diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4978 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1410,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004259850905218318,
+      "grad_norm": 1.125,
+      "learning_rate": 2.8169014084507043e-07,
+      "loss": 1.820163607597351,
+      "step": 2
+    },
+    {
+      "epoch": 0.008519701810436636,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.450704225352114e-07,
+      "loss": 1.8241215944290161,
+      "step": 4
+    },
+    {
+      "epoch": 0.012779552715654952,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.4084507042253523e-06,
+      "loss": 1.7736045122146606,
+      "step": 6
+    },
+    {
+      "epoch": 0.01703940362087327,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.971830985915493e-06,
+      "loss": 1.9067131280899048,
+      "step": 8
+    },
+    {
+      "epoch": 0.021299254526091587,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.535211267605634e-06,
+      "loss": 1.8296759128570557,
+      "step": 10
+    },
+    {
+      "epoch": 0.025559105431309903,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.0985915492957746e-06,
+      "loss": 1.698632001876831,
+      "step": 12
+    },
+    {
+      "epoch": 0.029818956336528223,
+      "grad_norm": 1.734375,
+      "learning_rate": 3.6619718309859158e-06,
+      "loss": 1.7496201992034912,
+      "step": 14
+    },
+    {
+      "epoch": 0.03407880724174654,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.225352112676057e-06,
+      "loss": 1.7501145601272583,
+      "step": 16
+    },
+    {
+      "epoch": 0.038338658146964855,
+      "grad_norm": 1.3984375,
+      "learning_rate": 4.788732394366197e-06,
+      "loss": 1.8459789752960205,
+      "step": 18
+    },
+    {
+      "epoch": 0.042598509052183174,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.352112676056338e-06,
+      "loss": 1.8062622547149658,
+      "step": 20
+    },
+    {
+      "epoch": 0.046858359957401494,
+      "grad_norm": 1.9140625,
+      "learning_rate": 5.915492957746479e-06,
+      "loss": 1.6213020086288452,
+      "step": 22
+    },
+    {
+      "epoch": 0.051118210862619806,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.478873239436621e-06,
+      "loss": 1.6766525506973267,
+      "step": 24
+    },
+    {
+      "epoch": 0.055378061767838126,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.042253521126761e-06,
+      "loss": 1.6821609735488892,
+      "step": 26
+    },
+    {
+      "epoch": 0.059637912673056445,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.6056338028169015e-06,
+      "loss": 1.6819344758987427,
+      "step": 28
+    },
+    {
+      "epoch": 0.06389776357827476,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.169014084507043e-06,
+      "loss": 1.668290376663208,
+      "step": 30
+    },
+    {
+      "epoch": 0.06815761448349308,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.732394366197183e-06,
+      "loss": 1.7105621099472046,
+      "step": 32
+    },
+    {
+      "epoch": 0.0724174653887114,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.295774647887325e-06,
+      "loss": 1.640573501586914,
+      "step": 34
+    },
+    {
+      "epoch": 0.07667731629392971,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.859154929577466e-06,
+      "loss": 1.7408779859542847,
+      "step": 36
+    },
+    {
+      "epoch": 0.08093716719914804,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.0422535211267606e-05,
+      "loss": 1.7062400579452515,
+      "step": 38
+    },
+    {
+      "epoch": 0.08519701810436635,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0985915492957748e-05,
+      "loss": 1.6299972534179688,
+      "step": 40
+    },
+    {
+      "epoch": 0.08945686900958466,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.1549295774647888e-05,
+      "loss": 1.7348750829696655,
+      "step": 42
+    },
+    {
+      "epoch": 0.09371671991480299,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.211267605633803e-05,
+      "loss": 1.5983866453170776,
+      "step": 44
+    },
+    {
+      "epoch": 0.0979765708200213,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.2676056338028171e-05,
+      "loss": 1.6268177032470703,
+      "step": 46
+    },
+    {
+      "epoch": 0.10223642172523961,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.323943661971831e-05,
+      "loss": 1.6371753215789795,
+      "step": 48
+    },
+    {
+      "epoch": 0.10649627263045794,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.380281690140845e-05,
+      "loss": 1.4547175168991089,
+      "step": 50
+    },
+    {
+      "epoch": 0.11075612353567625,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4366197183098594e-05,
+      "loss": 1.5619021654129028,
+      "step": 52
+    },
+    {
+      "epoch": 0.11501597444089456,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.4929577464788734e-05,
+      "loss": 1.5433554649353027,
+      "step": 54
+    },
+    {
+      "epoch": 0.11927582534611289,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.5492957746478872e-05,
+      "loss": 1.4989991188049316,
+      "step": 56
+    },
+    {
+      "epoch": 0.1235356762513312,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.6056338028169017e-05,
+      "loss": 1.549402117729187,
+      "step": 58
+    },
+    {
+      "epoch": 0.12779552715654952,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.6619718309859155e-05,
+      "loss": 1.434061050415039,
+      "step": 60
+    },
+    {
+      "epoch": 0.13205537806176784,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.7183098591549297e-05,
+      "loss": 1.5171629190444946,
+      "step": 62
+    },
+    {
+      "epoch": 0.13631522896698617,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.774647887323944e-05,
+      "loss": 1.3912996053695679,
+      "step": 64
+    },
+    {
+      "epoch": 0.14057507987220447,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.830985915492958e-05,
+      "loss": 1.4431164264678955,
+      "step": 66
+    },
+    {
+      "epoch": 0.1448349307774228,
+      "grad_norm": 0.625,
+      "learning_rate": 1.887323943661972e-05,
+      "loss": 1.385206699371338,
+      "step": 68
+    },
+    {
+      "epoch": 0.14909478168264112,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.943661971830986e-05,
+      "loss": 1.4210329055786133,
+      "step": 70
+    },
+    {
+      "epoch": 0.15335463258785942,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2e-05,
+      "loss": 1.4347320795059204,
+      "step": 72
+    },
+    {
+      "epoch": 0.15761448349307774,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9999911923960593e-05,
+      "loss": 1.4454214572906494,
+      "step": 74
+    },
+    {
+      "epoch": 0.16187433439829607,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9999647697781703e-05,
+      "loss": 1.34307861328125,
+      "step": 76
+    },
+    {
+      "epoch": 0.16613418530351437,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9999207327281333e-05,
+      "loss": 1.3641051054000854,
+      "step": 78
+    },
+    {
+      "epoch": 0.1703940362087327,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.9998590822156014e-05,
+      "loss": 1.3956571817398071,
+      "step": 80
+    },
+    {
+      "epoch": 0.17465388711395102,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.999779819598057e-05,
+      "loss": 1.3472541570663452,
+      "step": 82
+    },
+    {
+      "epoch": 0.17891373801916932,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.999682946620784e-05,
+      "loss": 1.3230623006820679,
+      "step": 84
+    },
+    {
+      "epoch": 0.18317358892438765,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.999568465416831e-05,
+      "loss": 1.3147826194763184,
+      "step": 86
+    },
+    {
+      "epoch": 0.18743343982960597,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9994363785069595e-05,
+      "loss": 1.3436126708984375,
+      "step": 88
+    },
+    {
+      "epoch": 0.19169329073482427,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.9992866887995928e-05,
+      "loss": 1.2736291885375977,
+      "step": 90
+    },
+    {
+      "epoch": 0.1959531416400426,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.99911939959075e-05,
+      "loss": 1.4134427309036255,
+      "step": 92
+    },
+    {
+      "epoch": 0.20021299254526093,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9989345145639734e-05,
+      "loss": 1.3293910026550293,
+      "step": 94
+    },
+    {
+      "epoch": 0.20447284345047922,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.998732037790249e-05,
+      "loss": 1.3805630207061768,
+      "step": 96
+    },
+    {
+      "epoch": 0.20873269435569755,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9985119737279156e-05,
+      "loss": 1.3184444904327393,
+      "step": 98
+    },
+    {
+      "epoch": 0.21299254526091588,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9982743272225637e-05,
+      "loss": 1.2683892250061035,
+      "step": 100
+    },
+    {
+      "epoch": 0.21725239616613418,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9980191035069357e-05,
+      "loss": 1.2953828573226929,
+      "step": 102
+    },
+    {
+      "epoch": 0.2215122470713525,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9977463082008048e-05,
+      "loss": 1.3181703090667725,
+      "step": 104
+    },
+    {
+      "epoch": 0.22577209797657083,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.997455947310854e-05,
+      "loss": 1.3661359548568726,
+      "step": 106
+    },
+    {
+      "epoch": 0.23003194888178913,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.997148027230541e-05,
+      "loss": 1.277463436126709,
+      "step": 108
+    },
+    {
+      "epoch": 0.23429179978700745,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9968225547399624e-05,
+      "loss": 1.244004249572754,
+      "step": 110
+    },
+    {
+      "epoch": 0.23855165069222578,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9964795370057004e-05,
+      "loss": 1.247603416442871,
+      "step": 112
+    },
+    {
+      "epoch": 0.24281150159744408,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.996118981580665e-05,
+      "loss": 1.2560734748840332,
+      "step": 114
+    },
+    {
+      "epoch": 0.2470713525026624,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.99574089640393e-05,
+      "loss": 1.2559478282928467,
+      "step": 116
+    },
+    {
+      "epoch": 0.25133120340788073,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9953452898005564e-05,
+      "loss": 1.2884942293167114,
+      "step": 118
+    },
+    {
+      "epoch": 0.25559105431309903,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.9949321704814103e-05,
+      "loss": 1.2233449220657349,
+      "step": 120
+    },
+    {
+      "epoch": 0.2598509052183174,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9945015475429704e-05,
+      "loss": 1.282986044883728,
+      "step": 122
+    },
+    {
+      "epoch": 0.2641107561235357,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9940534304671266e-05,
+      "loss": 1.2828234434127808,
+      "step": 124
+    },
+    {
+      "epoch": 0.268370607028754,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9935878291209737e-05,
+      "loss": 1.2087836265563965,
+      "step": 126
+    },
+    {
+      "epoch": 0.27263045793397234,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9931047537565917e-05,
+      "loss": 1.2843331098556519,
+      "step": 128
+    },
+    {
+      "epoch": 0.27689030883919064,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9926042150108228e-05,
+      "loss": 1.337433099746704,
+      "step": 130
+    },
+    {
+      "epoch": 0.28115015974440893,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9920862239050333e-05,
+      "loss": 1.244279146194458,
+      "step": 132
+    },
+    {
+      "epoch": 0.2854100106496273,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9915507918448763e-05,
+      "loss": 1.191616415977478,
+      "step": 134
+    },
+    {
+      "epoch": 0.2896698615548456,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9909979306200337e-05,
+      "loss": 1.3028016090393066,
+      "step": 136
+    },
+    {
+      "epoch": 0.2939297124600639,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9904276524039633e-05,
+      "loss": 1.2914124727249146,
+      "step": 138
+    },
+    {
+      "epoch": 0.29818956336528224,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9898399697536263e-05,
+      "loss": 1.318739652633667,
+      "step": 140
+    },
+    {
+      "epoch": 0.30244941427050054,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9892348956092136e-05,
+      "loss": 1.2955219745635986,
+      "step": 142
+    },
+    {
+      "epoch": 0.30670926517571884,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9886124432938582e-05,
+      "loss": 1.2421311140060425,
+      "step": 144
+    },
+    {
+      "epoch": 0.3109691160809372,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.9879726265133432e-05,
+      "loss": 1.2619209289550781,
+      "step": 146
+    },
+    {
+      "epoch": 0.3152289669861555,
+      "grad_norm": 1.875,
+      "learning_rate": 1.987315459355801e-05,
+      "loss": 1.2909760475158691,
+      "step": 148
+    },
+    {
+      "epoch": 0.3194888178913738,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9866409562914022e-05,
+      "loss": 1.2878739833831787,
+      "step": 150
+    },
+    {
+      "epoch": 0.32374866879659214,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.985949132172036e-05,
+      "loss": 1.2502833604812622,
+      "step": 152
+    },
+    {
+      "epoch": 0.32800851970181044,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9852400022309845e-05,
+      "loss": 1.2763196229934692,
+      "step": 154
+    },
+    {
+      "epoch": 0.33226837060702874,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.984513582082587e-05,
+      "loss": 1.3070451021194458,
+      "step": 156
+    },
+    {
+      "epoch": 0.3365282215122471,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9837698877218955e-05,
+      "loss": 1.2980157136917114,
+      "step": 158
+    },
+    {
+      "epoch": 0.3407880724174654,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.983008935524324e-05,
+      "loss": 1.2509483098983765,
+      "step": 160
+    },
+    {
+      "epoch": 0.3450479233226837,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9822307422452862e-05,
+      "loss": 1.1746965646743774,
+      "step": 162
+    },
+    {
+      "epoch": 0.34930777422790205,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9814353250198275e-05,
+      "loss": 1.1982603073120117,
+      "step": 164
+    },
+    {
+      "epoch": 0.35356762513312034,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9806227013622483e-05,
+      "loss": 1.2470253705978394,
+      "step": 166
+    },
+    {
+      "epoch": 0.35782747603833864,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9797928891657156e-05,
+      "loss": 1.2994393110275269,
+      "step": 168
+    },
+    {
+      "epoch": 0.362087326943557,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9789459067018733e-05,
+      "loss": 1.2793241739273071,
+      "step": 170
+    },
+    {
+      "epoch": 0.3663471778487753,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9780817726204363e-05,
+      "loss": 1.2065937519073486,
+      "step": 172
+    },
+    {
+      "epoch": 0.3706070287539936,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.9772005059487815e-05,
+      "loss": 1.290358543395996,
+      "step": 174
+    },
+    {
+      "epoch": 0.37486687965921195,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9763021260915283e-05,
+      "loss": 1.2577617168426514,
+      "step": 176
+    },
+    {
+      "epoch": 0.37912673056443025,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.9753866528301128e-05,
+      "loss": 1.1449509859085083,
+      "step": 178
+    },
+    {
+      "epoch": 0.38338658146964855,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9744541063223477e-05,
+      "loss": 1.2480742931365967,
+      "step": 180
+    },
+    {
+      "epoch": 0.3876464323748669,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.9735045071019853e-05,
+      "loss": 1.245701789855957,
+      "step": 182
+    },
+    {
+      "epoch": 0.3919062832800852,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9725378760782593e-05,
+      "loss": 1.2594965696334839,
+      "step": 184
+    },
+    {
+      "epoch": 0.3961661341853035,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9715542345354285e-05,
+      "loss": 1.112318992614746,
+      "step": 186
+    },
+    {
+      "epoch": 0.40042598509052185,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9705536041323056e-05,
+      "loss": 1.240614891052246,
+      "step": 188
+    },
+    {
+      "epoch": 0.40468583599574015,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.9695360069017798e-05,
+      "loss": 1.219789981842041,
+      "step": 190
+    },
+    {
+      "epoch": 0.40894568690095845,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9685014652503368e-05,
+      "loss": 1.2046880722045898,
+      "step": 192
+    },
+    {
+      "epoch": 0.4132055378061768,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9674500019575585e-05,
+      "loss": 1.3095979690551758,
+      "step": 194
+    },
+    {
+      "epoch": 0.4174653887113951,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.966381640175625e-05,
+      "loss": 1.236427903175354,
+      "step": 196
+    },
+    {
+      "epoch": 0.4217252396166134,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9652964034288054e-05,
+      "loss": 1.2324111461639404,
+      "step": 198
+    },
+    {
+      "epoch": 0.42598509052183176,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9641943156129385e-05,
+      "loss": 1.296306848526001,
+      "step": 200
+    },
+    {
+      "epoch": 0.43024494142705005,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.9630754009949062e-05,
+      "loss": 1.2453858852386475,
+      "step": 202
+    },
+    {
+      "epoch": 0.43450479233226835,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.961939684212101e-05,
+      "loss": 1.2566311359405518,
+      "step": 204
+    },
+    {
+      "epoch": 0.4387646432374867,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9607871902718817e-05,
+      "loss": 1.1997429132461548,
+      "step": 206
+    },
+    {
+      "epoch": 0.443024494142705,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.959617944551024e-05,
+      "loss": 1.2592103481292725,
+      "step": 208
+    },
+    {
+      "epoch": 0.4472843450479233,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9584319727951608e-05,
+      "loss": 1.2379335165023804,
+      "step": 210
+    },
+    {
+      "epoch": 0.45154419595314166,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.9572293011182153e-05,
+      "loss": 1.2336797714233398,
+      "step": 212
+    },
+    {
+      "epoch": 0.45580404685835996,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.9560099560018276e-05,
+      "loss": 1.2236690521240234,
+      "step": 214
+    },
+    {
+      "epoch": 0.46006389776357826,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.95477396429477e-05,
+      "loss": 1.2254308462142944,
+      "step": 216
+    },
+    {
+      "epoch": 0.4643237486687966,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9535213532123554e-05,
+      "loss": 1.1878576278686523,
+      "step": 218
+    },
+    {
+      "epoch": 0.4685835995740149,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.95225215033584e-05,
+      "loss": 1.1797441244125366,
+      "step": 220
+    },
+    {
+      "epoch": 0.4728434504792332,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.950966383611815e-05,
+      "loss": 1.2125729322433472,
+      "step": 222
+    },
+    {
+      "epoch": 0.47710330138445156,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.9496640813515896e-05,
+      "loss": 1.3302891254425049,
+      "step": 224
+    },
+    {
+      "epoch": 0.48136315228966986,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.9483452722305705e-05,
+      "loss": 1.206823468208313,
+      "step": 226
+    },
+    {
+      "epoch": 0.48562300319488816,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.9470099852876284e-05,
+      "loss": 1.1781184673309326,
+      "step": 228
+    },
+    {
+      "epoch": 0.4898828541001065,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9456582499244607e-05,
+      "loss": 1.1779451370239258,
+      "step": 230
+    },
+    {
+      "epoch": 0.4941427050053248,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.9442900959049402e-05,
+      "loss": 1.2148332595825195,
+      "step": 232
+    },
+    {
+      "epoch": 0.4984025559105431,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.942905553354465e-05,
+      "loss": 1.2204636335372925,
+      "step": 234
+    },
+    {
+      "epoch": 0.5026624068157615,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9415046527592905e-05,
+      "loss": 1.19704008102417,
+      "step": 236
+    },
+    {
+      "epoch": 0.5069222577209798,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9400874249658606e-05,
+      "loss": 1.230122685432434,
+      "step": 238
+    },
+    {
+      "epoch": 0.5111821086261981,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9386539011801288e-05,
+      "loss": 1.2034316062927246,
+      "step": 240
+    },
+    {
+      "epoch": 0.5154419595314164,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9372041129668688e-05,
+      "loss": 1.310076355934143,
+      "step": 242
+    },
+    {
+      "epoch": 0.5197018104366348,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.935738092248981e-05,
+      "loss": 1.2447279691696167,
+      "step": 244
+    },
+    {
+      "epoch": 0.5239616613418531,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9342558713067916e-05,
+      "loss": 1.249355673789978,
+      "step": 246
+    },
+    {
+      "epoch": 0.5282215122470714,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.932757482777337e-05,
+      "loss": 1.2074915170669556,
+      "step": 248
+    },
+    {
+      "epoch": 0.5324813631522897,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.9312429596536493e-05,
+      "loss": 1.2255703210830688,
+      "step": 250
+    },
+    {
+      "epoch": 0.536741214057508,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.929712335284028e-05,
+      "loss": 1.174513339996338,
+      "step": 252
+    },
+    {
+      "epoch": 0.5410010649627263,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9281656433713065e-05,
+      "loss": 1.186693787574768,
+      "step": 254
+    },
+    {
+      "epoch": 0.5452609158679447,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9266029179721087e-05,
+      "loss": 1.1817073822021484,
+      "step": 256
+    },
+    {
+      "epoch": 0.549520766773163,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.925024193496101e-05,
+      "loss": 1.2056523561477661,
+      "step": 258
+    },
+    {
+      "epoch": 0.5537806176783813,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.923429504705233e-05,
+      "loss": 1.2128726243972778,
+      "step": 260
+    },
+    {
+      "epoch": 0.5580404685835996,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.921818886712973e-05,
+      "loss": 1.2569330930709839,
+      "step": 262
+    },
+    {
+      "epoch": 0.5623003194888179,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.9201923749835343e-05,
+      "loss": 1.2348226308822632,
+      "step": 264
+    },
+    {
+      "epoch": 0.5665601703940362,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9185500053310947e-05,
+      "loss": 1.2059990167617798,
+      "step": 266
+    },
+    {
+      "epoch": 0.5708200212992546,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.916891813919008e-05,
+      "loss": 1.2185040712356567,
+      "step": 268
+    },
+    {
+      "epoch": 0.5750798722044729,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.915217837259007e-05,
+      "loss": 1.1790536642074585,
+      "step": 270
+    },
+    {
+      "epoch": 0.5793397231096912,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9135281122104e-05,
+      "loss": 1.2463805675506592,
+      "step": 272
+    },
+    {
+      "epoch": 0.5835995740149095,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9118226759792606e-05,
+      "loss": 1.2532294988632202,
+      "step": 274
+    },
+    {
+      "epoch": 0.5878594249201278,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.910101566117605e-05,
+      "loss": 1.257046103477478,
+      "step": 276
+    },
+    {
+      "epoch": 0.5921192758253461,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.908364820522568e-05,
+      "loss": 1.2225136756896973,
+      "step": 278
+    },
+    {
+      "epoch": 0.5963791267305645,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.906612477435569e-05,
+      "loss": 1.1029458045959473,
+      "step": 280
+    },
+    {
+      "epoch": 0.6006389776357828,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9048445754414667e-05,
+      "loss": 1.2707245349884033,
+      "step": 282
+    },
+    {
+      "epoch": 0.6048988285410011,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.903061153467713e-05,
+      "loss": 1.2144739627838135,
+      "step": 284
+    },
+    {
+      "epoch": 0.6091586794462194,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9012622507834943e-05,
+      "loss": 1.267938494682312,
+      "step": 286
+    },
+    {
+      "epoch": 0.6134185303514377,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.8994479069988654e-05,
+      "loss": 1.158606767654419,
+      "step": 288
+    },
+    {
+      "epoch": 0.617678381256656,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.897618162063881e-05,
+      "loss": 1.3172389268875122,
+      "step": 290
+    },
+    {
+      "epoch": 0.6219382321618744,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.895773056267712e-05,
+      "loss": 1.1937861442565918,
+      "step": 292
+    },
+    {
+      "epoch": 0.6261980830670927,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.893912630237762e-05,
+      "loss": 1.2631405591964722,
+      "step": 294
+    },
+    {
+      "epoch": 0.630457933972311,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8920369249387697e-05,
+      "loss": 1.2060716152191162,
+      "step": 296
+    },
+    {
+      "epoch": 0.6347177848775293,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.8901459816719085e-05,
+      "loss": 1.1968940496444702,
+      "step": 298
+    },
+    {
+      "epoch": 0.6389776357827476,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8882398420738766e-05,
+      "loss": 1.2191784381866455,
+      "step": 300
+    },
+    {
+      "epoch": 0.6432374866879659,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.8863185481159815e-05,
+      "loss": 1.1612508296966553,
+      "step": 302
+    },
+    {
+      "epoch": 0.6474973375931843,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.8843821421032124e-05,
+      "loss": 1.2527700662612915,
+      "step": 304
+    },
+    {
+      "epoch": 0.6517571884984026,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.8824306666733133e-05,
+      "loss": 1.2492940425872803,
+      "step": 306
+    },
+    {
+      "epoch": 0.6560170394036209,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.880464164795841e-05,
+      "loss": 1.2013378143310547,
+      "step": 308
+    },
+    {
+      "epoch": 0.6602768903088392,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8784826797712195e-05,
+      "loss": 1.1568100452423096,
+      "step": 310
+    },
+    {
+      "epoch": 0.6645367412140575,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.8764862552297865e-05,
+      "loss": 1.2038320302963257,
+      "step": 312
+    },
+    {
+      "epoch": 0.6687965921192758,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.874474935130835e-05,
+      "loss": 1.1380341053009033,
+      "step": 314
+    },
+    {
+      "epoch": 0.6730564430244942,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8724487637616405e-05,
+      "loss": 1.1762800216674805,
+      "step": 316
+    },
+    {
+      "epoch": 0.6773162939297125,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8704077857364913e-05,
+      "loss": 1.2073653936386108,
+      "step": 318
+    },
+    {
+      "epoch": 0.6815761448349308,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.8683520459957027e-05,
+      "loss": 1.2313804626464844,
+      "step": 320
+    },
+    {
+      "epoch": 0.6858359957401491,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.8662815898046277e-05,
+      "loss": 1.1902257204055786,
+      "step": 322
+    },
+    {
+      "epoch": 0.6900958466453674,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8641964627526623e-05,
+      "loss": 1.177359938621521,
+      "step": 324
+    },
+    {
+      "epoch": 0.6943556975505857,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8620967107522393e-05,
+      "loss": 1.1724225282669067,
+      "step": 326
+    },
+    {
+      "epoch": 0.6986155484558041,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.8599823800378183e-05,
+      "loss": 1.2089399099349976,
+      "step": 328
+    },
+    {
+      "epoch": 0.7028753993610224,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.8578535171648683e-05,
+      "loss": 1.1843369007110596,
+      "step": 330
+    },
+    {
+      "epoch": 0.7071352502662407,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8557101690088423e-05,
+      "loss": 1.2020741701126099,
+      "step": 332
+    },
+    {
+      "epoch": 0.711395101171459,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.853552382764144e-05,
+      "loss": 1.2178312540054321,
+      "step": 334
+    },
+    {
+      "epoch": 0.7156549520766773,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.8513802059430906e-05,
+      "loss": 1.3234312534332275,
+      "step": 336
+    },
+    {
+      "epoch": 0.7199148029818956,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8491936863748638e-05,
+      "loss": 1.2087228298187256,
+      "step": 338
+    },
+    {
+      "epoch": 0.724174653887114,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.8469928722044607e-05,
+      "loss": 1.213451862335205,
+      "step": 340
+    },
+    {
+      "epoch": 0.7284345047923323,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8447778118916297e-05,
+      "loss": 1.2303829193115234,
+      "step": 342
+    },
+    {
+      "epoch": 0.7326943556975506,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8425485542098055e-05,
+      "loss": 1.226832389831543,
+      "step": 344
+    },
+    {
+      "epoch": 0.7369542066027689,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.8403051482450352e-05,
+      "loss": 1.1740474700927734,
+      "step": 346
+    },
+    {
+      "epoch": 0.7412140575079872,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.8380476433948967e-05,
+      "loss": 1.234078288078308,
+      "step": 348
+    },
+    {
+      "epoch": 0.7454739084132055,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.8357760893674118e-05,
+      "loss": 1.1722126007080078,
+      "step": 350
+    },
+    {
+      "epoch": 0.7497337593184239,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.8334905361799505e-05,
+      "loss": 1.1766189336776733,
+      "step": 352
+    },
+    {
+      "epoch": 0.7539936102236422,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.8311910341581317e-05,
+      "loss": 1.2379800081253052,
+      "step": 354
+    },
+    {
+      "epoch": 0.7582534611288605,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.8288776339347123e-05,
+      "loss": 1.3067048788070679,
+      "step": 356
+    },
+    {
+      "epoch": 0.7625133120340788,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.826550386448475e-05,
+      "loss": 1.249035120010376,
+      "step": 358
+    },
+    {
+      "epoch": 0.7667731629392971,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8242093429431054e-05,
+      "loss": 1.153590440750122,
+      "step": 360
+    },
+    {
+      "epoch": 0.7710330138445154,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8218545549660627e-05,
+      "loss": 1.2533551454544067,
+      "step": 362
+    },
+    {
+      "epoch": 0.7752928647497338,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.8194860743674475e-05,
+      "loss": 1.2324496507644653,
+      "step": 364
+    },
+    {
+      "epoch": 0.7795527156549521,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.817103953298857e-05,
+      "loss": 1.1727190017700195,
+      "step": 366
+    },
+    {
+      "epoch": 0.7838125665601704,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.814708244212238e-05,
+      "loss": 1.202933430671692,
+      "step": 368
+    },
+    {
+      "epoch": 0.7880724174653887,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8122989998587334e-05,
+      "loss": 1.1554442644119263,
+      "step": 370
+    },
+    {
+      "epoch": 0.792332268370607,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.8098762732875176e-05,
+      "loss": 1.2414895296096802,
+      "step": 372
+    },
+    {
+      "epoch": 0.7965921192758253,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.807440117844632e-05,
+      "loss": 1.196638584136963,
+      "step": 374
+    },
+    {
+      "epoch": 0.8008519701810437,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8049905871718057e-05,
+      "loss": 1.1261889934539795,
+      "step": 376
+    },
+    {
+      "epoch": 0.805111821086262,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.80252773520528e-05,
+      "loss": 1.1604783535003662,
+      "step": 378
+    },
+    {
+      "epoch": 0.8093716719914803,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.8000516161746158e-05,
+      "loss": 1.2182838916778564,
+      "step": 380
+    },
+    {
+      "epoch": 0.8136315228966986,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7975622846015025e-05,
+      "loss": 1.118109941482544,
+      "step": 382
+    },
+    {
+      "epoch": 0.8178913738019169,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.795059795298557e-05,
+      "loss": 1.171169638633728,
+      "step": 384
+    },
+    {
+      "epoch": 0.8221512247071352,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.792544203368116e-05,
+      "loss": 1.2411582469940186,
+      "step": 386
+    },
+    {
+      "epoch": 0.8264110756123536,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.7900155642010224e-05,
+      "loss": 1.2089407444000244,
+      "step": 388
+    },
+    {
+      "epoch": 0.8306709265175719,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.7874739334754072e-05,
+      "loss": 1.1152267456054688,
+      "step": 390
+    },
+    {
+      "epoch": 0.8349307774227902,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7849193671554625e-05,
+      "loss": 1.1154651641845703,
+      "step": 392
+    },
+    {
+      "epoch": 0.8391906283280085,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.7823519214902093e-05,
+      "loss": 1.1766343116760254,
+      "step": 394
+    },
+    {
+      "epoch": 0.8434504792332268,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7797716530122588e-05,
+      "loss": 1.061235785484314,
+      "step": 396
+    },
+    {
+      "epoch": 0.8477103301384451,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.777178618536568e-05,
+      "loss": 1.1713075637817383,
+      "step": 398
+    },
+    {
+      "epoch": 0.8519701810436635,
+      "grad_norm": 0.5,
+      "learning_rate": 1.774572875159189e-05,
+      "loss": 1.1845115423202515,
+      "step": 400
+    },
+    {
+      "epoch": 0.8562300319488818,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.77195448025601e-05,
+      "loss": 1.1638309955596924,
+      "step": 402
+    },
+    {
+      "epoch": 0.8604898828541001,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.7693234914814955e-05,
+      "loss": 1.1716026067733765,
+      "step": 404
+    },
+    {
+      "epoch": 0.8647497337593184,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.7666799667674123e-05,
+      "loss": 1.2654908895492554,
+      "step": 406
+    },
+    {
+      "epoch": 0.8690095846645367,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.7640239643215574e-05,
+      "loss": 1.219713568687439,
+      "step": 408
+    },
+    {
+      "epoch": 0.873269435569755,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.761355542626475e-05,
+      "loss": 1.3021862506866455,
+      "step": 410
+    },
+    {
+      "epoch": 0.8775292864749734,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.758674760438168e-05,
+      "loss": 1.2263654470443726,
+      "step": 412
+    },
+    {
+      "epoch": 0.8817891373801917,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.7559816767848063e-05,
+      "loss": 1.212475061416626,
+      "step": 414
+    },
+    {
+      "epoch": 0.88604898828541,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.7532763509654257e-05,
+      "loss": 1.1879194974899292,
+      "step": 416
+    },
+    {
+      "epoch": 0.8903088391906283,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.7505588425486216e-05,
+      "loss": 1.1831648349761963,
+      "step": 418
+    },
+    {
+      "epoch": 0.8945686900958466,
+      "grad_norm": 0.75,
+      "learning_rate": 1.7478292113712387e-05,
+      "loss": 1.2148785591125488,
+      "step": 420
+    },
+    {
+      "epoch": 0.898828541001065,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.7450875175370532e-05,
+      "loss": 1.1164610385894775,
+      "step": 422
+    },
+    {
+      "epoch": 0.9030883919062833,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.7423338214154476e-05,
+      "loss": 1.1142120361328125,
+      "step": 424
+    },
+    {
+      "epoch": 0.9073482428115016,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.7395681836400856e-05,
+      "loss": 1.1910451650619507,
+      "step": 426
+    },
+    {
+      "epoch": 0.9116080937167199,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.736790665107571e-05,
+      "loss": 1.1739782094955444,
+      "step": 428
+    },
+    {
+      "epoch": 0.9158679446219382,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.7340013269761127e-05,
+      "loss": 1.156859278678894,
+      "step": 430
+    },
+    {
+      "epoch": 0.9201277955271565,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.731200230664174e-05,
+      "loss": 1.1874017715454102,
+      "step": 432
+    },
+    {
+      "epoch": 0.9243876464323749,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7283874378491226e-05,
+      "loss": 1.1091042757034302,
+      "step": 434
+    },
+    {
+      "epoch": 0.9286474973375932,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.7255630104658707e-05,
+      "loss": 1.2352137565612793,
+      "step": 436
+    },
+    {
+      "epoch": 0.9329073482428115,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.7227270107055134e-05,
+      "loss": 1.1861075162887573,
+      "step": 438
+    },
+    {
+      "epoch": 0.9371671991480298,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.7198795010139565e-05,
+      "loss": 1.1825189590454102,
+      "step": 440
+    },
+    {
+      "epoch": 0.9414270500532481,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.7170205440905444e-05,
+      "loss": 1.1355581283569336,
+      "step": 442
+    },
+    {
+      "epoch": 0.9456869009584664,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.7141502028866782e-05,
+      "loss": 1.1689903736114502,
+      "step": 444
+    },
+    {
+      "epoch": 0.9499467518636848,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.7112685406044283e-05,
+      "loss": 1.1596486568450928,
+      "step": 446
+    },
+    {
+      "epoch": 0.9542066027689031,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7083756206951457e-05,
+      "loss": 1.2047994136810303,
+      "step": 448
+    },
+    {
+      "epoch": 0.9584664536741214,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.705471506858062e-05,
+      "loss": 1.1618238687515259,
+      "step": 450
+    },
+    {
+      "epoch": 0.9627263045793397,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.702556263038888e-05,
+      "loss": 1.158185601234436,
+      "step": 452
+    },
+    {
+      "epoch": 0.966986155484558,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.699629953428407e-05,
+      "loss": 1.182377576828003,
+      "step": 454
+    },
+    {
+      "epoch": 0.9712460063897763,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.696692642461058e-05,
+      "loss": 1.1322381496429443,
+      "step": 456
+    },
+    {
+      "epoch": 0.9755058572949947,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.693744394813521e-05,
+      "loss": 1.1632483005523682,
+      "step": 458
+    },
+    {
+      "epoch": 0.979765708200213,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.6907852754032897e-05,
+      "loss": 1.1526472568511963,
+      "step": 460
+    },
+    {
+      "epoch": 0.9840255591054313,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.6878153493872438e-05,
+      "loss": 1.212873101234436,
+      "step": 462
+    },
+    {
+      "epoch": 0.9882854100106496,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.684834682160213e-05,
+      "loss": 1.1329699754714966,
+      "step": 464
+    },
+    {
+      "epoch": 0.9925452609158679,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.681843339353539e-05,
+      "loss": 1.2156919240951538,
+      "step": 466
+    },
+    {
+      "epoch": 0.9968051118210862,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.678841386833628e-05,
+      "loss": 1.1542456150054932,
+      "step": 468
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.675828890700502e-05,
+      "loss": 1.2353312969207764,
+      "step": 470
+    },
+    {
+      "epoch": 1.0042598509052183,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.6728059172863434e-05,
+      "loss": 1.0001182556152344,
+      "step": 472
+    },
+    {
+      "epoch": 1.0085197018104366,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.6697725331540337e-05,
+      "loss": 1.0624783039093018,
+      "step": 474
+    },
+    {
+      "epoch": 1.012779552715655,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.666728805095688e-05,
+      "loss": 1.0327106714248657,
+      "step": 476
+    },
+    {
+      "epoch": 1.0170394036208732,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.663674800131184e-05,
+      "loss": 0.9451117515563965,
+      "step": 478
+    },
+    {
+      "epoch": 1.0212992545260915,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.660610585506687e-05,
+      "loss": 1.0440083742141724,
+      "step": 480
+    },
+    {
+      "epoch": 1.0255591054313098,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.65753622869317e-05,
+      "loss": 1.0152018070220947,
+      "step": 482
+    },
+    {
+      "epoch": 1.0298189563365283,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.6544517973849247e-05,
+      "loss": 1.005466341972351,
+      "step": 484
+    },
+    {
+      "epoch": 1.0340788072417466,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.6513573594980753e-05,
+      "loss": 1.0366133451461792,
+      "step": 486
+    },
+    {
+      "epoch": 1.038338658146965,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.6482529831690805e-05,
+      "loss": 1.0141640901565552,
+      "step": 488
+    },
+    {
+      "epoch": 1.0425985090521832,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.6451387367532316e-05,
+      "loss": 0.9699265956878662,
+      "step": 490
+    },
+    {
+      "epoch": 1.0468583599574015,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.642014688823153e-05,
+      "loss": 1.0510493516921997,
+      "step": 492
+    },
+    {
+      "epoch": 1.0511182108626198,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.6388809081672854e-05,
+      "loss": 0.9604178071022034,
+      "step": 494
+    },
+    {
+      "epoch": 1.055378061767838,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.635737463788377e-05,
+      "loss": 1.0432132482528687,
+      "step": 496
+    },
+    {
+      "epoch": 1.0596379126730564,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.63258442490196e-05,
+      "loss": 0.9879993796348572,
+      "step": 498
+    },
+    {
+      "epoch": 1.0638977635782747,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.6294218609348292e-05,
+      "loss": 0.989188015460968,
+      "step": 500
+    },
+    {
+      "epoch": 1.068157614483493,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.626249841523511e-05,
+      "loss": 1.0020525455474854,
+      "step": 502
+    },
+    {
+      "epoch": 1.0724174653887113,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.623068436512732e-05,
+      "loss": 1.0052525997161865,
+      "step": 504
+    },
+    {
+      "epoch": 1.0766773162939298,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.619877715953882e-05,
+      "loss": 1.030534029006958,
+      "step": 506
+    },
+    {
+      "epoch": 1.0809371671991481,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.616677750103467e-05,
+      "loss": 1.1060301065444946,
+      "step": 508
+    },
+    {
+      "epoch": 1.0851970181043664,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6134686094215663e-05,
+      "loss": 0.976584255695343,
+      "step": 510
+    },
+    {
+      "epoch": 1.0894568690095847,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.6102503645702813e-05,
+      "loss": 0.9827842712402344,
+      "step": 512
+    },
+    {
+      "epoch": 1.093716719914803,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6070230864121765e-05,
+      "loss": 0.9941512942314148,
+      "step": 514
+    },
+    {
+      "epoch": 1.0979765708200213,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.603786846008722e-05,
+      "loss": 0.9916704893112183,
+      "step": 516
+    },
+    {
+      "epoch": 1.1022364217252396,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.6005417146187262e-05,
+      "loss": 0.9758402109146118,
+      "step": 518
+    },
+    {
+      "epoch": 1.106496272630458,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.59728776369677e-05,
+      "loss": 1.0404748916625977,
+      "step": 520
+    },
+    {
+      "epoch": 1.1107561235356762,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.5940250648916305e-05,
+      "loss": 1.0453121662139893,
+      "step": 522
+    },
+    {
+      "epoch": 1.1150159744408945,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5907536900447046e-05,
+      "loss": 0.9855365753173828,
+      "step": 524
+    },
+    {
+      "epoch": 1.1192758253461128,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.587473711188428e-05,
+      "loss": 0.9536042213439941,
+      "step": 526
+    },
+    {
+      "epoch": 1.123535676251331,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.5841852005446866e-05,
+      "loss": 1.0321474075317383,
+      "step": 528
+    },
+    {
+      "epoch": 1.1277955271565494,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.5808882305232284e-05,
+      "loss": 0.9870891571044922,
+      "step": 530
+    },
+    {
+      "epoch": 1.132055378061768,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.5775828737200708e-05,
+      "loss": 1.0295721292495728,
+      "step": 532
+    },
+    {
+      "epoch": 1.1363152289669862,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.5742692029158956e-05,
+      "loss": 0.9844380021095276,
+      "step": 534
+    },
+    {
+      "epoch": 1.1405750798722045,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.5709472910744535e-05,
+      "loss": 1.0327517986297607,
+      "step": 536
+    },
+    {
+      "epoch": 1.1448349307774228,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.567617211340955e-05,
+      "loss": 1.015358328819275,
+      "step": 538
+    },
+    {
+      "epoch": 1.1490947816826411,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.5642790370404572e-05,
+      "loss": 0.9756560325622559,
+      "step": 540
+    },
+    {
+      "epoch": 1.1533546325878594,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5609328416762538e-05,
+      "loss": 0.9365598559379578,
+      "step": 542
+    },
+    {
+      "epoch": 1.1576144834930777,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.557578698928253e-05,
+      "loss": 1.0103881359100342,
+      "step": 544
+    },
+    {
+      "epoch": 1.161874334398296,
+      "grad_norm": 1.25,
+      "learning_rate": 1.5542166826513576e-05,
+      "loss": 0.972273051738739,
+      "step": 546
+    },
+    {
+      "epoch": 1.1661341853035143,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.550846866873837e-05,
+      "loss": 1.016709804534912,
+      "step": 548
+    },
+    {
+      "epoch": 1.1703940362087326,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.5474693257956976e-05,
+      "loss": 1.0085346698760986,
+      "step": 550
+    },
+    {
+      "epoch": 1.174653887113951,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.5440841337870497e-05,
+      "loss": 0.9733573794364929,
+      "step": 552
+    },
+    {
+      "epoch": 1.1789137380191694,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.54069136538647e-05,
+      "loss": 1.0755430459976196,
+      "step": 554
+    },
+    {
+      "epoch": 1.1831735889243877,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5372910952993588e-05,
+      "loss": 1.1009080410003662,
+      "step": 556
+    },
+    {
+      "epoch": 1.187433439829606,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.5338833983962972e-05,
+      "loss": 0.9900831580162048,
+      "step": 558
+    },
+    {
+      "epoch": 1.1916932907348243,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.5304683497113963e-05,
+      "loss": 1.0196163654327393,
+      "step": 560
+    },
+    {
+      "epoch": 1.1959531416400426,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5270460244406467e-05,
+      "loss": 1.012573480606079,
+      "step": 562
+    },
+    {
+      "epoch": 1.200212992545261,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.523616497940263e-05,
+      "loss": 0.9771086573600769,
+      "step": 564
+    },
+    {
+      "epoch": 1.2044728434504792,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.5201798457250227e-05,
+      "loss": 1.048552393913269,
+      "step": 566
+    },
+    {
+      "epoch": 1.2087326943556975,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.5167361434666044e-05,
+      "loss": 0.9992176294326782,
+      "step": 568
+    },
+    {
+      "epoch": 1.2129925452609158,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.5132854669919223e-05,
+      "loss": 0.9978233575820923,
+      "step": 570
+    },
+    {
+      "epoch": 1.2172523961661341,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.5098278922814555e-05,
+      "loss": 0.9533007740974426,
+      "step": 572
+    },
+    {
+      "epoch": 1.2215122470713524,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5063634954675758e-05,
+      "loss": 1.05226469039917,
+      "step": 574
+    },
+    {
+      "epoch": 1.225772097976571,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.5028923528328712e-05,
+      "loss": 1.0399305820465088,
+      "step": 576
+    },
+    {
+      "epoch": 1.230031948881789,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.499414540808465e-05,
+      "loss": 1.0016851425170898,
+      "step": 578
+    },
+    {
+      "epoch": 1.2342917997870075,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.4959301359723348e-05,
+      "loss": 0.9840012192726135,
+      "step": 580
+    },
+    {
+      "epoch": 1.2385516506922258,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.4924392150476247e-05,
+      "loss": 0.9944754838943481,
+      "step": 582
+    },
+    {
+      "epoch": 1.2428115015974441,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.488941854900957e-05,
+      "loss": 1.0165916681289673,
+      "step": 584
+    },
+    {
+      "epoch": 1.2470713525026624,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.4854381325407393e-05,
+      "loss": 1.0132522583007812,
+      "step": 586
+    },
+    {
+      "epoch": 1.2513312034078807,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.4819281251154686e-05,
+      "loss": 1.0050327777862549,
+      "step": 588
+    },
+    {
+      "epoch": 1.255591054313099,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4784119099120338e-05,
+      "loss": 1.0251795053482056,
+      "step": 590
+    },
+    {
+      "epoch": 1.2598509052183173,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4748895643540109e-05,
+      "loss": 1.0751694440841675,
+      "step": 592
+    },
+    {
+      "epoch": 1.2641107561235356,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.4713611659999618e-05,
+      "loss": 0.9825355410575867,
+      "step": 594
+    },
+    {
+      "epoch": 1.268370607028754,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4678267925417254e-05,
+      "loss": 1.0707963705062866,
+      "step": 596
+    },
+    {
+      "epoch": 1.2726304579339724,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.4642865218027047e-05,
+      "loss": 0.9957266449928284,
+      "step": 598
+    },
+    {
+      "epoch": 1.2768903088391905,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4607404317361566e-05,
+      "loss": 0.9615009427070618,
+      "step": 600
+    },
+    {
+      "epoch": 1.281150159744409,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.4571886004234728e-05,
+      "loss": 0.9674661159515381,
+      "step": 602
+    },
+    {
+      "epoch": 1.2854100106496273,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.4536311060724622e-05,
+      "loss": 1.0484395027160645,
+      "step": 604
+    },
+    {
+      "epoch": 1.2896698615548456,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4500680270156285e-05,
+      "loss": 1.0365606546401978,
+      "step": 606
+    },
+    {
+      "epoch": 1.293929712460064,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.4464994417084434e-05,
+      "loss": 0.9990244507789612,
+      "step": 608
+    },
+    {
+      "epoch": 1.2981895633652822,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.442925428727623e-05,
+      "loss": 0.9851716756820679,
+      "step": 610
+    },
+    {
+      "epoch": 1.3024494142705005,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4393460667693937e-05,
+      "loss": 1.04202139377594,
+      "step": 612
+    },
+    {
+      "epoch": 1.3067092651757188,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4357614346477622e-05,
+      "loss": 1.027279019355774,
+      "step": 614
+    },
+    {
+      "epoch": 1.3109691160809371,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4321716112927777e-05,
+      "loss": 0.987048864364624,
+      "step": 616
+    },
+    {
+      "epoch": 1.3152289669861554,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.4285766757487966e-05,
+      "loss": 0.9855493307113647,
+      "step": 618
+    },
+    {
+      "epoch": 1.3194888178913737,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.4249767071727406e-05,
+      "loss": 0.9904506802558899,
+      "step": 620
+    },
+    {
+      "epoch": 1.323748668796592,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.4213717848323523e-05,
+      "loss": 1.0218966007232666,
+      "step": 622
+    },
+    {
+      "epoch": 1.3280085197018106,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4177619881044527e-05,
+      "loss": 1.0790345668792725,
+      "step": 624
+    },
+    {
+      "epoch": 1.3322683706070286,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.4141473964731915e-05,
+      "loss": 0.9817459583282471,
+      "step": 626
+    },
+    {
+      "epoch": 1.3365282215122471,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.4105280895282972e-05,
+      "loss": 0.9989323019981384,
+      "step": 628
+    },
+    {
+      "epoch": 1.3407880724174654,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4069041469633251e-05,
+      "loss": 0.9893896579742432,
+      "step": 630
+    },
+    {
+      "epoch": 1.3450479233226837,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4032756485739025e-05,
+      "loss": 1.0551786422729492,
+      "step": 632
+    },
+    {
+      "epoch": 1.349307774227902,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.3996426742559712e-05,
+      "loss": 0.9710611701011658,
+      "step": 634
+    },
+    {
+      "epoch": 1.3535676251331203,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3960053040040284e-05,
+      "loss": 0.9912348985671997,
+      "step": 636
+    },
+    {
+      "epoch": 1.3578274760383386,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.3923636179093653e-05,
+      "loss": 0.9898839592933655,
+      "step": 638
+    },
+    {
+      "epoch": 1.362087326943557,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.3887176961583043e-05,
+      "loss": 0.9827688932418823,
+      "step": 640
+    },
+    {
+      "epoch": 1.3663471778487752,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.3850676190304322e-05,
+      "loss": 0.9909817576408386,
+      "step": 642
+    },
+    {
+      "epoch": 1.3706070287539935,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.3814134668968331e-05,
+      "loss": 1.054543375968933,
+      "step": 644
+    },
+    {
+      "epoch": 1.374866879659212,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3777553202183186e-05,
+      "loss": 1.0329318046569824,
+      "step": 646
+    },
+    {
+      "epoch": 1.3791267305644301,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3740932595436574e-05,
+      "loss": 0.9633697867393494,
+      "step": 648
+    },
+    {
+      "epoch": 1.3833865814696487,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.3704273655077986e-05,
+      "loss": 0.9963237643241882,
+      "step": 650
+    },
+    {
+      "epoch": 1.387646432374867,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.3667577188300994e-05,
+      "loss": 0.9815477728843689,
+      "step": 652
+    },
+    {
+      "epoch": 1.3919062832800853,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.3630844003125466e-05,
+      "loss": 0.9907623529434204,
+      "step": 654
+    },
+    {
+      "epoch": 1.3961661341853036,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.3594074908379759e-05,
+      "loss": 1.0208251476287842,
+      "step": 656
+    },
+    {
+      "epoch": 1.4004259850905219,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3557270713682944e-05,
+      "loss": 1.0356820821762085,
+      "step": 658
+    },
+    {
+      "epoch": 1.4046858359957402,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.3520432229426933e-05,
+      "loss": 0.9749876260757446,
+      "step": 660
+    },
+    {
+      "epoch": 1.4089456869009584,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.3483560266758688e-05,
+      "loss": 0.9936063885688782,
+      "step": 662
+    },
+    {
+      "epoch": 1.4132055378061767,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.344665563756231e-05,
+      "loss": 0.9732779264450073,
+      "step": 664
+    },
+    {
+      "epoch": 1.417465388711395,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.3409719154441194e-05,
+      "loss": 1.0013114213943481,
+      "step": 666
+    },
+    {
+      "epoch": 1.4217252396166133,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3372751630700127e-05,
+      "loss": 0.9554456472396851,
+      "step": 668
+    },
+    {
+      "epoch": 1.4259850905218316,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.3335753880327378e-05,
+      "loss": 1.013936161994934,
+      "step": 670
+    },
+    {
+      "epoch": 1.4302449414270502,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.3298726717976776e-05,
+      "loss": 0.9992549419403076,
+      "step": 672
+    },
+    {
+      "epoch": 1.4345047923322682,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3261670958949778e-05,
+      "loss": 0.9462178945541382,
+      "step": 674
+    },
+    {
+      "epoch": 1.4387646432374868,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.32245874191775e-05,
+      "loss": 1.0281662940979004,
+      "step": 676
+    },
+    {
+      "epoch": 1.443024494142705,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3187476915202772e-05,
+      "loss": 1.080543875694275,
+      "step": 678
+    },
+    {
+      "epoch": 1.4472843450479234,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.315034026416215e-05,
+      "loss": 0.9739678502082825,
+      "step": 680
+    },
+    {
+      "epoch": 1.4515441959531417,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3113178283767916e-05,
+      "loss": 1.058993935585022,
+      "step": 682
+    },
+    {
+      "epoch": 1.45580404685836,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.3075991792290087e-05,
+      "loss": 0.9686048626899719,
+      "step": 684
+    },
+    {
+      "epoch": 1.4600638977635783,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.303878160853838e-05,
+      "loss": 1.0147266387939453,
+      "step": 686
+    },
+    {
+      "epoch": 1.4643237486687966,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.3001548551844198e-05,
+      "loss": 0.9957408905029297,
+      "step": 688
+    },
+    {
+      "epoch": 1.4685835995740149,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2964293442042583e-05,
+      "loss": 1.0241044759750366,
+      "step": 690
+    },
+    {
+      "epoch": 1.4728434504792332,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.2927017099454174e-05,
+      "loss": 0.9795851707458496,
+      "step": 692
+    },
+    {
+      "epoch": 1.4771033013844517,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.288972034486711e-05,
+      "loss": 1.0294287204742432,
+      "step": 694
+    },
+    {
+      "epoch": 1.4813631522896697,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.2852403999519009e-05,
+      "loss": 1.04385507106781,
+      "step": 696
+    },
+    {
+      "epoch": 1.4856230031948883,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.2815068885078848e-05,
+      "loss": 1.0272358655929565,
+      "step": 698
+    },
+    {
+      "epoch": 1.4898828541001066,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2777715823628869e-05,
+      "loss": 1.0158170461654663,
+      "step": 700
+    },
+    {
+      "epoch": 1.4941427050053249,
+      "grad_norm": 3.25,
+      "learning_rate": 1.274034563764651e-05,
+      "loss": 1.0180643796920776,
+      "step": 702
+    },
+    {
+      "epoch": 1.4984025559105432,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2702959149986263e-05,
+      "loss": 0.9400717616081238,
+      "step": 704
+    },
+    {
+      "epoch": 1.5026624068157615,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.2665557183861563e-05,
+      "loss": 1.0463502407073975,
+      "step": 706
+    },
+    {
+      "epoch": 1.5069222577209798,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.2628140562826674e-05,
+      "loss": 0.9682408571243286,
+      "step": 708
+    },
+    {
+      "epoch": 1.511182108626198,
+      "grad_norm": 2.0,
+      "learning_rate": 1.2590710110758542e-05,
+      "loss": 0.9887672066688538,
+      "step": 710
+    },
+    {
+      "epoch": 1.5154419595314164,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.2553266651838664e-05,
+      "loss": 0.9790204167366028,
+      "step": 712
+    },
+    {
+      "epoch": 1.5197018104366347,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.2515811010534926e-05,
+      "loss": 1.048098087310791,
+      "step": 714
+    },
+    {
+      "epoch": 1.5239616613418532,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.2478344011583463e-05,
+      "loss": 0.9795944094657898,
+      "step": 716
+    },
+    {
+      "epoch": 1.5282215122470713,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2440866479970501e-05,
+      "loss": 1.1048449277877808,
+      "step": 718
+    },
+    {
+      "epoch": 1.5324813631522898,
+      "grad_norm": 0.5,
+      "learning_rate": 1.2403379240914171e-05,
+      "loss": 1.1177802085876465,
+      "step": 720
+    },
+    {
+      "epoch": 1.5367412140575079,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2365883119846358e-05,
+      "loss": 0.9775673747062683,
+      "step": 722
+    },
+    {
+      "epoch": 1.5410010649627264,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2328378942394526e-05,
+      "loss": 1.0114903450012207,
+      "step": 724
+    },
+    {
+      "epoch": 1.5452609158679447,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.2290867534363523e-05,
+      "loss": 0.8942697048187256,
+      "step": 726
+    },
+    {
+      "epoch": 1.549520766773163,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.225334972171742e-05,
+      "loss": 1.0175893306732178,
+      "step": 728
+    },
+    {
+      "epoch": 1.5537806176783813,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2215826330561296e-05,
+      "loss": 1.027167797088623,
+      "step": 730
+    },
+    {
+      "epoch": 1.5580404685835996,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2178298187123072e-05,
+      "loss": 0.9816038608551025,
+      "step": 732
+    },
+    {
+      "epoch": 1.5623003194888179,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2140766117735309e-05,
+      "loss": 1.032719612121582,
+      "step": 734
+    },
+    {
+      "epoch": 1.5665601703940362,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.210323094881701e-05,
+      "loss": 1.034834384918213,
+      "step": 736
+    },
+    {
+      "epoch": 1.5708200212992547,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2065693506855438e-05,
+      "loss": 0.9969629645347595,
+      "step": 738
+    },
+    {
+      "epoch": 1.5750798722044728,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.2028154618387888e-05,
+      "loss": 0.9849120378494263,
+      "step": 740
+    },
+    {
+      "epoch": 1.5793397231096913,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.1990615109983516e-05,
+      "loss": 1.042475700378418,
+      "step": 742
+    },
+    {
+      "epoch": 1.5835995740149094,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.1953075808225134e-05,
+      "loss": 0.9866620898246765,
+      "step": 744
+    },
+    {
+      "epoch": 1.5878594249201279,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.191553753969099e-05,
+      "loss": 1.012044906616211,
+      "step": 746
+    },
+    {
+      "epoch": 1.592119275825346,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.1878001130936596e-05,
+      "loss": 0.9986369609832764,
+      "step": 748
+    },
+    {
+      "epoch": 1.5963791267305645,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.1840467408476495e-05,
+      "loss": 0.9213187098503113,
+      "step": 750
+    },
+    {
+      "epoch": 1.6006389776357828,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.1802937198766105e-05,
+      "loss": 1.0106372833251953,
+      "step": 752
+    },
+    {
+      "epoch": 1.604898828541001,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.1765411328183474e-05,
+      "loss": 1.030936360359192,
+      "step": 754
+    },
+    {
+      "epoch": 1.6091586794462194,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.172789062301112e-05,
+      "loss": 0.9806941747665405,
+      "step": 756
+    },
+    {
+      "epoch": 1.6134185303514377,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.1690375909417817e-05,
+      "loss": 1.0275804996490479,
+      "step": 758
+    },
+    {
+      "epoch": 1.617678381256656,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.165286801344042e-05,
+      "loss": 0.9850936532020569,
+      "step": 760
+    },
+    {
+      "epoch": 1.6219382321618743,
+      "grad_norm": 0.5,
+      "learning_rate": 1.1615367760965652e-05,
+      "loss": 0.9898473620414734,
+      "step": 762
+    },
+    {
+      "epoch": 1.6261980830670928,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.1577875977711943e-05,
+      "loss": 0.9166401624679565,
+      "step": 764
+    },
+    {
+      "epoch": 1.6304579339723109,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.1540393489211246e-05,
+      "loss": 1.0430673360824585,
+      "step": 766
+    },
+    {
+      "epoch": 1.6347177848775294,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.1502921120790832e-05,
+      "loss": 0.9577189683914185,
+      "step": 768
+    },
+    {
+      "epoch": 1.6389776357827475,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.146545969755516e-05,
+      "loss": 1.046250581741333,
+      "step": 770
+    },
+    {
+      "epoch": 1.643237486687966,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.1428010044367676e-05,
+      "loss": 0.9433165788650513,
+      "step": 772
+    },
+    {
+      "epoch": 1.6474973375931843,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.139057298583266e-05,
+      "loss": 1.0670093297958374,
+      "step": 774
+    },
+    {
+      "epoch": 1.6517571884984026,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1353149346277071e-05,
+      "loss": 1.0747054815292358,
+      "step": 776
+    },
+    {
+      "epoch": 1.6560170394036209,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.1315739949732403e-05,
+      "loss": 0.9709829092025757,
+      "step": 778
+    },
+    {
+      "epoch": 1.6602768903088392,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.1278345619916519e-05,
+      "loss": 0.9456352591514587,
+      "step": 780
+    },
+    {
+      "epoch": 1.6645367412140575,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1240967180215546e-05,
+      "loss": 0.9906789064407349,
+      "step": 782
+    },
+    {
+      "epoch": 1.6687965921192758,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.1203605453665706e-05,
+      "loss": 0.9821105599403381,
+      "step": 784
+    },
+    {
+      "epoch": 1.6730564430244943,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.1166261262935234e-05,
+      "loss": 0.9833469390869141,
+      "step": 786
+    },
+    {
+      "epoch": 1.6773162939297124,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.1128935430306224e-05,
+      "loss": 0.9621955752372742,
+      "step": 788
+    },
+    {
+      "epoch": 1.681576144834931,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.1091628777656558e-05,
+      "loss": 1.0520445108413696,
+      "step": 790
+    },
+    {
+      "epoch": 1.685835995740149,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.105434212644179e-05,
+      "loss": 1.00932776927948,
+      "step": 792
+    },
+    {
+      "epoch": 1.6900958466453675,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.1017076297677054e-05,
+      "loss": 1.0297919511795044,
+      "step": 794
+    },
+    {
+      "epoch": 1.6943556975505856,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.0979832111919002e-05,
+      "loss": 0.9979485273361206,
+      "step": 796
+    },
+    {
+      "epoch": 1.698615548455804,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.0942610389247735e-05,
+      "loss": 1.0736881494522095,
+      "step": 798
+    },
+    {
+      "epoch": 1.7028753993610224,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.0905411949248722e-05,
+      "loss": 0.9282923936843872,
+      "step": 800
+    },
+    {
+      "epoch": 1.7071352502662407,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.086823761099478e-05,
+      "loss": 0.9751482605934143,
+      "step": 802
+    },
+    {
+      "epoch": 1.711395101171459,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0831088193028037e-05,
+      "loss": 1.050538182258606,
+      "step": 804
+    },
+    {
+      "epoch": 1.7156549520766773,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.0793964513341892e-05,
+      "loss": 0.8432624936103821,
+      "step": 806
+    },
+    {
+      "epoch": 1.7199148029818956,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0756867389363007e-05,
+      "loss": 1.0375334024429321,
+      "step": 808
+    },
+    {
+      "epoch": 1.7241746538871139,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.0719797637933325e-05,
+      "loss": 1.0315303802490234,
+      "step": 810
+    },
+    {
+      "epoch": 1.7284345047923324,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.068275607529207e-05,
+      "loss": 1.0522267818450928,
+      "step": 812
+    },
+    {
+      "epoch": 1.7326943556975505,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0645743517057766e-05,
+      "loss": 0.9734303951263428,
+      "step": 814
+    },
+    {
+      "epoch": 1.736954206602769,
+      "grad_norm": 0.5,
+      "learning_rate": 1.0608760778210305e-05,
+      "loss": 0.956866979598999,
+      "step": 816
+    },
+    {
+      "epoch": 1.741214057507987,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.057180867307298e-05,
+      "loss": 0.9936807751655579,
+      "step": 818
+    },
+    {
+      "epoch": 1.7454739084132056,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.053488801529455e-05,
+      "loss": 1.0136879682540894,
+      "step": 820
+    },
+    {
+      "epoch": 1.749733759318424,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.0497999617831352e-05,
+      "loss": 0.9871541261672974,
+      "step": 822
+    },
+    {
+      "epoch": 1.7539936102236422,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.0461144292929361e-05,
+      "loss": 0.990527868270874,
+      "step": 824
+    },
+    {
+      "epoch": 1.7582534611288605,
+      "grad_norm": 1.25,
+      "learning_rate": 1.0424322852106354e-05,
+      "loss": 1.039267897605896,
+      "step": 826
+    },
+    {
+      "epoch": 1.7625133120340788,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.0387536106133989e-05,
+      "loss": 1.069290280342102,
+      "step": 828
+    },
+    {
+      "epoch": 1.766773162939297,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.0350784865019987e-05,
+      "loss": 0.9809550642967224,
+      "step": 830
+    },
+    {
+      "epoch": 1.7710330138445154,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.0314069937990298e-05,
+      "loss": 0.9778714776039124,
+      "step": 832
+    },
+    {
+      "epoch": 1.775292864749734,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.027739213347125e-05,
+      "loss": 0.9834805130958557,
+      "step": 834
+    },
+    {
+      "epoch": 1.779552715654952,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.0240752259071792e-05,
+      "loss": 1.0088666677474976,
+      "step": 836
+    },
+    {
+      "epoch": 1.7838125665601705,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.0204151121565675e-05,
+      "loss": 0.958771288394928,
+      "step": 838
+    },
+    {
+      "epoch": 1.7880724174653886,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.0167589526873706e-05,
+      "loss": 1.0218135118484497,
+      "step": 840
+    },
+    {
+      "epoch": 1.792332268370607,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0131068280045997e-05,
+      "loss": 1.027093768119812,
+      "step": 842
+    },
+    {
+      "epoch": 1.7965921192758252,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.0094588185244248e-05,
+      "loss": 1.0224063396453857,
+      "step": 844
+    },
+    {
+      "epoch": 1.8008519701810437,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0058150045724022e-05,
+      "loss": 1.0029429197311401,
+      "step": 846
+    },
+    {
+      "epoch": 1.805111821086262,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.002175466381707e-05,
+      "loss": 0.9247448444366455,
+      "step": 848
+    },
+    {
+      "epoch": 1.8093716719914803,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.985402840913666e-06,
+      "loss": 1.0142505168914795,
+      "step": 850
+    },
+    {
+      "epoch": 1.8136315228966986,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.949095377444961e-06,
+      "loss": 0.9922072291374207,
+      "step": 852
+    },
+    {
+      "epoch": 1.817891373801917,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.912833072865345e-06,
+      "loss": 1.0142475366592407,
+      "step": 854
+    },
+    {
+      "epoch": 1.8221512247071352,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.876616725634864e-06,
+      "loss": 0.9615439176559448,
+      "step": 856
+    },
+    {
+      "epoch": 1.8264110756123535,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.840447133201623e-06,
+      "loss": 0.9525391459465027,
+      "step": 858
+    },
+    {
+      "epoch": 1.830670926517572,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.804325091984234e-06,
+      "loss": 1.0148870944976807,
+      "step": 860
+    },
+    {
+      "epoch": 1.83493077742279,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.768251397354276e-06,
+      "loss": 0.9480238556861877,
+      "step": 862
+    },
+    {
+      "epoch": 1.8391906283280086,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9.73222684361879e-06,
+      "loss": 1.0006335973739624,
+      "step": 864
+    },
+    {
+      "epoch": 1.8434504792332267,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.696252224002783e-06,
+      "loss": 0.9949763417243958,
+      "step": 866
+    },
+    {
+      "epoch": 1.8477103301384452,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.660328330631752e-06,
+      "loss": 0.9215972423553467,
+      "step": 868
+    },
+    {
+      "epoch": 1.8519701810436635,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.624455954514258e-06,
+      "loss": 0.9809213280677795,
+      "step": 870
+    },
+    {
+      "epoch": 1.8562300319488818,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.588635885524514e-06,
+      "loss": 0.9926417469978333,
+      "step": 872
+    },
+    {
+      "epoch": 1.8604898828541,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.552868912384962e-06,
+      "loss": 1.0632542371749878,
+      "step": 874
+    },
+    {
+      "epoch": 1.8647497337593184,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.517155822648938e-06,
+      "loss": 1.0109238624572754,
+      "step": 876
+    },
+    {
+      "epoch": 1.8690095846645367,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.481497402683325e-06,
+      "loss": 0.9798102974891663,
+      "step": 878
+    },
+    {
+      "epoch": 1.873269435569755,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.445894437651211e-06,
+      "loss": 1.0319267511367798,
+      "step": 880
+    },
+    {
+      "epoch": 1.8775292864749735,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.41034771149464e-06,
+      "loss": 1.0264664888381958,
+      "step": 882
+    },
+    {
+      "epoch": 1.8817891373801916,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.374858006917328e-06,
+      "loss": 1.014363169670105,
+      "step": 884
+    },
+    {
+      "epoch": 1.8860489882854101,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.339426105367428e-06,
+      "loss": 0.9093478322029114,
+      "step": 886
+    },
+    {
+      "epoch": 1.8903088391906282,
+      "grad_norm": 0.4921875,
+      "learning_rate": 9.304052787020326e-06,
+      "loss": 0.9274075627326965,
+      "step": 888
+    },
+    {
+      "epoch": 1.8945686900958467,
+      "grad_norm": 0.484375,
+      "learning_rate": 9.26873883076147e-06,
+      "loss": 1.0133625268936157,
+      "step": 890
+    },
+    {
+      "epoch": 1.898828541001065,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.233485014169214e-06,
+      "loss": 1.0017213821411133,
+      "step": 892
+    },
+    {
+      "epoch": 1.9030883919062833,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.198292113497686e-06,
+      "loss": 1.0270392894744873,
+      "step": 894
+    },
+    {
+      "epoch": 1.9073482428115016,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.163160903659715e-06,
+      "loss": 0.9596851468086243,
+      "step": 896
+    },
+    {
+      "epoch": 1.91160809371672,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.128092158209762e-06,
+      "loss": 1.0112367868423462,
+      "step": 898
+    },
+    {
+      "epoch": 1.9158679446219382,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.093086649326873e-06,
+      "loss": 1.0255273580551147,
+      "step": 900
+    },
+    {
+      "epoch": 1.9201277955271565,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.05814514779769e-06,
+      "loss": 0.9857978820800781,
+      "step": 902
+    },
+    {
+      "epoch": 1.924387646432375,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.02326842299949e-06,
+      "loss": 0.9689497351646423,
+      "step": 904
+    },
+    {
+      "epoch": 1.928647497337593,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.98845724288321e-06,
+      "loss": 1.039789080619812,
+      "step": 906
+    },
+    {
+      "epoch": 1.9329073482428116,
+      "grad_norm": 0.38671875,
+      "learning_rate": 8.95371237395657e-06,
+      "loss": 0.9854997396469116,
+      "step": 908
+    },
+    {
+      "epoch": 1.9371671991480297,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.919034581267185e-06,
+      "loss": 0.863085150718689,
+      "step": 910
+    },
+    {
+      "epoch": 1.9414270500532482,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.884424628385713e-06,
+      "loss": 0.9845672845840454,
+      "step": 912
+    },
+    {
+      "epoch": 1.9456869009584663,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.849883277389046e-06,
+      "loss": 0.9393159747123718,
+      "step": 914
+    },
+    {
+      "epoch": 1.9499467518636848,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.815411288843531e-06,
+      "loss": 1.019649624824524,
+      "step": 916
+    },
+    {
+      "epoch": 1.9542066027689031,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.781009421788226e-06,
+      "loss": 1.001703143119812,
+      "step": 918
+    },
+    {
+      "epoch": 1.9584664536741214,
+      "grad_norm": 0.478515625,
+      "learning_rate": 8.74667843371818e-06,
+      "loss": 1.0148494243621826,
+      "step": 920
+    },
+    {
+      "epoch": 1.9627263045793397,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.712419080567753e-06,
+      "loss": 0.9227896332740784,
+      "step": 922
+    },
+    {
+      "epoch": 1.966986155484558,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.678232116693988e-06,
+      "loss": 0.9633923768997192,
+      "step": 924
+    },
+    {
+      "epoch": 1.9712460063897763,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.644118294859972e-06,
+      "loss": 0.9478899836540222,
+      "step": 926
+    },
+    {
+      "epoch": 1.9755058572949946,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.610078366218287e-06,
+      "loss": 1.0053707361221313,
+      "step": 928
+    },
+    {
+      "epoch": 1.9797657082002131,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.576113080294459e-06,
+      "loss": 0.9101724624633789,
+      "step": 930
+    },
+    {
+      "epoch": 1.9840255591054312,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.542223184970451e-06,
+      "loss": 1.0562280416488647,
+      "step": 932
+    },
+    {
+      "epoch": 1.9882854100106497,
+      "grad_norm": 0.494140625,
+      "learning_rate": 8.5084094264682e-06,
+      "loss": 0.9434181451797485,
+      "step": 934
+    },
+    {
+      "epoch": 1.9925452609158678,
+      "grad_norm": 0.75390625,
+      "learning_rate": 8.474672549333191e-06,
+      "loss": 1.0050933361053467,
+      "step": 936
+    },
+    {
+      "epoch": 1.9968051118210863,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.441013296418057e-06,
+      "loss": 1.014664649963379,
+      "step": 938
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.407432408866211e-06,
+      "loss": 0.9959129095077515,
+      "step": 940
+    },
+    {
+      "epoch": 2.0042598509052185,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.373930626095557e-06,
+      "loss": 0.8519871830940247,
+      "step": 942
+    },
+    {
+      "epoch": 2.0085197018104366,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.340508685782173e-06,
+      "loss": 0.8345139622688293,
+      "step": 944
+    },
+    {
+      "epoch": 2.012779552715655,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.30716732384409e-06,
+      "loss": 0.8144312500953674,
+      "step": 946
+    },
+    {
+      "epoch": 2.017039403620873,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.27390727442509e-06,
+      "loss": 0.8890555500984192,
+      "step": 948
+    },
+    {
+      "epoch": 2.0212992545260917,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.24072926987853e-06,
+      "loss": 0.9254493117332458,
+      "step": 950
+    },
+    {
+      "epoch": 2.02555910543131,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.207634040751218e-06,
+      "loss": 0.8629240989685059,
+      "step": 952
+    },
+    {
+      "epoch": 2.0298189563365283,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.174622315767325e-06,
+      "loss": 0.9234293699264526,
+      "step": 954
+    },
+    {
+      "epoch": 2.0340788072417464,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.141694821812352e-06,
+      "loss": 0.8904613852500916,
+      "step": 956
+    },
+    {
+      "epoch": 2.038338658146965,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.108852283917116e-06,
+      "loss": 0.8621918559074402,
+      "step": 958
+    },
+    {
+      "epoch": 2.042598509052183,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.07609542524178e-06,
+      "loss": 0.8154129981994629,
+      "step": 960
+    },
+    {
+      "epoch": 2.0468583599574015,
+      "grad_norm": 0.388671875,
+      "learning_rate": 8.043424967059936e-06,
+      "loss": 0.7776715755462646,
+      "step": 962
+    },
+    {
+      "epoch": 2.0511182108626196,
+      "grad_norm": 0.490234375,
+      "learning_rate": 8.010841628742728e-06,
+      "loss": 0.9061995148658752,
+      "step": 964
+    },
+    {
+      "epoch": 2.055378061767838,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.978346127742999e-06,
+      "loss": 0.8465338349342346,
+      "step": 966
+    },
+    {
+      "epoch": 2.0596379126730566,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.945939179579509e-06,
+      "loss": 0.8736094832420349,
+      "step": 968
+    },
+    {
+      "epoch": 2.0638977635782747,
+      "grad_norm": 0.625,
+      "learning_rate": 7.913621497821174e-06,
+      "loss": 0.8707051277160645,
+      "step": 970
+    },
+    {
+      "epoch": 2.0681576144834932,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.881393794071338e-06,
+      "loss": 0.8639911413192749,
+      "step": 972
+    },
+    {
+      "epoch": 2.0724174653887113,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.84925677795213e-06,
+      "loss": 0.8878441452980042,
+      "step": 974
+    },
+    {
+      "epoch": 2.07667731629393,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.817211157088823e-06,
+      "loss": 0.8697801232337952,
+      "step": 976
+    },
+    {
+      "epoch": 2.080937167199148,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.785257637094265e-06,
+      "loss": 0.9132862687110901,
+      "step": 978
+    },
+    {
+      "epoch": 2.0851970181043664,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.753396921553325e-06,
+      "loss": 0.9692128896713257,
+      "step": 980
+    },
+    {
+      "epoch": 2.0894568690095845,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.721629712007408e-06,
+      "loss": 0.8117839694023132,
+      "step": 982
+    },
+    {
+      "epoch": 2.093716719914803,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.68995670793902e-06,
+      "loss": 0.8652855157852173,
+      "step": 984
+    },
+    {
+      "epoch": 2.097976570820021,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.658378606756342e-06,
+      "loss": 0.8414390683174133,
+      "step": 986
+    },
+    {
+      "epoch": 2.1022364217252396,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.626896103777901e-06,
+      "loss": 0.975841224193573,
+      "step": 988
+    },
+    {
+      "epoch": 2.106496272630458,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.595509892217243e-06,
+      "loss": 0.8733735680580139,
+      "step": 990
+    },
+    {
+      "epoch": 2.110756123535676,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.5642206631676655e-06,
+      "loss": 0.8554286360740662,
+      "step": 992
+    },
+    {
+      "epoch": 2.1150159744408947,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.533029105587006e-06,
+      "loss": 0.8418490290641785,
+      "step": 994
+    },
+    {
+      "epoch": 2.119275825346113,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.501935906282486e-06,
+      "loss": 0.8224227428436279,
+      "step": 996
+    },
+    {
+      "epoch": 2.1235356762513313,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.4709417498955676e-06,
+      "loss": 0.8177536129951477,
+      "step": 998
+    },
+    {
+      "epoch": 2.1277955271565494,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.440047318886882e-06,
+      "loss": 0.8571978211402893,
+      "step": 1000
+    },
+    {
+      "epoch": 2.132055378061768,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.409253293521209e-06,
+      "loss": 0.8153810501098633,
+      "step": 1002
+    },
+    {
+      "epoch": 2.136315228966986,
+      "grad_norm": 0.75,
+      "learning_rate": 7.378560351852502e-06,
+      "loss": 0.8463664650917053,
+      "step": 1004
+    },
+    {
+      "epoch": 2.1405750798722045,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.347969169708937e-06,
+      "loss": 0.8821415305137634,
+      "step": 1006
+    },
+    {
+      "epoch": 2.1448349307774226,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.317480420678064e-06,
+      "loss": 0.9291739463806152,
+      "step": 1008
+    },
+    {
+      "epoch": 2.149094781682641,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.2870947760919486e-06,
+      "loss": 0.772282063961029,
+      "step": 1010
+    },
+    {
+      "epoch": 2.1533546325878596,
+      "grad_norm": 0.48046875,
+      "learning_rate": 7.256812905012389e-06,
+      "loss": 0.8808871507644653,
+      "step": 1012
+    },
+    {
+      "epoch": 2.1576144834930777,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.2266354742162095e-06,
+      "loss": 0.8753591179847717,
+      "step": 1014
+    },
+    {
+      "epoch": 2.1618743343982962,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.196563148180554e-06,
+      "loss": 0.8668736815452576,
+      "step": 1016
+    },
+    {
+      "epoch": 2.1661341853035143,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.166596589068271e-06,
+      "loss": 0.8054329752922058,
+      "step": 1018
+    },
+    {
+      "epoch": 2.170394036208733,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.136736456713316e-06,
+      "loss": 0.8477678298950195,
+      "step": 1020
+    },
+    {
+      "epoch": 2.174653887113951,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.1069834086062366e-06,
+      "loss": 0.7990297079086304,
+      "step": 1022
+    },
+    {
+      "epoch": 2.1789137380191694,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.077338099879696e-06,
+      "loss": 0.87007075548172,
+      "step": 1024
+    },
+    {
+      "epoch": 2.1831735889243875,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.047801183294034e-06,
+      "loss": 0.8642436265945435,
+      "step": 1026
+    },
+    {
+      "epoch": 2.187433439829606,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.018373309222914e-06,
+      "loss": 0.9285038113594055,
+      "step": 1028
+    },
+    {
+      "epoch": 2.191693290734824,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.989055125638982e-06,
+      "loss": 0.861228346824646,
+      "step": 1030
+    },
+    {
+      "epoch": 2.1959531416400426,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.95984727809961e-06,
+      "loss": 0.8538438677787781,
+      "step": 1032
+    },
+    {
+      "epoch": 2.2002129925452607,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.930750409732679e-06,
+      "loss": 0.9002454876899719,
+      "step": 1034
+    },
+    {
+      "epoch": 2.2044728434504792,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.901765161222426e-06,
+      "loss": 0.8670125603675842,
+      "step": 1036
+    },
+    {
+      "epoch": 2.2087326943556977,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.8728921707953225e-06,
+      "loss": 0.8965079188346863,
+      "step": 1038
+    },
+    {
+      "epoch": 2.212992545260916,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.844132074206028e-06,
+      "loss": 0.9224348664283752,
+      "step": 1040
+    },
+    {
+      "epoch": 2.2172523961661343,
+      "grad_norm": 0.859375,
+      "learning_rate": 6.815485504723389e-06,
+      "loss": 0.8701453804969788,
+      "step": 1042
+    },
+    {
+      "epoch": 2.2215122470713524,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.786953093116508e-06,
+      "loss": 0.8607268333435059,
+      "step": 1044
+    },
+    {
+      "epoch": 2.225772097976571,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.758535467640832e-06,
+      "loss": 0.8190476298332214,
+      "step": 1046
+    },
+    {
+      "epoch": 2.230031948881789,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.730233254024339e-06,
+      "loss": 0.7581257820129395,
+      "step": 1048
+    },
+    {
+      "epoch": 2.2342917997870075,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.702047075453751e-06,
+      "loss": 0.867296576499939,
+      "step": 1050
+    },
+    {
+      "epoch": 2.2385516506922256,
+      "grad_norm": 0.8203125,
+      "learning_rate": 6.673977552560804e-06,
+      "loss": 0.8582524061203003,
+      "step": 1052
+    },
+    {
+      "epoch": 2.242811501597444,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.646025303408606e-06,
+      "loss": 0.8571093082427979,
+      "step": 1054
+    },
+    {
+      "epoch": 2.247071352502662,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.618190943478003e-06,
+      "loss": 0.867594301700592,
+      "step": 1056
+    },
+    {
+      "epoch": 2.2513312034078807,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.590475085654047e-06,
+      "loss": 0.7921283841133118,
+      "step": 1058
+    },
+    {
+      "epoch": 2.255591054313099,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.56287834021248e-06,
+      "loss": 0.8648768067359924,
+      "step": 1060
+    },
+    {
+      "epoch": 2.2598509052183173,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.535401314806309e-06,
+      "loss": 0.8589727282524109,
+      "step": 1062
+    },
+    {
+      "epoch": 2.264110756123536,
+      "grad_norm": 0.48828125,
+      "learning_rate": 6.508044614452431e-06,
+      "loss": 0.8992140889167786,
+      "step": 1064
+    },
+    {
+      "epoch": 2.268370607028754,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.480808841518298e-06,
+      "loss": 0.8318920135498047,
+      "step": 1066
+    },
+    {
+      "epoch": 2.2726304579339724,
+      "grad_norm": 0.80859375,
+      "learning_rate": 6.453694595708667e-06,
+      "loss": 0.8525157570838928,
+      "step": 1068
+    },
+    {
+      "epoch": 2.2768903088391905,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.426702474052381e-06,
+      "loss": 0.8547057509422302,
+      "step": 1070
+    },
+    {
+      "epoch": 2.281150159744409,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.39983307088923e-06,
+      "loss": 0.8861790299415588,
+      "step": 1072
+    },
+    {
+      "epoch": 2.285410010649627,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.373086977856868e-06,
+      "loss": 0.8417205810546875,
+      "step": 1074
+    },
+    {
+      "epoch": 2.2896698615548456,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.346464783877784e-06,
+      "loss": 0.9120402932167053,
+      "step": 1076
+    },
+    {
+      "epoch": 2.2939297124600637,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.319967075146332e-06,
+      "loss": 0.8795621991157532,
+      "step": 1078
+    },
+    {
+      "epoch": 2.2981895633652822,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.293594435115814e-06,
+      "loss": 0.884793758392334,
+      "step": 1080
+    },
+    {
+      "epoch": 2.3024494142705008,
+      "grad_norm": 0.482421875,
+      "learning_rate": 6.267347444485654e-06,
+      "loss": 0.9300041198730469,
+      "step": 1082
+    },
+    {
+      "epoch": 2.306709265175719,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.2412266811886005e-06,
+      "loss": 0.8604894876480103,
+      "step": 1084
+    },
+    {
+      "epoch": 2.3109691160809374,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.215232720377991e-06,
+      "loss": 0.8723834753036499,
+      "step": 1086
+    },
+    {
+      "epoch": 2.3152289669861554,
+      "grad_norm": 0.482421875,
+      "learning_rate": 6.189366134415112e-06,
+      "loss": 0.8270635008811951,
+      "step": 1088
+    },
+    {
+      "epoch": 2.319488817891374,
+      "grad_norm": 0.84375,
+      "learning_rate": 6.163627492856572e-06,
+      "loss": 0.873162567615509,
+      "step": 1090
+    },
+    {
+      "epoch": 2.323748668796592,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.13801736244177e-06,
+      "loss": 0.8589759469032288,
+      "step": 1092
+    },
+    {
+      "epoch": 2.3280085197018106,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.112536307080422e-06,
+      "loss": 0.8987887501716614,
+      "step": 1094
+    },
+    {
+      "epoch": 2.3322683706070286,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.087184887840139e-06,
+      "loss": 0.8068996071815491,
+      "step": 1096
+    },
+    {
+      "epoch": 2.336528221512247,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.061963662934072e-06,
+      "loss": 0.802603006362915,
+      "step": 1098
+    },
+    {
+      "epoch": 2.3407880724174652,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.036873187708615e-06,
+      "loss": 0.8735599517822266,
+      "step": 1100
+    },
+    {
+      "epoch": 2.3450479233226837,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.011914014631191e-06,
+      "loss": 0.7997400760650635,
+      "step": 1102
+    },
+    {
+      "epoch": 2.349307774227902,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.987086693278083e-06,
+      "loss": 0.8167827129364014,
+      "step": 1104
+    },
+    {
+      "epoch": 2.3535676251331203,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.962391770322319e-06,
+      "loss": 0.853871762752533,
+      "step": 1106
+    },
+    {
+      "epoch": 2.357827476038339,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.937829789521655e-06,
+      "loss": 0.8599565029144287,
+      "step": 1108
+    },
+    {
+      "epoch": 2.362087326943557,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.913401291706585e-06,
+      "loss": 0.8716296553611755,
+      "step": 1110
+    },
+    {
+      "epoch": 2.3663471778487755,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.889106814768442e-06,
+      "loss": 0.8599187731742859,
+      "step": 1112
+    },
+    {
+      "epoch": 2.3706070287539935,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.864946893647556e-06,
+      "loss": 0.8140082955360413,
+      "step": 1114
+    },
+    {
+      "epoch": 2.374866879659212,
+      "grad_norm": 0.8125,
+      "learning_rate": 5.8409220603214704e-06,
+      "loss": 0.8339557647705078,
+      "step": 1116
+    },
+    {
+      "epoch": 2.37912673056443,
+      "grad_norm": 0.447265625,
+      "learning_rate": 5.817032843793222e-06,
+      "loss": 0.828040599822998,
+      "step": 1118
+    },
+    {
+      "epoch": 2.3833865814696487,
+      "grad_norm": 0.80078125,
+      "learning_rate": 5.793279770079706e-06,
+      "loss": 0.869745135307312,
+      "step": 1120
+    },
+    {
+      "epoch": 2.3876464323748667,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.769663362200086e-06,
+      "loss": 0.8580743074417114,
+      "step": 1122
+    },
+    {
+      "epoch": 2.3919062832800853,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.746184140164285e-06,
+      "loss": 0.8233816027641296,
+      "step": 1124
+    },
+    {
+      "epoch": 2.3961661341853033,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.722842620961518e-06,
+      "loss": 0.8709803819656372,
+      "step": 1126
+    },
+    {
+      "epoch": 2.400425985090522,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.69963931854893e-06,
+      "loss": 0.9047388434410095,
+      "step": 1128
+    },
+    {
+      "epoch": 2.40468583599574,
+      "grad_norm": 1.890625,
+      "learning_rate": 5.676574743840264e-06,
+      "loss": 0.8460210561752319,
+      "step": 1130
+    },
+    {
+      "epoch": 2.4089456869009584,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.6536494046946174e-06,
+      "loss": 0.9233959913253784,
+      "step": 1132
+    },
+    {
+      "epoch": 2.413205537806177,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.630863805905259e-06,
+      "loss": 0.8673460483551025,
+      "step": 1134
+    },
+    {
+      "epoch": 2.417465388711395,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.608218449188516e-06,
+      "loss": 0.9345380663871765,
+      "step": 1136
+    },
+    {
+      "epoch": 2.4217252396166136,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.585713833172718e-06,
+      "loss": 0.9205328226089478,
+      "step": 1138
+    },
+    {
+      "epoch": 2.4259850905218316,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.563350453387224e-06,
+      "loss": 0.7964959740638733,
+      "step": 1140
+    },
+    {
+      "epoch": 2.43024494142705,
+      "grad_norm": 0.49609375,
+      "learning_rate": 5.541128802251518e-06,
+      "loss": 0.811107873916626,
+      "step": 1142
+    },
+    {
+      "epoch": 2.4345047923322682,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.5190493690643566e-06,
+      "loss": 0.787108838558197,
+      "step": 1144
+    },
+    {
+      "epoch": 2.4387646432374868,
+      "grad_norm": 17.375,
+      "learning_rate": 5.497112639992993e-06,
+      "loss": 0.9524678587913513,
+      "step": 1146
+    },
+    {
+      "epoch": 2.443024494142705,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.4753190980624834e-06,
+      "loss": 0.7983661890029907,
+      "step": 1148
+    },
+    {
+      "epoch": 2.4472843450479234,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.453669223145042e-06,
+      "loss": 0.8548356294631958,
+      "step": 1150
+    },
+    {
+      "epoch": 2.451544195953142,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.432163491949475e-06,
+      "loss": 0.9406867027282715,
+      "step": 1152
+    },
+    {
+      "epoch": 2.45580404685836,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.410802378010691e-06,
+      "loss": 0.8843863010406494,
+      "step": 1154
+    },
+    {
+      "epoch": 2.460063897763578,
+      "grad_norm": 1.125,
+      "learning_rate": 5.389586351679272e-06,
+      "loss": 0.8945695757865906,
+      "step": 1156
+    },
+    {
+      "epoch": 2.4643237486687966,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.368515880111107e-06,
+      "loss": 0.8422825336456299,
+      "step": 1158
+    },
+    {
+      "epoch": 2.468583599574015,
+      "grad_norm": 0.84375,
+      "learning_rate": 5.347591427257115e-06,
+      "loss": 0.940974771976471,
+      "step": 1160
+    },
+    {
+      "epoch": 2.472843450479233,
+      "grad_norm": 0.484375,
+      "learning_rate": 5.326813453853033e-06,
+      "loss": 0.849635124206543,
+      "step": 1162
+    },
+    {
+      "epoch": 2.4771033013844517,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.3061824174092645e-06,
+      "loss": 0.8890464901924133,
+      "step": 1164
+    },
+    {
+      "epoch": 2.4813631522896697,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.2856987722008004e-06,
+      "loss": 0.8929099440574646,
+      "step": 1166
+    },
+    {
+      "epoch": 2.4856230031948883,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.265362969257225e-06,
+      "loss": 0.811379075050354,
+      "step": 1168
+    },
+    {
+      "epoch": 2.4898828541001063,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.245175456352788e-06,
+      "loss": 0.8641577959060669,
+      "step": 1170
+    },
+    {
+      "epoch": 2.494142705005325,
+      "grad_norm": 0.79296875,
+      "learning_rate": 5.225136677996531e-06,
+      "loss": 0.8928030133247375,
+      "step": 1172
+    },
+    {
+      "epoch": 2.498402555910543,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.205247075422514e-06,
+      "loss": 0.8985695838928223,
+      "step": 1174
+    },
+    {
+      "epoch": 2.5026624068157615,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.185507086580094e-06,
+      "loss": 0.825484037399292,
+      "step": 1176
+    },
+    {
+      "epoch": 2.50692225772098,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.165917146124276e-06,
+      "loss": 0.8295057415962219,
+      "step": 1178
+    },
+    {
+      "epoch": 2.511182108626198,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.146477685406154e-06,
+      "loss": 0.8838967084884644,
+      "step": 1180
+    },
+    {
+      "epoch": 2.515441959531416,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.127189132463404e-06,
+      "loss": 0.8865538239479065,
+      "step": 1182
+    },
+    {
+      "epoch": 2.5197018104366347,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.108051912010865e-06,
+      "loss": 0.839586079120636,
+      "step": 1184
+    },
+    {
+      "epoch": 2.523961661341853,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.089066445431183e-06,
+      "loss": 0.8859398365020752,
+      "step": 1186
+    },
+    {
+      "epoch": 2.5282215122470713,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.070233150765531e-06,
+      "loss": 0.8947311639785767,
+      "step": 1188
+    },
+    {
+      "epoch": 2.5324813631522898,
+      "grad_norm": 0.7265625,
+      "learning_rate": 5.051552442704411e-06,
+      "loss": 0.8858785033226013,
+      "step": 1190
+    },
+    {
+      "epoch": 2.536741214057508,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.033024732578518e-06,
+      "loss": 0.8125391006469727,
+      "step": 1192
+    },
+    {
+      "epoch": 2.5410010649627264,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.014650428349683e-06,
+      "loss": 0.8854655027389526,
+      "step": 1194
+    },
+    {
+      "epoch": 2.545260915867945,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.996429934601895e-06,
+      "loss": 0.8128026127815247,
+      "step": 1196
+    },
+    {
+      "epoch": 2.549520766773163,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.978363652532383e-06,
+      "loss": 0.8405453562736511,
+      "step": 1198
+    },
+    {
+      "epoch": 2.553780617678381,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.960451979942789e-06,
+      "loss": 0.959130585193634,
+      "step": 1200
+    },
+    {
+      "epoch": 2.5580404685835996,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.942695311230408e-06,
+      "loss": 0.7953628897666931,
+      "step": 1202
+    },
+    {
+      "epoch": 2.562300319488818,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.925094037379507e-06,
+      "loss": 0.7662147283554077,
+      "step": 1204
+    },
+    {
+      "epoch": 2.566560170394036,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.907648545952706e-06,
+      "loss": 0.8059567213058472,
+      "step": 1206
+    },
+    {
+      "epoch": 2.5708200212992547,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.890359221082451e-06,
+      "loss": 0.9323968291282654,
+      "step": 1208
+    },
+    {
+      "epoch": 2.5750798722044728,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.87322644346256e-06,
+      "loss": 0.8555244207382202,
+      "step": 1210
+    },
+    {
+      "epoch": 2.5793397231096913,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.856250590339828e-06,
+      "loss": 0.8834168314933777,
+      "step": 1212
+    },
+    {
+      "epoch": 2.5835995740149094,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.8394320355057355e-06,
+      "loss": 0.9264652132987976,
+      "step": 1214
+    },
+    {
+      "epoch": 2.587859424920128,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.822771149288208e-06,
+      "loss": 0.8777004480361938,
+      "step": 1216
+    },
+    {
+      "epoch": 2.592119275825346,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.80626829854346e-06,
+      "loss": 0.8719021081924438,
+      "step": 1218
+    },
+    {
+      "epoch": 2.5963791267305645,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.789923846647926e-06,
+      "loss": 0.8178717494010925,
+      "step": 1220
+    },
+    {
+      "epoch": 2.600638977635783,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.773738153490252e-06,
+      "loss": 0.9265211820602417,
+      "step": 1222
+    },
+    {
+      "epoch": 2.604898828541001,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.75771157546338e-06,
+      "loss": 0.9368568658828735,
+      "step": 1224
+    },
+    {
+      "epoch": 2.609158679446219,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.7418444654566845e-06,
+      "loss": 0.8576453328132629,
+      "step": 1226
+    },
+    {
+      "epoch": 2.6134185303514377,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.726137172848219e-06,
+      "loss": 0.8761544823646545,
+      "step": 1228
+    },
+    {
+      "epoch": 2.617678381256656,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.710590043497013e-06,
+      "loss": 0.8335532546043396,
+      "step": 1230
+    },
+    {
+      "epoch": 2.6219382321618743,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.695203419735462e-06,
+      "loss": 0.8370216488838196,
+      "step": 1232
+    },
+    {
+      "epoch": 2.626198083067093,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.6799776403617875e-06,
+      "loss": 0.844344973564148,
+      "step": 1234
+    },
+    {
+      "epoch": 2.630457933972311,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.6649130406325726e-06,
+      "loss": 0.9227336645126343,
+      "step": 1236
+    },
+    {
+      "epoch": 2.6347177848775294,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.650009952255389e-06,
+      "loss": 0.8581987023353577,
+      "step": 1238
+    },
+    {
+      "epoch": 2.6389776357827475,
+      "grad_norm": 1.4140625,
+      "learning_rate": 4.635268703381484e-06,
+      "loss": 0.8400872349739075,
+      "step": 1240
+    },
+    {
+      "epoch": 2.643237486687966,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.620689618598568e-06,
+      "loss": 0.8582939505577087,
+      "step": 1242
+    },
+    {
+      "epoch": 2.647497337593184,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.606273018923644e-06,
+      "loss": 0.9465791583061218,
+      "step": 1244
+    },
+    {
+      "epoch": 2.6517571884984026,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.592019221795969e-06,
+      "loss": 0.8765722513198853,
+      "step": 1246
+    },
+    {
+      "epoch": 2.656017039403621,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.57792854107004e-06,
+      "loss": 0.8363880515098572,
+      "step": 1248
+    },
+    {
+      "epoch": 2.660276890308839,
+      "grad_norm": 0.625,
+      "learning_rate": 4.564001287008695e-06,
+      "loss": 0.9550789594650269,
+      "step": 1250
+    },
+    {
+      "epoch": 2.6645367412140573,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.550237766276277e-06,
+      "loss": 0.8685204982757568,
+      "step": 1252
+    },
+    {
+      "epoch": 2.668796592119276,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.536638281931887e-06,
+      "loss": 0.8562531471252441,
+      "step": 1254
+    },
+    {
+      "epoch": 2.6730564430244943,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.523203133422705e-06,
+      "loss": 0.8494756817817688,
+      "step": 1256
+    },
+    {
+      "epoch": 2.6773162939297124,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.509932616577394e-06,
+      "loss": 0.899334728717804,
+      "step": 1258
+    },
+    {
+      "epoch": 2.681576144834931,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.4968270235995995e-06,
+      "loss": 0.8993234634399414,
+      "step": 1260
+    },
+    {
+      "epoch": 2.685835995740149,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.483886643061501e-06,
+      "loss": 0.8710511326789856,
+      "step": 1262
+    },
+    {
+      "epoch": 2.6900958466453675,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.471111759897464e-06,
+      "loss": 0.8785170912742615,
+      "step": 1264
+    },
+    {
+      "epoch": 2.6943556975505856,
+      "grad_norm": 1.6796875,
+      "learning_rate": 4.45850265539777e-06,
+      "loss": 0.8322556018829346,
+      "step": 1266
+    },
+    {
+      "epoch": 2.698615548455804,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.4460596072024114e-06,
+      "loss": 0.8931154012680054,
+      "step": 1268
+    },
+    {
+      "epoch": 2.702875399361022,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.433782889294992e-06,
+      "loss": 0.8494217395782471,
+      "step": 1270
+    },
+    {
+      "epoch": 2.7071352502662407,
+      "grad_norm": 1.0,
+      "learning_rate": 4.4216727719966825e-06,
+      "loss": 0.8380395174026489,
+      "step": 1272
+    },
+    {
+      "epoch": 2.711395101171459,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.409729521960276e-06,
+      "loss": 0.89030522108078,
+      "step": 1274
+    },
+    {
+      "epoch": 2.7156549520766773,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.397953402164312e-06,
+      "loss": 0.8374807238578796,
+      "step": 1276
+    },
+    {
+      "epoch": 2.7199148029818954,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.386344671907289e-06,
+      "loss": 0.8489108085632324,
+      "step": 1278
+    },
+    {
+      "epoch": 2.724174653887114,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.374903586801952e-06,
+      "loss": 0.9123415946960449,
+      "step": 1280
+    },
+    {
+      "epoch": 2.7284345047923324,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.363630398769667e-06,
+      "loss": 0.8218215107917786,
+      "step": 1282
+    },
+    {
+      "epoch": 2.7326943556975505,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.352525356034874e-06,
+      "loss": 0.8863515853881836,
+      "step": 1284
+    },
+    {
+      "epoch": 2.736954206602769,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.341588703119615e-06,
+      "loss": 0.8300585150718689,
+      "step": 1286
+    },
+    {
+      "epoch": 2.741214057507987,
+      "grad_norm": 1.96875,
+      "learning_rate": 4.330820680838162e-06,
+      "loss": 0.8769638538360596,
+      "step": 1288
+    },
+    {
+      "epoch": 2.7454739084132056,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.320221526291701e-06,
+      "loss": 0.8285037279129028,
+      "step": 1290
+    },
+    {
+      "epoch": 2.749733759318424,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.309791472863121e-06,
+      "loss": 0.8763640522956848,
+      "step": 1292
+    },
+    {
+      "epoch": 2.753993610223642,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.2995307502118735e-06,
+      "loss": 0.8772199749946594,
+      "step": 1294
+    },
+    {
+      "epoch": 2.7582534611288603,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.289439584268909e-06,
+      "loss": 0.9215362668037415,
+      "step": 1296
+    },
+    {
+      "epoch": 2.762513312034079,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.279518197231709e-06,
+      "loss": 0.781262218952179,
+      "step": 1298
+    },
+    {
+      "epoch": 2.7667731629392973,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.2697668075593955e-06,
+      "loss": 0.8765016198158264,
+      "step": 1300
+    },
+    {
+      "epoch": 2.7710330138445154,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.260185629967912e-06,
+      "loss": 0.9362728595733643,
+      "step": 1302
+    },
+    {
+      "epoch": 2.775292864749734,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.250774875425303e-06,
+      "loss": 0.9154821634292603,
+      "step": 1304
+    },
+    {
+      "epoch": 2.779552715654952,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.241534751147065e-06,
+      "loss": 0.8477409482002258,
+      "step": 1306
+    },
+    {
+      "epoch": 2.7838125665601705,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.232465460591584e-06,
+      "loss": 0.8393524289131165,
+      "step": 1308
+    },
+    {
+      "epoch": 2.7880724174653886,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.223567203455662e-06,
+      "loss": 1.008418083190918,
+      "step": 1310
+    },
+    {
+      "epoch": 2.792332268370607,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.214840175670106e-06,
+      "loss": 0.8442473411560059,
+      "step": 1312
+    },
+    {
+      "epoch": 2.796592119275825,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.2062845693954315e-06,
+      "loss": 0.8037950992584229,
+      "step": 1314
+    },
+    {
+      "epoch": 2.8008519701810437,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.197900573017613e-06,
+      "loss": 0.9449222683906555,
+      "step": 1316
+    },
+    {
+      "epoch": 2.8051118210862622,
+      "grad_norm": 1.71875,
+      "learning_rate": 4.189688371143952e-06,
+      "loss": 0.9002764225006104,
+      "step": 1318
+    },
+    {
+      "epoch": 2.8093716719914803,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.181648144599001e-06,
+      "loss": 0.9393887519836426,
+      "step": 1320
+    },
+    {
+      "epoch": 2.8136315228966984,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.17378007042059e-06,
+      "loss": 0.9079067707061768,
+      "step": 1322
+    },
+    {
+      "epoch": 2.817891373801917,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.166084321855919e-06,
+      "loss": 0.9096862077713013,
+      "step": 1324
+    },
+    {
+      "epoch": 2.8221512247071354,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.158561068357751e-06,
+      "loss": 0.9297478199005127,
+      "step": 1326
+    },
+    {
+      "epoch": 2.8264110756123535,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.151210475580683e-06,
+      "loss": 0.7916253805160522,
+      "step": 1328
+    },
+    {
+      "epoch": 2.830670926517572,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.144032705377486e-06,
+      "loss": 0.7997951507568359,
+      "step": 1330
+    },
+    {
+      "epoch": 2.83493077742279,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.137027915795555e-06,
+      "loss": 0.8519441485404968,
+      "step": 1332
+    },
+    {
+      "epoch": 2.8391906283280086,
+      "grad_norm": 0.5,
+      "learning_rate": 4.130196261073419e-06,
+      "loss": 0.8826906681060791,
+      "step": 1334
+    },
+    {
+      "epoch": 2.8434504792332267,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.123537891637352e-06,
+      "loss": 0.9272059798240662,
+      "step": 1336
+    },
+    {
+      "epoch": 2.847710330138445,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.117052954098055e-06,
+      "loss": 0.8484643697738647,
+      "step": 1338
+    },
+    {
+      "epoch": 2.8519701810436633,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.110741591247433e-06,
+      "loss": 0.8416329026222229,
+      "step": 1340
+    },
+    {
+      "epoch": 2.856230031948882,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.1046039420554465e-06,
+      "loss": 0.7760175466537476,
+      "step": 1342
+    },
+    {
+      "epoch": 2.8604898828541003,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.0986401416670535e-06,
+      "loss": 0.8102884888648987,
+      "step": 1344
+    },
+    {
+      "epoch": 2.8647497337593184,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.092850321399232e-06,
+      "loss": 0.8387068510055542,
+      "step": 1346
+    },
+    {
+      "epoch": 2.8690095846645365,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.0872346087380924e-06,
+      "loss": 0.8194290399551392,
+      "step": 1348
+    },
+    {
+      "epoch": 2.873269435569755,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.0817931273360686e-06,
+      "loss": 0.8201800584793091,
+      "step": 1350
+    },
+    {
+      "epoch": 2.8775292864749735,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.076525997009191e-06,
+      "loss": 0.8607783913612366,
+      "step": 1352
+    },
+    {
+      "epoch": 2.8817891373801916,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.0714333337344565e-06,
+      "loss": 0.8402998447418213,
+      "step": 1354
+    },
+    {
+      "epoch": 2.88604898828541,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.066515249647266e-06,
+      "loss": 0.8459066152572632,
+      "step": 1356
+    },
+    {
+      "epoch": 2.890308839190628,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.061771853038961e-06,
+      "loss": 0.8821372985839844,
+      "step": 1358
+    },
+    {
+      "epoch": 2.8945686900958467,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.05720324835444e-06,
+      "loss": 0.8806694746017456,
+      "step": 1360
+    },
+    {
+      "epoch": 2.8988285410010652,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.052809536189854e-06,
+      "loss": 0.9144023656845093,
+      "step": 1362
+    },
+    {
+      "epoch": 2.9030883919062833,
+      "grad_norm": 1.25,
+      "learning_rate": 4.048590813290395e-06,
+      "loss": 0.8206741213798523,
+      "step": 1364
+    },
+    {
+      "epoch": 2.9073482428115014,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.044547172548163e-06,
+      "loss": 0.8178682923316956,
+      "step": 1366
+    },
+    {
+      "epoch": 2.91160809371672,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.040678703000123e-06,
+      "loss": 0.8602153658866882,
+      "step": 1368
+    },
+    {
+      "epoch": 2.9158679446219384,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.036985489826145e-06,
+      "loss": 0.8640646934509277,
+      "step": 1370
+    },
+    {
+      "epoch": 2.9201277955271565,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.033467614347126e-06,
+      "loss": 0.8199517130851746,
+      "step": 1372
+    },
+    {
+      "epoch": 2.924387646432375,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.030125154023201e-06,
+      "loss": 0.8197891712188721,
+      "step": 1374
+    },
+    {
+      "epoch": 2.928647497337593,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.026958182452037e-06,
+      "loss": 0.9104003310203552,
+      "step": 1376
+    },
+    {
+      "epoch": 2.9329073482428116,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.023966769367212e-06,
+      "loss": 0.9194601774215698,
+      "step": 1378
+    },
+    {
+      "epoch": 2.9371671991480297,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.021150980636679e-06,
+      "loss": 0.8961591720581055,
+      "step": 1380
+    },
+    {
+      "epoch": 2.9414270500532482,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.01851087826132e-06,
+      "loss": 0.9252373576164246,
+      "step": 1382
+    },
+    {
+      "epoch": 2.9456869009584663,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.016046520373573e-06,
+      "loss": 0.8721043467521667,
+      "step": 1384
+    },
+    {
+      "epoch": 2.949946751863685,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.013757961236162e-06,
+      "loss": 0.8307064771652222,
+      "step": 1386
+    },
+    {
+      "epoch": 2.9542066027689033,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.011645251240889e-06,
+      "loss": 0.8707568049430847,
+      "step": 1388
+    },
+    {
+      "epoch": 2.9584664536741214,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.009708436907538e-06,
+      "loss": 0.826111376285553,
+      "step": 1390
+    },
+    {
+      "epoch": 2.9627263045793395,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.007947560882844e-06,
+      "loss": 0.8147805333137512,
+      "step": 1392
+    },
+    {
+      "epoch": 2.966986155484558,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.006362661939552e-06,
+      "loss": 0.9187301993370056,
+      "step": 1394
+    },
+    {
+      "epoch": 2.9712460063897765,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.0049537749755685e-06,
+      "loss": 0.7911329865455627,
+      "step": 1396
+    },
+    {
+      "epoch": 2.9755058572949946,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.003720931013188e-06,
+      "loss": 0.8627235293388367,
+      "step": 1398
+    },
+    {
+      "epoch": 2.979765708200213,
+      "grad_norm": 1.75,
+      "learning_rate": 4.002664157198416e-06,
+      "loss": 0.7859567999839783,
+      "step": 1400
+    },
+    {
+      "epoch": 2.984025559105431,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.001783476800364e-06,
+      "loss": 0.8008852005004883,
+      "step": 1402
+    },
+    {
+      "epoch": 2.9882854100106497,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.001078909210742e-06,
+      "loss": 0.8270421624183655,
+      "step": 1404
+    },
+    {
+      "epoch": 2.992545260915868,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.000550469943431e-06,
+      "loss": 0.8458138704299927,
+      "step": 1406
+    },
+    {
+      "epoch": 2.9968051118210863,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.0001981706341416e-06,
+      "loss": 0.8161488771438599,
+      "step": 1408
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.000022019040156e-06,
+      "loss": 0.8455473780632019,
+      "step": 1410
+    },
+    {
+      "epoch": 3.0,
+      "step": 1410,
+      "total_flos": 4.4194007019526554e+18,
+      "train_loss": 1.0514326345836018,
+      "train_runtime": 11229.2105,
+      "train_samples_per_second": 4.014,
+      "train_steps_per_second": 0.126
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1410,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 99999,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.4194007019526554e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}