": 1.198506176386096e-05,
+      "loss": 2.2744,
+      "step": 2475
+    },
+    {
+      "epoch": 0.704088342802301,
+      "grad_norm": 1.4810882806777954,
+      "learning_rate": 1.1973570812984775e-05,
+      "loss": 2.105,
+      "step": 2476
+    },
+    {
+      "epoch": 0.7043727080457591,
+      "grad_norm": 1.6196544170379639,
+      "learning_rate": 1.1962079862108592e-05,
+      "loss": 2.1092,
+      "step": 2477
+    },
+    {
+      "epoch": 0.7046570732892172,
+      "grad_norm": 1.455788254737854,
+      "learning_rate": 1.1950588911232406e-05,
+      "loss": 1.75,
+      "step": 2478
+    },
+    {
+      "epoch": 0.7049414385326753,
+      "grad_norm": 1.5926382541656494,
+      "learning_rate": 1.193909796035622e-05,
+      "loss": 1.8092,
+      "step": 2479
+    },
+    {
+      "epoch": 0.7052258037761334,
+      "grad_norm": 1.7150598764419556,
+      "learning_rate": 1.1927607009480035e-05,
+      "loss": 1.7442,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7055101690195916,
+      "grad_norm": 1.7297258377075195,
+      "learning_rate": 1.1916116058603851e-05,
+      "loss": 2.6048,
+      "step": 2481
+    },
+    {
+      "epoch": 0.7057945342630497,
+      "grad_norm": 1.6914091110229492,
+      "learning_rate": 1.1904625107727666e-05,
+      "loss": 2.3913,
+      "step": 2482
+    },
+    {
+      "epoch": 0.7060788995065078,
+      "grad_norm": 1.6037940979003906,
+      "learning_rate": 1.189313415685148e-05,
+      "loss": 2.2043,
+      "step": 2483
+    },
+    {
+      "epoch": 0.7063632647499659,
+      "grad_norm": 1.5348221063613892,
+      "learning_rate": 1.1881643205975295e-05,
+      "loss": 2.2177,
+      "step": 2484
+    },
+    {
+      "epoch": 0.706647629993424,
+      "grad_norm": 1.5680232048034668,
+      "learning_rate": 1.187015225509911e-05,
+      "loss": 2.2766,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7069319952368822,
+      "grad_norm": 1.5366462469100952,
+      "learning_rate": 1.1858661304222925e-05,
+      "loss": 1.98,
+      "step": 2486
+    },
+    {
+      "epoch": 0.7072163604803403,
+      "grad_norm": 1.4476945400238037,
+      "learning_rate": 1.184717035334674e-05,
+      "loss": 1.8068,
+      "step": 2487
+    },
+    {
+      "epoch": 0.7075007257237984,
+      "grad_norm": 1.6980016231536865,
+      "learning_rate": 1.1835679402470554e-05,
+      "loss": 1.6514,
+      "step": 2488
+    },
+    {
+      "epoch": 0.7077850909672565,
+      "grad_norm": 1.753934383392334,
+      "learning_rate": 1.1824188451594369e-05,
+      "loss": 2.7271,
+      "step": 2489
+    },
+    {
+      "epoch": 0.7080694562107146,
+      "grad_norm": 1.530096411705017,
+      "learning_rate": 1.1812697500718185e-05,
+      "loss": 2.0594,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7083538214541728,
+      "grad_norm": 1.539100170135498,
+      "learning_rate": 1.1801206549842001e-05,
+      "loss": 2.0438,
+      "step": 2491
+    },
+    {
+      "epoch": 0.7086381866976309,
+      "grad_norm": 1.5810004472732544,
+      "learning_rate": 1.1789715598965816e-05,
+      "loss": 2.2391,
+      "step": 2492
+    },
+    {
+      "epoch": 0.708922551941089,
+      "grad_norm": 1.5543794631958008,
+      "learning_rate": 1.1778224648089632e-05,
+      "loss": 2.0453,
+      "step": 2493
+    },
+    {
+      "epoch": 0.7092069171845471,
+      "grad_norm": 1.462825894355774,
+      "learning_rate": 1.1766733697213446e-05,
+      "loss": 1.8446,
+      "step": 2494
+    },
+    {
+      "epoch": 0.7094912824280052,
+      "grad_norm": 1.5893888473510742,
+      "learning_rate": 1.1755242746337261e-05,
+      "loss": 1.6117,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7097756476714634,
+      "grad_norm": 1.5945665836334229,
+      "learning_rate": 1.1743751795461075e-05,
+      "loss": 1.6783,
+      "step": 2496
+    },
+    {
+      "epoch": 0.7100600129149215,
+      "grad_norm": 1.7469823360443115,
+      "learning_rate": 1.173226084458489e-05,
+      "loss": 2.6954,
+      "step": 2497
+    },
+    {
+      "epoch": 0.7103443781583796,
+      "grad_norm": 1.4920471906661987,
+      "learning_rate": 1.1720769893708706e-05,
+      "loss": 2.2442,
+      "step": 2498
+    },
+    {
+      "epoch": 0.7106287434018377,
+      "grad_norm": 1.5059243440628052,
+      "learning_rate": 1.170927894283252e-05,
+      "loss": 2.3696,
+      "step": 2499
+    },
+    {
+      "epoch": 0.7109131086452958,
+      "grad_norm": 1.5132187604904175,
+      "learning_rate": 1.1697787991956335e-05,
+      "loss": 2.0736,
+      "step": 2500
+    },
+    {
+      "epoch": 0.711197473888754,
+      "grad_norm": 1.6220301389694214,
+      "learning_rate": 1.168629704108015e-05,
+      "loss": 2.0143,
+      "step": 2501
+    },
+    {
+      "epoch": 0.7114818391322121,
+      "grad_norm": 1.5194025039672852,
+      "learning_rate": 1.1674806090203966e-05,
+      "loss": 1.9679,
+      "step": 2502
+    },
+    {
+      "epoch": 0.7117662043756702,
+      "grad_norm": 1.4100217819213867,
+      "learning_rate": 1.166331513932778e-05,
+      "loss": 1.7964,
+      "step": 2503
+    },
+    {
+      "epoch": 0.7120505696191283,
+      "grad_norm": 1.6122227907180786,
+      "learning_rate": 1.1651824188451595e-05,
+      "loss": 1.7849,
+      "step": 2504
+    },
+    {
+      "epoch": 0.7123349348625864,
+      "grad_norm": 1.7167197465896606,
+      "learning_rate": 1.164033323757541e-05,
+      "loss": 2.6167,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7126193001060446,
+      "grad_norm": 1.4899656772613525,
+      "learning_rate": 1.1628842286699224e-05,
+      "loss": 2.3321,
+      "step": 2506
+    },
+    {
+      "epoch": 0.7129036653495027,
+      "grad_norm": 1.5323848724365234,
+      "learning_rate": 1.161735133582304e-05,
+      "loss": 2.3636,
+      "step": 2507
+    },
+    {
+      "epoch": 0.7131880305929608,
+      "grad_norm": 1.7162936925888062,
+      "learning_rate": 1.1605860384946855e-05,
+      "loss": 2.2844,
+      "step": 2508
+    },
+    {
+      "epoch": 0.7134723958364189,
+      "grad_norm": 1.5394136905670166,
+      "learning_rate": 1.159436943407067e-05,
+      "loss": 2.2002,
+      "step": 2509
+    },
+    {
+      "epoch": 0.7137567610798771,
+      "grad_norm": 1.561732292175293,
+      "learning_rate": 1.1582878483194487e-05,
+      "loss": 1.8321,
+      "step": 2510
+    },
+    {
+      "epoch": 0.7140411263233352,
+      "grad_norm": 1.4912663698196411,
+      "learning_rate": 1.1571387532318301e-05,
+      "loss": 1.6438,
+      "step": 2511
+    },
+    {
+      "epoch": 0.7143254915667933,
+      "grad_norm": 1.7454664707183838,
+      "learning_rate": 1.1559896581442116e-05,
+      "loss": 2.0865,
+      "step": 2512
+    },
+    {
+      "epoch": 0.7146098568102514,
+      "grad_norm": 1.7953159809112549,
+      "learning_rate": 1.154840563056593e-05,
+      "loss": 2.5611,
+      "step": 2513
+    },
+    {
+      "epoch": 0.7148942220537094,
+      "grad_norm": 1.570483922958374,
+      "learning_rate": 1.1536914679689747e-05,
+      "loss": 2.1698,
+      "step": 2514
+    },
+    {
+      "epoch": 0.7151785872971677,
+      "grad_norm": 1.4322417974472046,
+      "learning_rate": 1.1525423728813561e-05,
+      "loss": 2.1472,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7154629525406258,
+      "grad_norm": 1.5140742063522339,
+      "learning_rate": 1.1513932777937376e-05,
+      "loss": 2.2549,
+      "step": 2516
+    },
+    {
+      "epoch": 0.7157473177840838,
+      "grad_norm": 1.485280990600586,
+      "learning_rate": 1.150244182706119e-05,
+      "loss": 2.0911,
+      "step": 2517
+    },
+    {
+      "epoch": 0.7160316830275419,
+      "grad_norm": 1.444021224975586,
+      "learning_rate": 1.1490950876185005e-05,
+      "loss": 1.8068,
+      "step": 2518
+    },
+    {
+      "epoch": 0.716316048271,
+      "grad_norm": 1.5001165866851807,
+      "learning_rate": 1.147945992530882e-05,
+      "loss": 1.7745,
+      "step": 2519
+    },
+    {
+      "epoch": 0.7166004135144582,
+      "grad_norm": 1.7177432775497437,
+      "learning_rate": 1.1467968974432635e-05,
+      "loss": 1.7376,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7168847787579163,
+      "grad_norm": 1.9216365814208984,
+      "learning_rate": 1.145647802355645e-05,
+      "loss": 2.783,
+      "step": 2521
+    },
+    {
+      "epoch": 0.7171691440013744,
+      "grad_norm": 1.6360090970993042,
+      "learning_rate": 1.1444987072680264e-05,
+      "loss": 2.4359,
+      "step": 2522
+    },
+    {
+      "epoch": 0.7174535092448325,
+      "grad_norm": 1.5208964347839355,
+      "learning_rate": 1.143349612180408e-05,
+      "loss": 2.3083,
+      "step": 2523
+    },
+    {
+      "epoch": 0.7177378744882906,
+      "grad_norm": 1.394752860069275,
+      "learning_rate": 1.1422005170927895e-05,
+      "loss": 2.1679,
+      "step": 2524
+    },
+    {
+      "epoch": 0.7180222397317488,
+      "grad_norm": 1.508109211921692,
+      "learning_rate": 1.141051422005171e-05,
+      "loss": 2.1873,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7183066049752069,
+      "grad_norm": 1.7108025550842285,
+      "learning_rate": 1.1399023269175524e-05,
+      "loss": 1.8141,
+      "step": 2526
+    },
+    {
+      "epoch": 0.718590970218665,
+      "grad_norm": 1.727073311805725,
+      "learning_rate": 1.1387532318299338e-05,
+      "loss": 1.7839,
+      "step": 2527
+    },
+    {
+      "epoch": 0.7188753354621231,
+      "grad_norm": 1.7177026271820068,
+      "learning_rate": 1.1376041367423156e-05,
+      "loss": 1.8421,
+      "step": 2528
+    },
+    {
+      "epoch": 0.7191597007055812,
+      "grad_norm": 1.9634953737258911,
+      "learning_rate": 1.136455041654697e-05,
+      "loss": 2.5981,
+      "step": 2529
+    },
+    {
+      "epoch": 0.7194440659490394,
+      "grad_norm": 1.5275763273239136,
+      "learning_rate": 1.1353059465670785e-05,
+      "loss": 2.3648,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7197284311924975,
+      "grad_norm": 1.5566397905349731,
+      "learning_rate": 1.1341568514794601e-05,
+      "loss": 2.3055,
+      "step": 2531
+    },
+    {
+      "epoch": 0.7200127964359556,
+      "grad_norm": 1.6546510457992554,
+      "learning_rate": 1.1330077563918416e-05,
+      "loss": 2.1899,
+      "step": 2532
+    },
+    {
+      "epoch": 0.7202971616794137,
+      "grad_norm": 1.5946438312530518,
+      "learning_rate": 1.131858661304223e-05,
+      "loss": 2.1639,
+      "step": 2533
+    },
+    {
+      "epoch": 0.7205815269228718,
+      "grad_norm": 1.4848626852035522,
+      "learning_rate": 1.1307095662166045e-05,
+      "loss": 2.0955,
+      "step": 2534
+    },
+    {
+      "epoch": 0.72086589216633,
+      "grad_norm": 1.4589171409606934,
+      "learning_rate": 1.1295604711289861e-05,
+      "loss": 1.7393,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7211502574097881,
+      "grad_norm": 1.6969494819641113,
+      "learning_rate": 1.1284113760413676e-05,
+      "loss": 1.3365,
+      "step": 2536
+    },
+    {
+      "epoch": 0.7214346226532462,
+      "grad_norm": 1.7486660480499268,
+      "learning_rate": 1.127262280953749e-05,
+      "loss": 2.695,
+      "step": 2537
+    },
+    {
+      "epoch": 0.7217189878967043,
+      "grad_norm": 1.4772002696990967,
+      "learning_rate": 1.1261131858661305e-05,
+      "loss": 2.4941,
+      "step": 2538
+    },
+    {
+      "epoch": 0.7220033531401624,
+      "grad_norm": 1.4565849304199219,
+      "learning_rate": 1.124964090778512e-05,
+      "loss": 2.2789,
+      "step": 2539
+    },
+    {
+      "epoch": 0.7222877183836206,
+      "grad_norm": 1.3858977556228638,
+      "learning_rate": 1.1238149956908935e-05,
+      "loss": 2.2516,
+      "step": 2540
+    },
+    {
+      "epoch": 0.7225720836270787,
+      "grad_norm": 1.4847400188446045,
+      "learning_rate": 1.122665900603275e-05,
+      "loss": 2.1125,
+      "step": 2541
+    },
+    {
+      "epoch": 0.7228564488705368,
+      "grad_norm": 1.653690218925476,
+      "learning_rate": 1.1215168055156564e-05,
+      "loss": 1.8551,
+      "step": 2542
+    },
+    {
+      "epoch": 0.7231408141139949,
+      "grad_norm": 1.6189554929733276,
+      "learning_rate": 1.1203677104280379e-05,
+      "loss": 1.6745,
+      "step": 2543
+    },
+    {
+      "epoch": 0.723425179357453,
+      "grad_norm": 1.6225018501281738,
+      "learning_rate": 1.1192186153404195e-05,
+      "loss": 1.881,
+      "step": 2544
+    },
+    {
+      "epoch": 0.7237095446009112,
+      "grad_norm": 1.7607388496398926,
+      "learning_rate": 1.118069520252801e-05,
+      "loss": 2.5443,
+      "step": 2545
+    },
+    {
+      "epoch": 0.7239939098443693,
+      "grad_norm": 1.4778530597686768,
+      "learning_rate": 1.1169204251651826e-05,
+      "loss": 2.347,
+      "step": 2546
+    },
+    {
+      "epoch": 0.7242782750878274,
+      "grad_norm": 1.4404771327972412,
+      "learning_rate": 1.115771330077564e-05,
+      "loss": 2.4381,
+      "step": 2547
+    },
+    {
+      "epoch": 0.7245626403312855,
+      "grad_norm": 1.7989895343780518,
+      "learning_rate": 1.1146222349899456e-05,
+      "loss": 2.2025,
+      "step": 2548
+    },
+    {
+      "epoch": 0.7248470055747436,
+      "grad_norm": 1.5871349573135376,
+      "learning_rate": 1.1134731399023271e-05,
+      "loss": 2.1601,
+      "step": 2549
+    },
+    {
+      "epoch": 0.7251313708182018,
+      "grad_norm": 1.521226167678833,
+      "learning_rate": 1.1123240448147085e-05,
+      "loss": 1.7278,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7254157360616599,
+      "grad_norm": 1.6469875574111938,
+      "learning_rate": 1.11117494972709e-05,
+      "loss": 1.6268,
+      "step": 2551
+    },
+    {
+      "epoch": 0.725700101305118,
+      "grad_norm": 1.732232689857483,
+      "learning_rate": 1.1100258546394716e-05,
+      "loss": 1.616,
+      "step": 2552
+    },
+    {
+      "epoch": 0.7259844665485761,
+      "grad_norm": 1.7754976749420166,
+      "learning_rate": 1.108876759551853e-05,
+      "loss": 2.5168,
+      "step": 2553
+    },
+    {
+      "epoch": 0.7262688317920342,
+      "grad_norm": 1.471461296081543,
+      "learning_rate": 1.1077276644642345e-05,
+      "loss": 2.3877,
+      "step": 2554
+    },
+    {
+      "epoch": 0.7265531970354924,
+      "grad_norm": 1.427992820739746,
+      "learning_rate": 1.106578569376616e-05,
+      "loss": 2.1713,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7268375622789505,
+      "grad_norm": 1.6102306842803955,
+      "learning_rate": 1.1054294742889976e-05,
+      "loss": 2.1996,
+      "step": 2556
+    },
+    {
+      "epoch": 0.7271219275224086,
+      "grad_norm": 1.6785222291946411,
+      "learning_rate": 1.104280379201379e-05,
+      "loss": 2.0858,
+      "step": 2557
+    },
+    {
+      "epoch": 0.7274062927658667,
+      "grad_norm": 1.556443214416504,
+      "learning_rate": 1.1031312841137605e-05,
+      "loss": 1.9964,
+      "step": 2558
+    },
+    {
+      "epoch": 0.7276906580093249,
+      "grad_norm": 1.5980945825576782,
+      "learning_rate": 1.101982189026142e-05,
+      "loss": 1.9399,
+      "step": 2559
+    },
+    {
+      "epoch": 0.727975023252783,
+      "grad_norm": 1.6788307428359985,
+      "learning_rate": 1.1008330939385234e-05,
+      "loss": 1.7525,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7282593884962411,
+      "grad_norm": 1.6237465143203735,
+      "learning_rate": 1.099683998850905e-05,
+      "loss": 2.7299,
+      "step": 2561
+    },
+    {
+      "epoch": 0.7285437537396992,
+      "grad_norm": 1.56391179561615,
+      "learning_rate": 1.0985349037632864e-05,
+      "loss": 2.5252,
+      "step": 2562
+    },
+    {
+      "epoch": 0.7288281189831572,
+      "grad_norm": 1.4667418003082275,
+      "learning_rate": 1.0973858086756679e-05,
+      "loss": 2.1989,
+      "step": 2563
+    },
+    {
+      "epoch": 0.7291124842266155,
+      "grad_norm": 1.421413540840149,
+      "learning_rate": 1.0962367135880493e-05,
+      "loss": 2.199,
+      "step": 2564
+    },
+    {
+      "epoch": 0.7293968494700735,
+      "grad_norm": 1.588476538658142,
+      "learning_rate": 1.0950876185004311e-05,
+      "loss": 2.1981,
+      "step": 2565
+    },
+    {
+      "epoch": 0.7296812147135316,
+      "grad_norm": 1.5892815589904785,
+      "learning_rate": 1.0939385234128126e-05,
+      "loss": 1.884,
+      "step": 2566
+    },
+    {
+      "epoch": 0.7299655799569897,
+      "grad_norm": 1.5493581295013428,
+      "learning_rate": 1.092789428325194e-05,
+      "loss": 1.8578,
+      "step": 2567
+    },
+    {
+      "epoch": 0.7302499452004478,
+      "grad_norm": 1.7328555583953857,
+      "learning_rate": 1.0916403332375755e-05,
+      "loss": 1.6838,
+      "step": 2568
+    },
+    {
+      "epoch": 0.730534310443906,
+      "grad_norm": 1.8126568794250488,
+      "learning_rate": 1.0904912381499571e-05,
+      "loss": 2.7335,
+      "step": 2569
+    },
+    {
+      "epoch": 0.7308186756873641,
+      "grad_norm": 1.505807638168335,
+      "learning_rate": 1.0893421430623386e-05,
+      "loss": 2.3911,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7311030409308222,
+      "grad_norm": 1.6133103370666504,
+      "learning_rate": 1.08819304797472e-05,
+      "loss": 2.2435,
+      "step": 2571
+    },
+    {
+      "epoch": 0.7313874061742803,
+      "grad_norm": 1.4566494226455688,
+      "learning_rate": 1.0870439528871015e-05,
+      "loss": 2.0838,
+      "step": 2572
+    },
+    {
+      "epoch": 0.7316717714177384,
+      "grad_norm": 1.5375003814697266,
+      "learning_rate": 1.085894857799483e-05,
+      "loss": 1.9548,
+      "step": 2573
+    },
+    {
+      "epoch": 0.7319561366611966,
+      "grad_norm": 1.5447187423706055,
+      "learning_rate": 1.0847457627118645e-05,
+      "loss": 2.0166,
+      "step": 2574
+    },
+    {
+      "epoch": 0.7322405019046547,
+      "grad_norm": 1.5107059478759766,
+      "learning_rate": 1.083596667624246e-05,
+      "loss": 1.6546,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7325248671481128,
+      "grad_norm": 1.5727344751358032,
+      "learning_rate": 1.0824475725366274e-05,
+      "loss": 1.6626,
+      "step": 2576
+    },
+    {
+      "epoch": 0.7328092323915709,
+      "grad_norm": 1.8032009601593018,
+      "learning_rate": 1.0812984774490089e-05,
+      "loss": 2.8995,
+      "step": 2577
+    },
+    {
+      "epoch": 0.733093597635029,
+      "grad_norm": 1.4891507625579834,
+      "learning_rate": 1.0801493823613905e-05,
+      "loss": 2.3765,
+      "step": 2578
+    },
+    {
+      "epoch": 0.7333779628784872,
+      "grad_norm": 1.535091519355774,
+      "learning_rate": 1.079000287273772e-05,
+      "loss": 2.2866,
+      "step": 2579
+    },
+    {
+      "epoch": 0.7336623281219453,
+      "grad_norm": 1.510955572128296,
+      "learning_rate": 1.0778511921861534e-05,
+      "loss": 2.2064,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7339466933654034,
+      "grad_norm": 1.803688406944275,
+      "learning_rate": 1.0767020970985348e-05,
+      "loss": 2.2372,
+      "step": 2581
+    },
+    {
+      "epoch": 0.7342310586088615,
+      "grad_norm": 1.6097159385681152,
+      "learning_rate": 1.0755530020109165e-05,
+      "loss": 1.8109,
+      "step": 2582
+    },
+    {
+      "epoch": 0.7345154238523196,
+      "grad_norm": 1.478907585144043,
+      "learning_rate": 1.074403906923298e-05,
+      "loss": 1.6932,
+      "step": 2583
+    },
+    {
+      "epoch": 0.7347997890957778,
+      "grad_norm": 1.6126059293746948,
+      "learning_rate": 1.0732548118356795e-05,
+      "loss": 1.6228,
+      "step": 2584
+    },
+    {
+      "epoch": 0.7350841543392359,
+      "grad_norm": 1.7208243608474731,
+      "learning_rate": 1.0721057167480611e-05,
+      "loss": 2.7996,
+      "step": 2585
+    },
+    {
+      "epoch": 0.735368519582694,
+      "grad_norm": 1.4276622533798218,
+      "learning_rate": 1.0709566216604426e-05,
+      "loss": 2.3561,
+      "step": 2586
+    },
+    {
+      "epoch": 0.7356528848261521,
+      "grad_norm": 1.5668575763702393,
+      "learning_rate": 1.069807526572824e-05,
+      "loss": 2.2804,
+      "step": 2587
+    },
+    {
+      "epoch": 0.7359372500696102,
+      "grad_norm": 1.4144526720046997,
+      "learning_rate": 1.0686584314852055e-05,
+      "loss": 2.1127,
+      "step": 2588
+    },
+    {
+      "epoch": 0.7362216153130684,
+      "grad_norm": 1.6393625736236572,
+      "learning_rate": 1.067509336397587e-05,
+      "loss": 2.2331,
+      "step": 2589
+    },
+    {
+      "epoch": 0.7365059805565265,
+      "grad_norm": 1.552177906036377,
+      "learning_rate": 1.0663602413099686e-05,
+      "loss": 1.876,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7367903457999846,
+      "grad_norm": 1.5180331468582153,
+      "learning_rate": 1.06521114622235e-05,
+      "loss": 1.9381,
+      "step": 2591
+    },
+    {
+      "epoch": 0.7370747110434427,
+      "grad_norm": 1.8120715618133545,
+      "learning_rate": 1.0640620511347315e-05,
+      "loss": 1.7459,
+      "step": 2592
+    },
+    {
+      "epoch": 0.7373590762869008,
+      "grad_norm": 1.9727323055267334,
+      "learning_rate": 1.0629129560471129e-05,
+      "loss": 2.7591,
+      "step": 2593
+    },
+    {
+      "epoch": 0.737643441530359,
+      "grad_norm": 1.6499255895614624,
+      "learning_rate": 1.0617638609594945e-05,
+      "loss": 2.582,
+      "step": 2594
+    },
+    {
+      "epoch": 0.7379278067738171,
+      "grad_norm": 1.4931328296661377,
+      "learning_rate": 1.060614765871876e-05,
+      "loss": 2.2723,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7382121720172752,
+      "grad_norm": 1.4535672664642334,
+      "learning_rate": 1.0594656707842574e-05,
+      "loss": 2.1424,
+      "step": 2596
+    },
+    {
+      "epoch": 0.7384965372607333,
+      "grad_norm": 1.6324405670166016,
+      "learning_rate": 1.0583165756966389e-05,
+      "loss": 2.2187,
+      "step": 2597
+    },
+    {
+      "epoch": 0.7387809025041914,
+      "grad_norm": 1.4527398347854614,
+      "learning_rate": 1.0571674806090203e-05,
+      "loss": 1.8947,
+      "step": 2598
+    },
+    {
+      "epoch": 0.7390652677476496,
+      "grad_norm": 1.51231050491333,
+      "learning_rate": 1.056018385521402e-05,
+      "loss": 1.6234,
+      "step": 2599
+    },
+    {
+      "epoch": 0.7393496329911077,
+      "grad_norm": 1.6000384092330933,
+      "learning_rate": 1.0548692904337834e-05,
+      "loss": 1.8085,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7396339982345658,
+      "grad_norm": 1.772000789642334,
+      "learning_rate": 1.0537201953461649e-05,
+      "loss": 2.6206,
+      "step": 2601
+    },
+    {
+      "epoch": 0.7399183634780239,
+      "grad_norm": 1.5927568674087524,
+      "learning_rate": 1.0525711002585466e-05,
+      "loss": 2.4873,
+      "step": 2602
+    },
+    {
+      "epoch": 0.740202728721482,
+      "grad_norm": 1.451069951057434,
+      "learning_rate": 1.0514220051709281e-05,
+      "loss": 2.0906,
+      "step": 2603
+    },
+    {
+      "epoch": 0.7404870939649402,
+      "grad_norm": 1.6845619678497314,
+      "learning_rate": 1.0502729100833095e-05,
+      "loss": 2.1397,
+      "step": 2604
+    },
+    {
+      "epoch": 0.7407714592083983,
+      "grad_norm": 1.546721339225769,
+      "learning_rate": 1.049123814995691e-05,
+      "loss": 1.8256,
+      "step": 2605
+    },
+    {
+      "epoch": 0.7410558244518564,
+      "grad_norm": 1.4821118116378784,
+      "learning_rate": 1.0479747199080726e-05,
+      "loss": 1.842,
+      "step": 2606
+    },
+    {
+      "epoch": 0.7413401896953145,
+      "grad_norm": 1.5237010717391968,
+      "learning_rate": 1.046825624820454e-05,
+      "loss": 1.6978,
+      "step": 2607
+    },
+    {
+      "epoch": 0.7416245549387727,
+      "grad_norm": 1.6505897045135498,
+      "learning_rate": 1.0456765297328355e-05,
+      "loss": 1.6463,
+      "step": 2608
+    },
+    {
+      "epoch": 0.7419089201822308,
+      "grad_norm": 1.657488226890564,
+      "learning_rate": 1.044527434645217e-05,
+      "loss": 2.9463,
+      "step": 2609
+    },
+    {
+      "epoch": 0.7421932854256889,
+      "grad_norm": 1.5707652568817139,
+      "learning_rate": 1.0433783395575984e-05,
+      "loss": 2.5416,
+      "step": 2610
+    },
+    {
+      "epoch": 0.742477650669147,
+      "grad_norm": 1.4029492139816284,
+      "learning_rate": 1.04222924446998e-05,
+      "loss": 2.2495,
+      "step": 2611
+    },
+    {
+      "epoch": 0.742762015912605,
+      "grad_norm": 1.707994818687439,
+      "learning_rate": 1.0410801493823615e-05,
+      "loss": 2.0821,
+      "step": 2612
+    },
+    {
+      "epoch": 0.7430463811560633,
+      "grad_norm": 1.5762234926223755,
+      "learning_rate": 1.039931054294743e-05,
+      "loss": 2.0927,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7433307463995213,
+      "grad_norm": 1.567622423171997,
+      "learning_rate": 1.0387819592071244e-05,
+      "loss": 1.8988,
+      "step": 2614
+    },
+    {
+      "epoch": 0.7436151116429794,
+      "grad_norm": 1.5860332250595093,
+      "learning_rate": 1.037632864119506e-05,
+      "loss": 1.9538,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7438994768864375,
+      "grad_norm": 1.6096969842910767,
+      "learning_rate": 1.0364837690318874e-05,
+      "loss": 1.8601,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7441838421298956,
+      "grad_norm": 1.703778862953186,
+      "learning_rate": 1.0353346739442689e-05,
+      "loss": 2.7728,
+      "step": 2617
+    },
+    {
+      "epoch": 0.7444682073733538,
+      "grad_norm": 1.4619109630584717,
+      "learning_rate": 1.0341855788566503e-05,
+      "loss": 2.373,
+      "step": 2618
+    },
+    {
+      "epoch": 0.7447525726168119,
+      "grad_norm": 1.4133814573287964,
+      "learning_rate": 1.0330364837690318e-05,
+      "loss": 2.2617,
+      "step": 2619
+    },
+    {
+      "epoch": 0.74503693786027,
+      "grad_norm": 1.5261253118515015,
+      "learning_rate": 1.0318873886814136e-05,
+      "loss": 2.1614,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7453213031037281,
+      "grad_norm": 1.649577021598816,
+      "learning_rate": 1.030738293593795e-05,
+      "loss": 2.1312,
+      "step": 2621
+    },
+    {
+      "epoch": 0.7456056683471862,
+      "grad_norm": 1.4830260276794434,
+      "learning_rate": 1.0295891985061765e-05,
+      "loss": 1.6617,
+      "step": 2622
+    },
+    {
+      "epoch": 0.7458900335906444,
+      "grad_norm": 1.6154065132141113,
+      "learning_rate": 1.0284401034185581e-05,
+      "loss": 1.8633,
+      "step": 2623
+    },
+    {
+      "epoch": 0.7461743988341025,
+      "grad_norm": 1.634304404258728,
+      "learning_rate": 1.0272910083309395e-05,
+      "loss": 1.88,
+      "step": 2624
+    },
+    {
+      "epoch": 0.7464587640775606,
+      "grad_norm": 1.7923297882080078,
+      "learning_rate": 1.026141913243321e-05,
+      "loss": 2.8546,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7467431293210187,
+      "grad_norm": 1.4359058141708374,
+      "learning_rate": 1.0249928181557024e-05,
+      "loss": 2.2801,
+      "step": 2626
+    },
+    {
+      "epoch": 0.7470274945644768,
+      "grad_norm": 1.4793764352798462,
+      "learning_rate": 1.023843723068084e-05,
+      "loss": 2.3711,
+      "step": 2627
+    },
+    {
+      "epoch": 0.747311859807935,
+      "grad_norm": 1.487067699432373,
+      "learning_rate": 1.0226946279804655e-05,
+      "loss": 2.1701,
+      "step": 2628
+    },
+    {
+      "epoch": 0.7475962250513931,
+      "grad_norm": 1.7713996171951294,
+      "learning_rate": 1.021545532892847e-05,
+      "loss": 2.283,
+      "step": 2629
+    },
+    {
+      "epoch": 0.7478805902948512,
+      "grad_norm": 1.474770188331604,
+      "learning_rate": 1.0203964378052284e-05,
+      "loss": 1.8782,
+      "step": 2630
+    },
+    {
+      "epoch": 0.7481649555383093,
+      "grad_norm": 1.6031092405319214,
+      "learning_rate": 1.0192473427176099e-05,
+      "loss": 1.89,
+      "step": 2631
+    },
+    {
+      "epoch": 0.7484493207817674,
+      "grad_norm": 1.7248785495758057,
+      "learning_rate": 1.0180982476299915e-05,
+      "loss": 1.8929,
+      "step": 2632
+    },
+    {
+      "epoch": 0.7487336860252256,
+      "grad_norm": 1.6073254346847534,
+      "learning_rate": 1.016949152542373e-05,
+      "loss": 2.6679,
+      "step": 2633
+    },
+    {
+      "epoch": 0.7490180512686837,
+      "grad_norm": 1.4250975847244263,
+      "learning_rate": 1.0158000574547544e-05,
+      "loss": 2.393,
+      "step": 2634
+    },
+    {
+      "epoch": 0.7493024165121418,
+      "grad_norm": 1.3941985368728638,
+      "learning_rate": 1.0146509623671358e-05,
+      "loss": 2.203,
+      "step": 2635
+    },
+    {
+      "epoch": 0.7495867817555999,
+      "grad_norm": 1.540351390838623,
+      "learning_rate": 1.0135018672795175e-05,
+      "loss": 2.0514,
+      "step": 2636
+    },
+    {
+      "epoch": 0.749871146999058,
+      "grad_norm": 1.4758607149124146,
+      "learning_rate": 1.0123527721918989e-05,
+      "loss": 2.1415,
+      "step": 2637
+    },
+    {
+      "epoch": 0.7501555122425162,
+      "grad_norm": 1.5170818567276,
+      "learning_rate": 1.0112036771042804e-05,
+      "loss": 1.7861,
+      "step": 2638
+    },
+    {
+      "epoch": 0.7504398774859743,
+      "grad_norm": 1.5630824565887451,
+      "learning_rate": 1.0100545820166621e-05,
+      "loss": 1.7773,
+      "step": 2639
+    },
+    {
+      "epoch": 0.7507242427294324,
+      "grad_norm": 1.5273665189743042,
+      "learning_rate": 1.0089054869290436e-05,
+      "loss": 1.7984,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7510086079728905,
+      "grad_norm": 1.8208073377609253,
+      "learning_rate": 1.007756391841425e-05,
+      "loss": 2.9004,
+      "step": 2641
+    },
+    {
+      "epoch": 0.7512929732163486,
+      "grad_norm": 1.5931475162506104,
+      "learning_rate": 1.0066072967538065e-05,
+      "loss": 2.4002,
+      "step": 2642
+    },
+    {
+      "epoch": 0.7515773384598068,
+      "grad_norm": 1.410474181175232,
+      "learning_rate": 1.005458201666188e-05,
+      "loss": 2.0302,
+      "step": 2643
+    },
+    {
+      "epoch": 0.7518617037032649,
+      "grad_norm": 1.5305248498916626,
+      "learning_rate": 1.0043091065785696e-05,
+      "loss": 2.0728,
+      "step": 2644
+    },
+    {
+      "epoch": 0.752146068946723,
+      "grad_norm": 1.7503458261489868,
+      "learning_rate": 1.003160011490951e-05,
+      "loss": 2.1175,
+      "step": 2645
+    },
+    {
+      "epoch": 0.7524304341901811,
+      "grad_norm": 1.5125000476837158,
+      "learning_rate": 1.0020109164033325e-05,
+      "loss": 1.8522,
+      "step": 2646
+    },
+    {
+      "epoch": 0.7527147994336392,
+      "grad_norm": 1.5034544467926025,
+      "learning_rate": 1.0008618213157139e-05,
+      "loss": 1.7503,
+      "step": 2647
+    },
+    {
+      "epoch": 0.7529991646770974,
+      "grad_norm": 1.5701968669891357,
+      "learning_rate": 9.997127262280955e-06,
+      "loss": 1.7253,
+      "step": 2648
+    },
+    {
+      "epoch": 0.7532835299205555,
+      "grad_norm": 1.7175260782241821,
+      "learning_rate": 9.98563631140477e-06,
+      "loss": 2.6342,
+      "step": 2649
+    },
+    {
+      "epoch": 0.7535678951640136,
+      "grad_norm": 1.4930824041366577,
+      "learning_rate": 9.974145360528584e-06,
+      "loss": 2.4192,
+      "step": 2650
+    },
+    {
+      "epoch": 0.7538522604074717,
+      "grad_norm": 1.5410807132720947,
+      "learning_rate": 9.962654409652399e-06,
+      "loss": 2.3138,
+      "step": 2651
+    },
+    {
+      "epoch": 0.7541366256509299,
+      "grad_norm": 1.4877644777297974,
+      "learning_rate": 9.951163458776213e-06,
+      "loss": 2.0459,
+      "step": 2652
+    },
+    {
+      "epoch": 0.754420990894388,
+      "grad_norm": 1.711830973625183,
+      "learning_rate": 9.93967250790003e-06,
+      "loss": 2.1824,
+      "step": 2653
+    },
+    {
+      "epoch": 0.7547053561378461,
+      "grad_norm": 1.4872764348983765,
+      "learning_rate": 9.928181557023846e-06,
+      "loss": 1.7254,
+      "step": 2654
+    },
+    {
+      "epoch": 0.7549897213813042,
+      "grad_norm": 1.5591570138931274,
+      "learning_rate": 9.91669060614766e-06,
+      "loss": 1.9,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7552740866247623,
+      "grad_norm": 1.5367933511734009,
+      "learning_rate": 9.905199655271475e-06,
+      "loss": 1.6445,
+      "step": 2656
+    },
+    {
+      "epoch": 0.7555584518682205,
+      "grad_norm": 1.752350091934204,
+      "learning_rate": 9.89370870439529e-06,
+      "loss": 2.8917,
+      "step": 2657
+    },
+    {
+      "epoch": 0.7558428171116786,
+      "grad_norm": 1.5216292142868042,
+      "learning_rate": 9.882217753519104e-06,
+      "loss": 2.4243,
+      "step": 2658
+    },
+    {
+      "epoch": 0.7561271823551367,
+      "grad_norm": 1.5508999824523926,
+      "learning_rate": 9.87072680264292e-06,
+      "loss": 2.1663,
+      "step": 2659
+    },
+    {
+      "epoch": 0.7564115475985947,
+      "grad_norm": 1.4459530115127563,
+      "learning_rate": 9.859235851766734e-06,
+      "loss": 2.1307,
+      "step": 2660
+    },
+    {
+      "epoch": 0.7566959128420528,
+      "grad_norm": 1.627680778503418,
+      "learning_rate": 9.847744900890549e-06,
+      "loss": 2.0607,
+      "step": 2661
+    },
+    {
+      "epoch": 0.756980278085511,
+      "grad_norm": 1.4881747961044312,
+      "learning_rate": 9.836253950014365e-06,
+      "loss": 1.9483,
+      "step": 2662
+    },
+    {
+      "epoch": 0.7572646433289691,
+      "grad_norm": 1.6278051137924194,
+      "learning_rate": 9.82476299913818e-06,
+      "loss": 1.794,
+      "step": 2663
+    },
+    {
+      "epoch": 0.7575490085724272,
+      "grad_norm": 1.6977025270462036,
+      "learning_rate": 9.813272048261994e-06,
+      "loss": 1.6891,
+      "step": 2664
+    },
+    {
+      "epoch": 0.7578333738158853,
+      "grad_norm": 1.795324444770813,
+      "learning_rate": 9.80178109738581e-06,
+      "loss": 2.4871,
+      "step": 2665
+    },
+    {
+      "epoch": 0.7581177390593434,
+      "grad_norm": 1.48861563205719,
+      "learning_rate": 9.790290146509625e-06,
+      "loss": 2.4192,
+      "step": 2666
+    },
+    {
+      "epoch": 0.7584021043028016,
+      "grad_norm": 1.4271656274795532,
+      "learning_rate": 9.77879919563344e-06,
+      "loss": 2.1096,
+      "step": 2667
+    },
+    {
+      "epoch": 0.7586864695462597,
+      "grad_norm": 1.4701792001724243,
+      "learning_rate": 9.767308244757254e-06,
+      "loss": 2.0263,
+      "step": 2668
+    },
+    {
+      "epoch": 0.7589708347897178,
+      "grad_norm": 1.5513591766357422,
+      "learning_rate": 9.75581729388107e-06,
+      "loss": 2.2207,
+      "step": 2669
+    },
+    {
+      "epoch": 0.7592552000331759,
+      "grad_norm": 1.4835832118988037,
+      "learning_rate": 9.744326343004884e-06,
+      "loss": 1.9584,
+      "step": 2670
+    },
+    {
+      "epoch": 0.759539565276634,
+      "grad_norm": 1.5424950122833252,
+      "learning_rate": 9.7328353921287e-06,
+      "loss": 2.0076,
+      "step": 2671
+    },
+    {
+      "epoch": 0.7598239305200922,
+      "grad_norm": 1.5847523212432861,
+      "learning_rate": 9.721344441252515e-06,
+      "loss": 1.712,
+      "step": 2672
+    },
+    {
+      "epoch": 0.7601082957635503,
+      "grad_norm": 1.8330718278884888,
+      "learning_rate": 9.70985349037633e-06,
+      "loss": 2.7035,
+      "step": 2673
+    },
+    {
+      "epoch": 0.7603926610070084,
+      "grad_norm": 1.533942699432373,
+      "learning_rate": 9.698362539500144e-06,
+      "loss": 2.1846,
+      "step": 2674
+    },
+    {
+      "epoch": 0.7606770262504665,
+      "grad_norm": 1.5470412969589233,
+      "learning_rate": 9.68687158862396e-06,
+      "loss": 2.2428,
+      "step": 2675
+    },
+    {
+      "epoch": 0.7609613914939246,
+      "grad_norm": 1.4431195259094238,
+      "learning_rate": 9.675380637747775e-06,
+      "loss": 2.2741,
+      "step": 2676
+    },
+    {
+      "epoch": 0.7612457567373828,
+      "grad_norm": 1.7669216394424438,
+      "learning_rate": 9.66388968687159e-06,
+      "loss": 2.05,
+      "step": 2677
+    },
+    {
+      "epoch": 0.7615301219808409,
+      "grad_norm": 1.4941767454147339,
+      "learning_rate": 9.652398735995404e-06,
+      "loss": 2.0978,
+      "step": 2678
+    },
+    {
+      "epoch": 0.761814487224299,
+      "grad_norm": 1.4513996839523315,
+      "learning_rate": 9.640907785119218e-06,
+      "loss": 1.7291,
+      "step": 2679
+    },
+    {
+      "epoch": 0.7620988524677571,
+      "grad_norm": 1.507140874862671,
+      "learning_rate": 9.629416834243034e-06,
+      "loss": 1.7039,
+      "step": 2680
+    },
+    {
+      "epoch": 0.7623832177112152,
+      "grad_norm": 1.6555514335632324,
+      "learning_rate": 9.61792588336685e-06,
+      "loss": 2.5856,
+      "step": 2681
+    },
+    {
+      "epoch": 0.7626675829546734,
+      "grad_norm": 1.4625468254089355,
+      "learning_rate": 9.606434932490665e-06,
+      "loss": 2.4481,
+      "step": 2682
+    },
+    {
+      "epoch": 0.7629519481981315,
+      "grad_norm": 1.55473792552948,
+      "learning_rate": 9.59494398161448e-06,
+      "loss": 2.0131,
+      "step": 2683
+    },
+    {
+      "epoch": 0.7632363134415896,
+      "grad_norm": 1.5071228742599487,
+      "learning_rate": 9.583453030738294e-06,
+      "loss": 2.2608,
+      "step": 2684
+    },
+    {
+      "epoch": 0.7635206786850477,
+      "grad_norm": 1.4757564067840576,
+      "learning_rate": 9.571962079862109e-06,
+      "loss": 1.9793,
+      "step": 2685
+    },
+    {
+      "epoch": 0.7638050439285058,
+      "grad_norm": 1.4441161155700684,
+      "learning_rate": 9.560471128985925e-06,
+      "loss": 1.7435,
+      "step": 2686
+    },
+    {
+      "epoch": 0.764089409171964,
+      "grad_norm": 1.4009804725646973,
+      "learning_rate": 9.54898017810974e-06,
+      "loss": 1.7192,
+      "step": 2687
+    },
+    {
+      "epoch": 0.7643737744154221,
+      "grad_norm": 1.5814915895462036,
+      "learning_rate": 9.537489227233554e-06,
+      "loss": 1.6818,
+      "step": 2688
+    },
+    {
+      "epoch": 0.7646581396588802,
+      "grad_norm": 1.7841424942016602,
+      "learning_rate": 9.525998276357368e-06,
+      "loss": 2.8207,
+      "step": 2689
+    },
+    {
+      "epoch": 0.7649425049023383,
+      "grad_norm": 1.4650540351867676,
+      "learning_rate": 9.514507325481185e-06,
+      "loss": 2.3041,
+      "step": 2690
+    },
+    {
+      "epoch": 0.7652268701457964,
+      "grad_norm": 1.519219994544983,
+      "learning_rate": 9.503016374604999e-06,
+      "loss": 2.0641,
+      "step": 2691
+    },
+    {
+      "epoch": 0.7655112353892546,
+      "grad_norm": 1.5591269731521606,
+      "learning_rate": 9.491525423728815e-06,
+      "loss": 2.3594,
+      "step": 2692
+    },
+    {
+      "epoch": 0.7657956006327127,
+      "grad_norm": 1.680830717086792,
+      "learning_rate": 9.48003447285263e-06,
+      "loss": 2.3668,
+      "step": 2693
+    },
+    {
+      "epoch": 0.7660799658761708,
+      "grad_norm": 1.5440500974655151,
+      "learning_rate": 9.468543521976444e-06,
+      "loss": 1.9979,
+      "step": 2694
+    },
+    {
+      "epoch": 0.7663643311196289,
+      "grad_norm": 1.6357736587524414,
+      "learning_rate": 9.457052571100259e-06,
+      "loss": 1.729,
+      "step": 2695
+    },
+    {
+      "epoch": 0.766648696363087,
+      "grad_norm": 1.5347177982330322,
+      "learning_rate": 9.445561620224075e-06,
+      "loss": 1.8021,
+      "step": 2696
+    },
+    {
+      "epoch": 0.7669330616065452,
+      "grad_norm": 1.7074389457702637,
+      "learning_rate": 9.43407066934789e-06,
+      "loss": 2.7968,
+      "step": 2697
+    },
+    {
+      "epoch": 0.7672174268500033,
+      "grad_norm": 1.4788157939910889,
+      "learning_rate": 9.422579718471704e-06,
+      "loss": 2.5411,
+      "step": 2698
+    },
+    {
+      "epoch": 0.7675017920934614,
+      "grad_norm": 1.428257703781128,
+      "learning_rate": 9.41108876759552e-06,
+      "loss": 2.3632,
+      "step": 2699
+    },
+    {
+      "epoch": 0.7677861573369195,
+      "grad_norm": 1.4511548280715942,
+      "learning_rate": 9.399597816719335e-06,
+      "loss": 2.102,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7680705225803777,
+      "grad_norm": 1.6660867929458618,
+      "learning_rate": 9.388106865843149e-06,
+      "loss": 1.9556,
+      "step": 2701
+    },
+    {
+      "epoch": 0.7683548878238358,
+      "grad_norm": 1.6087857484817505,
+      "learning_rate": 9.376615914966965e-06,
+      "loss": 1.9751,
+      "step": 2702
+    },
+    {
+      "epoch": 0.7686392530672939,
+      "grad_norm": 1.5220601558685303,
+      "learning_rate": 9.36512496409078e-06,
+      "loss": 1.8027,
+      "step": 2703
+    },
+    {
+      "epoch": 0.768923618310752,
+      "grad_norm": 1.727441668510437,
+      "learning_rate": 9.353634013214594e-06,
+      "loss": 1.7694,
+      "step": 2704
+    },
+    {
+      "epoch": 0.76920798355421,
+      "grad_norm": 1.7294251918792725,
+      "learning_rate": 9.342143062338409e-06,
+      "loss": 2.6689,
+      "step": 2705
+    },
+    {
+      "epoch": 0.7694923487976683,
+      "grad_norm": 1.4868618249893188,
+      "learning_rate": 9.330652111462223e-06,
+      "loss": 2.4436,
+      "step": 2706
+    },
+    {
+      "epoch": 0.7697767140411264,
+      "grad_norm": 1.4418073892593384,
+      "learning_rate": 9.31916116058604e-06,
+      "loss": 2.1285,
+      "step": 2707
+    },
+    {
+      "epoch": 0.7700610792845844,
+      "grad_norm": 1.4674290418624878,
+      "learning_rate": 9.307670209709856e-06,
+      "loss": 1.9759,
+      "step": 2708
+    },
+    {
+      "epoch": 0.7703454445280425,
+      "grad_norm": 1.6296724081039429,
+      "learning_rate": 9.29617925883367e-06,
+      "loss": 2.1011,
+      "step": 2709
+    },
+    {
+      "epoch": 0.7706298097715006,
+      "grad_norm": 1.5555256605148315,
+      "learning_rate": 9.284688307957485e-06,
+      "loss": 1.7846,
+      "step": 2710
+    },
+    {
+      "epoch": 0.7709141750149588,
+      "grad_norm": 1.4738327264785767,
+      "learning_rate": 9.273197357081299e-06,
+      "loss": 1.5781,
+      "step": 2711
+    },
+    {
+      "epoch": 0.7711985402584169,
+      "grad_norm": 1.5029698610305786,
+      "learning_rate": 9.261706406205114e-06,
+      "loss": 1.5587,
+      "step": 2712
+    },
+    {
+      "epoch": 0.771482905501875,
+      "grad_norm": 1.7328534126281738,
+      "learning_rate": 9.25021545532893e-06,
+      "loss": 2.6136,
+      "step": 2713
+    },
+    {
+      "epoch": 0.7717672707453331,
+      "grad_norm": 1.5154101848602295,
+      "learning_rate": 9.238724504452744e-06,
+      "loss": 2.4293,
+      "step": 2714
+    },
+    {
+      "epoch": 0.7720516359887912,
+      "grad_norm": 1.420172929763794,
+      "learning_rate": 9.227233553576559e-06,
+      "loss": 2.1027,
+      "step": 2715
+    },
+    {
+      "epoch": 0.7723360012322494,
+      "grad_norm": 1.5309847593307495,
+      "learning_rate": 9.215742602700373e-06,
+      "loss": 2.3935,
+      "step": 2716
+    },
+    {
+      "epoch": 0.7726203664757075,
+      "grad_norm": 1.5749589204788208,
+      "learning_rate": 9.20425165182419e-06,
+      "loss": 1.9977,
+      "step": 2717
+    },
+    {
+      "epoch": 0.7729047317191656,
+      "grad_norm": 1.5268434286117554,
+      "learning_rate": 9.192760700948004e-06,
+      "loss": 1.7747,
+      "step": 2718
+    },
+    {
+      "epoch": 0.7731890969626237,
+      "grad_norm": 1.5345548391342163,
+      "learning_rate": 9.18126975007182e-06,
+      "loss": 1.7534,
+      "step": 2719
+    },
+    {
+      "epoch": 0.7734734622060818,
+      "grad_norm": 1.646863341331482,
+      "learning_rate": 9.169778799195635e-06,
+      "loss": 1.7113,
+      "step": 2720
+    },
+    {
+      "epoch": 0.77375782744954,
+      "grad_norm": 1.6630712747573853,
+      "learning_rate": 9.15828784831945e-06,
+      "loss": 2.6318,
+      "step": 2721
+    },
+    {
+      "epoch": 0.7740421926929981,
+      "grad_norm": 1.4465360641479492,
+      "learning_rate": 9.146796897443264e-06,
+      "loss": 2.5043,
+      "step": 2722
+    },
+    {
+      "epoch": 0.7743265579364562,
+      "grad_norm": 1.4710736274719238,
+      "learning_rate": 9.13530594656708e-06,
+      "loss": 2.1993,
+      "step": 2723
+    },
+    {
+      "epoch": 0.7746109231799143,
+      "grad_norm": 1.5619592666625977,
+      "learning_rate": 9.123814995690894e-06,
+      "loss": 2.1542,
+      "step": 2724
+    },
+    {
+      "epoch": 0.7748952884233724,
+      "grad_norm": 1.507546305656433,
+      "learning_rate": 9.112324044814709e-06,
+      "loss": 2.048,
+      "step": 2725
+    },
+    {
+      "epoch": 0.7751796536668306,
+      "grad_norm": 1.5877704620361328,
+      "learning_rate": 9.100833093938523e-06,
+      "loss": 1.8837,
+      "step": 2726
+    },
+    {
+      "epoch": 0.7754640189102887,
+      "grad_norm": 1.5766810178756714,
+      "learning_rate": 9.08934214306234e-06,
+      "loss": 1.7013,
+      "step": 2727
+    },
+    {
+      "epoch": 0.7757483841537468,
+      "grad_norm": 1.535407304763794,
+      "learning_rate": 9.077851192186154e-06,
+      "loss": 1.726,
+      "step": 2728
+    },
+    {
+      "epoch": 0.7760327493972049,
+      "grad_norm": 1.723103404045105,
+      "learning_rate": 9.06636024130997e-06,
+      "loss": 2.5775,
+      "step": 2729
+    },
+    {
+      "epoch": 0.776317114640663,
+      "grad_norm": 1.5206818580627441,
+      "learning_rate": 9.054869290433785e-06,
+      "loss": 2.4745,
+      "step": 2730
+    },
+    {
+      "epoch": 0.7766014798841212,
+      "grad_norm": 1.486801266670227,
+      "learning_rate": 9.0433783395576e-06,
+      "loss": 2.3211,
+      "step": 2731
+    },
+    {
+      "epoch": 0.7768858451275793,
+      "grad_norm": 1.546639323234558,
+      "learning_rate": 9.031887388681414e-06,
+      "loss": 2.2065,
+      "step": 2732
+    },
+    {
+      "epoch": 0.7771702103710374,
+      "grad_norm": 1.6100459098815918,
+      "learning_rate": 9.020396437805228e-06,
+      "loss": 2.1838,
+      "step": 2733
+    },
+    {
+      "epoch": 0.7774545756144955,
+      "grad_norm": 1.537714958190918,
+      "learning_rate": 9.008905486929044e-06,
+      "loss": 1.845,
+      "step": 2734
+    },
+    {
+      "epoch": 0.7777389408579536,
+      "grad_norm": 1.5195116996765137,
+      "learning_rate": 8.997414536052859e-06,
+      "loss": 1.8668,
+      "step": 2735
+    },
+    {
+      "epoch": 0.7780233061014118,
+      "grad_norm": 1.630244255065918,
+      "learning_rate": 8.985923585176675e-06,
+      "loss": 1.8279,
+      "step": 2736
+    },
+    {
+      "epoch": 0.7783076713448699,
+      "grad_norm": 1.72090744972229,
+      "learning_rate": 8.97443263430049e-06,
+      "loss": 2.7764,
+      "step": 2737
+    },
+    {
+      "epoch": 0.778592036588328,
+      "grad_norm": 1.5594855546951294,
+      "learning_rate": 8.962941683424304e-06,
+      "loss": 2.2368,
+      "step": 2738
+    },
+    {
+      "epoch": 0.7788764018317861,
+      "grad_norm": 1.4783302545547485,
+      "learning_rate": 8.951450732548119e-06,
+      "loss": 2.2075,
+      "step": 2739
+    },
+    {
+      "epoch": 0.7791607670752442,
+      "grad_norm": 1.5036635398864746,
+      "learning_rate": 8.939959781671935e-06,
+      "loss": 2.1962,
+      "step": 2740
+    },
+    {
+      "epoch": 0.7794451323187024,
+      "grad_norm": 1.511878252029419,
+      "learning_rate": 8.92846883079575e-06,
+      "loss": 1.9705,
+      "step": 2741
+    },
+    {
+      "epoch": 0.7797294975621605,
+      "grad_norm": 1.6692416667938232,
+      "learning_rate": 8.916977879919564e-06,
+      "loss": 2.0001,
+      "step": 2742
+    },
+    {
+      "epoch": 0.7800138628056186,
+      "grad_norm": 1.5080974102020264,
+      "learning_rate": 8.905486929043378e-06,
+      "loss": 1.5839,
+      "step": 2743
+    },
+    {
+      "epoch": 0.7802982280490767,
+      "grad_norm": 1.500978946685791,
+      "learning_rate": 8.893995978167193e-06,
+      "loss": 1.7545,
+      "step": 2744
+    },
+    {
+      "epoch": 0.7805825932925348,
+      "grad_norm": 1.763393521308899,
+      "learning_rate": 8.882505027291009e-06,
+      "loss": 2.7148,
+      "step": 2745
+    },
+    {
+      "epoch": 0.780866958535993,
+      "grad_norm": 1.525657057762146,
+      "learning_rate": 8.871014076414825e-06,
+      "loss": 2.4186,
+      "step": 2746
+    },
+    {
+      "epoch": 0.7811513237794511,
+      "grad_norm": 1.4492183923721313,
+      "learning_rate": 8.85952312553864e-06,
+      "loss": 2.1651,
+      "step": 2747
+    },
+    {
+      "epoch": 0.7814356890229092,
+      "grad_norm": 1.5383533239364624,
+      "learning_rate": 8.848032174662454e-06,
+      "loss": 2.1932,
+      "step": 2748
+    },
+    {
+      "epoch": 0.7817200542663673,
+      "grad_norm": 1.5642298460006714,
+      "learning_rate": 8.836541223786269e-06,
+      "loss": 2.1424,
+      "step": 2749
+    },
+    {
+      "epoch": 0.7820044195098255,
+      "grad_norm": 1.4735430479049683,
+      "learning_rate": 8.825050272910083e-06,
+      "loss": 1.7743,
+      "step": 2750
+    },
+    {
+      "epoch": 0.7822887847532836,
+      "grad_norm": 1.368130087852478,
+      "learning_rate": 8.8135593220339e-06,
+      "loss": 1.8134,
+      "step": 2751
+    },
+    {
+      "epoch": 0.7825731499967417,
+      "grad_norm": 1.6192513704299927,
+      "learning_rate": 8.802068371157714e-06,
+      "loss": 1.7028,
+      "step": 2752
+    },
+    {
+      "epoch": 0.7828575152401998,
+      "grad_norm": 1.9017709493637085,
+      "learning_rate": 8.790577420281528e-06,
+      "loss": 2.8195,
+      "step": 2753
+    },
+    {
+      "epoch": 0.7831418804836578,
+      "grad_norm": 1.5791484117507935,
+      "learning_rate": 8.779086469405343e-06,
+      "loss": 2.4216,
+      "step": 2754
+    },
+    {
+      "epoch": 0.783426245727116,
+      "grad_norm": 1.4580869674682617,
+      "learning_rate": 8.767595518529159e-06,
+      "loss": 2.0641,
+      "step": 2755
+    },
+    {
+      "epoch": 0.7837106109705742,
+      "grad_norm": 1.4634650945663452,
+      "learning_rate": 8.756104567652974e-06,
+      "loss": 2.1984,
+      "step": 2756
+    },
+    {
+      "epoch": 0.7839949762140322,
+      "grad_norm": 1.5937823057174683,
+      "learning_rate": 8.74461361677679e-06,
+      "loss": 1.9328,
+      "step": 2757
+    },
+    {
+      "epoch": 0.7842793414574903,
+      "grad_norm": 1.5349475145339966,
+      "learning_rate": 8.733122665900604e-06,
+      "loss": 1.7728,
+      "step": 2758
+    },
+    {
+      "epoch": 0.7845637067009484,
+      "grad_norm": 1.4689794778823853,
+      "learning_rate": 8.721631715024419e-06,
+      "loss": 1.7538,
+      "step": 2759
+    },
+    {
+      "epoch": 0.7848480719444066,
+      "grad_norm": 1.59420645236969,
+      "learning_rate": 8.710140764148233e-06,
+      "loss": 1.587,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7851324371878647,
+      "grad_norm": 1.7142568826675415,
+      "learning_rate": 8.69864981327205e-06,
+      "loss": 2.6669,
+      "step": 2761
+    },
+    {
+      "epoch": 0.7854168024313228,
+      "grad_norm": 1.4829474687576294,
+      "learning_rate": 8.687158862395864e-06,
+      "loss": 2.5419,
+      "step": 2762
+    },
+    {
+      "epoch": 0.7857011676747809,
+      "grad_norm": 1.5407049655914307,
+      "learning_rate": 8.675667911519678e-06,
+      "loss": 2.2923,
+      "step": 2763
+    },
+    {
+      "epoch": 0.785985532918239,
+      "grad_norm": 1.489302158355713,
+      "learning_rate": 8.664176960643495e-06,
+      "loss": 2.3303,
+      "step": 2764
+    },
+    {
+      "epoch": 0.7862698981616972,
+      "grad_norm": 1.5373653173446655,
+      "learning_rate": 8.652686009767309e-06,
+      "loss": 2.0711,
+      "step": 2765
+    },
+    {
+      "epoch": 0.7865542634051553,
+      "grad_norm": 1.6663075685501099,
+      "learning_rate": 8.641195058891124e-06,
+      "loss": 1.8473,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7868386286486134,
+      "grad_norm": 1.595422387123108,
+      "learning_rate": 8.62970410801494e-06,
+      "loss": 1.7362,
+      "step": 2767
+    },
+    {
+      "epoch": 0.7871229938920715,
+      "grad_norm": 1.601948857307434,
+      "learning_rate": 8.618213157138754e-06,
+      "loss": 1.6511,
+      "step": 2768
+    },
+    {
+      "epoch": 0.7874073591355296,
+      "grad_norm": 1.8554433584213257,
+      "learning_rate": 8.606722206262569e-06,
+      "loss": 2.7364,
+      "step": 2769
+    },
+    {
+      "epoch": 0.7876917243789878,
+      "grad_norm": 1.4515372514724731,
+      "learning_rate": 8.595231255386383e-06,
+      "loss": 2.6112,
+      "step": 2770
+    },
+    {
+      "epoch": 0.7879760896224459,
+      "grad_norm": 1.4031703472137451,
+      "learning_rate": 8.583740304510198e-06,
+      "loss": 2.2281,
+      "step": 2771
+    },
+    {
+      "epoch": 0.788260454865904,
+      "grad_norm": 1.527431845664978,
+      "learning_rate": 8.572249353634014e-06,
+      "loss": 2.2233,
+      "step": 2772
+    },
+    {
+      "epoch": 0.7885448201093621,
+      "grad_norm": 1.635292649269104,
+      "learning_rate": 8.56075840275783e-06,
+      "loss": 2.1587,
+      "step": 2773
+    },
+    {
+      "epoch": 0.7888291853528202,
+      "grad_norm": 1.6808167695999146,
+      "learning_rate": 8.549267451881645e-06,
+      "loss": 1.8091,
+      "step": 2774
+    },
+    {
+      "epoch": 0.7891135505962784,
+      "grad_norm": 1.5428953170776367,
+      "learning_rate": 8.537776501005459e-06,
+      "loss": 1.9059,
+      "step": 2775
+    },
+    {
+      "epoch": 0.7893979158397365,
+      "grad_norm": 1.5486587285995483,
+      "learning_rate": 8.526285550129274e-06,
+      "loss": 1.6714,
+      "step": 2776
+    },
+    {
+      "epoch": 0.7896822810831946,
+      "grad_norm": 1.6755555868148804,
+      "learning_rate": 8.514794599253088e-06,
+      "loss": 2.6297,
+      "step": 2777
+    },
+    {
+      "epoch": 0.7899666463266527,
+      "grad_norm": 1.4705214500427246,
+      "learning_rate": 8.503303648376904e-06,
+      "loss": 2.2428,
+      "step": 2778
+    },
+    {
+      "epoch": 0.7902510115701108,
+      "grad_norm": 1.4628565311431885,
+      "learning_rate": 8.491812697500719e-06,
+      "loss": 2.0421,
+      "step": 2779
+    },
+    {
+      "epoch": 0.790535376813569,
+      "grad_norm": 1.5100350379943848,
+      "learning_rate": 8.480321746624533e-06,
+      "loss": 2.2767,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7908197420570271,
+      "grad_norm": 1.6801701784133911,
+      "learning_rate": 8.468830795748348e-06,
+      "loss": 1.9363,
+      "step": 2781
+    },
+    {
+      "epoch": 0.7911041073004852,
+      "grad_norm": 1.6616177558898926,
+      "learning_rate": 8.457339844872164e-06,
+      "loss": 2.0598,
+      "step": 2782
+    },
+    {
+      "epoch": 0.7913884725439433,
+      "grad_norm": 1.5380514860153198,
+      "learning_rate": 8.445848893995979e-06,
+      "loss": 1.7708,
+      "step": 2783
+    },
+    {
+      "epoch": 0.7916728377874014,
+      "grad_norm": 1.5327441692352295,
+      "learning_rate": 8.434357943119795e-06,
+      "loss": 1.7078,
+      "step": 2784
+    },
+    {
+      "epoch": 0.7919572030308596,
+      "grad_norm": 1.597601294517517,
+      "learning_rate": 8.42286699224361e-06,
+      "loss": 2.5618,
+      "step": 2785
+    },
+    {
+      "epoch": 0.7922415682743177,
+      "grad_norm": 1.663575291633606,
+      "learning_rate": 8.411376041367424e-06,
+      "loss": 2.422,
+      "step": 2786
+    },
+    {
+      "epoch": 0.7925259335177758,
+      "grad_norm": 1.4837387800216675,
+      "learning_rate": 8.399885090491238e-06,
+      "loss": 2.3795,
+      "step": 2787
+    },
+    {
+      "epoch": 0.7928102987612339,
+      "grad_norm": 1.405079960823059,
+      "learning_rate": 8.388394139615054e-06,
+      "loss": 2.2017,
+      "step": 2788
+    },
+    {
+      "epoch": 0.793094664004692,
+      "grad_norm": 1.6203116178512573,
+      "learning_rate": 8.376903188738869e-06,
+      "loss": 2.1513,
+      "step": 2789
+    },
+    {
+      "epoch": 0.7933790292481502,
+      "grad_norm": 1.4136089086532593,
+      "learning_rate": 8.365412237862683e-06,
+      "loss": 1.6959,
+      "step": 2790
+    },
+    {
+      "epoch": 0.7936633944916083,
+      "grad_norm": 1.5652225017547607,
+      "learning_rate": 8.353921286986498e-06,
+      "loss": 1.7965,
+      "step": 2791
+    },
+    {
+      "epoch": 0.7939477597350664,
+      "grad_norm": 1.5910707712173462,
+      "learning_rate": 8.342430336110314e-06,
+      "loss": 1.6998,
+      "step": 2792
+    },
+    {
+      "epoch": 0.7942321249785245,
+      "grad_norm": 1.7475180625915527,
+      "learning_rate": 8.330939385234129e-06,
+      "loss": 2.6038,
+      "step": 2793
+    },
+    {
+      "epoch": 0.7945164902219826,
+      "grad_norm": 1.538881540298462,
+      "learning_rate": 8.319448434357945e-06,
+      "loss": 2.217,
+      "step": 2794
+    },
+    {
+      "epoch": 0.7948008554654408,
+      "grad_norm": 1.5075610876083374,
+      "learning_rate": 8.30795748348176e-06,
+      "loss": 2.4162,
+      "step": 2795
+    },
+    {
+      "epoch": 0.7950852207088989,
+      "grad_norm": 1.4764316082000732,
+      "learning_rate": 8.296466532605574e-06,
+      "loss": 2.1782,
+      "step": 2796
+    },
+    {
+      "epoch": 0.795369585952357,
+      "grad_norm": 1.5661760568618774,
+      "learning_rate": 8.284975581729388e-06,
+      "loss": 2.1276,
+      "step": 2797
+    },
+    {
+      "epoch": 0.7956539511958151,
+      "grad_norm": 1.4711788892745972,
+      "learning_rate": 8.273484630853203e-06,
+      "loss": 1.7211,
+      "step": 2798
+    },
+    {
+      "epoch": 0.7959383164392733,
+      "grad_norm": 1.5930848121643066,
+      "learning_rate": 8.261993679977019e-06,
+      "loss": 1.6124,
+      "step": 2799
+    },
+    {
+      "epoch": 0.7962226816827314,
+      "grad_norm": 1.6392720937728882,
+      "learning_rate": 8.250502729100833e-06,
+      "loss": 1.8286,
+      "step": 2800
+    },
+    {
+      "epoch": 0.7965070469261895,
+      "grad_norm": 1.7386542558670044,
+      "learning_rate": 8.23901177822465e-06,
+      "loss": 2.7341,
+      "step": 2801
+    },
+    {
+      "epoch": 0.7967914121696476,
+      "grad_norm": 1.4948952198028564,
+      "learning_rate": 8.227520827348464e-06,
+      "loss": 2.4624,
+      "step": 2802
+    },
+    {
+      "epoch": 0.7970757774131056,
+      "grad_norm": 1.3838168382644653,
+      "learning_rate": 8.216029876472279e-06,
+      "loss": 2.3202,
+      "step": 2803
+    },
+    {
+      "epoch": 0.7973601426565639,
+      "grad_norm": 1.4264721870422363,
+      "learning_rate": 8.204538925596093e-06,
+      "loss": 2.2274,
+      "step": 2804
+    },
+    {
+      "epoch": 0.797644507900022,
+      "grad_norm": 1.5542453527450562,
+      "learning_rate": 8.19304797471991e-06,
+      "loss": 2.0876,
+      "step": 2805
+    },
+    {
+      "epoch": 0.79792887314348,
+      "grad_norm": 1.5131659507751465,
+      "learning_rate": 8.181557023843724e-06,
+      "loss": 1.8612,
+      "step": 2806
+    },
+    {
+      "epoch": 0.7982132383869381,
+      "grad_norm": 1.5086268186569214,
+      "learning_rate": 8.170066072967538e-06,
+      "loss": 1.7961,
+      "step": 2807
+    },
+    {
+      "epoch": 0.7984976036303962,
+      "grad_norm": 1.5793956518173218,
+      "learning_rate": 8.158575122091353e-06,
+      "loss": 1.6068,
+      "step": 2808
+    },
+    {
+      "epoch": 0.7987819688738544,
+      "grad_norm": 1.7621535062789917,
+      "learning_rate": 8.147084171215169e-06,
+      "loss": 2.4022,
+      "step": 2809
+    },
+    {
+      "epoch": 0.7990663341173125,
+      "grad_norm": 1.5189746618270874,
+      "learning_rate": 8.135593220338983e-06,
+      "loss": 2.3452,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7993506993607706,
+      "grad_norm": 1.418518304824829,
+      "learning_rate": 8.1241022694628e-06,
+      "loss": 1.9866,
+      "step": 2811
+    },
+    {
+      "epoch": 0.7996350646042287,
+      "grad_norm": 1.5955320596694946,
+      "learning_rate": 8.112611318586614e-06,
+      "loss": 2.1267,
+      "step": 2812
+    },
+    {
+      "epoch": 0.7999194298476868,
+      "grad_norm": 1.7210025787353516,
+      "learning_rate": 8.101120367710429e-06,
+      "loss": 1.9749,
+      "step": 2813
+    },
+    {
+      "epoch": 0.800203795091145,
+      "grad_norm": 1.5319161415100098,
+      "learning_rate": 8.089629416834243e-06,
+      "loss": 1.9589,
+      "step": 2814
+    },
+    {
+      "epoch": 0.8004881603346031,
+      "grad_norm": 1.6114243268966675,
+      "learning_rate": 8.07813846595806e-06,
+      "loss": 1.6376,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8007725255780612,
+      "grad_norm": 1.6235712766647339,
+      "learning_rate": 8.066647515081874e-06,
+      "loss": 1.7676,
+      "step": 2816
+    },
+    {
+      "epoch": 0.8010568908215193,
+      "grad_norm": 1.5523582696914673,
+      "learning_rate": 8.055156564205688e-06,
+      "loss": 2.7133,
+      "step": 2817
+    },
+    {
+      "epoch": 0.8013412560649774,
+      "grad_norm": 1.4934026002883911,
+      "learning_rate": 8.043665613329503e-06,
+      "loss": 2.5263,
+      "step": 2818
+    },
+    {
+      "epoch": 0.8016256213084356,
+      "grad_norm": 1.4201688766479492,
+      "learning_rate": 8.032174662453317e-06,
+      "loss": 2.214,
+      "step": 2819
+    },
+    {
+      "epoch": 0.8019099865518937,
+      "grad_norm": 1.5365192890167236,
+      "learning_rate": 8.020683711577134e-06,
+      "loss": 2.2509,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8021943517953518,
+      "grad_norm": 1.5041007995605469,
+      "learning_rate": 8.00919276070095e-06,
+      "loss": 2.1733,
+      "step": 2821
+    },
+    {
+      "epoch": 0.8024787170388099,
+      "grad_norm": 1.4607468843460083,
+      "learning_rate": 7.997701809824764e-06,
+      "loss": 1.8385,
+      "step": 2822
+    },
+    {
+      "epoch": 0.802763082282268,
+      "grad_norm": 1.6158239841461182,
+      "learning_rate": 7.986210858948579e-06,
+      "loss": 1.8724,
+      "step": 2823
+    },
+    {
+      "epoch": 0.8030474475257262,
+      "grad_norm": 1.6089496612548828,
+      "learning_rate": 7.974719908072393e-06,
+      "loss": 1.746,
+      "step": 2824
+    },
+    {
+      "epoch": 0.8033318127691843,
+      "grad_norm": 1.6502569913864136,
+      "learning_rate": 7.963228957196208e-06,
+      "loss": 2.6797,
+      "step": 2825
+    },
+    {
+      "epoch": 0.8036161780126424,
+      "grad_norm": 1.4397003650665283,
+      "learning_rate": 7.951738006320024e-06,
+      "loss": 2.2595,
+      "step": 2826
+    },
+    {
+      "epoch": 0.8039005432561005,
+      "grad_norm": 1.5017975568771362,
+      "learning_rate": 7.940247055443838e-06,
+      "loss": 2.3404,
+      "step": 2827
+    },
+    {
+      "epoch": 0.8041849084995586,
+      "grad_norm": 1.4788340330123901,
+      "learning_rate": 7.928756104567653e-06,
+      "loss": 2.2218,
+      "step": 2828
+    },
+    {
+      "epoch": 0.8044692737430168,
+      "grad_norm": 1.5507515668869019,
+      "learning_rate": 7.917265153691469e-06,
+      "loss": 2.0377,
+      "step": 2829
+    },
+    {
+      "epoch": 0.8047536389864749,
+      "grad_norm": 1.5333271026611328,
+      "learning_rate": 7.905774202815284e-06,
+      "loss": 1.948,
+      "step": 2830
+    },
+    {
+      "epoch": 0.805038004229933,
+      "grad_norm": 1.4439997673034668,
+      "learning_rate": 7.894283251939098e-06,
+      "loss": 1.7283,
+      "step": 2831
+    },
+    {
+      "epoch": 0.8053223694733911,
+      "grad_norm": 1.5981254577636719,
+      "learning_rate": 7.882792301062914e-06,
+      "loss": 1.899,
+      "step": 2832
+    },
+    {
+      "epoch": 0.8056067347168492,
+      "grad_norm": 1.6823999881744385,
+      "learning_rate": 7.871301350186729e-06,
+      "loss": 2.8227,
+      "step": 2833
+    },
+    {
+      "epoch": 0.8058910999603074,
+      "grad_norm": 1.593544840812683,
+      "learning_rate": 7.859810399310543e-06,
+      "loss": 2.4295,
+      "step": 2834
+    },
+    {
+      "epoch": 0.8061754652037655,
+      "grad_norm": 1.3846356868743896,
+      "learning_rate": 7.848319448434358e-06,
+      "loss": 2.0042,
+      "step": 2835
+    },
+    {
+      "epoch": 0.8064598304472236,
+      "grad_norm": 1.4523168802261353,
+      "learning_rate": 7.836828497558174e-06,
+      "loss": 2.0895,
+      "step": 2836
+    },
+    {
+      "epoch": 0.8067441956906817,
+      "grad_norm": 1.5731124877929688,
+      "learning_rate": 7.825337546681988e-06,
+      "loss": 2.2216,
+      "step": 2837
+    },
+    {
+      "epoch": 0.8070285609341398,
+      "grad_norm": 1.5912824869155884,
+      "learning_rate": 7.813846595805805e-06,
+      "loss": 2.0229,
+      "step": 2838
+    },
+    {
+      "epoch": 0.807312926177598,
+      "grad_norm": 1.5180526971817017,
+      "learning_rate": 7.80235564492962e-06,
+      "loss": 1.8176,
+      "step": 2839
+    },
+    {
+      "epoch": 0.8075972914210561,
+      "grad_norm": 1.5623652935028076,
+      "learning_rate": 7.790864694053434e-06,
+      "loss": 1.8123,
+      "step": 2840
+    },
+    {
+      "epoch": 0.8078816566645142,
+      "grad_norm": 1.7535358667373657,
+      "learning_rate": 7.779373743177248e-06,
+      "loss": 2.7814,
+      "step": 2841
+    },
+    {
+      "epoch": 0.8081660219079723,
+      "grad_norm": 1.5314868688583374,
+      "learning_rate": 7.767882792301064e-06,
+      "loss": 2.2653,
+      "step": 2842
+    },
+    {
+      "epoch": 0.8084503871514305,
+      "grad_norm": 1.4723985195159912,
+      "learning_rate": 7.756391841424879e-06,
+      "loss": 2.2257,
+      "step": 2843
+    },
+    {
+      "epoch": 0.8087347523948886,
+      "grad_norm": 1.5310198068618774,
+      "learning_rate": 7.744900890548693e-06,
+      "loss": 2.0548,
+      "step": 2844
+    },
+    {
+      "epoch": 0.8090191176383467,
+      "grad_norm": 1.6289122104644775,
+      "learning_rate": 7.733409939672508e-06,
+      "loss": 2.1181,
+      "step": 2845
+    },
+    {
+      "epoch": 0.8093034828818048,
+      "grad_norm": 1.669812798500061,
+      "learning_rate": 7.721918988796322e-06,
+      "loss": 1.6431,
+      "step": 2846
+    },
+    {
+      "epoch": 0.8095878481252629,
+      "grad_norm": 1.6252498626708984,
+      "learning_rate": 7.710428037920139e-06,
+      "loss": 1.7344,
+      "step": 2847
+    },
+    {
+      "epoch": 0.8098722133687211,
+      "grad_norm": 1.6311345100402832,
+      "learning_rate": 7.698937087043955e-06,
+      "loss": 1.8725,
+      "step": 2848
+    },
+    {
+      "epoch": 0.8101565786121792,
+      "grad_norm": 1.6725844144821167,
+      "learning_rate": 7.68744613616777e-06,
+      "loss": 2.5774,
+      "step": 2849
+    },
+    {
+      "epoch": 0.8104409438556373,
+      "grad_norm": 1.5067769289016724,
+      "learning_rate": 7.675955185291584e-06,
+      "loss": 2.2774,
+      "step": 2850
+    },
+    {
+      "epoch": 0.8107253090990953,
+      "grad_norm": 1.5363556146621704,
+      "learning_rate": 7.664464234415398e-06,
+      "loss": 2.1134,
+      "step": 2851
+    },
+    {
+      "epoch": 0.8110096743425534,
+      "grad_norm": 1.4846501350402832,
+      "learning_rate": 7.652973283539213e-06,
+      "loss": 2.0919,
+      "step": 2852
+    },
+    {
+      "epoch": 0.8112940395860117,
+      "grad_norm": 1.6034198999404907,
+      "learning_rate": 7.641482332663029e-06,
+      "loss": 2.0541,
+      "step": 2853
+    },
+    {
+      "epoch": 0.8115784048294697,
+      "grad_norm": 1.49429452419281,
+      "learning_rate": 7.629991381786843e-06,
+      "loss": 1.6646,
+      "step": 2854
+    },
+    {
+      "epoch": 0.8118627700729278,
+      "grad_norm": 1.4957629442214966,
+      "learning_rate": 7.618500430910658e-06,
+      "loss": 1.8006,
+      "step": 2855
+    },
+    {
+      "epoch": 0.8121471353163859,
+      "grad_norm": 1.5566238164901733,
+      "learning_rate": 7.607009480034473e-06,
+      "loss": 1.7759,
+      "step": 2856
+    },
+    {
+      "epoch": 0.812431500559844,
+      "grad_norm": 1.7334346771240234,
+      "learning_rate": 7.595518529158289e-06,
+      "loss": 2.607,
+      "step": 2857
+    },
+    {
+      "epoch": 0.8127158658033022,
+      "grad_norm": 1.5797719955444336,
+      "learning_rate": 7.584027578282104e-06,
+      "loss": 2.4472,
+      "step": 2858
+    },
+    {
+      "epoch": 0.8130002310467603,
+      "grad_norm": 1.4985685348510742,
+      "learning_rate": 7.572536627405918e-06,
+      "loss": 2.4235,
+      "step": 2859
+    },
+    {
+      "epoch": 0.8132845962902184,
+      "grad_norm": 1.4722380638122559,
+      "learning_rate": 7.561045676529734e-06,
+      "loss": 2.1274,
+      "step": 2860
+    },
+    {
+      "epoch": 0.8135689615336765,
+      "grad_norm": 1.4994633197784424,
+      "learning_rate": 7.549554725653548e-06,
+      "loss": 2.074,
+      "step": 2861
+    },
+    {
+      "epoch": 0.8138533267771346,
+      "grad_norm": 1.689273476600647,
+      "learning_rate": 7.538063774777364e-06,
+      "loss": 2.0419,
+      "step": 2862
+    },
+    {
+      "epoch": 0.8141376920205928,
+      "grad_norm": 1.5861849784851074,
+      "learning_rate": 7.526572823901178e-06,
+      "loss": 1.9478,
+      "step": 2863
+    },
+    {
+      "epoch": 0.8144220572640509,
+      "grad_norm": 1.5588481426239014,
+      "learning_rate": 7.5150818730249935e-06,
+      "loss": 1.7538,
+      "step": 2864
+    },
+    {
+      "epoch": 0.814706422507509,
+      "grad_norm": 1.675533413887024,
+      "learning_rate": 7.503590922148808e-06,
+      "loss": 2.7183,
+      "step": 2865
+    },
+    {
+      "epoch": 0.8149907877509671,
+      "grad_norm": 1.4782207012176514,
+      "learning_rate": 7.492099971272624e-06,
+      "loss": 2.4106,
+      "step": 2866
+    },
+    {
+      "epoch": 0.8152751529944252,
+      "grad_norm": 1.4612501859664917,
+      "learning_rate": 7.480609020396439e-06,
+      "loss": 2.4035,
+      "step": 2867
+    },
+    {
+      "epoch": 0.8155595182378834,
+      "grad_norm": 1.5237157344818115,
+      "learning_rate": 7.469118069520254e-06,
+      "loss": 2.203,
+      "step": 2868
+    },
+    {
+      "epoch": 0.8158438834813415,
+      "grad_norm": 1.601366639137268,
+      "learning_rate": 7.4576271186440685e-06,
+      "loss": 2.0569,
+      "step": 2869
+    },
+    {
+      "epoch": 0.8161282487247996,
+      "grad_norm": 1.5333240032196045,
+      "learning_rate": 7.446136167767883e-06,
+      "loss": 1.932,
+      "step": 2870
+    },
+    {
+      "epoch": 0.8164126139682577,
+      "grad_norm": 1.6135165691375732,
+      "learning_rate": 7.434645216891698e-06,
+      "loss": 1.879,
+      "step": 2871
+    },
+    {
+      "epoch": 0.8166969792117158,
+      "grad_norm": 1.542605996131897,
+      "learning_rate": 7.423154266015513e-06,
+      "loss": 1.7374,
+      "step": 2872
+    },
+    {
+      "epoch": 0.816981344455174,
+      "grad_norm": 1.7298074960708618,
+      "learning_rate": 7.411663315139328e-06,
+      "loss": 2.7893,
+      "step": 2873
+    },
+    {
+      "epoch": 0.8172657096986321,
+      "grad_norm": 1.439632773399353,
+      "learning_rate": 7.400172364263143e-06,
+      "loss": 2.1772,
+      "step": 2874
+    },
+    {
+      "epoch": 0.8175500749420902,
+      "grad_norm": 1.3654167652130127,
+      "learning_rate": 7.388681413386959e-06,
+      "loss": 2.2241,
+      "step": 2875
+    },
+    {
+      "epoch": 0.8178344401855483,
+      "grad_norm": 1.4207921028137207,
+      "learning_rate": 7.377190462510773e-06,
+      "loss": 2.1142,
+      "step": 2876
+    },
+    {
+      "epoch": 0.8181188054290064,
+      "grad_norm": 1.5437437295913696,
+      "learning_rate": 7.365699511634589e-06,
+      "loss": 2.0925,
+      "step": 2877
+    },
+    {
+      "epoch": 0.8184031706724646,
+      "grad_norm": 1.6208105087280273,
+      "learning_rate": 7.354208560758403e-06,
+      "loss": 1.9039,
+      "step": 2878
+    },
+    {
+      "epoch": 0.8186875359159227,
+      "grad_norm": 1.5807172060012817,
+      "learning_rate": 7.3427176098822185e-06,
+      "loss": 1.8186,
+      "step": 2879
+    },
+    {
+      "epoch": 0.8189719011593808,
+      "grad_norm": 1.6478073596954346,
+      "learning_rate": 7.331226659006033e-06,
+      "loss": 1.7717,
+      "step": 2880
+    },
+    {
+      "epoch": 0.8192562664028389,
+      "grad_norm": 1.6527962684631348,
+      "learning_rate": 7.319735708129848e-06,
+      "loss": 2.7834,
+      "step": 2881
+    },
+    {
+      "epoch": 0.819540631646297,
+      "grad_norm": 1.4940738677978516,
+      "learning_rate": 7.308244757253663e-06,
+      "loss": 2.2835,
+      "step": 2882
+    },
+    {
+      "epoch": 0.8198249968897552,
+      "grad_norm": 1.490037202835083,
+      "learning_rate": 7.296753806377478e-06,
+      "loss": 2.1039,
+      "step": 2883
+    },
+    {
+      "epoch": 0.8201093621332133,
+      "grad_norm": 1.4153006076812744,
+      "learning_rate": 7.2852628555012936e-06,
+      "loss": 2.1536,
+      "step": 2884
+    },
+    {
+      "epoch": 0.8203937273766714,
+      "grad_norm": 1.5724902153015137,
+      "learning_rate": 7.273771904625109e-06,
+      "loss": 1.9262,
+      "step": 2885
+    },
+    {
+      "epoch": 0.8206780926201295,
+      "grad_norm": 1.4987521171569824,
+      "learning_rate": 7.262280953748923e-06,
+      "loss": 1.9364,
+      "step": 2886
+    },
+    {
+      "epoch": 0.8209624578635876,
+      "grad_norm": 1.650689721107483,
+      "learning_rate": 7.250790002872739e-06,
+      "loss": 1.7343,
+      "step": 2887
+    },
+    {
+      "epoch": 0.8212468231070458,
+      "grad_norm": 1.5347603559494019,
+      "learning_rate": 7.239299051996553e-06,
+      "loss": 1.9424,
+      "step": 2888
+    },
+    {
+      "epoch": 0.8215311883505039,
+      "grad_norm": 1.7783067226409912,
+      "learning_rate": 7.227808101120369e-06,
+      "loss": 2.81,
+      "step": 2889
+    },
+    {
+      "epoch": 0.821815553593962,
+      "grad_norm": 1.4985435009002686,
+      "learning_rate": 7.216317150244183e-06,
+      "loss": 2.3066,
+      "step": 2890
+    },
+    {
+      "epoch": 0.8220999188374201,
+      "grad_norm": 1.5006333589553833,
+      "learning_rate": 7.204826199367998e-06,
+      "loss": 2.191,
+      "step": 2891
+    },
+    {
+      "epoch": 0.8223842840808783,
+      "grad_norm": 1.4915610551834106,
+      "learning_rate": 7.193335248491813e-06,
+      "loss": 2.0556,
+      "step": 2892
+    },
+    {
+      "epoch": 0.8226686493243364,
+      "grad_norm": 1.5882519483566284,
+      "learning_rate": 7.1818442976156274e-06,
+      "loss": 2.0908,
+      "step": 2893
+    },
+    {
+      "epoch": 0.8229530145677945,
+      "grad_norm": 1.4619516134262085,
+      "learning_rate": 7.170353346739444e-06,
+      "loss": 1.9773,
+      "step": 2894
+    },
+    {
+      "epoch": 0.8232373798112526,
+      "grad_norm": 1.5004616975784302,
+      "learning_rate": 7.158862395863259e-06,
+      "loss": 1.9638,
+      "step": 2895
+    },
+    {
+      "epoch": 0.8235217450547107,
+      "grad_norm": 1.5177069902420044,
+      "learning_rate": 7.1473714449870735e-06,
+      "loss": 1.6566,
+      "step": 2896
+    },
+    {
+      "epoch": 0.8238061102981689,
+      "grad_norm": 1.724021315574646,
+      "learning_rate": 7.135880494110888e-06,
+      "loss": 2.7129,
+      "step": 2897
+    },
+    {
+      "epoch": 0.824090475541627,
+      "grad_norm": 1.537143349647522,
+      "learning_rate": 7.124389543234703e-06,
+      "loss": 2.3473,
+      "step": 2898
+    },
+    {
+      "epoch": 0.824374840785085,
+      "grad_norm": 1.3727134466171265,
+      "learning_rate": 7.112898592358518e-06,
+      "loss": 2.4072,
+      "step": 2899
+    },
+    {
+      "epoch": 0.8246592060285431,
+      "grad_norm": 1.4793521165847778,
+      "learning_rate": 7.101407641482333e-06,
+      "loss": 2.1391,
+      "step": 2900
+    },
+    {
+      "epoch": 0.8249435712720012,
+      "grad_norm": 1.4474788904190063,
+      "learning_rate": 7.089916690606148e-06,
+      "loss": 2.084,
+      "step": 2901
+    },
+    {
+      "epoch": 0.8252279365154594,
+      "grad_norm": 1.6021233797073364,
+      "learning_rate": 7.078425739729963e-06,
+      "loss": 1.8724,
+      "step": 2902
+    },
+    {
+      "epoch": 0.8255123017589175,
+      "grad_norm": 1.5263599157333374,
+      "learning_rate": 7.066934788853778e-06,
+      "loss": 1.7535,
+      "step": 2903
+    },
+    {
+      "epoch": 0.8257966670023756,
+      "grad_norm": 1.5314630270004272,
+      "learning_rate": 7.055443837977594e-06,
+      "loss": 1.7294,
+      "step": 2904
+    },
+    {
+      "epoch": 0.8260810322458337,
+      "grad_norm": 1.6846193075180054,
+      "learning_rate": 7.043952887101408e-06,
+      "loss": 2.6378,
+      "step": 2905
+    },
+    {
+      "epoch": 0.8263653974892918,
+      "grad_norm": 1.6511030197143555,
+      "learning_rate": 7.0324619362252235e-06,
+      "loss": 2.4237,
+      "step": 2906
+    },
+    {
+      "epoch": 0.82664976273275,
+      "grad_norm": 1.453942894935608,
+      "learning_rate": 7.020970985349038e-06,
+      "loss": 2.3405,
+      "step": 2907
+    },
+    {
+      "epoch": 0.8269341279762081,
+      "grad_norm": 1.4035779237747192,
+      "learning_rate": 7.009480034472853e-06,
+      "loss": 2.0887,
+      "step": 2908
+    },
+    {
+      "epoch": 0.8272184932196662,
+      "grad_norm": 1.5787320137023926,
+      "learning_rate": 6.997989083596668e-06,
+      "loss": 2.1067,
+      "step": 2909
+    },
+    {
+      "epoch": 0.8275028584631243,
+      "grad_norm": 1.6145730018615723,
+      "learning_rate": 6.986498132720483e-06,
+      "loss": 1.9487,
+      "step": 2910
+    },
+    {
+      "epoch": 0.8277872237065824,
+      "grad_norm": 1.4792077541351318,
+      "learning_rate": 6.975007181844298e-06,
+      "loss": 1.8149,
+      "step": 2911
+    },
+    {
+      "epoch": 0.8280715889500406,
+      "grad_norm": 1.5759203433990479,
+      "learning_rate": 6.963516230968114e-06,
+      "loss": 1.8977,
+      "step": 2912
+    },
+    {
+      "epoch": 0.8283559541934987,
+      "grad_norm": 1.6006420850753784,
+      "learning_rate": 6.952025280091928e-06,
+      "loss": 2.8743,
+      "step": 2913
+    },
+    {
+      "epoch": 0.8286403194369568,
+      "grad_norm": 1.4774819612503052,
+      "learning_rate": 6.940534329215744e-06,
+      "loss": 2.3842,
+      "step": 2914
+    },
+    {
+      "epoch": 0.8289246846804149,
+      "grad_norm": 1.4915317296981812,
+      "learning_rate": 6.929043378339558e-06,
+      "loss": 2.2381,
+      "step": 2915
+    },
+    {
+      "epoch": 0.829209049923873,
+      "grad_norm": 1.4956070184707642,
+      "learning_rate": 6.9175524274633736e-06,
+      "loss": 2.2751,
+      "step": 2916
+    },
+    {
+      "epoch": 0.8294934151673312,
+      "grad_norm": 1.5358397960662842,
+      "learning_rate": 6.906061476587188e-06,
+      "loss": 2.0869,
+      "step": 2917
+    },
+    {
+      "epoch": 0.8297777804107893,
+      "grad_norm": 1.4759082794189453,
+      "learning_rate": 6.8945705257110026e-06,
+      "loss": 1.6791,
+      "step": 2918
+    },
+    {
+      "epoch": 0.8300621456542474,
+      "grad_norm": 1.5984234809875488,
+      "learning_rate": 6.883079574834818e-06,
+      "loss": 1.7364,
+      "step": 2919
+    },
+    {
+      "epoch": 0.8303465108977055,
+      "grad_norm": 1.6775093078613281,
+      "learning_rate": 6.871588623958632e-06,
+      "loss": 1.734,
+      "step": 2920
+    },
+    {
+      "epoch": 0.8306308761411636,
+      "grad_norm": 1.687727928161621,
+      "learning_rate": 6.860097673082449e-06,
+      "loss": 2.5652,
+      "step": 2921
+    },
+    {
+      "epoch": 0.8309152413846218,
+      "grad_norm": 1.4916083812713623,
+      "learning_rate": 6.848606722206264e-06,
+      "loss": 2.2731,
+      "step": 2922
+    },
+    {
+      "epoch": 0.8311996066280799,
+      "grad_norm": 1.407136082649231,
+      "learning_rate": 6.8371157713300784e-06,
+      "loss": 2.1255,
+      "step": 2923
+    },
+    {
+      "epoch": 0.831483971871538,
+      "grad_norm": 1.390185832977295,
+      "learning_rate": 6.825624820453893e-06,
+      "loss": 2.043,
+      "step": 2924
+    },
+    {
+      "epoch": 0.8317683371149961,
+      "grad_norm": 1.599908471107483,
+      "learning_rate": 6.814133869577708e-06,
+      "loss": 2.3225,
+      "step": 2925
+    },
+    {
+      "epoch": 0.8320527023584542,
+      "grad_norm": 1.5953835248947144,
+      "learning_rate": 6.802642918701523e-06,
+      "loss": 1.7706,
+      "step": 2926
+    },
+    {
+      "epoch": 0.8323370676019124,
+      "grad_norm": 1.5901626348495483,
+      "learning_rate": 6.791151967825338e-06,
+      "loss": 1.7223,
+      "step": 2927
+    },
+    {
+      "epoch": 0.8326214328453705,
+      "grad_norm": 1.617637038230896,
+      "learning_rate": 6.779661016949153e-06,
+      "loss": 1.9098,
+      "step": 2928
+    },
+    {
+      "epoch": 0.8329057980888286,
+      "grad_norm": 1.7057212591171265,
+      "learning_rate": 6.768170066072968e-06,
+      "loss": 2.6873,
+      "step": 2929
+    },
+    {
+      "epoch": 0.8331901633322867,
+      "grad_norm": 1.514479637145996,
+      "learning_rate": 6.7566791151967825e-06,
+      "loss": 2.2949,
+      "step": 2930
+    },
+    {
+      "epoch": 0.8334745285757448,
+      "grad_norm": 1.5152493715286255,
+      "learning_rate": 6.745188164320599e-06,
+      "loss": 2.2854,
+      "step": 2931
+    },
+    {
+      "epoch": 0.833758893819203,
+      "grad_norm": 1.4838507175445557,
+      "learning_rate": 6.733697213444413e-06,
+      "loss": 1.9852,
+      "step": 2932
+    },
+    {
+      "epoch": 0.8340432590626611,
+      "grad_norm": 1.5288348197937012,
+      "learning_rate": 6.7222062625682285e-06,
+      "loss": 1.994,
+      "step": 2933
+    },
+    {
+      "epoch": 0.8343276243061192,
+      "grad_norm": 1.4767318964004517,
+      "learning_rate": 6.710715311692043e-06,
+      "loss": 1.6974,
+      "step": 2934
+    },
+    {
+      "epoch": 0.8346119895495773,
+      "grad_norm": 1.5459532737731934,
+      "learning_rate": 6.699224360815858e-06,
+      "loss": 1.9026,
+      "step": 2935
+    },
+    {
+      "epoch": 0.8348963547930354,
+      "grad_norm": 1.5643906593322754,
+      "learning_rate": 6.687733409939673e-06,
+      "loss": 1.7265,
+      "step": 2936
+    },
+    {
+      "epoch": 0.8351807200364936,
+      "grad_norm": 1.6530026197433472,
+      "learning_rate": 6.676242459063488e-06,
+      "loss": 2.6701,
+      "step": 2937
+    },
+    {
+      "epoch": 0.8354650852799517,
+      "grad_norm": 1.464538812637329,
+      "learning_rate": 6.664751508187303e-06,
+      "loss": 2.3489,
+      "step": 2938
+    },
+    {
+      "epoch": 0.8357494505234098,
+      "grad_norm": 1.469451904296875,
+      "learning_rate": 6.653260557311117e-06,
+      "loss": 2.242,
+      "step": 2939
+    },
+    {
+      "epoch": 0.8360338157668679,
+      "grad_norm": 1.4956061840057373,
+      "learning_rate": 6.641769606434933e-06,
+      "loss": 2.0947,
+      "step": 2940
+    },
+    {
+      "epoch": 0.8363181810103261,
+      "grad_norm": 1.5736510753631592,
+      "learning_rate": 6.630278655558749e-06,
+      "loss": 2.1535,
+      "step": 2941
+    },
+    {
+      "epoch": 0.8366025462537842,
+      "grad_norm": 1.4406914710998535,
+      "learning_rate": 6.618787704682563e-06,
+      "loss": 1.8535,
+      "step": 2942
+    },
+    {
+      "epoch": 0.8368869114972423,
+      "grad_norm": 1.5601533651351929,
+      "learning_rate": 6.6072967538063786e-06,
+      "loss": 1.9428,
+      "step": 2943
+    },
+    {
+      "epoch": 0.8371712767407004,
+      "grad_norm": 1.4611488580703735,
+      "learning_rate": 6.595805802930193e-06,
+      "loss": 1.6424,
+      "step": 2944
+    },
+    {
+      "epoch": 0.8374556419841584,
+      "grad_norm": 1.6867340803146362,
+      "learning_rate": 6.5843148520540075e-06,
+      "loss": 2.6731,
+      "step": 2945
+    },
+    {
+      "epoch": 0.8377400072276167,
+      "grad_norm": 1.4799604415893555,
+      "learning_rate": 6.572823901177823e-06,
+      "loss": 2.3161,
+      "step": 2946
+    },
+    {
+      "epoch": 0.8380243724710748,
+      "grad_norm": 1.4663941860198975,
+      "learning_rate": 6.561332950301637e-06,
+      "loss": 2.2613,
+      "step": 2947
+    },
+    {
+      "epoch": 0.8383087377145328,
+      "grad_norm": 1.496065616607666,
+      "learning_rate": 6.549841999425453e-06,
+      "loss": 1.9605,
+      "step": 2948
+    },
+    {
+      "epoch": 0.8385931029579909,
+      "grad_norm": 1.416529893875122,
+      "learning_rate": 6.538351048549268e-06,
+      "loss": 2.041,
+      "step": 2949
+    },
+    {
+      "epoch": 0.838877468201449,
+      "grad_norm": 1.5188525915145874,
+      "learning_rate": 6.5268600976730834e-06,
+      "loss": 2.0049,
+      "step": 2950
+    },
+    {
+      "epoch": 0.8391618334449072,
+      "grad_norm": 1.6561343669891357,
+      "learning_rate": 6.515369146796898e-06,
+      "loss": 1.7761,
+      "step": 2951
+    },
+    {
+      "epoch": 0.8394461986883653,
+      "grad_norm": 1.8842443227767944,
+      "learning_rate": 6.503878195920713e-06,
+      "loss": 1.8444,
+      "step": 2952
+    },
+    {
+      "epoch": 0.8397305639318234,
+      "grad_norm": 1.5928751230239868,
+      "learning_rate": 6.492387245044528e-06,
+      "loss": 2.7766,
+      "step": 2953
+    },
+    {
+      "epoch": 0.8400149291752815,
+      "grad_norm": 1.551367998123169,
+      "learning_rate": 6.480896294168343e-06,
+      "loss": 2.3994,
+      "step": 2954
+    },
+    {
+      "epoch": 0.8402992944187396,
+      "grad_norm": 1.4871084690093994,
+      "learning_rate": 6.469405343292158e-06,
+      "loss": 2.2015,
+      "step": 2955
+    },
+    {
+      "epoch": 0.8405836596621978,
+      "grad_norm": 1.6148115396499634,
+      "learning_rate": 6.457914392415973e-06,
+      "loss": 2.1799,
+      "step": 2956
+    },
+    {
+      "epoch": 0.8408680249056559,
+      "grad_norm": 1.5156315565109253,
+      "learning_rate": 6.4464234415397874e-06,
+      "loss": 2.031,
+      "step": 2957
+    },
+    {
+      "epoch": 0.841152390149114,
+      "grad_norm": 1.4778313636779785,
+      "learning_rate": 6.434932490663604e-06,
+      "loss": 1.9324,
+      "step": 2958
+    },
+    {
+      "epoch": 0.8414367553925721,
+      "grad_norm": 1.5155595541000366,
+      "learning_rate": 6.423441539787418e-06,
+      "loss": 1.7774,
+      "step": 2959
+    },
+    {
+      "epoch": 0.8417211206360302,
+      "grad_norm": 1.6561177968978882,
+      "learning_rate": 6.4119505889112335e-06,
+      "loss": 1.724,
+      "step": 2960
+    },
+    {
+      "epoch": 0.8420054858794884,
+      "grad_norm": 2.0607149600982666,
+      "learning_rate": 6.400459638035048e-06,
+      "loss": 2.829,
+      "step": 2961
+    },
+    {
+      "epoch": 0.8422898511229465,
+      "grad_norm": 1.4684460163116455,
+      "learning_rate": 6.388968687158863e-06,
+      "loss": 2.5184,
+      "step": 2962
+    },
+    {
+      "epoch": 0.8425742163664046,
+      "grad_norm": 1.4178595542907715,
+      "learning_rate": 6.377477736282678e-06,
+      "loss": 2.165,
+      "step": 2963
+    },
+    {
+      "epoch": 0.8428585816098627,
+      "grad_norm": 1.4885823726654053,
+      "learning_rate": 6.365986785406492e-06,
+      "loss": 2.027,
+      "step": 2964
+    },
+    {
+      "epoch": 0.8431429468533208,
+      "grad_norm": 1.5183528661727905,
+      "learning_rate": 6.354495834530308e-06,
+      "loss": 1.9819,
+      "step": 2965
+    },
+    {
+      "epoch": 0.843427312096779,
+      "grad_norm": 1.3794087171554565,
+      "learning_rate": 6.343004883654122e-06,
+      "loss": 1.5737,
+      "step": 2966
+    },
+    {
+      "epoch": 0.8437116773402371,
+      "grad_norm": 1.5426974296569824,
+      "learning_rate": 6.3315139327779375e-06,
+      "loss": 1.6998,
+      "step": 2967
+    },
+    {
+      "epoch": 0.8439960425836952,
+      "grad_norm": 1.6509177684783936,
+      "learning_rate": 6.320022981901754e-06,
+      "loss": 1.8888,
+      "step": 2968
+    },
+    {
+      "epoch": 0.8442804078271533,
+      "grad_norm": 1.6451570987701416,
+      "learning_rate": 6.308532031025568e-06,
+      "loss": 2.8388,
+      "step": 2969
+    },
+    {
+      "epoch": 0.8445647730706114,
+      "grad_norm": 1.47929048538208,
+      "learning_rate": 6.297041080149383e-06,
+      "loss": 2.4065,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8448491383140696,
+      "grad_norm": 1.3882609605789185,
+      "learning_rate": 6.285550129273198e-06,
+      "loss": 2.1397,
+      "step": 2971
+    },
+    {
+      "epoch": 0.8451335035575277,
+      "grad_norm": 1.4089328050613403,
+      "learning_rate": 6.2740591783970125e-06,
+      "loss": 2.13,
+      "step": 2972
+    },
+    {
+      "epoch": 0.8454178688009858,
+      "grad_norm": 1.5933607816696167,
+      "learning_rate": 6.262568227520828e-06,
+      "loss": 2.1391,
+      "step": 2973
+    },
+    {
+      "epoch": 0.8457022340444439,
+      "grad_norm": 1.4820233583450317,
+      "learning_rate": 6.251077276644642e-06,
+      "loss": 1.9708,
+      "step": 2974
+    },
+    {
+      "epoch": 0.845986599287902,
+      "grad_norm": 1.4918299913406372,
+      "learning_rate": 6.239586325768458e-06,
+      "loss": 1.8649,
+      "step": 2975
+    },
+    {
+      "epoch": 0.8462709645313602,
+      "grad_norm": 1.7036802768707275,
+      "learning_rate": 6.228095374892272e-06,
+      "loss": 1.8209,
+      "step": 2976
+    },
+    {
+      "epoch": 0.8465553297748183,
+      "grad_norm": 1.7224177122116089,
+      "learning_rate": 6.216604424016088e-06,
+      "loss": 2.6908,
+      "step": 2977
+    },
+    {
+      "epoch": 0.8468396950182764,
+      "grad_norm": 1.5237011909484863,
+      "learning_rate": 6.205113473139903e-06,
+      "loss": 2.4184,
+      "step": 2978
+    },
+    {
+      "epoch": 0.8471240602617345,
+      "grad_norm": 1.4036983251571655,
+      "learning_rate": 6.193622522263718e-06,
+      "loss": 2.0392,
+      "step": 2979
+    },
+    {
+      "epoch": 0.8474084255051926,
+      "grad_norm": 1.3745940923690796,
+      "learning_rate": 6.182131571387533e-06,
+      "loss": 2.136,
+      "step": 2980
+    },
+    {
+      "epoch": 0.8476927907486508,
+      "grad_norm": 1.3899210691452026,
+      "learning_rate": 6.170640620511348e-06,
+      "loss": 1.8753,
+      "step": 2981
+    },
+    {
+      "epoch": 0.8479771559921089,
+      "grad_norm": 1.5072089433670044,
+      "learning_rate": 6.159149669635163e-06,
+      "loss": 1.8412,
+      "step": 2982
+    },
+    {
+      "epoch": 0.848261521235567,
+      "grad_norm": 1.4439475536346436,
+      "learning_rate": 6.147658718758978e-06,
+      "loss": 1.5971,
+      "step": 2983
+    },
+    {
+      "epoch": 0.8485458864790251,
+      "grad_norm": 1.7163138389587402,
+      "learning_rate": 6.1361677678827924e-06,
+      "loss": 1.6792,
+      "step": 2984
+    },
+    {
+      "epoch": 0.8488302517224832,
+      "grad_norm": 1.6258455514907837,
+      "learning_rate": 6.124676817006607e-06,
+      "loss": 2.6053,
+      "step": 2985
+    },
+    {
+      "epoch": 0.8491146169659414,
+      "grad_norm": 1.390105962753296,
+      "learning_rate": 6.113185866130423e-06,
+      "loss": 2.3118,
+      "step": 2986
+    },
+    {
+      "epoch": 0.8493989822093995,
+      "grad_norm": 1.5217338800430298,
+      "learning_rate": 6.1016949152542385e-06,
+      "loss": 2.1152,
+      "step": 2987
+    },
+    {
+      "epoch": 0.8496833474528576,
+      "grad_norm": 1.5048413276672363,
+      "learning_rate": 6.090203964378053e-06,
+      "loss": 2.2262,
+      "step": 2988
+    },
+    {
+      "epoch": 0.8499677126963157,
+      "grad_norm": 1.4718562364578247,
+      "learning_rate": 6.078713013501868e-06,
+      "loss": 2.1621,
+      "step": 2989
+    },
+    {
+      "epoch": 0.8502520779397739,
+      "grad_norm": 1.3779963254928589,
+      "learning_rate": 6.067222062625683e-06,
+      "loss": 1.6434,
+      "step": 2990
+    },
+    {
+      "epoch": 0.850536443183232,
+      "grad_norm": 1.5905683040618896,
+      "learning_rate": 6.055731111749497e-06,
+      "loss": 1.7679,
+      "step": 2991
+    },
+    {
+      "epoch": 0.8508208084266901,
+      "grad_norm": 1.6249622106552124,
+      "learning_rate": 6.044240160873313e-06,
+      "loss": 1.661,
+      "step": 2992
+    },
+    {
+      "epoch": 0.8511051736701482,
+      "grad_norm": 1.6707121133804321,
+      "learning_rate": 6.032749209997127e-06,
+      "loss": 2.5263,
+      "step": 2993
+    },
+    {
+      "epoch": 0.8513895389136062,
+      "grad_norm": 1.5046759843826294,
+      "learning_rate": 6.0212582591209425e-06,
+      "loss": 2.2967,
+      "step": 2994
+    },
+    {
+      "epoch": 0.8516739041570645,
+      "grad_norm": 1.5674413442611694,
+      "learning_rate": 6.009767308244759e-06,
+      "loss": 2.0386,
+      "step": 2995
+    },
+    {
+      "epoch": 0.8519582694005225,
+      "grad_norm": 1.4231250286102295,
+      "learning_rate": 5.998276357368573e-06,
+      "loss": 2.318,
+      "step": 2996
+    },
+    {
+      "epoch": 0.8522426346439806,
+      "grad_norm": 1.5579756498336792,
+      "learning_rate": 5.986785406492388e-06,
+      "loss": 2.0154,
+      "step": 2997
+    },
+    {
+      "epoch": 0.8525269998874387,
+      "grad_norm": 1.404220700263977,
+      "learning_rate": 5.975294455616203e-06,
+      "loss": 1.7704,
+      "step": 2998
+    },
+    {
+      "epoch": 0.8528113651308968,
+      "grad_norm": 1.4892497062683105,
+      "learning_rate": 5.9638035047400175e-06,
+      "loss": 1.7247,
+      "step": 2999
+    },
+    {
+      "epoch": 0.853095730374355,
+      "grad_norm": 1.6335406303405762,
+      "learning_rate": 5.952312553863833e-06,
+      "loss": 1.5733,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8533800956178131,
+      "grad_norm": 1.781394124031067,
+      "learning_rate": 5.940821602987647e-06,
+      "loss": 2.7558,
+      "step": 3001
+    },
+    {
+      "epoch": 0.8536644608612712,
+      "grad_norm": 1.4314653873443604,
+      "learning_rate": 5.929330652111463e-06,
+      "loss": 2.3129,
+      "step": 3002
+    },
+    {
+      "epoch": 0.8539488261047293,
+      "grad_norm": 1.5230185985565186,
+      "learning_rate": 5.917839701235277e-06,
+      "loss": 2.3062,
+      "step": 3003
+    },
+    {
+      "epoch": 0.8542331913481874,
+      "grad_norm": 1.4675681591033936,
+      "learning_rate": 5.9063487503590925e-06,
+      "loss": 2.0741,
+      "step": 3004
+    },
+    {
+      "epoch": 0.8545175565916456,
+      "grad_norm": 1.5174206495285034,
+      "learning_rate": 5.894857799482908e-06,
+      "loss": 2.0766,
+      "step": 3005
+    },
+    {
+      "epoch": 0.8548019218351037,
+      "grad_norm": 1.480103850364685,
+      "learning_rate": 5.883366848606723e-06,
+      "loss": 1.6913,
+      "step": 3006
+    },
+    {
+      "epoch": 0.8550862870785618,
+      "grad_norm": 1.5968501567840576,
+      "learning_rate": 5.871875897730538e-06,
+      "loss": 1.7072,
+      "step": 3007
+    },
+    {
+      "epoch": 0.8553706523220199,
+      "grad_norm": 1.528976321220398,
+      "learning_rate": 5.860384946854353e-06,
+      "loss": 1.6712,
+      "step": 3008
+    },
+    {
+      "epoch": 0.855655017565478,
+      "grad_norm": 1.6280978918075562,
+      "learning_rate": 5.8488939959781676e-06,
+      "loss": 2.822,
+      "step": 3009
+    },
+    {
+      "epoch": 0.8559393828089362,
+      "grad_norm": 1.4882813692092896,
+      "learning_rate": 5.837403045101983e-06,
+      "loss": 2.0284,
+      "step": 3010
+    },
+    {
+      "epoch": 0.8562237480523943,
+      "grad_norm": 1.4951153993606567,
+      "learning_rate": 5.825912094225797e-06,
+      "loss": 2.2519,
+      "step": 3011
+    },
+    {
+      "epoch": 0.8565081132958524,
+      "grad_norm": 1.4803876876831055,
+      "learning_rate": 5.814421143349612e-06,
+      "loss": 2.2105,
+      "step": 3012
+    },
+    {
+      "epoch": 0.8567924785393105,
+      "grad_norm": 1.4944387674331665,
+      "learning_rate": 5.802930192473427e-06,
+      "loss": 2.0987,
+      "step": 3013
+    },
+    {
+      "epoch": 0.8570768437827686,
+      "grad_norm": 1.4236083030700684,
+      "learning_rate": 5.7914392415972434e-06,
+      "loss": 1.8932,
+      "step": 3014
+    },
+    {
+      "epoch": 0.8573612090262268,
+      "grad_norm": 1.4983583688735962,
+      "learning_rate": 5.779948290721058e-06,
+      "loss": 1.6304,
+      "step": 3015
+    },
+    {
+      "epoch": 0.8576455742696849,
+      "grad_norm": 1.554520606994629,
+      "learning_rate": 5.768457339844873e-06,
+      "loss": 1.5071,
+      "step": 3016
+    },
+    {
+      "epoch": 0.857929939513143,
+      "grad_norm": 1.8354604244232178,
+      "learning_rate": 5.756966388968688e-06,
+      "loss": 2.5812,
+      "step": 3017
+    },
+    {
+      "epoch": 0.8582143047566011,
+      "grad_norm": 1.4702214002609253,
+      "learning_rate": 5.745475438092502e-06,
+      "loss": 2.4053,
+      "step": 3018
+    },
+    {
+      "epoch": 0.8584986700000592,
+      "grad_norm": 1.5303442478179932,
+      "learning_rate": 5.733984487216318e-06,
+      "loss": 2.3453,
+      "step": 3019
+    },
+    {
+      "epoch": 0.8587830352435174,
+      "grad_norm": 1.5670809745788574,
+      "learning_rate": 5.722493536340132e-06,
+      "loss": 2.2886,
+      "step": 3020
+    },
+    {
+      "epoch": 0.8590674004869755,
+      "grad_norm": 1.6706324815750122,
+      "learning_rate": 5.7110025854639475e-06,
+      "loss": 1.9736,
+      "step": 3021
+    },
+    {
+      "epoch": 0.8593517657304336,
+      "grad_norm": 1.497307300567627,
+      "learning_rate": 5.699511634587762e-06,
+      "loss": 1.6649,
+      "step": 3022
+    },
+    {
+      "epoch": 0.8596361309738917,
+      "grad_norm": 1.4424810409545898,
+      "learning_rate": 5.688020683711578e-06,
+      "loss": 1.7273,
+      "step": 3023
+    },
+    {
+      "epoch": 0.8599204962173498,
+      "grad_norm": 1.5607413053512573,
+      "learning_rate": 5.676529732835393e-06,
+      "loss": 1.7178,
+      "step": 3024
+    },
+    {
+      "epoch": 0.860204861460808,
+      "grad_norm": 1.6604523658752441,
+      "learning_rate": 5.665038781959208e-06,
+      "loss": 2.5183,
+      "step": 3025
+    },
+    {
+      "epoch": 0.8604892267042661,
+      "grad_norm": 1.4753596782684326,
+      "learning_rate": 5.6535478310830225e-06,
+      "loss": 2.4315,
+      "step": 3026
+    },
+    {
+      "epoch": 0.8607735919477242,
+      "grad_norm": 1.5106394290924072,
+      "learning_rate": 5.642056880206838e-06,
+      "loss": 2.2461,
+      "step": 3027
+    },
+    {
+      "epoch": 0.8610579571911823,
+      "grad_norm": 1.4064358472824097,
+      "learning_rate": 5.630565929330652e-06,
+      "loss": 1.9892,
+      "step": 3028
+    },
+    {
+      "epoch": 0.8613423224346404,
+      "grad_norm": 1.5657269954681396,
+      "learning_rate": 5.619074978454468e-06,
+      "loss": 2.0241,
+      "step": 3029
+    },
+    {
+      "epoch": 0.8616266876780986,
+      "grad_norm": 1.3951325416564941,
+      "learning_rate": 5.607584027578282e-06,
+      "loss": 1.962,
+      "step": 3030
+    },
+    {
+      "epoch": 0.8619110529215567,
+      "grad_norm": 1.3956987857818604,
+      "learning_rate": 5.5960930767020975e-06,
+      "loss": 1.7142,
+      "step": 3031
+    },
+    {
+      "epoch": 0.8621954181650148,
+      "grad_norm": 1.8063596487045288,
+      "learning_rate": 5.584602125825913e-06,
+      "loss": 1.7746,
+      "step": 3032
+    },
+    {
+      "epoch": 0.8624797834084729,
+      "grad_norm": 1.7616422176361084,
+      "learning_rate": 5.573111174949728e-06,
+      "loss": 2.6338,
+      "step": 3033
+    },
+    {
+      "epoch": 0.8627641486519311,
+      "grad_norm": 1.576029896736145,
+      "learning_rate": 5.561620224073543e-06,
+      "loss": 2.487,
+      "step": 3034
+    },
+    {
+      "epoch": 0.8630485138953892,
+      "grad_norm": 1.469588279724121,
+      "learning_rate": 5.550129273197358e-06,
+      "loss": 2.3708,
+      "step": 3035
+    },
+    {
+      "epoch": 0.8633328791388473,
+      "grad_norm": 1.5429444313049316,
+      "learning_rate": 5.5386383223211726e-06,
+      "loss": 2.1626,
+      "step": 3036
+    },
+    {
+      "epoch": 0.8636172443823054,
+      "grad_norm": 1.5128809213638306,
+      "learning_rate": 5.527147371444988e-06,
+      "loss": 1.9956,
+      "step": 3037
+    },
+    {
+      "epoch": 0.8639016096257635,
+      "grad_norm": 1.458830714225769,
+      "learning_rate": 5.515656420568802e-06,
+      "loss": 1.9701,
+      "step": 3038
+    },
+    {
+      "epoch": 0.8641859748692217,
+      "grad_norm": 1.585025429725647,
+      "learning_rate": 5.504165469692617e-06,
+      "loss": 1.8178,
+      "step": 3039
+    },
+    {
+      "epoch": 0.8644703401126798,
+      "grad_norm": 1.6396241188049316,
+      "learning_rate": 5.492674518816432e-06,
+      "loss": 1.7052,
+      "step": 3040
+    },
+    {
+      "epoch": 0.8647547053561379,
+      "grad_norm": 1.7168043851852417,
+      "learning_rate": 5.481183567940247e-06,
+      "loss": 2.8315,
+      "step": 3041
+    },
+    {
+      "epoch": 0.865039070599596,
+      "grad_norm": 1.431821346282959,
+      "learning_rate": 5.469692617064063e-06,
+      "loss": 2.2904,
+      "step": 3042
+    },
+    {
+      "epoch": 0.865323435843054,
+      "grad_norm": 1.5637871026992798,
+      "learning_rate": 5.458201666187877e-06,
+      "loss": 2.2041,
+      "step": 3043
+    },
+    {
+      "epoch": 0.8656078010865123,
+      "grad_norm": 1.4506059885025024,
+      "learning_rate": 5.446710715311693e-06,
+      "loss": 2.2017,
+      "step": 3044
+    },
+    {
+      "epoch": 0.8658921663299703,
+      "grad_norm": 1.5410045385360718,
+      "learning_rate": 5.435219764435507e-06,
+      "loss": 1.9422,
+      "step": 3045
+    },
+    {
+      "epoch": 0.8661765315734284,
+      "grad_norm": 1.4460687637329102,
+      "learning_rate": 5.423728813559323e-06,
+      "loss": 1.7594,
+      "step": 3046
+    },
+    {
+      "epoch": 0.8664608968168865,
+      "grad_norm": 1.5550706386566162,
+      "learning_rate": 5.412237862683137e-06,
+      "loss": 1.7515,
+      "step": 3047
+    },
+    {
+      "epoch": 0.8667452620603446,
+      "grad_norm": 1.5752551555633545,
+      "learning_rate": 5.4007469118069524e-06,
+      "loss": 1.644,
+      "step": 3048
+    },
+    {
+      "epoch": 0.8670296273038028,
+      "grad_norm": 1.8624625205993652,
+      "learning_rate": 5.389255960930767e-06,
+      "loss": 2.817,
+      "step": 3049
+    },
+    {
+      "epoch": 0.8673139925472609,
+      "grad_norm": 1.416124939918518,
+      "learning_rate": 5.377765010054582e-06,
+      "loss": 2.1468,
+      "step": 3050
+    },
+    {
+      "epoch": 0.867598357790719,
+      "grad_norm": 1.4211061000823975,
+      "learning_rate": 5.366274059178398e-06,
+      "loss": 2.0979,
+      "step": 3051
+    },
+    {
+      "epoch": 0.8678827230341771,
+      "grad_norm": 1.4626717567443848,
+      "learning_rate": 5.354783108302213e-06,
+      "loss": 2.0164,
+      "step": 3052
+    },
+    {
+      "epoch": 0.8681670882776352,
+      "grad_norm": 1.486775279045105,
+      "learning_rate": 5.3432921574260275e-06,
+      "loss": 2.0579,
+      "step": 3053
+    },
+    {
+      "epoch": 0.8684514535210934,
+      "grad_norm": 1.5586199760437012,
+      "learning_rate": 5.331801206549843e-06,
+      "loss": 2.1097,
+      "step": 3054
+    },
+    {
+      "epoch": 0.8687358187645515,
+      "grad_norm": 1.5133928060531616,
+      "learning_rate": 5.320310255673657e-06,
+      "loss": 1.8172,
+      "step": 3055
+    },
+    {
+      "epoch": 0.8690201840080096,
+      "grad_norm": 1.5888675451278687,
+      "learning_rate": 5.308819304797473e-06,
+      "loss": 1.7316,
+      "step": 3056
+    },
+    {
+      "epoch": 0.8693045492514677,
+      "grad_norm": 1.7980897426605225,
+      "learning_rate": 5.297328353921287e-06,
+      "loss": 2.8514,
+      "step": 3057
+    },
+    {
+      "epoch": 0.8695889144949258,
+      "grad_norm": 1.4976806640625,
+      "learning_rate": 5.285837403045102e-06,
+      "loss": 2.5206,
+      "step": 3058
+    },
+    {
+      "epoch": 0.869873279738384,
+      "grad_norm": 1.5453336238861084,
+      "learning_rate": 5.274346452168917e-06,
+      "loss": 2.4075,
+      "step": 3059
+    },
+    {
+      "epoch": 0.8701576449818421,
+      "grad_norm": 1.3958678245544434,
+      "learning_rate": 5.262855501292733e-06,
+      "loss": 1.9729,
+      "step": 3060
+    },
+    {
+      "epoch": 0.8704420102253002,
+      "grad_norm": 1.5089367628097534,
+      "learning_rate": 5.251364550416548e-06,
+      "loss": 2.061,
+      "step": 3061
+    },
+    {
+      "epoch": 0.8707263754687583,
+      "grad_norm": 1.5372157096862793,
+      "learning_rate": 5.239873599540363e-06,
+      "loss": 1.8647,
+      "step": 3062
+    },
+    {
+      "epoch": 0.8710107407122164,
+      "grad_norm": 1.4671186208724976,
+      "learning_rate": 5.2283826486641775e-06,
+      "loss": 1.6414,
+      "step": 3063
+    },
+    {
+      "epoch": 0.8712951059556746,
+      "grad_norm": 1.5816073417663574,
+      "learning_rate": 5.216891697787992e-06,
+      "loss": 1.8357,
+      "step": 3064
+    },
+    {
+      "epoch": 0.8715794711991327,
+      "grad_norm": 1.703484296798706,
+      "learning_rate": 5.205400746911807e-06,
+      "loss": 2.8886,
+      "step": 3065
+    },
+    {
+      "epoch": 0.8718638364425908,
+      "grad_norm": 1.511811375617981,
+      "learning_rate": 5.193909796035622e-06,
+      "loss": 2.3373,
+      "step": 3066
+    },
+    {
+      "epoch": 0.8721482016860489,
+      "grad_norm": 1.5414661169052124,
+      "learning_rate": 5.182418845159437e-06,
+      "loss": 2.1505,
+      "step": 3067
+    },
+    {
+      "epoch": 0.872432566929507,
+      "grad_norm": 1.5230942964553833,
+      "learning_rate": 5.170927894283252e-06,
+      "loss": 2.1039,
+      "step": 3068
+    },
+    {
+      "epoch": 0.8727169321729652,
+      "grad_norm": 1.4703636169433594,
+      "learning_rate": 5.159436943407068e-06,
+      "loss": 2.0667,
+      "step": 3069
+    },
+    {
+      "epoch": 0.8730012974164233,
+      "grad_norm": 1.432626485824585,
+      "learning_rate": 5.147945992530882e-06,
+      "loss": 1.9027,
+      "step": 3070
+    },
+    {
+      "epoch": 0.8732856626598814,
+      "grad_norm": 1.5644500255584717,
+      "learning_rate": 5.136455041654698e-06,
+      "loss": 1.756,
+      "step": 3071
+    },
+    {
+      "epoch": 0.8735700279033395,
+      "grad_norm": 1.6351237297058105,
+      "learning_rate": 5.124964090778512e-06,
+      "loss": 1.7897,
+      "step": 3072
+    },
+    {
+      "epoch": 0.8738543931467976,
+      "grad_norm": 1.7098308801651,
+      "learning_rate": 5.113473139902328e-06,
+      "loss": 2.7434,
+      "step": 3073
+    },
+    {
+      "epoch": 0.8741387583902558,
+      "grad_norm": 1.4680440425872803,
+      "learning_rate": 5.101982189026142e-06,
+      "loss": 2.1812,
+      "step": 3074
+    },
+    {
+      "epoch": 0.8744231236337139,
+      "grad_norm": 1.4263336658477783,
+      "learning_rate": 5.0904912381499574e-06,
+      "loss": 2.2004,
+      "step": 3075
+    },
+    {
+      "epoch": 0.874707488877172,
+      "grad_norm": 1.4721750020980835,
+      "learning_rate": 5.079000287273772e-06,
+      "loss": 2.1114,
+      "step": 3076
+    },
+    {
+      "epoch": 0.8749918541206301,
+      "grad_norm": 1.6234413385391235,
+      "learning_rate": 5.067509336397587e-06,
+      "loss": 2.1707,
+      "step": 3077
+    },
+    {
+      "epoch": 0.8752762193640882,
+      "grad_norm": 1.5787984132766724,
+      "learning_rate": 5.056018385521402e-06,
+      "loss": 1.9247,
+      "step": 3078
+    },
+    {
+      "epoch": 0.8755605846075464,
+      "grad_norm": 1.4935133457183838,
+      "learning_rate": 5.044527434645218e-06,
+      "loss": 1.78,
+      "step": 3079
+    },
+    {
+      "epoch": 0.8758449498510045,
+      "grad_norm": 1.624784231185913,
+      "learning_rate": 5.0330364837690325e-06,
+      "loss": 1.7151,
+      "step": 3080
+    },
+    {
+      "epoch": 0.8761293150944626,
+      "grad_norm": 1.7499675750732422,
+      "learning_rate": 5.021545532892848e-06,
+      "loss": 2.8374,
+      "step": 3081
+    },
+    {
+      "epoch": 0.8764136803379207,
+      "grad_norm": 1.4725526571273804,
+      "learning_rate": 5.010054582016662e-06,
+      "loss": 2.3646,
+      "step": 3082
+    },
+    {
+      "epoch": 0.8766980455813789,
+      "grad_norm": 1.5129636526107788,
+      "learning_rate": 4.998563631140478e-06,
+      "loss": 2.2392,
+      "step": 3083
+    },
+    {
+      "epoch": 0.876982410824837,
+      "grad_norm": 1.4263098239898682,
+      "learning_rate": 4.987072680264292e-06,
+      "loss": 2.2247,
+      "step": 3084
+    },
+    {
+      "epoch": 0.8772667760682951,
+      "grad_norm": 1.5456427335739136,
+      "learning_rate": 4.975581729388107e-06,
+      "loss": 2.068,
+      "step": 3085
+    },
+    {
+      "epoch": 0.8775511413117532,
+      "grad_norm": 1.4494200944900513,
+      "learning_rate": 4.964090778511923e-06,
+      "loss": 1.893,
+      "step": 3086
+    },
+    {
+      "epoch": 0.8778355065552113,
+      "grad_norm": 1.5122965574264526,
+      "learning_rate": 4.952599827635737e-06,
+      "loss": 1.6924,
+      "step": 3087
+    },
+    {
+      "epoch": 0.8781198717986695,
+      "grad_norm": 1.593127965927124,
+      "learning_rate": 4.941108876759552e-06,
+      "loss": 1.7498,
+      "step": 3088
+    },
+    {
+      "epoch": 0.8784042370421276,
+      "grad_norm": 1.7108339071273804,
+      "learning_rate": 4.929617925883367e-06,
+      "loss": 2.7341,
+      "step": 3089
+    },
+    {
+      "epoch": 0.8786886022855857,
+      "grad_norm": 1.4968798160552979,
+      "learning_rate": 4.9181269750071825e-06,
+      "loss": 2.4878,
+      "step": 3090
+    },
+    {
+      "epoch": 0.8789729675290437,
+      "grad_norm": 1.514430284500122,
+      "learning_rate": 4.906636024130997e-06,
+      "loss": 2.2634,
+      "step": 3091
+    },
+    {
+      "epoch": 0.8792573327725018,
+      "grad_norm": 1.5105502605438232,
+      "learning_rate": 4.895145073254812e-06,
+      "loss": 2.2145,
+      "step": 3092
+    },
+    {
+      "epoch": 0.87954169801596,
+      "grad_norm": 1.5424399375915527,
+      "learning_rate": 4.883654122378627e-06,
+      "loss": 2.1025,
+      "step": 3093
+    },
+    {
+      "epoch": 0.8798260632594181,
+      "grad_norm": 1.4774196147918701,
+      "learning_rate": 4.872163171502442e-06,
+      "loss": 1.7497,
+      "step": 3094
+    },
+    {
+      "epoch": 0.8801104285028762,
+      "grad_norm": 1.5772838592529297,
+      "learning_rate": 4.8606722206262575e-06,
+      "loss": 1.6591,
+      "step": 3095
+    },
+    {
+      "epoch": 0.8803947937463343,
+      "grad_norm": 1.5816643238067627,
+      "learning_rate": 4.849181269750072e-06,
+      "loss": 1.6053,
+      "step": 3096
+    },
+    {
+      "epoch": 0.8806791589897924,
+      "grad_norm": 1.8602535724639893,
+      "learning_rate": 4.837690318873887e-06,
+      "loss": 2.8624,
+      "step": 3097
+    },
+    {
+      "epoch": 0.8809635242332506,
+      "grad_norm": 1.4142849445343018,
+      "learning_rate": 4.826199367997702e-06,
+      "loss": 2.1765,
+      "step": 3098
+    },
+    {
+      "epoch": 0.8812478894767087,
+      "grad_norm": 1.418143391609192,
+      "learning_rate": 4.814708417121517e-06,
+      "loss": 2.142,
+      "step": 3099
+    },
+    {
+      "epoch": 0.8815322547201668,
+      "grad_norm": 1.5086448192596436,
+      "learning_rate": 4.8032174662453326e-06,
+      "loss": 2.2362,
+      "step": 3100
+    },
+    {
+      "epoch": 0.8818166199636249,
+      "grad_norm": 1.5796101093292236,
+      "learning_rate": 4.791726515369147e-06,
+      "loss": 2.2314,
+      "step": 3101
+    },
+    {
+      "epoch": 0.882100985207083,
+      "grad_norm": 1.4776320457458496,
+      "learning_rate": 4.780235564492962e-06,
+      "loss": 2.0073,
+      "step": 3102
+    },
+    {
+      "epoch": 0.8823853504505412,
+      "grad_norm": 1.5670294761657715,
+      "learning_rate": 4.768744613616777e-06,
+      "loss": 1.7092,
+      "step": 3103
+    },
+    {
+      "epoch": 0.8826697156939993,
+      "grad_norm": 1.5772371292114258,
+      "learning_rate": 4.757253662740592e-06,
+      "loss": 1.6023,
+      "step": 3104
+    },
+    {
+      "epoch": 0.8829540809374574,
+      "grad_norm": 1.6653498411178589,
+      "learning_rate": 4.745762711864408e-06,
+      "loss": 2.7154,
+      "step": 3105
+    },
+    {
+      "epoch": 0.8832384461809155,
+      "grad_norm": 1.4277734756469727,
+      "learning_rate": 4.734271760988222e-06,
+      "loss": 2.4472,
+      "step": 3106
+    },
+    {
+      "epoch": 0.8835228114243736,
+      "grad_norm": 1.4376863241195679,
+      "learning_rate": 4.7227808101120374e-06,
+      "loss": 2.1116,
+      "step": 3107
+    },
+    {
+      "epoch": 0.8838071766678318,
+      "grad_norm": 1.4949778318405151,
+      "learning_rate": 4.711289859235852e-06,
+      "loss": 2.2258,
+      "step": 3108
+    },
+    {
+      "epoch": 0.8840915419112899,
+      "grad_norm": 1.6463311910629272,
+      "learning_rate": 4.699798908359667e-06,
+      "loss": 2.1172,
+      "step": 3109
+    },
+    {
+      "epoch": 0.884375907154748,
+      "grad_norm": 1.5631059408187866,
+      "learning_rate": 4.688307957483483e-06,
+      "loss": 1.8706,
+      "step": 3110
+    },
+    {
+      "epoch": 0.8846602723982061,
+      "grad_norm": 1.5554920434951782,
+      "learning_rate": 4.676817006607297e-06,
+      "loss": 1.8467,
+      "step": 3111
+    },
+    {
+      "epoch": 0.8849446376416642,
+      "grad_norm": 1.6312092542648315,
+      "learning_rate": 4.665326055731112e-06,
+      "loss": 1.6004,
+      "step": 3112
+    },
+    {
+      "epoch": 0.8852290028851224,
+      "grad_norm": 1.807760238647461,
+      "learning_rate": 4.653835104854928e-06,
+      "loss": 2.7044,
+      "step": 3113
+    },
+    {
+      "epoch": 0.8855133681285805,
+      "grad_norm": 1.6211071014404297,
+      "learning_rate": 4.642344153978742e-06,
+      "loss": 2.3121,
+      "step": 3114
+    },
+    {
+      "epoch": 0.8857977333720386,
+      "grad_norm": 1.6699950695037842,
+      "learning_rate": 4.630853203102557e-06,
+      "loss": 2.2832,
+      "step": 3115
+    },
+    {
+      "epoch": 0.8860820986154967,
+      "grad_norm": 1.495404839515686,
+      "learning_rate": 4.619362252226372e-06,
+      "loss": 2.0071,
+      "step": 3116
+    },
+    {
+      "epoch": 0.8863664638589548,
+      "grad_norm": 1.4680708646774292,
+      "learning_rate": 4.607871301350187e-06,
+      "loss": 2.1704,
+      "step": 3117
+    },
+    {
+      "epoch": 0.886650829102413,
+      "grad_norm": 1.5451794862747192,
+      "learning_rate": 4.596380350474002e-06,
+      "loss": 1.8087,
+      "step": 3118
+    },
+    {
+      "epoch": 0.8869351943458711,
+      "grad_norm": 1.470167636871338,
+      "learning_rate": 4.584889399597817e-06,
+      "loss": 1.7698,
+      "step": 3119
+    },
+    {
+      "epoch": 0.8872195595893292,
+      "grad_norm": 1.5515146255493164,
+      "learning_rate": 4.573398448721632e-06,
+      "loss": 1.6454,
+      "step": 3120
+    },
+    {
+      "epoch": 0.8875039248327873,
+      "grad_norm": 1.7271732091903687,
+      "learning_rate": 4.561907497845447e-06,
+      "loss": 2.6316,
+      "step": 3121
+    },
+    {
+      "epoch": 0.8877882900762454,
+      "grad_norm": 1.494756817817688,
+      "learning_rate": 4.550416546969262e-06,
+      "loss": 2.4271,
+      "step": 3122
+    },
+    {
+      "epoch": 0.8880726553197036,
+      "grad_norm": 1.4232012033462524,
+      "learning_rate": 4.538925596093077e-06,
+      "loss": 2.2557,
+      "step": 3123
+    },
+    {
+      "epoch": 0.8883570205631617,
+      "grad_norm": 1.4829121828079224,
+      "learning_rate": 4.527434645216892e-06,
+      "loss": 2.0524,
+      "step": 3124
+    },
+    {
+      "epoch": 0.8886413858066198,
+      "grad_norm": 1.5503113269805908,
+      "learning_rate": 4.515943694340707e-06,
+      "loss": 1.9053,
+      "step": 3125
+    },
+    {
+      "epoch": 0.8889257510500779,
+      "grad_norm": 1.4779075384140015,
+      "learning_rate": 4.504452743464522e-06,
+      "loss": 1.8903,
+      "step": 3126
+    },
+    {
+      "epoch": 0.889210116293536,
+      "grad_norm": 1.5646677017211914,
+      "learning_rate": 4.4929617925883376e-06,
+      "loss": 1.7303,
+      "step": 3127
+    },
+    {
+      "epoch": 0.8894944815369942,
+      "grad_norm": 1.709120273590088,
+      "learning_rate": 4.481470841712152e-06,
+      "loss": 1.7812,
+      "step": 3128
+    },
+    {
+      "epoch": 0.8897788467804523,
+      "grad_norm": 1.6658837795257568,
+      "learning_rate": 4.469979890835967e-06,
+      "loss": 2.8185,
+      "step": 3129
+    },
+    {
+      "epoch": 0.8900632120239104,
+      "grad_norm": 1.4445295333862305,
+      "learning_rate": 4.458488939959782e-06,
+      "loss": 2.2779,
+      "step": 3130
+    },
+    {
+      "epoch": 0.8903475772673685,
+      "grad_norm": 1.5250149965286255,
+      "learning_rate": 4.446997989083596e-06,
+      "loss": 2.2863,
+      "step": 3131
+    },
+    {
+      "epoch": 0.8906319425108267,
+      "grad_norm": 1.364835262298584,
+      "learning_rate": 4.435507038207413e-06,
+      "loss": 2.0326,
+      "step": 3132
+    },
+    {
+      "epoch": 0.8909163077542848,
+      "grad_norm": 1.4369089603424072,
+      "learning_rate": 4.424016087331227e-06,
+      "loss": 2.0469,
+      "step": 3133
+    },
+    {
+      "epoch": 0.8912006729977429,
+      "grad_norm": 1.4388717412948608,
+      "learning_rate": 4.4125251364550416e-06,
+      "loss": 1.7271,
+      "step": 3134
+    },
+    {
+      "epoch": 0.891485038241201,
+      "grad_norm": 1.51015305519104,
+      "learning_rate": 4.401034185578857e-06,
+      "loss": 1.7673,
+      "step": 3135
+    },
+    {
+      "epoch": 0.891769403484659,
+      "grad_norm": 1.6226184368133545,
+      "learning_rate": 4.389543234702671e-06,
+      "loss": 1.6646,
+      "step": 3136
+    },
+    {
+      "epoch": 0.8920537687281173,
+      "grad_norm": 1.579350233078003,
+      "learning_rate": 4.378052283826487e-06,
+      "loss": 2.6567,
+      "step": 3137
+    },
+    {
+      "epoch": 0.8923381339715754,
+      "grad_norm": 1.427051067352295,
+      "learning_rate": 4.366561332950302e-06,
+      "loss": 2.104,
+      "step": 3138
+    },
+    {
+      "epoch": 0.8926224992150334,
+      "grad_norm": 1.3372541666030884,
+      "learning_rate": 4.355070382074117e-06,
+      "loss": 2.2061,
+      "step": 3139
+    },
+    {
+      "epoch": 0.8929068644584915,
+      "grad_norm": 1.4767935276031494,
+      "learning_rate": 4.343579431197932e-06,
+      "loss": 2.2415,
+      "step": 3140
+    },
+    {
+      "epoch": 0.8931912297019496,
+      "grad_norm": 1.5126450061798096,
+      "learning_rate": 4.332088480321747e-06,
+      "loss": 2.0232,
+      "step": 3141
+    },
+    {
+      "epoch": 0.8934755949454078,
+      "grad_norm": 1.5773214101791382,
+      "learning_rate": 4.320597529445562e-06,
+      "loss": 2.0683,
+      "step": 3142
+    },
+    {
+      "epoch": 0.8937599601888659,
+      "grad_norm": 1.64472496509552,
+      "learning_rate": 4.309106578569377e-06,
+      "loss": 1.9053,
+      "step": 3143
+    },
+    {
+      "epoch": 0.894044325432324,
+      "grad_norm": 1.5076712369918823,
+      "learning_rate": 4.297615627693192e-06,
+      "loss": 1.7959,
+      "step": 3144
+    },
+    {
+      "epoch": 0.8943286906757821,
+      "grad_norm": 1.8122453689575195,
+      "learning_rate": 4.286124676817007e-06,
+      "loss": 2.8296,
+      "step": 3145
+    },
+    {
+      "epoch": 0.8946130559192402,
+      "grad_norm": 1.409722924232483,
+      "learning_rate": 4.274633725940822e-06,
+      "loss": 2.3145,
+      "step": 3146
+    },
+    {
+      "epoch": 0.8948974211626984,
+      "grad_norm": 1.43026602268219,
+      "learning_rate": 4.263142775064637e-06,
+      "loss": 2.2631,
+      "step": 3147
+    },
+    {
+      "epoch": 0.8951817864061565,
+      "grad_norm": 1.4173415899276733,
+      "learning_rate": 4.251651824188452e-06,
+      "loss": 2.1985,
+      "step": 3148
+    },
+    {
+      "epoch": 0.8954661516496146,
+      "grad_norm": 1.4719431400299072,
+      "learning_rate": 4.240160873312267e-06,
+      "loss": 1.9246,
+      "step": 3149
+    },
+    {
+      "epoch": 0.8957505168930727,
+      "grad_norm": 1.5066853761672974,
+      "learning_rate": 4.228669922436082e-06,
+      "loss": 1.8011,
+      "step": 3150
+    },
+    {
+      "epoch": 0.8960348821365308,
+      "grad_norm": 1.6725085973739624,
+      "learning_rate": 4.217178971559897e-06,
+      "loss": 1.608,
+      "step": 3151
+    },
+    {
+      "epoch": 0.896319247379989,
+      "grad_norm": 1.6424719095230103,
+      "learning_rate": 4.205688020683712e-06,
+      "loss": 1.7574,
+      "step": 3152
+    },
+    {
+      "epoch": 0.8966036126234471,
+      "grad_norm": 1.7039910554885864,
+      "learning_rate": 4.194197069807527e-06,
+      "loss": 2.7038,
+      "step": 3153
+    },
+    {
+      "epoch": 0.8968879778669052,
+      "grad_norm": 1.482321858406067,
+      "learning_rate": 4.182706118931342e-06,
+      "loss": 2.2022,
+      "step": 3154
+    },
+    {
+      "epoch": 0.8971723431103633,
+      "grad_norm": 1.4395989179611206,
+      "learning_rate": 4.171215168055157e-06,
+      "loss": 2.3795,
+      "step": 3155
+    },
+    {
+      "epoch": 0.8974567083538214,
+      "grad_norm": 1.381151795387268,
+      "learning_rate": 4.159724217178972e-06,
+      "loss": 2.2011,
+      "step": 3156
+    },
+    {
+      "epoch": 0.8977410735972796,
+      "grad_norm": 1.4382047653198242,
+      "learning_rate": 4.148233266302787e-06,
+      "loss": 1.9888,
+      "step": 3157
+    },
+    {
+      "epoch": 0.8980254388407377,
+      "grad_norm": 1.566701889038086,
+      "learning_rate": 4.136742315426601e-06,
+      "loss": 1.9158,
+      "step": 3158
+    },
+    {
+      "epoch": 0.8983098040841958,
+      "grad_norm": 1.452852725982666,
+      "learning_rate": 4.125251364550417e-06,
+      "loss": 1.7248,
+      "step": 3159
+    },
+    {
+      "epoch": 0.8985941693276539,
+      "grad_norm": 1.540461778640747,
+      "learning_rate": 4.113760413674232e-06,
+      "loss": 1.8487,
+      "step": 3160
+    },
+    {
+      "epoch": 0.898878534571112,
+      "grad_norm": 2.046682119369507,
+      "learning_rate": 4.1022694627980466e-06,
+      "loss": 2.6915,
+      "step": 3161
+    },
+    {
+      "epoch": 0.8991628998145702,
+      "grad_norm": 1.4611966609954834,
+      "learning_rate": 4.090778511921862e-06,
+      "loss": 2.324,
+      "step": 3162
+    },
+    {
+      "epoch": 0.8994472650580283,
+      "grad_norm": 1.5413190126419067,
+      "learning_rate": 4.079287561045676e-06,
+      "loss": 2.1069,
+      "step": 3163
+    },
+    {
+      "epoch": 0.8997316303014864,
+      "grad_norm": 1.4921023845672607,
+      "learning_rate": 4.067796610169492e-06,
+      "loss": 2.1568,
+      "step": 3164
+    },
+    {
+      "epoch": 0.9000159955449445,
+      "grad_norm": 1.4238756895065308,
+      "learning_rate": 4.056305659293307e-06,
+      "loss": 2.0992,
+      "step": 3165
+    },
+    {
+      "epoch": 0.9003003607884026,
+      "grad_norm": 1.2893463373184204,
+      "learning_rate": 4.044814708417122e-06,
+      "loss": 1.7327,
+      "step": 3166
+    },
+    {
+      "epoch": 0.9005847260318608,
+      "grad_norm": 1.4908276796340942,
+      "learning_rate": 4.033323757540937e-06,
+      "loss": 1.6974,
+      "step": 3167
+    },
+    {
+      "epoch": 0.9008690912753189,
+      "grad_norm": 1.6396807432174683,
+      "learning_rate": 4.0218328066647514e-06,
+      "loss": 1.8493,
+      "step": 3168
+    },
+    {
+      "epoch": 0.901153456518777,
+      "grad_norm": 1.881540298461914,
+      "learning_rate": 4.010341855788567e-06,
+      "loss": 2.7723,
+      "step": 3169
+    },
+    {
+      "epoch": 0.9014378217622351,
+      "grad_norm": 1.3765074014663696,
+      "learning_rate": 3.998850904912382e-06,
+      "loss": 2.4163,
+      "step": 3170
+    },
+    {
+      "epoch": 0.9017221870056932,
+      "grad_norm": 1.4607453346252441,
+      "learning_rate": 3.987359954036197e-06,
+      "loss": 2.1944,
+      "step": 3171
+    },
+    {
+      "epoch": 0.9020065522491514,
+      "grad_norm": 1.5134079456329346,
+      "learning_rate": 3.975869003160012e-06,
+      "loss": 2.1223,
+      "step": 3172
+    },
+    {
+      "epoch": 0.9022909174926095,
+      "grad_norm": 1.612130880355835,
+      "learning_rate": 3.9643780522838265e-06,
+      "loss": 2.0902,
+      "step": 3173
+    },
+    {
+      "epoch": 0.9025752827360676,
+      "grad_norm": 1.4546161890029907,
+      "learning_rate": 3.952887101407642e-06,
+      "loss": 1.6594,
+      "step": 3174
+    },
+    {
+      "epoch": 0.9028596479795257,
+      "grad_norm": 1.5486887693405151,
+      "learning_rate": 3.941396150531457e-06,
+      "loss": 1.7874,
+      "step": 3175
+    },
+    {
+      "epoch": 0.9031440132229838,
+      "grad_norm": 1.671090006828308,
+      "learning_rate": 3.929905199655272e-06,
+      "loss": 1.7745,
+      "step": 3176
+    },
+    {
+      "epoch": 0.903428378466442,
+      "grad_norm": 1.8405499458312988,
+      "learning_rate": 3.918414248779087e-06,
+      "loss": 2.6482,
+      "step": 3177
+    },
+    {
+      "epoch": 0.9037127437099001,
+      "grad_norm": 1.4907866716384888,
+      "learning_rate": 3.906923297902902e-06,
+      "loss": 2.3249,
+      "step": 3178
+    },
+    {
+      "epoch": 0.9039971089533582,
+      "grad_norm": 1.5165135860443115,
+      "learning_rate": 3.895432347026717e-06,
+      "loss": 2.4738,
+      "step": 3179
+    },
+    {
+      "epoch": 0.9042814741968163,
+      "grad_norm": 1.5014480352401733,
+      "learning_rate": 3.883941396150532e-06,
+      "loss": 2.1646,
+      "step": 3180
+    },
+    {
+      "epoch": 0.9045658394402745,
+      "grad_norm": 1.5247701406478882,
+      "learning_rate": 3.872450445274347e-06,
+      "loss": 1.9272,
+      "step": 3181
+    },
+    {
+      "epoch": 0.9048502046837326,
+      "grad_norm": 1.585174798965454,
+      "learning_rate": 3.860959494398161e-06,
+      "loss": 1.9112,
+      "step": 3182
+    },
+    {
+      "epoch": 0.9051345699271907,
+      "grad_norm": 1.5576118230819702,
+      "learning_rate": 3.849468543521977e-06,
+      "loss": 1.716,
+      "step": 3183
+    },
+    {
+      "epoch": 0.9054189351706488,
+      "grad_norm": 1.6155881881713867,
+      "learning_rate": 3.837977592645792e-06,
+      "loss": 1.7258,
+      "step": 3184
+    },
+    {
+      "epoch": 0.9057033004141068,
+      "grad_norm": 1.6325162649154663,
+      "learning_rate": 3.826486641769606e-06,
+      "loss": 2.4844,
+      "step": 3185
+    },
+    {
+      "epoch": 0.9059876656575651,
+      "grad_norm": 1.38741934299469,
+      "learning_rate": 3.814995690893422e-06,
+      "loss": 2.1399,
+      "step": 3186
+    },
+    {
+      "epoch": 0.9062720309010232,
+      "grad_norm": 1.4506083726882935,
+      "learning_rate": 3.8035047400172366e-06,
+      "loss": 2.076,
+      "step": 3187
+    },
+    {
+      "epoch": 0.9065563961444812,
+      "grad_norm": 1.423922061920166,
+      "learning_rate": 3.792013789141052e-06,
+      "loss": 2.0409,
+      "step": 3188
+    },
+    {
+      "epoch": 0.9068407613879393,
+      "grad_norm": 1.591650366783142,
+      "learning_rate": 3.780522838264867e-06,
+      "loss": 2.1596,
+      "step": 3189
+    },
+    {
+      "epoch": 0.9071251266313974,
+      "grad_norm": 1.5626674890518188,
+      "learning_rate": 3.769031887388682e-06,
+      "loss": 1.9831,
+      "step": 3190
+    },
+    {
+      "epoch": 0.9074094918748556,
+      "grad_norm": 1.5110690593719482,
+      "learning_rate": 3.7575409365124967e-06,
+      "loss": 1.9154,
+      "step": 3191
+    },
+    {
+      "epoch": 0.9076938571183137,
+      "grad_norm": 1.5417031049728394,
+      "learning_rate": 3.746049985636312e-06,
+      "loss": 1.704,
+      "step": 3192
+    },
+    {
+      "epoch": 0.9079782223617718,
+      "grad_norm": 1.6698137521743774,
+      "learning_rate": 3.734559034760127e-06,
+      "loss": 2.7173,
+      "step": 3193
+    },
+    {
+      "epoch": 0.9082625876052299,
+      "grad_norm": 1.58273446559906,
+      "learning_rate": 3.7230680838839415e-06,
+      "loss": 2.3419,
+      "step": 3194
+    },
+    {
+      "epoch": 0.908546952848688,
+      "grad_norm": 1.4195032119750977,
+      "learning_rate": 3.7115771330077564e-06,
+      "loss": 2.184,
+      "step": 3195
+    },
+    {
+      "epoch": 0.9088313180921462,
+      "grad_norm": 1.9811179637908936,
+      "learning_rate": 3.7000861821315713e-06,
+      "loss": 2.1592,
+      "step": 3196
+    },
+    {
+      "epoch": 0.9091156833356043,
+      "grad_norm": 1.8835865259170532,
+      "learning_rate": 3.6885952312553867e-06,
+      "loss": 2.1639,
+      "step": 3197
+    },
+    {
+      "epoch": 0.9094000485790624,
+      "grad_norm": 1.4895811080932617,
+      "learning_rate": 3.6771042803792016e-06,
+      "loss": 1.8653,
+      "step": 3198
+    },
+    {
+      "epoch": 0.9096844138225205,
+      "grad_norm": 1.5817967653274536,
+      "learning_rate": 3.6656133295030165e-06,
+      "loss": 1.8198,
+      "step": 3199
+    },
+    {
+      "epoch": 0.9099687790659786,
+      "grad_norm": 1.4373753070831299,
+      "learning_rate": 3.6541223786268314e-06,
+      "loss": 1.5907,
+      "step": 3200
+    },
+    {
+      "epoch": 0.9102531443094368,
+      "grad_norm": 1.6298754215240479,
+      "learning_rate": 3.6426314277506468e-06,
+      "loss": 2.7107,
+      "step": 3201
+    },
+    {
+      "epoch": 0.9105375095528949,
+      "grad_norm": 1.4616210460662842,
+      "learning_rate": 3.6311404768744617e-06,
+      "loss": 2.2934,
+      "step": 3202
+    },
+    {
+      "epoch": 0.910821874796353,
+      "grad_norm": 1.4932990074157715,
+      "learning_rate": 3.6196495259982766e-06,
+      "loss": 1.951,
+      "step": 3203
+    },
+    {
+      "epoch": 0.9111062400398111,
+      "grad_norm": 1.4640005826950073,
+      "learning_rate": 3.6081585751220915e-06,
+      "loss": 1.9661,
+      "step": 3204
+    },
+    {
+      "epoch": 0.9113906052832692,
+      "grad_norm": 1.583443284034729,
+      "learning_rate": 3.5966676242459065e-06,
+      "loss": 2.0104,
+      "step": 3205
+    },
+    {
+      "epoch": 0.9116749705267274,
+      "grad_norm": 1.4007854461669922,
+      "learning_rate": 3.585176673369722e-06,
+      "loss": 1.9239,
+      "step": 3206
+    },
+    {
+      "epoch": 0.9119593357701855,
+      "grad_norm": 1.590077519416809,
+      "learning_rate": 3.5736857224935367e-06,
+      "loss": 1.8776,
+      "step": 3207
+    },
+    {
+      "epoch": 0.9122437010136436,
+      "grad_norm": 1.5077543258666992,
+      "learning_rate": 3.5621947716173517e-06,
+      "loss": 1.6225,
+      "step": 3208
+    },
+    {
+      "epoch": 0.9125280662571017,
+      "grad_norm": 1.8528907299041748,
+      "learning_rate": 3.5507038207411666e-06,
+      "loss": 2.6378,
+      "step": 3209
+    },
+    {
+      "epoch": 0.9128124315005598,
+      "grad_norm": 1.5426160097122192,
+      "learning_rate": 3.5392128698649815e-06,
+      "loss": 2.4834,
+      "step": 3210
+    },
+    {
+      "epoch": 0.913096796744018,
+      "grad_norm": 1.3771498203277588,
+      "learning_rate": 3.527721918988797e-06,
+      "loss": 2.1898,
+      "step": 3211
+    },
+    {
+      "epoch": 0.9133811619874761,
+      "grad_norm": 1.4479544162750244,
+      "learning_rate": 3.5162309681126118e-06,
+      "loss": 2.2442,
+      "step": 3212
+    },
+    {
+      "epoch": 0.9136655272309342,
+      "grad_norm": 1.5544335842132568,
+      "learning_rate": 3.5047400172364267e-06,
+      "loss": 1.9227,
+      "step": 3213
+    },
+    {
+      "epoch": 0.9139498924743923,
+      "grad_norm": 1.4518970251083374,
+      "learning_rate": 3.4932490663602416e-06,
+      "loss": 1.8666,
+      "step": 3214
+    },
+    {
+      "epoch": 0.9142342577178504,
+      "grad_norm": 1.4834986925125122,
+      "learning_rate": 3.481758115484057e-06,
+      "loss": 1.5869,
+      "step": 3215
+    },
+    {
+      "epoch": 0.9145186229613086,
+      "grad_norm": 1.5631589889526367,
+      "learning_rate": 3.470267164607872e-06,
+      "loss": 1.6409,
+      "step": 3216
+    },
+    {
+      "epoch": 0.9148029882047667,
+      "grad_norm": 1.7660410404205322,
+      "learning_rate": 3.4587762137316868e-06,
+      "loss": 2.7967,
+      "step": 3217
+    },
+    {
+      "epoch": 0.9150873534482248,
+      "grad_norm": 1.4638726711273193,
+      "learning_rate": 3.4472852628555013e-06,
+      "loss": 2.276,
+      "step": 3218
+    },
+    {
+      "epoch": 0.9153717186916829,
+      "grad_norm": 1.5217621326446533,
+      "learning_rate": 3.435794311979316e-06,
+      "loss": 2.2588,
+      "step": 3219
+    },
+    {
+      "epoch": 0.915656083935141,
+      "grad_norm": 1.4932063817977905,
+      "learning_rate": 3.424303361103132e-06,
+      "loss": 2.138,
+      "step": 3220
+    },
+    {
+      "epoch": 0.9159404491785992,
+      "grad_norm": 1.5329244136810303,
+      "learning_rate": 3.4128124102269465e-06,
+      "loss": 1.9642,
+      "step": 3221
+    },
+    {
+      "epoch": 0.9162248144220573,
+      "grad_norm": 1.458665132522583,
+      "learning_rate": 3.4013214593507614e-06,
+      "loss": 1.961,
+      "step": 3222
+    },
+    {
+      "epoch": 0.9165091796655154,
+      "grad_norm": 1.5886058807373047,
+      "learning_rate": 3.3898305084745763e-06,
+      "loss": 1.9216,
+      "step": 3223
+    },
+    {
+      "epoch": 0.9167935449089735,
+      "grad_norm": 1.5155012607574463,
+      "learning_rate": 3.3783395575983912e-06,
+      "loss": 1.754,
+      "step": 3224
+    },
+    {
+      "epoch": 0.9170779101524317,
+      "grad_norm": 1.733697772026062,
+      "learning_rate": 3.3668486067222066e-06,
+      "loss": 2.5523,
+      "step": 3225
+    },
+    {
+      "epoch": 0.9173622753958898,
+      "grad_norm": 1.5041455030441284,
+      "learning_rate": 3.3553576558460215e-06,
+      "loss": 2.2183,
+      "step": 3226
+    },
+    {
+      "epoch": 0.9176466406393479,
+      "grad_norm": 1.443900465965271,
+      "learning_rate": 3.3438667049698364e-06,
+      "loss": 2.2949,
+      "step": 3227
+    },
+    {
+      "epoch": 0.917931005882806,
+      "grad_norm": 1.4563335180282593,
+      "learning_rate": 3.3323757540936513e-06,
+      "loss": 2.2165,
+      "step": 3228
+    },
+    {
+      "epoch": 0.9182153711262641,
+      "grad_norm": 1.5183465480804443,
+      "learning_rate": 3.3208848032174667e-06,
+      "loss": 2.1452,
+      "step": 3229
+    },
+    {
+      "epoch": 0.9184997363697223,
+      "grad_norm": 1.4686776399612427,
+      "learning_rate": 3.3093938523412816e-06,
+      "loss": 1.9584,
+      "step": 3230
+    },
+    {
+      "epoch": 0.9187841016131804,
+      "grad_norm": 1.4295648336410522,
+      "learning_rate": 3.2979029014650965e-06,
+      "loss": 1.7519,
+      "step": 3231
+    },
+    {
+      "epoch": 0.9190684668566385,
+      "grad_norm": 1.6376121044158936,
+      "learning_rate": 3.2864119505889114e-06,
+      "loss": 1.6658,
+      "step": 3232
+    },
+    {
+      "epoch": 0.9193528321000966,
+      "grad_norm": 1.892962098121643,
+      "learning_rate": 3.2749209997127264e-06,
+      "loss": 2.629,
+      "step": 3233
+    },
+    {
+      "epoch": 0.9196371973435546,
+      "grad_norm": 1.4390010833740234,
+      "learning_rate": 3.2634300488365417e-06,
+      "loss": 2.4514,
+      "step": 3234
+    },
+    {
+      "epoch": 0.9199215625870129,
+      "grad_norm": 1.4880425930023193,
+      "learning_rate": 3.2519390979603566e-06,
+      "loss": 2.1385,
+      "step": 3235
+    },
+    {
+      "epoch": 0.920205927830471,
+      "grad_norm": 1.5567518472671509,
+      "learning_rate": 3.2404481470841716e-06,
+      "loss": 2.1202,
+      "step": 3236
+    },
+    {
+      "epoch": 0.920490293073929,
+      "grad_norm": 1.4381585121154785,
+      "learning_rate": 3.2289571962079865e-06,
+      "loss": 1.9348,
+      "step": 3237
+    },
+    {
+      "epoch": 0.9207746583173871,
+      "grad_norm": 1.4756149053573608,
+      "learning_rate": 3.217466245331802e-06,
+      "loss": 1.8763,
+      "step": 3238
+    },
+    {
+      "epoch": 0.9210590235608452,
+      "grad_norm": 1.4677278995513916,
+      "learning_rate": 3.2059752944556167e-06,
+      "loss": 1.5907,
+      "step": 3239
+    },
+    {
+      "epoch": 0.9213433888043034,
+      "grad_norm": 1.5422618389129639,
+      "learning_rate": 3.1944843435794317e-06,
+      "loss": 1.6591,
+      "step": 3240
+    },
+    {
+      "epoch": 0.9216277540477615,
+      "grad_norm": 1.696669101715088,
+      "learning_rate": 3.182993392703246e-06,
+      "loss": 2.6402,
+      "step": 3241
+    },
+    {
+      "epoch": 0.9219121192912196,
+      "grad_norm": 1.6080213785171509,
+      "learning_rate": 3.171502441827061e-06,
+      "loss": 2.5319,
+      "step": 3242
+    },
+    {
+      "epoch": 0.9221964845346777,
+      "grad_norm": 1.4872413873672485,
+      "learning_rate": 3.160011490950877e-06,
+      "loss": 2.1528,
+      "step": 3243
+    },
+    {
+      "epoch": 0.9224808497781358,
+      "grad_norm": 1.3897939920425415,
+      "learning_rate": 3.1485205400746913e-06,
+      "loss": 2.2001,
+      "step": 3244
+    },
+    {
+      "epoch": 0.922765215021594,
+      "grad_norm": 1.5631881952285767,
+      "learning_rate": 3.1370295891985063e-06,
+      "loss": 2.1586,
+      "step": 3245
+    },
+    {
+      "epoch": 0.9230495802650521,
+      "grad_norm": 1.603376865386963,
+      "learning_rate": 3.125538638322321e-06,
+      "loss": 1.8584,
+      "step": 3246
+    },
+    {
+      "epoch": 0.9233339455085102,
+      "grad_norm": 1.4845857620239258,
+      "learning_rate": 3.114047687446136e-06,
+      "loss": 1.6781,
+      "step": 3247
+    },
+    {
+      "epoch": 0.9236183107519683,
+      "grad_norm": 1.5035042762756348,
+      "learning_rate": 3.1025567365699515e-06,
+      "loss": 1.5507,
+      "step": 3248
+    },
+    {
+      "epoch": 0.9239026759954264,
+      "grad_norm": 1.5952107906341553,
+      "learning_rate": 3.0910657856937664e-06,
+      "loss": 2.6295,
+      "step": 3249
+    },
+    {
+      "epoch": 0.9241870412388846,
+      "grad_norm": 1.4013127088546753,
+      "learning_rate": 3.0795748348175813e-06,
+      "loss": 2.59,
+      "step": 3250
+    },
+    {
+      "epoch": 0.9244714064823427,
+      "grad_norm": 1.4684206247329712,
+      "learning_rate": 3.0680838839413962e-06,
+      "loss": 2.1826,
+      "step": 3251
+    },
+    {
+      "epoch": 0.9247557717258008,
+      "grad_norm": 1.4950830936431885,
+      "learning_rate": 3.0565929330652116e-06,
+      "loss": 2.2631,
+      "step": 3252
+    },
+    {
+      "epoch": 0.9250401369692589,
+      "grad_norm": 1.56867253780365,
+      "learning_rate": 3.0451019821890265e-06,
+      "loss": 2.0641,
+      "step": 3253
+    },
+    {
+      "epoch": 0.925324502212717,
+      "grad_norm": 1.5613471269607544,
+      "learning_rate": 3.0336110313128414e-06,
+      "loss": 1.8448,
+      "step": 3254
+    },
+    {
+      "epoch": 0.9256088674561752,
+      "grad_norm": 1.413219690322876,
+      "learning_rate": 3.0221200804366563e-06,
+      "loss": 2.0414,
+      "step": 3255
+    },
+    {
+      "epoch": 0.9258932326996333,
+      "grad_norm": 1.5721766948699951,
+      "learning_rate": 3.0106291295604712e-06,
+      "loss": 1.4875,
+      "step": 3256
+    },
+    {
+      "epoch": 0.9261775979430914,
+      "grad_norm": 1.5911824703216553,
+      "learning_rate": 2.9991381786842866e-06,
+      "loss": 2.5628,
+      "step": 3257
+    },
+    {
+      "epoch": 0.9264619631865495,
+      "grad_norm": 1.427064299583435,
+      "learning_rate": 2.9876472278081015e-06,
+      "loss": 2.5189,
+      "step": 3258
+    },
+    {
+      "epoch": 0.9267463284300076,
+      "grad_norm": 1.47379732131958,
+      "learning_rate": 2.9761562769319164e-06,
+      "loss": 2.258,
+      "step": 3259
+    },
+    {
+      "epoch": 0.9270306936734658,
+      "grad_norm": 1.4227964878082275,
+      "learning_rate": 2.9646653260557313e-06,
+      "loss": 1.9227,
+      "step": 3260
+    },
+    {
+      "epoch": 0.9273150589169239,
+      "grad_norm": 1.488387107849121,
+      "learning_rate": 2.9531743751795463e-06,
+      "loss": 1.8924,
+      "step": 3261
+    },
+    {
+      "epoch": 0.927599424160382,
+      "grad_norm": 1.4256445169448853,
+      "learning_rate": 2.9416834243033616e-06,
+      "loss": 1.6688,
+      "step": 3262
+    },
+    {
+      "epoch": 0.9278837894038401,
+      "grad_norm": 1.5635275840759277,
+      "learning_rate": 2.9301924734271765e-06,
+      "loss": 1.7883,
+      "step": 3263
+    },
+    {
+      "epoch": 0.9281681546472982,
+      "grad_norm": 1.5392005443572998,
+      "learning_rate": 2.9187015225509915e-06,
+      "loss": 1.6098,
+      "step": 3264
+    },
+    {
+      "epoch": 0.9284525198907564,
+      "grad_norm": 1.5983314514160156,
+      "learning_rate": 2.907210571674806e-06,
+      "loss": 2.6295,
+      "step": 3265
+    },
+    {
+      "epoch": 0.9287368851342145,
+      "grad_norm": 1.55398690700531,
+      "learning_rate": 2.8957196207986217e-06,
+      "loss": 2.2437,
+      "step": 3266
+    },
+    {
+      "epoch": 0.9290212503776726,
+      "grad_norm": 1.477097511291504,
+      "learning_rate": 2.8842286699224366e-06,
+      "loss": 2.2931,
+      "step": 3267
+    },
+    {
+      "epoch": 0.9293056156211307,
+      "grad_norm": 1.5535800457000732,
+      "learning_rate": 2.872737719046251e-06,
+      "loss": 2.1373,
+      "step": 3268
+    },
+    {
+      "epoch": 0.9295899808645888,
+      "grad_norm": 1.570197582244873,
+      "learning_rate": 2.861246768170066e-06,
+      "loss": 2.0627,
+      "step": 3269
+    },
+    {
+      "epoch": 0.929874346108047,
+      "grad_norm": 1.4615015983581543,
+      "learning_rate": 2.849755817293881e-06,
+      "loss": 1.8171,
+      "step": 3270
+    },
+    {
+      "epoch": 0.9301587113515051,
+      "grad_norm": 1.552833080291748,
+      "learning_rate": 2.8382648664176963e-06,
+      "loss": 1.7411,
+      "step": 3271
+    },
+    {
+      "epoch": 0.9304430765949632,
+      "grad_norm": 1.5648410320281982,
+      "learning_rate": 2.8267739155415112e-06,
+      "loss": 1.6361,
+      "step": 3272
+    },
+    {
+      "epoch": 0.9307274418384213,
+      "grad_norm": 1.7800242900848389,
+      "learning_rate": 2.815282964665326e-06,
+      "loss": 2.5435,
+      "step": 3273
+    },
+    {
+      "epoch": 0.9310118070818795,
+      "grad_norm": 1.5224663019180298,
+      "learning_rate": 2.803792013789141e-06,
+      "loss": 2.364,
+      "step": 3274
+    },
+    {
+      "epoch": 0.9312961723253376,
+      "grad_norm": 1.4838275909423828,
+      "learning_rate": 2.7923010629129564e-06,
+      "loss": 2.1583,
+      "step": 3275
+    },
+    {
+      "epoch": 0.9315805375687957,
+      "grad_norm": 1.468105435371399,
+      "learning_rate": 2.7808101120367714e-06,
+      "loss": 2.3038,
+      "step": 3276
+    },
+    {
+      "epoch": 0.9318649028122538,
+      "grad_norm": 1.6450117826461792,
+      "learning_rate": 2.7693191611605863e-06,
+      "loss": 2.2681,
+      "step": 3277
+    },
+    {
+      "epoch": 0.9321492680557119,
+      "grad_norm": 1.4350481033325195,
+      "learning_rate": 2.757828210284401e-06,
+      "loss": 1.7551,
+      "step": 3278
+    },
+    {
+      "epoch": 0.9324336332991701,
+      "grad_norm": 1.4676655530929565,
+      "learning_rate": 2.746337259408216e-06,
+      "loss": 1.7063,
+      "step": 3279
+    },
+    {
+      "epoch": 0.9327179985426282,
+      "grad_norm": 1.562709927558899,
+      "learning_rate": 2.7348463085320315e-06,
+      "loss": 1.7359,
+      "step": 3280
+    },
+    {
+      "epoch": 0.9330023637860863,
+      "grad_norm": 1.739317536354065,
+      "learning_rate": 2.7233553576558464e-06,
+      "loss": 2.5635,
+      "step": 3281
+    },
+    {
+      "epoch": 0.9332867290295443,
+      "grad_norm": 1.480629801750183,
+      "learning_rate": 2.7118644067796613e-06,
+      "loss": 2.3922,
+      "step": 3282
+    },
+    {
+      "epoch": 0.9335710942730024,
+      "grad_norm": 1.4481021165847778,
+      "learning_rate": 2.7003734559034762e-06,
+      "loss": 2.2305,
+      "step": 3283
+    },
+    {
+      "epoch": 0.9338554595164607,
+      "grad_norm": 1.4051294326782227,
+      "learning_rate": 2.688882505027291e-06,
+      "loss": 1.8938,
+      "step": 3284
+    },
+    {
+      "epoch": 0.9341398247599187,
+      "grad_norm": 1.5220015048980713,
+      "learning_rate": 2.6773915541511065e-06,
+      "loss": 2.045,
+      "step": 3285
+    },
+    {
+      "epoch": 0.9344241900033768,
+      "grad_norm": 1.5136582851409912,
+      "learning_rate": 2.6659006032749214e-06,
+      "loss": 1.8875,
+      "step": 3286
+    },
+    {
+      "epoch": 0.9347085552468349,
+      "grad_norm": 1.5294973850250244,
+      "learning_rate": 2.6544096523987363e-06,
+      "loss": 1.8228,
+      "step": 3287
+    },
+    {
+      "epoch": 0.934992920490293,
+      "grad_norm": 1.4998856782913208,
+      "learning_rate": 2.642918701522551e-06,
+      "loss": 1.5929,
+      "step": 3288
+    },
+    {
+      "epoch": 0.9352772857337512,
+      "grad_norm": 1.751221776008606,
+      "learning_rate": 2.6314277506463666e-06,
+      "loss": 2.6353,
+      "step": 3289
+    },
+    {
+      "epoch": 0.9355616509772093,
+      "grad_norm": 1.5697683095932007,
+      "learning_rate": 2.6199367997701815e-06,
+      "loss": 2.68,
+      "step": 3290
+    },
+    {
+      "epoch": 0.9358460162206674,
+      "grad_norm": 1.5043424367904663,
+      "learning_rate": 2.608445848893996e-06,
+      "loss": 2.1733,
+      "step": 3291
+    },
+    {
+      "epoch": 0.9361303814641255,
+      "grad_norm": 1.4452601671218872,
+      "learning_rate": 2.596954898017811e-06,
+      "loss": 2.0397,
+      "step": 3292
+    },
+    {
+      "epoch": 0.9364147467075836,
+      "grad_norm": 1.478661298751831,
+      "learning_rate": 2.585463947141626e-06,
+      "loss": 1.8285,
+      "step": 3293
+    },
+    {
+      "epoch": 0.9366991119510418,
+      "grad_norm": 1.4804670810699463,
+      "learning_rate": 2.573972996265441e-06,
+      "loss": 1.9527,
+      "step": 3294
+    },
+    {
+      "epoch": 0.9369834771944999,
+      "grad_norm": 1.5798107385635376,
+      "learning_rate": 2.562482045389256e-06,
+      "loss": 1.8947,
+      "step": 3295
+    },
+    {
+      "epoch": 0.937267842437958,
+      "grad_norm": 1.7154254913330078,
+      "learning_rate": 2.550991094513071e-06,
+      "loss": 1.7322,
+      "step": 3296
+    },
+    {
+      "epoch": 0.9375522076814161,
+      "grad_norm": 1.6413829326629639,
+      "learning_rate": 2.539500143636886e-06,
+      "loss": 2.9358,
+      "step": 3297
+    },
+    {
+      "epoch": 0.9378365729248742,
+      "grad_norm": 1.5562938451766968,
+      "learning_rate": 2.528009192760701e-06,
+      "loss": 2.3359,
+      "step": 3298
+    },
+    {
+      "epoch": 0.9381209381683324,
+      "grad_norm": 1.3978956937789917,
+      "learning_rate": 2.5165182418845162e-06,
+      "loss": 2.2585,
+      "step": 3299
+    },
+    {
+      "epoch": 0.9384053034117905,
+      "grad_norm": 1.5323485136032104,
+      "learning_rate": 2.505027291008331e-06,
+      "loss": 2.1701,
+      "step": 3300
+    },
+    {
+      "epoch": 0.9386896686552486,
+      "grad_norm": 1.5247138738632202,
+      "learning_rate": 2.493536340132146e-06,
+      "loss": 2.2914,
+      "step": 3301
+    },
+    {
+      "epoch": 0.9389740338987067,
+      "grad_norm": 1.4950916767120361,
+      "learning_rate": 2.4820453892559614e-06,
+      "loss": 1.9327,
+      "step": 3302
+    },
+    {
+      "epoch": 0.9392583991421648,
+      "grad_norm": 1.5644844770431519,
+      "learning_rate": 2.470554438379776e-06,
+      "loss": 1.7878,
+      "step": 3303
+    },
+    {
+      "epoch": 0.939542764385623,
+      "grad_norm": 1.7486357688903809,
+      "learning_rate": 2.4590634875035913e-06,
+      "loss": 1.5255,
+      "step": 3304
+    },
+    {
+      "epoch": 0.9398271296290811,
+      "grad_norm": 1.6616047620773315,
+      "learning_rate": 2.447572536627406e-06,
+      "loss": 2.4573,
+      "step": 3305
+    },
+    {
+      "epoch": 0.9401114948725392,
+      "grad_norm": 1.5083410739898682,
+      "learning_rate": 2.436081585751221e-06,
+      "loss": 2.2363,
+      "step": 3306
+    },
+    {
+      "epoch": 0.9403958601159973,
+      "grad_norm": 1.4322084188461304,
+      "learning_rate": 2.424590634875036e-06,
+      "loss": 2.1424,
+      "step": 3307
+    },
+    {
+      "epoch": 0.9406802253594554,
+      "grad_norm": 1.466874122619629,
+      "learning_rate": 2.413099683998851e-06,
+      "loss": 2.1491,
+      "step": 3308
+    },
+    {
+      "epoch": 0.9409645906029136,
+      "grad_norm": 1.532874584197998,
+      "learning_rate": 2.4016087331226663e-06,
+      "loss": 1.9099,
+      "step": 3309
+    },
+    {
+      "epoch": 0.9412489558463717,
+      "grad_norm": 1.3527659177780151,
+      "learning_rate": 2.390117782246481e-06,
+      "loss": 1.8746,
+      "step": 3310
+    },
+    {
+      "epoch": 0.9415333210898298,
+      "grad_norm": 1.550384521484375,
+      "learning_rate": 2.378626831370296e-06,
+      "loss": 1.8256,
+      "step": 3311
+    },
+    {
+      "epoch": 0.9418176863332879,
+      "grad_norm": 1.514704704284668,
+      "learning_rate": 2.367135880494111e-06,
+      "loss": 1.5771,
+      "step": 3312
+    },
+    {
+      "epoch": 0.942102051576746,
+      "grad_norm": 1.789668083190918,
+      "learning_rate": 2.355644929617926e-06,
+      "loss": 2.8358,
+      "step": 3313
+    },
+    {
+      "epoch": 0.9423864168202042,
+      "grad_norm": 1.4652148485183716,
+      "learning_rate": 2.3441539787417413e-06,
+      "loss": 2.312,
+      "step": 3314
+    },
+    {
+      "epoch": 0.9426707820636623,
+      "grad_norm": 1.5131518840789795,
+      "learning_rate": 2.332663027865556e-06,
+      "loss": 2.0083,
+      "step": 3315
+    },
+    {
+      "epoch": 0.9429551473071204,
+      "grad_norm": 1.439961314201355,
+      "learning_rate": 2.321172076989371e-06,
+      "loss": 2.076,
+      "step": 3316
+    },
+    {
+      "epoch": 0.9432395125505785,
+      "grad_norm": 1.473046898841858,
+      "learning_rate": 2.309681126113186e-06,
+      "loss": 1.969,
+      "step": 3317
+    },
+    {
+      "epoch": 0.9435238777940366,
+      "grad_norm": 1.4626710414886475,
+      "learning_rate": 2.298190175237001e-06,
+      "loss": 1.7359,
+      "step": 3318
+    },
+    {
+      "epoch": 0.9438082430374948,
+      "grad_norm": 1.4370124340057373,
+      "learning_rate": 2.286699224360816e-06,
+      "loss": 1.7492,
+      "step": 3319
+    },
+    {
+      "epoch": 0.9440926082809529,
+      "grad_norm": 1.5686242580413818,
+      "learning_rate": 2.275208273484631e-06,
+      "loss": 1.683,
+      "step": 3320
+    },
+    {
+      "epoch": 0.944376973524411,
+      "grad_norm": 1.6691089868545532,
+      "learning_rate": 2.263717322608446e-06,
+      "loss": 2.6069,
+      "step": 3321
+    },
+    {
+      "epoch": 0.9446613387678691,
+      "grad_norm": 1.55553138256073,
+      "learning_rate": 2.252226371732261e-06,
+      "loss": 2.3096,
+      "step": 3322
+    },
+    {
+      "epoch": 0.9449457040113273,
+      "grad_norm": 1.4959852695465088,
+      "learning_rate": 2.240735420856076e-06,
+      "loss": 2.4402,
+      "step": 3323
+    },
+    {
+      "epoch": 0.9452300692547854,
+      "grad_norm": 1.4532426595687866,
+      "learning_rate": 2.229244469979891e-06,
+      "loss": 2.0286,
+      "step": 3324
+    },
+    {
+      "epoch": 0.9455144344982435,
+      "grad_norm": 1.565828561782837,
+      "learning_rate": 2.2177535191037063e-06,
+      "loss": 2.217,
+      "step": 3325
+    },
+    {
+      "epoch": 0.9457987997417016,
+      "grad_norm": 1.489469289779663,
+      "learning_rate": 2.2062625682275208e-06,
+      "loss": 1.8138,
+      "step": 3326
+    },
+    {
+      "epoch": 0.9460831649851597,
+      "grad_norm": 1.5402168035507202,
+      "learning_rate": 2.1947716173513357e-06,
+      "loss": 1.8234,
+      "step": 3327
+    },
+    {
+      "epoch": 0.9463675302286179,
+      "grad_norm": 1.4881329536437988,
+      "learning_rate": 2.183280666475151e-06,
+      "loss": 1.4388,
+      "step": 3328
+    },
+    {
+      "epoch": 0.946651895472076,
+      "grad_norm": 1.8732259273529053,
+      "learning_rate": 2.171789715598966e-06,
+      "loss": 2.7893,
+      "step": 3329
+    },
+    {
+      "epoch": 0.946936260715534,
+      "grad_norm": 1.452468752861023,
+      "learning_rate": 2.160298764722781e-06,
+      "loss": 2.2695,
+      "step": 3330
+    },
+    {
+      "epoch": 0.9472206259589921,
+      "grad_norm": 1.4889235496520996,
+      "learning_rate": 2.148807813846596e-06,
+      "loss": 2.2459,
+      "step": 3331
+    },
+    {
+      "epoch": 0.9475049912024502,
+      "grad_norm": 1.4310044050216675,
+      "learning_rate": 2.137316862970411e-06,
+      "loss": 2.1622,
+      "step": 3332
+    },
+    {
+      "epoch": 0.9477893564459084,
+      "grad_norm": 1.490521788597107,
+      "learning_rate": 2.125825912094226e-06,
+      "loss": 1.9426,
+      "step": 3333
+    },
+    {
+      "epoch": 0.9480737216893665,
+      "grad_norm": 1.4765541553497314,
+      "learning_rate": 2.114334961218041e-06,
+      "loss": 1.8035,
+      "step": 3334
+    },
+    {
+      "epoch": 0.9483580869328246,
+      "grad_norm": 1.5888854265213013,
+      "learning_rate": 2.102844010341856e-06,
+      "loss": 1.8977,
+      "step": 3335
+    },
+    {
+      "epoch": 0.9486424521762827,
+      "grad_norm": 1.5300711393356323,
+      "learning_rate": 2.091353059465671e-06,
+      "loss": 1.8379,
+      "step": 3336
+    },
+    {
+      "epoch": 0.9489268174197408,
+      "grad_norm": 1.5477615594863892,
+      "learning_rate": 2.079862108589486e-06,
+      "loss": 2.6793,
+      "step": 3337
+    },
+    {
+      "epoch": 0.949211182663199,
+      "grad_norm": 1.5292421579360962,
+      "learning_rate": 2.0683711577133007e-06,
+      "loss": 2.3593,
+      "step": 3338
+    },
+    {
+      "epoch": 0.9494955479066571,
+      "grad_norm": 1.4530328512191772,
+      "learning_rate": 2.056880206837116e-06,
+      "loss": 2.1165,
+      "step": 3339
+    },
+    {
+      "epoch": 0.9497799131501152,
+      "grad_norm": 1.475620150566101,
+      "learning_rate": 2.045389255960931e-06,
+      "loss": 2.1348,
+      "step": 3340
+    },
+    {
+      "epoch": 0.9500642783935733,
+      "grad_norm": 1.4726742506027222,
+      "learning_rate": 2.033898305084746e-06,
+      "loss": 2.1875,
+      "step": 3341
+    },
+    {
+      "epoch": 0.9503486436370314,
+      "grad_norm": 1.4152458906173706,
+      "learning_rate": 2.022407354208561e-06,
+      "loss": 1.9168,
+      "step": 3342
+    },
+    {
+      "epoch": 0.9506330088804896,
+      "grad_norm": 1.4417401552200317,
+      "learning_rate": 2.0109164033323757e-06,
+      "loss": 1.778,
+      "step": 3343
+    },
+    {
+      "epoch": 0.9509173741239477,
+      "grad_norm": 1.5183541774749756,
+      "learning_rate": 1.999425452456191e-06,
+      "loss": 1.6503,
+      "step": 3344
+    },
+    {
+      "epoch": 0.9512017393674058,
+      "grad_norm": 1.5476274490356445,
+      "learning_rate": 1.987934501580006e-06,
+      "loss": 2.781,
+      "step": 3345
+    },
+    {
+      "epoch": 0.9514861046108639,
+      "grad_norm": 1.5080674886703491,
+      "learning_rate": 1.976443550703821e-06,
+      "loss": 2.0026,
+      "step": 3346
+    },
+    {
+      "epoch": 0.951770469854322,
+      "grad_norm": 1.437429666519165,
+      "learning_rate": 1.964952599827636e-06,
+      "loss": 2.3271,
+      "step": 3347
+    },
+    {
+      "epoch": 0.9520548350977802,
+      "grad_norm": 1.526536226272583,
+      "learning_rate": 1.953461648951451e-06,
+      "loss": 2.0797,
+      "step": 3348
+    },
+    {
+      "epoch": 0.9523392003412383,
+      "grad_norm": 1.5311537981033325,
+      "learning_rate": 1.941970698075266e-06,
+      "loss": 1.9877,
+      "step": 3349
+    },
+    {
+      "epoch": 0.9526235655846964,
+      "grad_norm": 1.4898496866226196,
+      "learning_rate": 1.9304797471990806e-06,
+      "loss": 1.8399,
+      "step": 3350
+    },
+    {
+      "epoch": 0.9529079308281545,
+      "grad_norm": 1.5804930925369263,
+      "learning_rate": 1.918988796322896e-06,
+      "loss": 1.8401,
+      "step": 3351
+    },
+    {
+      "epoch": 0.9531922960716126,
+      "grad_norm": 1.619555950164795,
+      "learning_rate": 1.907497845446711e-06,
+      "loss": 1.7878,
+      "step": 3352
+    },
+    {
+      "epoch": 0.9534766613150708,
+      "grad_norm": 1.7056188583374023,
+      "learning_rate": 1.896006894570526e-06,
+      "loss": 2.5584,
+      "step": 3353
+    },
+    {
+      "epoch": 0.9537610265585289,
+      "grad_norm": 1.5180672407150269,
+      "learning_rate": 1.884515943694341e-06,
+      "loss": 2.3481,
+      "step": 3354
+    },
+    {
+      "epoch": 0.954045391801987,
+      "grad_norm": 1.4261457920074463,
+      "learning_rate": 1.873024992818156e-06,
+      "loss": 2.0368,
+      "step": 3355
+    },
+    {
+      "epoch": 0.9543297570454451,
+      "grad_norm": 1.4864054918289185,
+      "learning_rate": 1.8615340419419707e-06,
+      "loss": 2.1562,
+      "step": 3356
+    },
+    {
+      "epoch": 0.9546141222889032,
+      "grad_norm": 1.6229420900344849,
+      "learning_rate": 1.8500430910657857e-06,
+      "loss": 1.8096,
+      "step": 3357
+    },
+    {
+      "epoch": 0.9548984875323614,
+      "grad_norm": 1.468389630317688,
+      "learning_rate": 1.8385521401896008e-06,
+      "loss": 1.7469,
+      "step": 3358
+    },
+    {
+      "epoch": 0.9551828527758195,
+      "grad_norm": 2.151934862136841,
+      "learning_rate": 1.8270611893134157e-06,
+      "loss": 1.9153,
+      "step": 3359
+    },
+    {
+      "epoch": 0.9554672180192776,
+      "grad_norm": 1.6833430528640747,
+      "learning_rate": 1.8155702384372309e-06,
+      "loss": 1.7895,
+      "step": 3360
+    },
+    {
+      "epoch": 0.9557515832627357,
+      "grad_norm": 1.6929632425308228,
+      "learning_rate": 1.8040792875610458e-06,
+      "loss": 2.6418,
+      "step": 3361
+    },
+    {
+      "epoch": 0.9560359485061938,
+      "grad_norm": 1.4506680965423584,
+      "learning_rate": 1.792588336684861e-06,
+      "loss": 2.3191,
+      "step": 3362
+    },
+    {
+      "epoch": 0.956320313749652,
+      "grad_norm": 1.452488899230957,
+      "learning_rate": 1.7810973858086758e-06,
+      "loss": 2.2384,
+      "step": 3363
+    },
+    {
+      "epoch": 0.9566046789931101,
+      "grad_norm": 1.4387120008468628,
+      "learning_rate": 1.7696064349324907e-06,
+      "loss": 1.9809,
+      "step": 3364
+    },
+    {
+      "epoch": 0.9568890442365682,
+      "grad_norm": 1.4808820486068726,
+      "learning_rate": 1.7581154840563059e-06,
+      "loss": 1.9542,
+      "step": 3365
+    },
+    {
+      "epoch": 0.9571734094800263,
+      "grad_norm": 1.5310330390930176,
+      "learning_rate": 1.7466245331801208e-06,
+      "loss": 1.9475,
+      "step": 3366
+    },
+    {
+      "epoch": 0.9574577747234844,
+      "grad_norm": 1.628447413444519,
+      "learning_rate": 1.735133582303936e-06,
+      "loss": 1.8739,
+      "step": 3367
+    },
+    {
+      "epoch": 0.9577421399669426,
+      "grad_norm": 1.5938904285430908,
+      "learning_rate": 1.7236426314277506e-06,
+      "loss": 1.7342,
+      "step": 3368
+    },
+    {
+      "epoch": 0.9580265052104007,
+      "grad_norm": 1.7595031261444092,
+      "learning_rate": 1.712151680551566e-06,
+      "loss": 2.6314,
+      "step": 3369
+    },
+    {
+      "epoch": 0.9583108704538588,
+      "grad_norm": 1.4485570192337036,
+      "learning_rate": 1.7006607296753807e-06,
+      "loss": 2.3092,
+      "step": 3370
+    },
+    {
+      "epoch": 0.9585952356973169,
+      "grad_norm": 1.341435432434082,
+      "learning_rate": 1.6891697787991956e-06,
+      "loss": 2.1334,
+      "step": 3371
+    },
+    {
+      "epoch": 0.9588796009407751,
+      "grad_norm": 1.397361159324646,
+      "learning_rate": 1.6776788279230107e-06,
+      "loss": 2.1447,
+      "step": 3372
+    },
+    {
+      "epoch": 0.9591639661842332,
+      "grad_norm": 1.4863638877868652,
+      "learning_rate": 1.6661878770468257e-06,
+      "loss": 2.0255,
+      "step": 3373
+    },
+    {
+      "epoch": 0.9594483314276913,
+      "grad_norm": 1.468654751777649,
+      "learning_rate": 1.6546969261706408e-06,
+      "loss": 1.7851,
+      "step": 3374
+    },
+    {
+      "epoch": 0.9597326966711494,
+      "grad_norm": 1.470919132232666,
+      "learning_rate": 1.6432059752944557e-06,
+      "loss": 1.7603,
+      "step": 3375
+    },
+    {
+      "epoch": 0.9600170619146075,
+      "grad_norm": 1.625359296798706,
+      "learning_rate": 1.6317150244182709e-06,
+      "loss": 1.7318,
+      "step": 3376
+    },
+    {
+      "epoch": 0.9603014271580657,
+      "grad_norm": 1.6317858695983887,
+      "learning_rate": 1.6202240735420858e-06,
+      "loss": 2.6952,
+      "step": 3377
+    },
+    {
+      "epoch": 0.9605857924015238,
+      "grad_norm": 1.432524561882019,
+      "learning_rate": 1.608733122665901e-06,
+      "loss": 2.3517,
+      "step": 3378
+    },
+    {
+      "epoch": 0.9608701576449818,
+      "grad_norm": 1.5162678956985474,
+      "learning_rate": 1.5972421717897158e-06,
+      "loss": 2.374,
+      "step": 3379
+    },
+    {
+      "epoch": 0.9611545228884399,
+      "grad_norm": 1.4993191957473755,
+      "learning_rate": 1.5857512209135305e-06,
+      "loss": 2.2357,
+      "step": 3380
+    },
+    {
+      "epoch": 0.961438888131898,
+      "grad_norm": 1.4974647760391235,
+      "learning_rate": 1.5742602700373457e-06,
+      "loss": 1.9876,
+      "step": 3381
+    },
+    {
+      "epoch": 0.9617232533753562,
+      "grad_norm": 1.4593394994735718,
+      "learning_rate": 1.5627693191611606e-06,
+      "loss": 1.7961,
+      "step": 3382
+    },
+    {
+      "epoch": 0.9620076186188143,
+      "grad_norm": 1.453155517578125,
+      "learning_rate": 1.5512783682849757e-06,
+      "loss": 1.6868,
+      "step": 3383
+    },
+    {
+      "epoch": 0.9622919838622724,
+      "grad_norm": 1.6229658126831055,
+      "learning_rate": 1.5397874174087906e-06,
+      "loss": 1.8026,
+      "step": 3384
+    },
+    {
+      "epoch": 0.9625763491057305,
+      "grad_norm": 1.7281709909439087,
+      "learning_rate": 1.5282964665326058e-06,
+      "loss": 2.781,
+      "step": 3385
+    },
+    {
+      "epoch": 0.9628607143491886,
+      "grad_norm": 1.487370491027832,
+      "learning_rate": 1.5168055156564207e-06,
+      "loss": 2.3773,
+      "step": 3386
+    },
+    {
+      "epoch": 0.9631450795926468,
+      "grad_norm": 1.5010656118392944,
+      "learning_rate": 1.5053145647802356e-06,
+      "loss": 2.1523,
+      "step": 3387
+    },
+    {
+      "epoch": 0.9634294448361049,
+      "grad_norm": 1.5002316236495972,
+      "learning_rate": 1.4938236139040508e-06,
+      "loss": 2.0299,
+      "step": 3388
+    },
+    {
+      "epoch": 0.963713810079563,
+      "grad_norm": 1.5855457782745361,
+      "learning_rate": 1.4823326630278657e-06,
+      "loss": 2.1472,
+      "step": 3389
+    },
+    {
+      "epoch": 0.9639981753230211,
+      "grad_norm": 1.4458882808685303,
+      "learning_rate": 1.4708417121516808e-06,
+      "loss": 1.7629,
+      "step": 3390
+    },
+    {
+      "epoch": 0.9642825405664792,
+      "grad_norm": 1.5790014266967773,
+      "learning_rate": 1.4593507612754957e-06,
+      "loss": 1.6007,
+      "step": 3391
+    },
+    {
+      "epoch": 0.9645669058099374,
+      "grad_norm": 1.5922648906707764,
+      "learning_rate": 1.4478598103993109e-06,
+      "loss": 1.9121,
+      "step": 3392
+    },
+    {
+      "epoch": 0.9648512710533955,
+      "grad_norm": 1.733428955078125,
+      "learning_rate": 1.4363688595231256e-06,
+      "loss": 2.5225,
+      "step": 3393
+    },
+    {
+      "epoch": 0.9651356362968536,
+      "grad_norm": 1.4528639316558838,
+      "learning_rate": 1.4248779086469405e-06,
+      "loss": 2.1728,
+      "step": 3394
+    },
+    {
+      "epoch": 0.9654200015403117,
+      "grad_norm": 1.411819338798523,
+      "learning_rate": 1.4133869577707556e-06,
+      "loss": 2.1298,
+      "step": 3395
+    },
+    {
+      "epoch": 0.9657043667837698,
+      "grad_norm": 1.5605361461639404,
+      "learning_rate": 1.4018960068945705e-06,
+      "loss": 2.3136,
+      "step": 3396
+    },
+    {
+      "epoch": 0.965988732027228,
+      "grad_norm": 1.5014349222183228,
+      "learning_rate": 1.3904050560183857e-06,
+      "loss": 2.0917,
+      "step": 3397
+    },
+    {
+      "epoch": 0.9662730972706861,
+      "grad_norm": 1.4370369911193848,
+      "learning_rate": 1.3789141051422006e-06,
+      "loss": 1.8056,
+      "step": 3398
+    },
+    {
+      "epoch": 0.9665574625141442,
+      "grad_norm": 1.5899243354797363,
+      "learning_rate": 1.3674231542660157e-06,
+      "loss": 1.5606,
+      "step": 3399
+    },
+    {
+      "epoch": 0.9668418277576023,
+      "grad_norm": 1.4517974853515625,
+      "learning_rate": 1.3559322033898307e-06,
+      "loss": 1.5532,
+      "step": 3400
+    },
+    {
+      "epoch": 0.9671261930010604,
+      "grad_norm": 1.7034136056900024,
+      "learning_rate": 1.3444412525136456e-06,
+      "loss": 2.5908,
+      "step": 3401
+    },
+    {
+      "epoch": 0.9674105582445186,
+      "grad_norm": 1.504889726638794,
+      "learning_rate": 1.3329503016374607e-06,
+      "loss": 2.4146,
+      "step": 3402
+    },
+    {
+      "epoch": 0.9676949234879767,
+      "grad_norm": 1.5229851007461548,
+      "learning_rate": 1.3214593507612754e-06,
+      "loss": 2.2832,
+      "step": 3403
+    },
+    {
+      "epoch": 0.9679792887314348,
+      "grad_norm": 1.3611642122268677,
+      "learning_rate": 1.3099683998850908e-06,
+      "loss": 2.2432,
+      "step": 3404
+    },
+    {
+      "epoch": 0.9682636539748929,
+      "grad_norm": 1.4854189157485962,
+      "learning_rate": 1.2984774490089055e-06,
+      "loss": 2.1156,
+      "step": 3405
+    },
+    {
+      "epoch": 0.968548019218351,
+      "grad_norm": 1.5415984392166138,
+      "learning_rate": 1.2869864981327206e-06,
+      "loss": 1.859,
+      "step": 3406
+    },
+    {
+      "epoch": 0.9688323844618092,
+      "grad_norm": 1.4857087135314941,
+      "learning_rate": 1.2754955472565355e-06,
+      "loss": 1.6399,
+      "step": 3407
+    },
+    {
+      "epoch": 0.9691167497052673,
+      "grad_norm": 1.531732201576233,
+      "learning_rate": 1.2640045963803504e-06,
+      "loss": 1.6132,
+      "step": 3408
+    },
+    {
+      "epoch": 0.9694011149487254,
+      "grad_norm": 1.7313965559005737,
+      "learning_rate": 1.2525136455041656e-06,
+      "loss": 2.5503,
+      "step": 3409
+    },
+    {
+      "epoch": 0.9696854801921835,
+      "grad_norm": 1.430211067199707,
+      "learning_rate": 1.2410226946279807e-06,
+      "loss": 2.2446,
+      "step": 3410
+    },
+    {
+      "epoch": 0.9699698454356416,
+      "grad_norm": 1.432405948638916,
+      "learning_rate": 1.2295317437517956e-06,
+      "loss": 2.2009,
+      "step": 3411
+    },
+    {
+      "epoch": 0.9702542106790998,
+      "grad_norm": 1.5122909545898438,
+      "learning_rate": 1.2180407928756105e-06,
+      "loss": 2.2101,
+      "step": 3412
+    },
+    {
+      "epoch": 0.9705385759225579,
+      "grad_norm": 1.592177391052246,
+      "learning_rate": 1.2065498419994255e-06,
+      "loss": 2.1943,
+      "step": 3413
+    },
+    {
+      "epoch": 0.970822941166016,
+      "grad_norm": 1.4828754663467407,
+      "learning_rate": 1.1950588911232406e-06,
+      "loss": 2.0317,
+      "step": 3414
+    },
+    {
+      "epoch": 0.9711073064094741,
+      "grad_norm": 1.4396264553070068,
+      "learning_rate": 1.1835679402470555e-06,
+      "loss": 1.7245,
+      "step": 3415
+    },
+    {
+      "epoch": 0.9713916716529322,
+      "grad_norm": 1.5686695575714111,
+      "learning_rate": 1.1720769893708707e-06,
+      "loss": 1.8843,
+      "step": 3416
+    },
+    {
+      "epoch": 0.9716760368963904,
+      "grad_norm": 1.7028121948242188,
+      "learning_rate": 1.1605860384946856e-06,
+      "loss": 2.6639,
+      "step": 3417
+    },
+    {
+      "epoch": 0.9719604021398485,
+      "grad_norm": 1.5416810512542725,
+      "learning_rate": 1.1490950876185005e-06,
+      "loss": 2.38,
+      "step": 3418
+    },
+    {
+      "epoch": 0.9722447673833066,
+      "grad_norm": 1.492241621017456,
+      "learning_rate": 1.1376041367423154e-06,
+      "loss": 2.2302,
+      "step": 3419
+    },
+    {
+      "epoch": 0.9725291326267647,
+      "grad_norm": 1.4459635019302368,
+      "learning_rate": 1.1261131858661306e-06,
+      "loss": 2.0947,
+      "step": 3420
+    },
+    {
+      "epoch": 0.9728134978702229,
+      "grad_norm": 1.649179458618164,
+      "learning_rate": 1.1146222349899455e-06,
+      "loss": 2.1094,
+      "step": 3421
+    },
+    {
+      "epoch": 0.973097863113681,
+      "grad_norm": 1.3947135210037231,
+      "learning_rate": 1.1031312841137604e-06,
+      "loss": 1.8,
+      "step": 3422
+    },
+    {
+      "epoch": 0.9733822283571391,
+      "grad_norm": 1.423638105392456,
+      "learning_rate": 1.0916403332375755e-06,
+      "loss": 1.7245,
+      "step": 3423
+    },
+    {
+      "epoch": 0.9736665936005972,
+      "grad_norm": 1.5806896686553955,
+      "learning_rate": 1.0801493823613904e-06,
+      "loss": 1.8189,
+      "step": 3424
+    },
+    {
+      "epoch": 0.9739509588440552,
+      "grad_norm": 1.8673456907272339,
+      "learning_rate": 1.0686584314852056e-06,
+      "loss": 2.8048,
+      "step": 3425
+    },
+    {
+      "epoch": 0.9742353240875135,
+      "grad_norm": 1.4717233180999756,
+      "learning_rate": 1.0571674806090205e-06,
+      "loss": 2.0696,
+      "step": 3426
+    },
+    {
+      "epoch": 0.9745196893309716,
+      "grad_norm": 1.3976044654846191,
+      "learning_rate": 1.0456765297328354e-06,
+      "loss": 2.1881,
+      "step": 3427
+    },
+    {
+      "epoch": 0.9748040545744296,
+      "grad_norm": 1.4841516017913818,
+      "learning_rate": 1.0341855788566503e-06,
+      "loss": 2.1457,
+      "step": 3428
+    },
+    {
+      "epoch": 0.9750884198178877,
+      "grad_norm": 1.4041268825531006,
+      "learning_rate": 1.0226946279804655e-06,
+      "loss": 1.9074,
+      "step": 3429
+    },
+    {
+      "epoch": 0.9753727850613458,
+      "grad_norm": 1.5698847770690918,
+      "learning_rate": 1.0112036771042804e-06,
+      "loss": 1.8711,
+      "step": 3430
+    },
+    {
+      "epoch": 0.975657150304804,
+      "grad_norm": 1.568758249282837,
+      "learning_rate": 9.997127262280955e-07,
+      "loss": 1.7552,
+      "step": 3431
+    },
+    {
+      "epoch": 0.9759415155482621,
+      "grad_norm": 1.540923833847046,
+      "learning_rate": 9.882217753519104e-07,
+      "loss": 1.724,
+      "step": 3432
+    },
+    {
+      "epoch": 0.9762258807917202,
+      "grad_norm": 1.6241703033447266,
+      "learning_rate": 9.767308244757256e-07,
+      "loss": 2.583,
+      "step": 3433
+    },
+    {
+      "epoch": 0.9765102460351783,
+      "grad_norm": 1.4737218618392944,
+      "learning_rate": 9.652398735995403e-07,
+      "loss": 2.2108,
+      "step": 3434
+    },
+    {
+      "epoch": 0.9767946112786364,
+      "grad_norm": 1.477249026298523,
+      "learning_rate": 9.537489227233554e-07,
+      "loss": 2.4118,
+      "step": 3435
+    },
+    {
+      "epoch": 0.9770789765220946,
+      "grad_norm": 1.4675695896148682,
+      "learning_rate": 9.422579718471705e-07,
+      "loss": 2.2902,
+      "step": 3436
+    },
+    {
+      "epoch": 0.9773633417655527,
+      "grad_norm": 1.5511778593063354,
+      "learning_rate": 9.307670209709854e-07,
+      "loss": 2.0277,
+      "step": 3437
+    },
+    {
+      "epoch": 0.9776477070090108,
+      "grad_norm": 1.3725906610488892,
+      "learning_rate": 9.192760700948004e-07,
+      "loss": 1.8596,
+      "step": 3438
+    },
+    {
+      "epoch": 0.9779320722524689,
+      "grad_norm": 1.5125075578689575,
+      "learning_rate": 9.077851192186154e-07,
+      "loss": 1.551,
+      "step": 3439
+    },
+    {
+      "epoch": 0.978216437495927,
+      "grad_norm": 1.536952257156372,
+      "learning_rate": 8.962941683424305e-07,
+      "loss": 1.6953,
+      "step": 3440
+    },
+    {
+      "epoch": 0.9785008027393852,
+      "grad_norm": 1.8010717630386353,
+      "learning_rate": 8.848032174662454e-07,
+      "loss": 2.8914,
+      "step": 3441
+    },
+    {
+      "epoch": 0.9787851679828433,
+      "grad_norm": 1.4743620157241821,
+      "learning_rate": 8.733122665900604e-07,
+      "loss": 2.3008,
+      "step": 3442
+    },
+    {
+      "epoch": 0.9790695332263014,
+      "grad_norm": 1.5010536909103394,
+      "learning_rate": 8.618213157138753e-07,
+      "loss": 1.9155,
+      "step": 3443
+    },
+    {
+      "epoch": 0.9793538984697595,
+      "grad_norm": 1.4484606981277466,
+      "learning_rate": 8.503303648376903e-07,
+      "loss": 2.0419,
+      "step": 3444
+    },
+    {
+      "epoch": 0.9796382637132176,
+      "grad_norm": 1.5516077280044556,
+      "learning_rate": 8.388394139615054e-07,
+      "loss": 2.034,
+      "step": 3445
+    },
+    {
+      "epoch": 0.9799226289566758,
+      "grad_norm": 1.445167899131775,
+      "learning_rate": 8.273484630853204e-07,
+      "loss": 1.7347,
+      "step": 3446
+    },
+    {
+      "epoch": 0.9802069942001339,
+      "grad_norm": 1.5177215337753296,
+      "learning_rate": 8.158575122091354e-07,
+      "loss": 1.7245,
+      "step": 3447
+    },
+    {
+      "epoch": 0.980491359443592,
+      "grad_norm": 1.4989851713180542,
+      "learning_rate": 8.043665613329505e-07,
+      "loss": 1.7891,
+      "step": 3448
+    },
+    {
+      "epoch": 0.9807757246870501,
+      "grad_norm": 1.808464527130127,
+      "learning_rate": 7.928756104567653e-07,
+      "loss": 2.9027,
+      "step": 3449
+    },
+    {
+      "epoch": 0.9810600899305082,
+      "grad_norm": 1.4827134609222412,
+      "learning_rate": 7.813846595805803e-07,
+      "loss": 2.3215,
+      "step": 3450
+    },
+    {
+      "epoch": 0.9813444551739664,
+      "grad_norm": 1.4429081678390503,
+      "learning_rate": 7.698937087043953e-07,
+      "loss": 2.2759,
+      "step": 3451
+    },
+    {
+      "epoch": 0.9816288204174245,
+      "grad_norm": 1.4872095584869385,
+      "learning_rate": 7.584027578282104e-07,
+      "loss": 2.0688,
+      "step": 3452
+    },
+    {
+      "epoch": 0.9819131856608826,
+      "grad_norm": 1.5652166604995728,
+      "learning_rate": 7.469118069520254e-07,
+      "loss": 2.0771,
+      "step": 3453
+    },
+    {
+      "epoch": 0.9821975509043407,
+      "grad_norm": 1.380708932876587,
+      "learning_rate": 7.354208560758404e-07,
+      "loss": 1.616,
+      "step": 3454
+    },
+    {
+      "epoch": 0.9824819161477988,
+      "grad_norm": 1.489150047302246,
+      "learning_rate": 7.239299051996554e-07,
+      "loss": 1.7372,
+      "step": 3455
+    },
+    {
+      "epoch": 0.982766281391257,
+      "grad_norm": 1.7191890478134155,
+      "learning_rate": 7.124389543234702e-07,
+      "loss": 1.7137,
+      "step": 3456
+    },
+    {
+      "epoch": 0.9830506466347151,
+      "grad_norm": 1.6472156047821045,
+      "learning_rate": 7.009480034472853e-07,
+      "loss": 2.8431,
+      "step": 3457
+    },
+    {
+      "epoch": 0.9833350118781732,
+      "grad_norm": 1.4679713249206543,
+      "learning_rate": 6.894570525711003e-07,
+      "loss": 2.2365,
+      "step": 3458
+    },
+    {
+      "epoch": 0.9836193771216313,
+      "grad_norm": 1.43304443359375,
+      "learning_rate": 6.779661016949153e-07,
+      "loss": 2.2726,
+      "step": 3459
+    },
+    {
+      "epoch": 0.9839037423650894,
+      "grad_norm": 1.4966747760772705,
+      "learning_rate": 6.664751508187304e-07,
+      "loss": 1.9848,
+      "step": 3460
+    },
+    {
+      "epoch": 0.9841881076085476,
+      "grad_norm": 1.5169103145599365,
+      "learning_rate": 6.549841999425454e-07,
+      "loss": 2.0286,
+      "step": 3461
+    },
+    {
+      "epoch": 0.9844724728520057,
+      "grad_norm": 1.6073898077011108,
+      "learning_rate": 6.434932490663603e-07,
+      "loss": 1.8712,
+      "step": 3462
+    },
+    {
+      "epoch": 0.9847568380954638,
+      "grad_norm": 1.4505630731582642,
+      "learning_rate": 6.320022981901752e-07,
+      "loss": 1.6837,
+      "step": 3463
+    },
+    {
+      "epoch": 0.9850412033389219,
+      "grad_norm": 1.5838265419006348,
+      "learning_rate": 6.205113473139904e-07,
+      "loss": 1.7908,
+      "step": 3464
+    },
+    {
+      "epoch": 0.9853255685823801,
+      "grad_norm": 1.5671101808547974,
+      "learning_rate": 6.090203964378053e-07,
+      "loss": 2.6081,
+      "step": 3465
+    },
+    {
+      "epoch": 0.9856099338258382,
+      "grad_norm": 1.4666972160339355,
+      "learning_rate": 5.975294455616203e-07,
+      "loss": 2.4855,
+      "step": 3466
+    },
+    {
+      "epoch": 0.9858942990692963,
+      "grad_norm": 1.4629732370376587,
+      "learning_rate": 5.860384946854353e-07,
+      "loss": 2.1884,
+      "step": 3467
+    },
+    {
+      "epoch": 0.9861786643127544,
+      "grad_norm": 1.4864407777786255,
+      "learning_rate": 5.745475438092502e-07,
+      "loss": 2.1033,
+      "step": 3468
+    },
+    {
+      "epoch": 0.9864630295562125,
+      "grad_norm": 1.498170256614685,
+      "learning_rate": 5.630565929330653e-07,
+      "loss": 2.0259,
+      "step": 3469
+    },
+    {
+      "epoch": 0.9867473947996707,
+      "grad_norm": 1.402068018913269,
+      "learning_rate": 5.515656420568802e-07,
+      "loss": 1.8348,
+      "step": 3470
+    },
+    {
+      "epoch": 0.9870317600431288,
+      "grad_norm": 1.5257517099380493,
+      "learning_rate": 5.400746911806952e-07,
+      "loss": 1.8354,
+      "step": 3471
+    },
+    {
+      "epoch": 0.9873161252865869,
+      "grad_norm": 1.5966296195983887,
+      "learning_rate": 5.285837403045103e-07,
+      "loss": 1.7858,
+      "step": 3472
+    },
+    {
+      "epoch": 0.987600490530045,
+      "grad_norm": 1.67741060256958,
+      "learning_rate": 5.170927894283252e-07,
+      "loss": 2.6827,
+      "step": 3473
+    },
+    {
+      "epoch": 0.987884855773503,
+      "grad_norm": 1.4048272371292114,
+      "learning_rate": 5.056018385521402e-07,
+      "loss": 2.5054,
+      "step": 3474
+    },
+    {
+      "epoch": 0.9881692210169613,
+      "grad_norm": 1.3851925134658813,
+      "learning_rate": 4.941108876759552e-07,
+      "loss": 2.1778,
+      "step": 3475
+    },
+    {
+      "epoch": 0.9884535862604193,
+      "grad_norm": 1.344907283782959,
+      "learning_rate": 4.826199367997701e-07,
+      "loss": 2.2011,
+      "step": 3476
+    },
+    {
+      "epoch": 0.9887379515038774,
+      "grad_norm": 1.4560372829437256,
+      "learning_rate": 4.711289859235852e-07,
+      "loss": 2.0771,
+      "step": 3477
+    },
+    {
+      "epoch": 0.9890223167473355,
+      "grad_norm": 1.4100260734558105,
+      "learning_rate": 4.596380350474002e-07,
+      "loss": 1.7795,
+      "step": 3478
+    },
+    {
+      "epoch": 0.9893066819907936,
+      "grad_norm": 1.4562536478042603,
+      "learning_rate": 4.481470841712152e-07,
+      "loss": 1.7464,
+      "step": 3479
+    },
+    {
+      "epoch": 0.9895910472342518,
+      "grad_norm": 1.5763298273086548,
+      "learning_rate": 4.366561332950302e-07,
+      "loss": 1.6895,
+      "step": 3480
+    },
+    {
+      "epoch": 0.9898754124777099,
+      "grad_norm": 1.698847770690918,
+      "learning_rate": 4.251651824188452e-07,
+      "loss": 2.9543,
+      "step": 3481
+    },
+    {
+      "epoch": 0.990159777721168,
+      "grad_norm": 1.475989818572998,
+      "learning_rate": 4.136742315426602e-07,
+      "loss": 2.3288,
+      "step": 3482
+    },
+    {
+      "epoch": 0.9904441429646261,
+      "grad_norm": 1.4876697063446045,
+      "learning_rate": 4.0218328066647523e-07,
+      "loss": 2.1684,
+      "step": 3483
+    },
+    {
+      "epoch": 0.9907285082080842,
+      "grad_norm": 1.5078527927398682,
+      "learning_rate": 3.9069232979029015e-07,
+      "loss": 2.2732,
+      "step": 3484
+    },
+    {
+      "epoch": 0.9910128734515424,
+      "grad_norm": 1.5910695791244507,
+      "learning_rate": 3.792013789141052e-07,
+      "loss": 2.215,
+      "step": 3485
+    },
+    {
+      "epoch": 0.9912972386950005,
+      "grad_norm": 1.6093436479568481,
+      "learning_rate": 3.677104280379202e-07,
+      "loss": 1.9322,
+      "step": 3486
+    },
+    {
+      "epoch": 0.9915816039384586,
+      "grad_norm": 1.5375397205352783,
+      "learning_rate": 3.562194771617351e-07,
+      "loss": 1.7414,
+      "step": 3487
+    },
+    {
+      "epoch": 0.9918659691819167,
+      "grad_norm": 1.5303829908370972,
+      "learning_rate": 3.4472852628555015e-07,
+      "loss": 1.4854,
+      "step": 3488
+    },
+    {
+      "epoch": 0.9921503344253748,
+      "grad_norm": 1.6482408046722412,
+      "learning_rate": 3.332375754093652e-07,
+      "loss": 2.4752,
+      "step": 3489
+    },
+    {
+      "epoch": 0.992434699668833,
+      "grad_norm": 1.387835144996643,
+      "learning_rate": 3.2174662453318015e-07,
+      "loss": 2.0685,
+      "step": 3490
+    },
+    {
+      "epoch": 0.9927190649122911,
+      "grad_norm": 1.428237795829773,
+      "learning_rate": 3.102556736569952e-07,
+      "loss": 2.1544,
+      "step": 3491
+    },
+    {
+      "epoch": 0.9930034301557492,
+      "grad_norm": 1.4636019468307495,
+      "learning_rate": 2.9876472278081015e-07,
+      "loss": 2.2585,
+      "step": 3492
+    },
+    {
+      "epoch": 0.9932877953992073,
+      "grad_norm": 1.5067780017852783,
+      "learning_rate": 2.872737719046251e-07,
+      "loss": 1.8351,
+      "step": 3493
+    },
+    {
+      "epoch": 0.9935721606426654,
+      "grad_norm": 1.483904480934143,
+      "learning_rate": 2.757828210284401e-07,
+      "loss": 1.6935,
+      "step": 3494
+    },
+    {
+      "epoch": 0.9938565258861236,
+      "grad_norm": 1.4138548374176025,
+      "learning_rate": 2.642918701522551e-07,
+      "loss": 1.8169,
+      "step": 3495
+    },
+    {
+      "epoch": 0.9941408911295817,
+      "grad_norm": 1.464961290359497,
+      "learning_rate": 2.528009192760701e-07,
+      "loss": 1.7828,
+      "step": 3496
+    },
+    {
+      "epoch": 0.9944252563730398,
+      "grad_norm": 1.8348134756088257,
+      "learning_rate": 2.4130996839988507e-07,
+      "loss": 2.8307,
+      "step": 3497
+    },
+    {
+      "epoch": 0.9947096216164979,
+      "grad_norm": 1.4368764162063599,
+      "learning_rate": 2.298190175237001e-07,
+      "loss": 2.3679,
+      "step": 3498
+    },
+    {
+      "epoch": 0.994993986859956,
+      "grad_norm": 1.479984998703003,
+      "learning_rate": 2.183280666475151e-07,
+      "loss": 2.2969,
+      "step": 3499
+    },
+    {
+      "epoch": 0.9952783521034142,
+      "grad_norm": 1.4134024381637573,
+      "learning_rate": 2.068371157713301e-07,
+      "loss": 2.2069,
+      "step": 3500
+    },
+    {
+      "epoch": 0.9955627173468723,
+      "grad_norm": 1.640987515449524,
+      "learning_rate": 1.9534616489514507e-07,
+      "loss": 1.9712,
+      "step": 3501
+    },
+    {
+      "epoch": 0.9958470825903304,
+      "grad_norm": 1.4915003776550293,
+      "learning_rate": 1.838552140189601e-07,
+      "loss": 1.6915,
+      "step": 3502
+    },
+    {
+      "epoch": 0.9961314478337885,
+      "grad_norm": 1.5747153759002686,
+      "learning_rate": 1.7236426314277507e-07,
+      "loss": 1.5998,
+      "step": 3503
+    },
+    {
+      "epoch": 0.9964158130772466,
+      "grad_norm": 1.5918612480163574,
+      "learning_rate": 1.6087331226659008e-07,
+      "loss": 1.7691,
+      "step": 3504
+    },
+    {
+      "epoch": 0.9967001783207048,
+      "grad_norm": 1.6233141422271729,
+      "learning_rate": 1.4938236139040508e-07,
+      "loss": 2.538,
+      "step": 3505
+    },
+    {
+      "epoch": 0.9969845435641629,
+      "grad_norm": 1.5444676876068115,
+      "learning_rate": 1.3789141051422005e-07,
+      "loss": 2.3451,
+      "step": 3506
+    },
+    {
+      "epoch": 0.997268908807621,
+      "grad_norm": 1.4213060140609741,
+      "learning_rate": 1.2640045963803505e-07,
+      "loss": 2.1811,
+      "step": 3507
+    },
+    {
+      "epoch": 0.9975532740510791,
+      "grad_norm": 1.5667362213134766,
+      "learning_rate": 1.1490950876185005e-07,
+      "loss": 2.0772,
+      "step": 3508
+    },
+    {
+      "epoch": 0.9978376392945372,
+      "grad_norm": 1.6838546991348267,
+      "learning_rate": 1.0341855788566505e-07,
+      "loss": 1.8718,
+      "step": 3509
+    },
+    {
+      "epoch": 0.9981220045379954,
+      "grad_norm": 1.4781723022460938,
+      "learning_rate": 9.192760700948005e-08,
+      "loss": 1.916,
+      "step": 3510
+    },
+    {
+      "epoch": 0.9984063697814535,
+      "grad_norm": 1.4474114179611206,
+      "learning_rate": 8.043665613329504e-08,
+      "loss": 1.7148,
+      "step": 3511
+    },
+    {
+      "epoch": 0.9986907350249116,
+      "grad_norm": 1.7419354915618896,
+      "learning_rate": 6.894570525711002e-08,
+      "loss": 1.6992,
+      "step": 3512
+    },
+    {
+      "epoch": 0.9989751002683697,
+      "grad_norm": 1.6991440057754517,
+      "learning_rate": 5.7454754380925025e-08,
+      "loss": 2.6591,
+      "step": 3513
+    },
+    {
+      "epoch": 0.9992594655118279,
+      "grad_norm": 1.4011856317520142,
+      "learning_rate": 4.5963803504740025e-08,
+      "loss": 2.2933,
+      "step": 3514
+    },
+    {
+      "epoch": 0.999543830755286,
+      "grad_norm": 1.5167193412780762,
+      "learning_rate": 3.447285262855501e-08,
+      "loss": 2.3063,
+      "step": 3515
+    },
+    {
+      "epoch": 0.9998281959987441,
+      "grad_norm": 1.423835039138794,
+      "learning_rate": 2.2981901752370013e-08,
+      "loss": 2.2698,
+      "step": 3516
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7970972061157227,
+      "learning_rate": 1.1490950876185006e-08,
+      "loss": 2.0559,
+      "step": 3517
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3517,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.433231803074478e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}