diff --git "a/checkpoint-2514/trainer_state.json" "b/checkpoint-2514/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2514/trainer_state.json"
@@ -0,0 +1,5899 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 2514,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002386634844868735,
+      "grad_norm": 87.45082092285156,
+      "learning_rate": 2.3809523809523808e-06,
+      "loss": 8.3691,
+      "step": 3
+    },
+    {
+      "epoch": 0.00477326968973747,
+      "grad_norm": 42.669490814208984,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 7.0093,
+      "step": 6
+    },
+    {
+      "epoch": 0.007159904534606206,
+      "grad_norm": 24.51582145690918,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 5.8097,
+      "step": 9
+    },
+    {
+      "epoch": 0.00954653937947494,
+      "grad_norm": 16.242338180541992,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 4.978,
+      "step": 12
+    },
+    {
+      "epoch": 0.011933174224343675,
+      "grad_norm": 11.578532218933105,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 4.8608,
+      "step": 15
+    },
+    {
+      "epoch": 0.014319809069212411,
+      "grad_norm": 14.094877243041992,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 4.6646,
+      "step": 18
+    },
+    {
+      "epoch": 0.016706443914081145,
+      "grad_norm": 8.234806060791016,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 4.428,
+      "step": 21
+    },
+    {
+      "epoch": 0.01909307875894988,
+      "grad_norm": 9.00922966003418,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 4.2484,
+      "step": 24
+    },
+    {
+      "epoch": 0.021479713603818614,
+      "grad_norm": 8.009718894958496,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 4.1826,
+      "step": 27
+    },
+    {
+      "epoch": 0.02386634844868735,
+      "grad_norm": 5.844778537750244,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 4.0247,
+      "step": 30
+    },
+    {
+      "epoch": 0.026252983293556086,
+      "grad_norm": 5.393738746643066,
+      "learning_rate": 2.6190476190476192e-05,
+      "loss": 3.6734,
+      "step": 33
+    },
+    {
+      "epoch": 0.028639618138424822,
+      "grad_norm": 5.676444053649902,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 3.621,
+      "step": 36
+    },
+    {
+      "epoch": 0.031026252983293555,
+      "grad_norm": 5.530320644378662,
+      "learning_rate": 3.095238095238095e-05,
+      "loss": 3.4785,
+      "step": 39
+    },
+    {
+      "epoch": 0.03341288782816229,
+      "grad_norm": 3.998183488845825,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 3.2492,
+      "step": 42
+    },
+    {
+      "epoch": 0.03579952267303103,
+      "grad_norm": 5.9084553718566895,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 3.2145,
+      "step": 45
+    },
+    {
+      "epoch": 0.03818615751789976,
+      "grad_norm": 4.704463958740234,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 3.1231,
+      "step": 48
+    },
+    {
+      "epoch": 0.0405727923627685,
+      "grad_norm": 4.222476482391357,
+      "learning_rate": 4.047619047619048e-05,
+      "loss": 2.9818,
+      "step": 51
+    },
+    {
+      "epoch": 0.04295942720763723,
+      "grad_norm": 6.013700008392334,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 3.0594,
+      "step": 54
+    },
+    {
+      "epoch": 0.045346062052505964,
+      "grad_norm": 5.095407009124756,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 2.9516,
+      "step": 57
+    },
+    {
+      "epoch": 0.0477326968973747,
+      "grad_norm": 4.415056228637695,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 2.8278,
+      "step": 60
+    },
+    {
+      "epoch": 0.050119331742243436,
+      "grad_norm": 4.036295413970947,
+      "learning_rate": 5e-05,
+      "loss": 2.7948,
+      "step": 63
+    },
+    {
+      "epoch": 0.05250596658711217,
+      "grad_norm": 4.796054840087891,
+      "learning_rate": 5.2380952380952384e-05,
+      "loss": 2.7323,
+      "step": 66
+    },
+    {
+      "epoch": 0.05489260143198091,
+      "grad_norm": 3.858863592147827,
+      "learning_rate": 5.4761904761904766e-05,
+      "loss": 2.5518,
+      "step": 69
+    },
+    {
+      "epoch": 0.057279236276849645,
+      "grad_norm": 2.7978243827819824,
+      "learning_rate": 5.714285714285714e-05,
+      "loss": 2.5852,
+      "step": 72
+    },
+    {
+      "epoch": 0.059665871121718374,
+      "grad_norm": 3.5701358318328857,
+      "learning_rate": 5.9523809523809524e-05,
+      "loss": 2.6322,
+      "step": 75
+    },
+    {
+      "epoch": 0.06205250596658711,
+      "grad_norm": 3.843608856201172,
+      "learning_rate": 6.19047619047619e-05,
+      "loss": 2.6322,
+      "step": 78
+    },
+    {
+      "epoch": 0.06443914081145585,
+      "grad_norm": 4.596922874450684,
+      "learning_rate": 6.428571428571429e-05,
+      "loss": 2.5773,
+      "step": 81
+    },
+    {
+      "epoch": 0.06682577565632458,
+      "grad_norm": 3.2659518718719482,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 2.5062,
+      "step": 84
+    },
+    {
+      "epoch": 0.06921241050119331,
+      "grad_norm": 2.8752944469451904,
+      "learning_rate": 6.904761904761905e-05,
+      "loss": 2.5073,
+      "step": 87
+    },
+    {
+      "epoch": 0.07159904534606205,
+      "grad_norm": 3.333749294281006,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 2.5105,
+      "step": 90
+    },
+    {
+      "epoch": 0.07398568019093078,
+      "grad_norm": 2.5306403636932373,
+      "learning_rate": 7.380952380952382e-05,
+      "loss": 2.5395,
+      "step": 93
+    },
+    {
+      "epoch": 0.07637231503579953,
+      "grad_norm": 2.989917278289795,
+      "learning_rate": 7.619047619047618e-05,
+      "loss": 2.4063,
+      "step": 96
+    },
+    {
+      "epoch": 0.07875894988066826,
+      "grad_norm": 3.6913046836853027,
+      "learning_rate": 7.857142857142858e-05,
+      "loss": 2.385,
+      "step": 99
+    },
+    {
+      "epoch": 0.081145584725537,
+      "grad_norm": 3.3200862407684326,
+      "learning_rate": 8.095238095238096e-05,
+      "loss": 2.3451,
+      "step": 102
+    },
+    {
+      "epoch": 0.08353221957040573,
+      "grad_norm": 3.1073644161224365,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 2.3661,
+      "step": 105
+    },
+    {
+      "epoch": 0.08591885441527446,
+      "grad_norm": 2.79823637008667,
+      "learning_rate": 8.571428571428571e-05,
+      "loss": 2.3642,
+      "step": 108
+    },
+    {
+      "epoch": 0.0883054892601432,
+      "grad_norm": 2.8037662506103516,
+      "learning_rate": 8.80952380952381e-05,
+      "loss": 2.3049,
+      "step": 111
+    },
+    {
+      "epoch": 0.09069212410501193,
+      "grad_norm": 2.33240008354187,
+      "learning_rate": 9.047619047619048e-05,
+      "loss": 2.4671,
+      "step": 114
+    },
+    {
+      "epoch": 0.09307875894988067,
+      "grad_norm": 2.2528834342956543,
+      "learning_rate": 9.285714285714286e-05,
+      "loss": 2.3363,
+      "step": 117
+    },
+    {
+      "epoch": 0.0954653937947494,
+      "grad_norm": 3.4800667762756348,
+      "learning_rate": 9.523809523809524e-05,
+      "loss": 2.3207,
+      "step": 120
+    },
+    {
+      "epoch": 0.09785202863961814,
+      "grad_norm": 2.395854949951172,
+      "learning_rate": 9.761904761904762e-05,
+      "loss": 2.3903,
+      "step": 123
+    },
+    {
+      "epoch": 0.10023866348448687,
+      "grad_norm": 3.5940475463867188,
+      "learning_rate": 0.0001,
+      "loss": 2.1336,
+      "step": 126
+    },
+    {
+      "epoch": 0.1026252983293556,
+      "grad_norm": 3.4392244815826416,
+      "learning_rate": 9.999961058466053e-05,
+      "loss": 2.3802,
+      "step": 129
+    },
+    {
+      "epoch": 0.10501193317422435,
+      "grad_norm": 2.595665216445923,
+      "learning_rate": 9.999844234470782e-05,
+      "loss": 2.2468,
+      "step": 132
+    },
+    {
+      "epoch": 0.10739856801909307,
+      "grad_norm": 2.5754427909851074,
+      "learning_rate": 9.999649529833915e-05,
+      "loss": 2.2207,
+      "step": 135
+    },
+    {
+      "epoch": 0.10978520286396182,
+      "grad_norm": 2.359457492828369,
+      "learning_rate": 9.999376947588288e-05,
+      "loss": 2.3501,
+      "step": 138
+    },
+    {
+      "epoch": 0.11217183770883055,
+      "grad_norm": 2.7424278259277344,
+      "learning_rate": 9.999026491979808e-05,
+      "loss": 2.3205,
+      "step": 141
+    },
+    {
+      "epoch": 0.11455847255369929,
+      "grad_norm": 2.8538436889648438,
+      "learning_rate": 9.99859816846739e-05,
+      "loss": 2.2671,
+      "step": 144
+    },
+    {
+      "epoch": 0.11694510739856802,
+      "grad_norm": 2.7669293880462646,
+      "learning_rate": 9.998091983722863e-05,
+      "loss": 2.2559,
+      "step": 147
+    },
+    {
+      "epoch": 0.11933174224343675,
+      "grad_norm": 2.62872314453125,
+      "learning_rate": 9.99750794563087e-05,
+      "loss": 2.1225,
+      "step": 150
+    },
+    {
+      "epoch": 0.12171837708830549,
+      "grad_norm": 3.0914721488952637,
+      "learning_rate": 9.996846063288747e-05,
+      "loss": 2.2146,
+      "step": 153
+    },
+    {
+      "epoch": 0.12410501193317422,
+      "grad_norm": 2.880335569381714,
+      "learning_rate": 9.996106347006379e-05,
+      "loss": 2.3611,
+      "step": 156
+    },
+    {
+      "epoch": 0.12649164677804295,
+      "grad_norm": 4.729184150695801,
+      "learning_rate": 9.99528880830604e-05,
+      "loss": 2.1814,
+      "step": 159
+    },
+    {
+      "epoch": 0.1288782816229117,
+      "grad_norm": 2.658954381942749,
+      "learning_rate": 9.994393459922218e-05,
+      "loss": 2.1278,
+      "step": 162
+    },
+    {
+      "epoch": 0.13126491646778043,
+      "grad_norm": 2.626692056655884,
+      "learning_rate": 9.993420315801406e-05,
+      "loss": 2.0775,
+      "step": 165
+    },
+    {
+      "epoch": 0.13365155131264916,
+      "grad_norm": 3.7431132793426514,
+      "learning_rate": 9.992369391101895e-05,
+      "loss": 2.1602,
+      "step": 168
+    },
+    {
+      "epoch": 0.1360381861575179,
+      "grad_norm": 2.7827000617980957,
+      "learning_rate": 9.991240702193532e-05,
+      "loss": 2.291,
+      "step": 171
+    },
+    {
+      "epoch": 0.13842482100238662,
+      "grad_norm": 2.846406936645508,
+      "learning_rate": 9.990034266657467e-05,
+      "loss": 2.1872,
+      "step": 174
+    },
+    {
+      "epoch": 0.14081145584725538,
+      "grad_norm": 3.0862011909484863,
+      "learning_rate": 9.988750103285883e-05,
+      "loss": 2.0943,
+      "step": 177
+    },
+    {
+      "epoch": 0.1431980906921241,
+      "grad_norm": 2.312624931335449,
+      "learning_rate": 9.987388232081694e-05,
+      "loss": 2.1579,
+      "step": 180
+    },
+    {
+      "epoch": 0.14558472553699284,
+      "grad_norm": 2.634638786315918,
+      "learning_rate": 9.985948674258243e-05,
+      "loss": 2.081,
+      "step": 183
+    },
+    {
+      "epoch": 0.14797136038186157,
+      "grad_norm": 3.84676194190979,
+      "learning_rate": 9.984431452238967e-05,
+      "loss": 2.2108,
+      "step": 186
+    },
+    {
+      "epoch": 0.15035799522673032,
+      "grad_norm": 3.062570810317993,
+      "learning_rate": 9.982836589657043e-05,
+      "loss": 2.1521,
+      "step": 189
+    },
+    {
+      "epoch": 0.15274463007159905,
+      "grad_norm": 2.0936121940612793,
+      "learning_rate": 9.981164111355035e-05,
+      "loss": 2.1847,
+      "step": 192
+    },
+    {
+      "epoch": 0.15513126491646778,
+      "grad_norm": 2.978933811187744,
+      "learning_rate": 9.979414043384485e-05,
+      "loss": 2.1839,
+      "step": 195
+    },
+    {
+      "epoch": 0.1575178997613365,
+      "grad_norm": 2.353105306625366,
+      "learning_rate": 9.977586413005531e-05,
+      "loss": 2.2415,
+      "step": 198
+    },
+    {
+      "epoch": 0.15990453460620524,
+      "grad_norm": 2.6958158016204834,
+      "learning_rate": 9.975681248686461e-05,
+      "loss": 2.2367,
+      "step": 201
+    },
+    {
+      "epoch": 0.162291169451074,
+      "grad_norm": 2.262739658355713,
+      "learning_rate": 9.973698580103285e-05,
+      "loss": 2.1569,
+      "step": 204
+    },
+    {
+      "epoch": 0.16467780429594273,
+      "grad_norm": 2.56048583984375,
+      "learning_rate": 9.971638438139266e-05,
+      "loss": 2.3021,
+      "step": 207
+    },
+    {
+      "epoch": 0.16706443914081145,
+      "grad_norm": 4.17548942565918,
+      "learning_rate": 9.96950085488444e-05,
+      "loss": 2.028,
+      "step": 210
+    },
+    {
+      "epoch": 0.16945107398568018,
+      "grad_norm": 4.712838649749756,
+      "learning_rate": 9.967285863635112e-05,
+      "loss": 2.2895,
+      "step": 213
+    },
+    {
+      "epoch": 0.1718377088305489,
+      "grad_norm": 2.642404556274414,
+      "learning_rate": 9.964993498893349e-05,
+      "loss": 2.1517,
+      "step": 216
+    },
+    {
+      "epoch": 0.17422434367541767,
+      "grad_norm": 3.9263668060302734,
+      "learning_rate": 9.962623796366429e-05,
+      "loss": 2.1385,
+      "step": 219
+    },
+    {
+      "epoch": 0.1766109785202864,
+      "grad_norm": 2.9646389484405518,
+      "learning_rate": 9.960176792966289e-05,
+      "loss": 2.1171,
+      "step": 222
+    },
+    {
+      "epoch": 0.17899761336515513,
+      "grad_norm": 2.0947139263153076,
+      "learning_rate": 9.95765252680896e-05,
+      "loss": 2.174,
+      "step": 225
+    },
+    {
+      "epoch": 0.18138424821002386,
+      "grad_norm": 2.083189010620117,
+      "learning_rate": 9.95505103721396e-05,
+      "loss": 2.1498,
+      "step": 228
+    },
+    {
+      "epoch": 0.18377088305489261,
+      "grad_norm": 2.691925048828125,
+      "learning_rate": 9.952372364703687e-05,
+      "loss": 2.2291,
+      "step": 231
+    },
+    {
+      "epoch": 0.18615751789976134,
+      "grad_norm": 3.0350098609924316,
+      "learning_rate": 9.949616551002787e-05,
+      "loss": 2.1164,
+      "step": 234
+    },
+    {
+      "epoch": 0.18854415274463007,
+      "grad_norm": 2.8906643390655518,
+      "learning_rate": 9.946783639037504e-05,
+      "loss": 2.0806,
+      "step": 237
+    },
+    {
+      "epoch": 0.1909307875894988,
+      "grad_norm": 3.903940439224243,
+      "learning_rate": 9.943873672935014e-05,
+      "loss": 2.1149,
+      "step": 240
+    },
+    {
+      "epoch": 0.19331742243436753,
+      "grad_norm": 2.354952096939087,
+      "learning_rate": 9.940886698022734e-05,
+      "loss": 2.0019,
+      "step": 243
+    },
+    {
+      "epoch": 0.1957040572792363,
+      "grad_norm": 2.914287805557251,
+      "learning_rate": 9.93782276082762e-05,
+      "loss": 2.1007,
+      "step": 246
+    },
+    {
+      "epoch": 0.19809069212410502,
+      "grad_norm": 2.2259702682495117,
+      "learning_rate": 9.934681909075434e-05,
+      "loss": 2.0461,
+      "step": 249
+    },
+    {
+      "epoch": 0.20047732696897375,
+      "grad_norm": 2.6093738079071045,
+      "learning_rate": 9.931464191690015e-05,
+      "loss": 2.125,
+      "step": 252
+    },
+    {
+      "epoch": 0.20286396181384247,
+      "grad_norm": 2.4364943504333496,
+      "learning_rate": 9.928169658792498e-05,
+      "loss": 2.1239,
+      "step": 255
+    },
+    {
+      "epoch": 0.2052505966587112,
+      "grad_norm": 4.371973514556885,
+      "learning_rate": 9.924798361700553e-05,
+      "loss": 2.0904,
+      "step": 258
+    },
+    {
+      "epoch": 0.20763723150357996,
+      "grad_norm": 2.1466689109802246,
+      "learning_rate": 9.92135035292757e-05,
+      "loss": 2.1023,
+      "step": 261
+    },
+    {
+      "epoch": 0.2100238663484487,
+      "grad_norm": 2.1940388679504395,
+      "learning_rate": 9.91782568618185e-05,
+      "loss": 2.1517,
+      "step": 264
+    },
+    {
+      "epoch": 0.21241050119331742,
+      "grad_norm": 2.100590229034424,
+      "learning_rate": 9.914224416365764e-05,
+      "loss": 2.1644,
+      "step": 267
+    },
+    {
+      "epoch": 0.21479713603818615,
+      "grad_norm": 2.8541860580444336,
+      "learning_rate": 9.910546599574902e-05,
+      "loss": 2.3204,
+      "step": 270
+    },
+    {
+      "epoch": 0.2171837708830549,
+      "grad_norm": 4.037013530731201,
+      "learning_rate": 9.906792293097194e-05,
+      "loss": 2.1066,
+      "step": 273
+    },
+    {
+      "epoch": 0.21957040572792363,
+      "grad_norm": 2.3364968299865723,
+      "learning_rate": 9.90296155541202e-05,
+      "loss": 2.1277,
+      "step": 276
+    },
+    {
+      "epoch": 0.22195704057279236,
+      "grad_norm": 2.3627822399139404,
+      "learning_rate": 9.899054446189304e-05,
+      "loss": 1.954,
+      "step": 279
+    },
+    {
+      "epoch": 0.2243436754176611,
+      "grad_norm": 2.593341588973999,
+      "learning_rate": 9.895071026288574e-05,
+      "loss": 2.0014,
+      "step": 282
+    },
+    {
+      "epoch": 0.22673031026252982,
+      "grad_norm": 2.349030017852783,
+      "learning_rate": 9.891011357758022e-05,
+      "loss": 1.9998,
+      "step": 285
+    },
+    {
+      "epoch": 0.22911694510739858,
+      "grad_norm": 2.243206024169922,
+      "learning_rate": 9.886875503833536e-05,
+      "loss": 2.1566,
+      "step": 288
+    },
+    {
+      "epoch": 0.2315035799522673,
+      "grad_norm": 2.4595718383789062,
+      "learning_rate": 9.882663528937717e-05,
+      "loss": 2.0575,
+      "step": 291
+    },
+    {
+      "epoch": 0.23389021479713604,
+      "grad_norm": 2.3169078826904297,
+      "learning_rate": 9.87837549867887e-05,
+      "loss": 1.9049,
+      "step": 294
+    },
+    {
+      "epoch": 0.23627684964200477,
+      "grad_norm": 2.022441864013672,
+      "learning_rate": 9.87401147984998e-05,
+      "loss": 2.1902,
+      "step": 297
+    },
+    {
+      "epoch": 0.2386634844868735,
+      "grad_norm": 2.343163251876831,
+      "learning_rate": 9.869571540427689e-05,
+      "loss": 2.0831,
+      "step": 300
+    },
+    {
+      "epoch": 0.24105011933174225,
+      "grad_norm": 2.565887212753296,
+      "learning_rate": 9.865055749571215e-05,
+      "loss": 2.1372,
+      "step": 303
+    },
+    {
+      "epoch": 0.24343675417661098,
+      "grad_norm": 2.249260902404785,
+      "learning_rate": 9.860464177621284e-05,
+      "loss": 1.8908,
+      "step": 306
+    },
+    {
+      "epoch": 0.2458233890214797,
+      "grad_norm": 2.274188280105591,
+      "learning_rate": 9.855796896099045e-05,
+      "loss": 2.156,
+      "step": 309
+    },
+    {
+      "epoch": 0.24821002386634844,
+      "grad_norm": 2.4920082092285156,
+      "learning_rate": 9.851053977704931e-05,
+      "loss": 1.9275,
+      "step": 312
+    },
+    {
+      "epoch": 0.25059665871121717,
+      "grad_norm": 2.18017578125,
+      "learning_rate": 9.846235496317555e-05,
+      "loss": 1.9534,
+      "step": 315
+    },
+    {
+      "epoch": 0.2529832935560859,
+      "grad_norm": 2.2087161540985107,
+      "learning_rate": 9.841341526992536e-05,
+      "loss": 2.0569,
+      "step": 318
+    },
+    {
+      "epoch": 0.2553699284009546,
+      "grad_norm": 2.050931215286255,
+      "learning_rate": 9.836372145961345e-05,
+      "loss": 2.117,
+      "step": 321
+    },
+    {
+      "epoch": 0.2577565632458234,
+      "grad_norm": 2.36757755279541,
+      "learning_rate": 9.83132743063011e-05,
+      "loss": 2.1293,
+      "step": 324
+    },
+    {
+      "epoch": 0.26014319809069214,
+      "grad_norm": 2.284879207611084,
+      "learning_rate": 9.826207459578411e-05,
+      "loss": 1.9431,
+      "step": 327
+    },
+    {
+      "epoch": 0.26252983293556087,
+      "grad_norm": 2.449686050415039,
+      "learning_rate": 9.821012312558058e-05,
+      "loss": 1.8996,
+      "step": 330
+    },
+    {
+      "epoch": 0.2649164677804296,
+      "grad_norm": 1.8972541093826294,
+      "learning_rate": 9.815742070491852e-05,
+      "loss": 2.0876,
+      "step": 333
+    },
+    {
+      "epoch": 0.26730310262529833,
+      "grad_norm": 2.3827104568481445,
+      "learning_rate": 9.810396815472314e-05,
+      "loss": 1.9801,
+      "step": 336
+    },
+    {
+      "epoch": 0.26968973747016706,
+      "grad_norm": 2.1279213428497314,
+      "learning_rate": 9.804976630760419e-05,
+      "loss": 2.0433,
+      "step": 339
+    },
+    {
+      "epoch": 0.2720763723150358,
+      "grad_norm": 2.0066232681274414,
+      "learning_rate": 9.799481600784286e-05,
+      "loss": 1.9296,
+      "step": 342
+    },
+    {
+      "epoch": 0.2744630071599045,
+      "grad_norm": 2.5718936920166016,
+      "learning_rate": 9.793911811137875e-05,
+      "loss": 2.0074,
+      "step": 345
+    },
+    {
+      "epoch": 0.27684964200477324,
+      "grad_norm": 2.3425660133361816,
+      "learning_rate": 9.788267348579648e-05,
+      "loss": 2.0321,
+      "step": 348
+    },
+    {
+      "epoch": 0.27923627684964203,
+      "grad_norm": 2.5972917079925537,
+      "learning_rate": 9.782548301031217e-05,
+      "loss": 2.0536,
+      "step": 351
+    },
+    {
+      "epoch": 0.28162291169451076,
+      "grad_norm": 2.15671706199646,
+      "learning_rate": 9.776754757575975e-05,
+      "loss": 1.9679,
+      "step": 354
+    },
+    {
+      "epoch": 0.2840095465393795,
+      "grad_norm": 2.41680645942688,
+      "learning_rate": 9.770886808457709e-05,
+      "loss": 2.029,
+      "step": 357
+    },
+    {
+      "epoch": 0.2863961813842482,
+      "grad_norm": 2.8900387287139893,
+      "learning_rate": 9.764944545079196e-05,
+      "loss": 2.1488,
+      "step": 360
+    },
+    {
+      "epoch": 0.28878281622911695,
+      "grad_norm": 2.2235772609710693,
+      "learning_rate": 9.758928060000778e-05,
+      "loss": 2.0734,
+      "step": 363
+    },
+    {
+      "epoch": 0.2911694510739857,
+      "grad_norm": 2.1928718090057373,
+      "learning_rate": 9.752837446938915e-05,
+      "loss": 1.907,
+      "step": 366
+    },
+    {
+      "epoch": 0.2935560859188544,
+      "grad_norm": 2.063851833343506,
+      "learning_rate": 9.746672800764735e-05,
+      "loss": 1.9107,
+      "step": 369
+    },
+    {
+      "epoch": 0.29594272076372313,
+      "grad_norm": 2.3695101737976074,
+      "learning_rate": 9.740434217502547e-05,
+      "loss": 1.9469,
+      "step": 372
+    },
+    {
+      "epoch": 0.29832935560859186,
+      "grad_norm": 2.556959867477417,
+      "learning_rate": 9.734121794328357e-05,
+      "loss": 2.0305,
+      "step": 375
+    },
+    {
+      "epoch": 0.30071599045346065,
+      "grad_norm": 2.1321029663085938,
+      "learning_rate": 9.727735629568336e-05,
+      "loss": 2.0459,
+      "step": 378
+    },
+    {
+      "epoch": 0.3031026252983294,
+      "grad_norm": 1.8276643753051758,
+      "learning_rate": 9.721275822697306e-05,
+      "loss": 1.9669,
+      "step": 381
+    },
+    {
+      "epoch": 0.3054892601431981,
+      "grad_norm": 2.0436830520629883,
+      "learning_rate": 9.714742474337186e-05,
+      "loss": 2.0737,
+      "step": 384
+    },
+    {
+      "epoch": 0.30787589498806683,
+      "grad_norm": 2.2903330326080322,
+      "learning_rate": 9.708135686255416e-05,
+      "loss": 2.1781,
+      "step": 387
+    },
+    {
+      "epoch": 0.31026252983293556,
+      "grad_norm": 2.438894510269165,
+      "learning_rate": 9.701455561363379e-05,
+      "loss": 1.9024,
+      "step": 390
+    },
+    {
+      "epoch": 0.3126491646778043,
+      "grad_norm": 2.9679908752441406,
+      "learning_rate": 9.6947022037148e-05,
+      "loss": 1.9496,
+      "step": 393
+    },
+    {
+      "epoch": 0.315035799522673,
+      "grad_norm": 2.693708896636963,
+      "learning_rate": 9.687875718504126e-05,
+      "loss": 1.9314,
+      "step": 396
+    },
+    {
+      "epoch": 0.31742243436754175,
+      "grad_norm": 3.0519704818725586,
+      "learning_rate": 9.680976212064874e-05,
+      "loss": 2.0026,
+      "step": 399
+    },
+    {
+      "epoch": 0.3198090692124105,
+      "grad_norm": 1.8532629013061523,
+      "learning_rate": 9.674003791867991e-05,
+      "loss": 2.0183,
+      "step": 402
+    },
+    {
+      "epoch": 0.3221957040572792,
+      "grad_norm": 2.272416353225708,
+      "learning_rate": 9.666958566520174e-05,
+      "loss": 2.0488,
+      "step": 405
+    },
+    {
+      "epoch": 0.324582338902148,
+      "grad_norm": 2.514479637145996,
+      "learning_rate": 9.659840645762175e-05,
+      "loss": 2.0881,
+      "step": 408
+    },
+    {
+      "epoch": 0.3269689737470167,
+      "grad_norm": 2.127908945083618,
+      "learning_rate": 9.652650140467093e-05,
+      "loss": 2.0058,
+      "step": 411
+    },
+    {
+      "epoch": 0.32935560859188545,
+      "grad_norm": 2.1128835678100586,
+      "learning_rate": 9.645387162638652e-05,
+      "loss": 2.0027,
+      "step": 414
+    },
+    {
+      "epoch": 0.3317422434367542,
+      "grad_norm": 2.1104748249053955,
+      "learning_rate": 9.638051825409453e-05,
+      "loss": 2.1834,
+      "step": 417
+    },
+    {
+      "epoch": 0.3341288782816229,
+      "grad_norm": 1.9978591203689575,
+      "learning_rate": 9.630644243039207e-05,
+      "loss": 1.9216,
+      "step": 420
+    },
+    {
+      "epoch": 0.33651551312649164,
+      "grad_norm": 2.1656672954559326,
+      "learning_rate": 9.623164530912963e-05,
+      "loss": 2.1389,
+      "step": 423
+    },
+    {
+      "epoch": 0.33890214797136037,
+      "grad_norm": 1.7347074747085571,
+      "learning_rate": 9.615612805539305e-05,
+      "loss": 1.8929,
+      "step": 426
+    },
+    {
+      "epoch": 0.3412887828162291,
+      "grad_norm": 2.3430702686309814,
+      "learning_rate": 9.607989184548543e-05,
+      "loss": 2.0085,
+      "step": 429
+    },
+    {
+      "epoch": 0.3436754176610978,
+      "grad_norm": 1.9454632997512817,
+      "learning_rate": 9.600293786690872e-05,
+      "loss": 2.0007,
+      "step": 432
+    },
+    {
+      "epoch": 0.3460620525059666,
+      "grad_norm": 1.9286153316497803,
+      "learning_rate": 9.592526731834537e-05,
+      "loss": 2.1495,
+      "step": 435
+    },
+    {
+      "epoch": 0.34844868735083534,
+      "grad_norm": 2.3506131172180176,
+      "learning_rate": 9.584688140963944e-05,
+      "loss": 1.8264,
+      "step": 438
+    },
+    {
+      "epoch": 0.35083532219570407,
+      "grad_norm": 2.0399367809295654,
+      "learning_rate": 9.576778136177798e-05,
+      "loss": 2.0056,
+      "step": 441
+    },
+    {
+      "epoch": 0.3532219570405728,
+      "grad_norm": 2.3488686084747314,
+      "learning_rate": 9.568796840687184e-05,
+      "loss": 1.8853,
+      "step": 444
+    },
+    {
+      "epoch": 0.3556085918854415,
+      "grad_norm": 2.3103203773498535,
+      "learning_rate": 9.560744378813659e-05,
+      "loss": 1.9849,
+      "step": 447
+    },
+    {
+      "epoch": 0.35799522673031026,
+      "grad_norm": 2.4852659702301025,
+      "learning_rate": 9.552620875987311e-05,
+      "loss": 1.9174,
+      "step": 450
+    },
+    {
+      "epoch": 0.360381861575179,
+      "grad_norm": 2.327434778213501,
+      "learning_rate": 9.544426458744804e-05,
+      "loss": 1.8321,
+      "step": 453
+    },
+    {
+      "epoch": 0.3627684964200477,
+      "grad_norm": 2.1217284202575684,
+      "learning_rate": 9.536161254727408e-05,
+      "loss": 1.8995,
+      "step": 456
+    },
+    {
+      "epoch": 0.36515513126491644,
+      "grad_norm": 2.4897334575653076,
+      "learning_rate": 9.527825392679012e-05,
+      "loss": 1.8949,
+      "step": 459
+    },
+    {
+      "epoch": 0.36754176610978523,
+      "grad_norm": 2.132058620452881,
+      "learning_rate": 9.51941900244412e-05,
+      "loss": 1.8962,
+      "step": 462
+    },
+    {
+      "epoch": 0.36992840095465396,
+      "grad_norm": 1.8956185579299927,
+      "learning_rate": 9.51094221496582e-05,
+      "loss": 1.9098,
+      "step": 465
+    },
+    {
+      "epoch": 0.3723150357995227,
+      "grad_norm": 2.0302867889404297,
+      "learning_rate": 9.502395162283759e-05,
+      "loss": 1.7664,
+      "step": 468
+    },
+    {
+      "epoch": 0.3747016706443914,
+      "grad_norm": 2.364084243774414,
+      "learning_rate": 9.493777977532072e-05,
+      "loss": 1.932,
+      "step": 471
+    },
+    {
+      "epoch": 0.37708830548926014,
+      "grad_norm": 2.0411782264709473,
+      "learning_rate": 9.485090794937319e-05,
+      "loss": 1.887,
+      "step": 474
+    },
+    {
+      "epoch": 0.3794749403341289,
+      "grad_norm": 2.052417755126953,
+      "learning_rate": 9.476333749816382e-05,
+      "loss": 1.8125,
+      "step": 477
+    },
+    {
+      "epoch": 0.3818615751789976,
+      "grad_norm": 2.2700257301330566,
+      "learning_rate": 9.467506978574371e-05,
+      "loss": 1.8167,
+      "step": 480
+    },
+    {
+      "epoch": 0.38424821002386633,
+      "grad_norm": 1.9266332387924194,
+      "learning_rate": 9.45861061870249e-05,
+      "loss": 1.9767,
+      "step": 483
+    },
+    {
+      "epoch": 0.38663484486873506,
+      "grad_norm": 2.934418201446533,
+      "learning_rate": 9.449644808775902e-05,
+      "loss": 1.8826,
+      "step": 486
+    },
+    {
+      "epoch": 0.38902147971360385,
+      "grad_norm": 1.8543293476104736,
+      "learning_rate": 9.44060968845156e-05,
+      "loss": 1.874,
+      "step": 489
+    },
+    {
+      "epoch": 0.3914081145584726,
+      "grad_norm": 2.2420761585235596,
+      "learning_rate": 9.431505398466045e-05,
+      "loss": 1.9835,
+      "step": 492
+    },
+    {
+      "epoch": 0.3937947494033413,
+      "grad_norm": 2.448664665222168,
+      "learning_rate": 9.42233208063336e-05,
+      "loss": 1.7932,
+      "step": 495
+    },
+    {
+      "epoch": 0.39618138424821003,
+      "grad_norm": 2.271794557571411,
+      "learning_rate": 9.413089877842736e-05,
+      "loss": 2.0198,
+      "step": 498
+    },
+    {
+      "epoch": 0.39856801909307876,
+      "grad_norm": 2.7900123596191406,
+      "learning_rate": 9.403778934056391e-05,
+      "loss": 1.9902,
+      "step": 501
+    },
+    {
+      "epoch": 0.4009546539379475,
+      "grad_norm": 1.9538331031799316,
+      "learning_rate": 9.394399394307303e-05,
+      "loss": 2.1558,
+      "step": 504
+    },
+    {
+      "epoch": 0.4033412887828162,
+      "grad_norm": 4.123261451721191,
+      "learning_rate": 9.384951404696933e-05,
+      "loss": 1.8325,
+      "step": 507
+    },
+    {
+      "epoch": 0.40572792362768495,
+      "grad_norm": 2.156297206878662,
+      "learning_rate": 9.375435112392969e-05,
+      "loss": 2.0274,
+      "step": 510
+    },
+    {
+      "epoch": 0.4081145584725537,
+      "grad_norm": 2.066129207611084,
+      "learning_rate": 9.365850665627016e-05,
+      "loss": 1.8711,
+      "step": 513
+    },
+    {
+      "epoch": 0.4105011933174224,
+      "grad_norm": 1.8385334014892578,
+      "learning_rate": 9.356198213692297e-05,
+      "loss": 1.8337,
+      "step": 516
+    },
+    {
+      "epoch": 0.4128878281622912,
+      "grad_norm": 2.1893041133880615,
+      "learning_rate": 9.346477906941331e-05,
+      "loss": 1.784,
+      "step": 519
+    },
+    {
+      "epoch": 0.4152744630071599,
+      "grad_norm": 2.0097122192382812,
+      "learning_rate": 9.336689896783573e-05,
+      "loss": 1.8381,
+      "step": 522
+    },
+    {
+      "epoch": 0.41766109785202865,
+      "grad_norm": 2.2761220932006836,
+      "learning_rate": 9.32683433568308e-05,
+      "loss": 1.876,
+      "step": 525
+    },
+    {
+      "epoch": 0.4200477326968974,
+      "grad_norm": 2.7852184772491455,
+      "learning_rate": 9.316911377156117e-05,
+      "loss": 1.9546,
+      "step": 528
+    },
+    {
+      "epoch": 0.4224343675417661,
+      "grad_norm": 1.915184497833252,
+      "learning_rate": 9.306921175768775e-05,
+      "loss": 1.8832,
+      "step": 531
+    },
+    {
+      "epoch": 0.42482100238663484,
+      "grad_norm": 2.2405636310577393,
+      "learning_rate": 9.29686388713456e-05,
+      "loss": 2.097,
+      "step": 534
+    },
+    {
+      "epoch": 0.42720763723150357,
+      "grad_norm": 1.9388247728347778,
+      "learning_rate": 9.286739667911972e-05,
+      "loss": 1.923,
+      "step": 537
+    },
+    {
+      "epoch": 0.4295942720763723,
+      "grad_norm": 1.7430307865142822,
+      "learning_rate": 9.276548675802059e-05,
+      "loss": 1.962,
+      "step": 540
+    },
+    {
+      "epoch": 0.431980906921241,
+      "grad_norm": 1.7537423372268677,
+      "learning_rate": 9.266291069545972e-05,
+      "loss": 1.8785,
+      "step": 543
+    },
+    {
+      "epoch": 0.4343675417661098,
+      "grad_norm": 1.791382074356079,
+      "learning_rate": 9.255967008922474e-05,
+      "loss": 1.9461,
+      "step": 546
+    },
+    {
+      "epoch": 0.43675417661097854,
+      "grad_norm": 1.7727302312850952,
+      "learning_rate": 9.245576654745471e-05,
+      "loss": 2.0444,
+      "step": 549
+    },
+    {
+      "epoch": 0.43914081145584727,
+      "grad_norm": 2.0288734436035156,
+      "learning_rate": 9.235120168861496e-05,
+      "loss": 1.7547,
+      "step": 552
+    },
+    {
+      "epoch": 0.441527446300716,
+      "grad_norm": 2.7705018520355225,
+      "learning_rate": 9.224597714147186e-05,
+      "loss": 1.9798,
+      "step": 555
+    },
+    {
+      "epoch": 0.4439140811455847,
+      "grad_norm": 1.8357696533203125,
+      "learning_rate": 9.214009454506753e-05,
+      "loss": 1.8111,
+      "step": 558
+    },
+    {
+      "epoch": 0.44630071599045346,
+      "grad_norm": 2.0281825065612793,
+      "learning_rate": 9.203355554869428e-05,
+      "loss": 1.8906,
+      "step": 561
+    },
+    {
+      "epoch": 0.4486873508353222,
+      "grad_norm": 2.804478406906128,
+      "learning_rate": 9.192636181186888e-05,
+      "loss": 1.722,
+      "step": 564
+    },
+    {
+      "epoch": 0.4510739856801909,
+      "grad_norm": 1.8126853704452515,
+      "learning_rate": 9.181851500430673e-05,
+      "loss": 1.7653,
+      "step": 567
+    },
+    {
+      "epoch": 0.45346062052505964,
+      "grad_norm": 1.9144084453582764,
+      "learning_rate": 9.171001680589588e-05,
+      "loss": 1.8009,
+      "step": 570
+    },
+    {
+      "epoch": 0.45584725536992843,
+      "grad_norm": 2.006549835205078,
+      "learning_rate": 9.160086890667086e-05,
+      "loss": 1.8403,
+      "step": 573
+    },
+    {
+      "epoch": 0.45823389021479716,
+      "grad_norm": 2.879286766052246,
+      "learning_rate": 9.14910730067863e-05,
+      "loss": 1.9097,
+      "step": 576
+    },
+    {
+      "epoch": 0.4606205250596659,
+      "grad_norm": 1.752193808555603,
+      "learning_rate": 9.138063081649051e-05,
+      "loss": 2.0254,
+      "step": 579
+    },
+    {
+      "epoch": 0.4630071599045346,
+      "grad_norm": 1.8021249771118164,
+      "learning_rate": 9.126954405609882e-05,
+      "loss": 1.8794,
+      "step": 582
+    },
+    {
+      "epoch": 0.46539379474940334,
+      "grad_norm": 1.8375767469406128,
+      "learning_rate": 9.115781445596676e-05,
+      "loss": 1.7489,
+      "step": 585
+    },
+    {
+      "epoch": 0.4677804295942721,
+      "grad_norm": 2.183716297149658,
+      "learning_rate": 9.104544375646313e-05,
+      "loss": 2.1149,
+      "step": 588
+    },
+    {
+      "epoch": 0.4701670644391408,
+      "grad_norm": 2.1002678871154785,
+      "learning_rate": 9.093243370794291e-05,
+      "loss": 1.8637,
+      "step": 591
+    },
+    {
+      "epoch": 0.47255369928400953,
+      "grad_norm": 2.2214958667755127,
+      "learning_rate": 9.081878607071996e-05,
+      "loss": 1.9605,
+      "step": 594
+    },
+    {
+      "epoch": 0.47494033412887826,
+      "grad_norm": 1.9720337390899658,
+      "learning_rate": 9.07045026150396e-05,
+      "loss": 1.8876,
+      "step": 597
+    },
+    {
+      "epoch": 0.477326968973747,
+      "grad_norm": 1.9128150939941406,
+      "learning_rate": 9.058958512105104e-05,
+      "loss": 1.8217,
+      "step": 600
+    },
+    {
+      "epoch": 0.4797136038186158,
+      "grad_norm": 2.2358174324035645,
+      "learning_rate": 9.047403537877971e-05,
+      "loss": 1.8511,
+      "step": 603
+    },
+    {
+      "epoch": 0.4821002386634845,
+      "grad_norm": 1.7367171049118042,
+      "learning_rate": 9.035785518809927e-05,
+      "loss": 1.9224,
+      "step": 606
+    },
+    {
+      "epoch": 0.48448687350835323,
+      "grad_norm": 2.0542192459106445,
+      "learning_rate": 9.024104635870368e-05,
+      "loss": 1.8781,
+      "step": 609
+    },
+    {
+      "epoch": 0.48687350835322196,
+      "grad_norm": 5.130634784698486,
+      "learning_rate": 9.012361071007891e-05,
+      "loss": 1.8786,
+      "step": 612
+    },
+    {
+      "epoch": 0.4892601431980907,
+      "grad_norm": 2.5925798416137695,
+      "learning_rate": 9.000555007147469e-05,
+      "loss": 1.893,
+      "step": 615
+    },
+    {
+      "epoch": 0.4916467780429594,
+      "grad_norm": 2.019935369491577,
+      "learning_rate": 8.988686628187597e-05,
+      "loss": 1.8782,
+      "step": 618
+    },
+    {
+      "epoch": 0.49403341288782815,
+      "grad_norm": 1.9143917560577393,
+      "learning_rate": 8.976756118997427e-05,
+      "loss": 1.9076,
+      "step": 621
+    },
+    {
+      "epoch": 0.4964200477326969,
+      "grad_norm": 2.5640761852264404,
+      "learning_rate": 8.964763665413893e-05,
+      "loss": 1.7351,
+      "step": 624
+    },
+    {
+      "epoch": 0.4988066825775656,
+      "grad_norm": 2.209772825241089,
+      "learning_rate": 8.952709454238808e-05,
+      "loss": 1.9958,
+      "step": 627
+    },
+    {
+      "epoch": 0.5011933174224343,
+      "grad_norm": 2.1098861694335938,
+      "learning_rate": 8.940593673235962e-05,
+      "loss": 1.9609,
+      "step": 630
+    },
+    {
+      "epoch": 0.5035799522673031,
+      "grad_norm": 2.109833002090454,
+      "learning_rate": 8.928416511128195e-05,
+      "loss": 1.7695,
+      "step": 633
+    },
+    {
+      "epoch": 0.5059665871121718,
+      "grad_norm": 1.9139786958694458,
+      "learning_rate": 8.916178157594453e-05,
+      "loss": 1.8483,
+      "step": 636
+    },
+    {
+      "epoch": 0.5083532219570406,
+      "grad_norm": 1.9831109046936035,
+      "learning_rate": 8.903878803266841e-05,
+      "loss": 2.0499,
+      "step": 639
+    },
+    {
+      "epoch": 0.5107398568019093,
+      "grad_norm": 1.6918339729309082,
+      "learning_rate": 8.891518639727649e-05,
+      "loss": 1.7375,
+      "step": 642
+    },
+    {
+      "epoch": 0.513126491646778,
+      "grad_norm": 2.150010585784912,
+      "learning_rate": 8.879097859506372e-05,
+      "loss": 1.9052,
+      "step": 645
+    },
+    {
+      "epoch": 0.5155131264916468,
+      "grad_norm": 1.9660903215408325,
+      "learning_rate": 8.866616656076696e-05,
+      "loss": 1.787,
+      "step": 648
+    },
+    {
+      "epoch": 0.5178997613365155,
+      "grad_norm": 2.0688490867614746,
+      "learning_rate": 8.854075223853508e-05,
+      "loss": 1.9195,
+      "step": 651
+    },
+    {
+      "epoch": 0.5202863961813843,
+      "grad_norm": 1.7813605070114136,
+      "learning_rate": 8.841473758189854e-05,
+      "loss": 1.779,
+      "step": 654
+    },
+    {
+      "epoch": 0.522673031026253,
+      "grad_norm": 2.305001735687256,
+      "learning_rate": 8.828812455373891e-05,
+      "loss": 1.9889,
+      "step": 657
+    },
+    {
+      "epoch": 0.5250596658711217,
+      "grad_norm": 1.8771156072616577,
+      "learning_rate": 8.816091512625843e-05,
+      "loss": 1.8912,
+      "step": 660
+    },
+    {
+      "epoch": 0.5274463007159904,
+      "grad_norm": 2.458179473876953,
+      "learning_rate": 8.803311128094918e-05,
+      "loss": 1.9203,
+      "step": 663
+    },
+    {
+      "epoch": 0.5298329355608592,
+      "grad_norm": 2.151304244995117,
+      "learning_rate": 8.790471500856228e-05,
+      "loss": 2.0082,
+      "step": 666
+    },
+    {
+      "epoch": 0.5322195704057279,
+      "grad_norm": 1.7993594408035278,
+      "learning_rate": 8.777572830907684e-05,
+      "loss": 2.0773,
+      "step": 669
+    },
+    {
+      "epoch": 0.5346062052505967,
+      "grad_norm": 1.6973276138305664,
+      "learning_rate": 8.764615319166886e-05,
+      "loss": 2.0038,
+      "step": 672
+    },
+    {
+      "epoch": 0.5369928400954654,
+      "grad_norm": 1.8245609998703003,
+      "learning_rate": 8.751599167467985e-05,
+      "loss": 1.7664,
+      "step": 675
+    },
+    {
+      "epoch": 0.5393794749403341,
+      "grad_norm": 1.7999241352081299,
+      "learning_rate": 8.738524578558547e-05,
+      "loss": 1.8544,
+      "step": 678
+    },
+    {
+      "epoch": 0.5417661097852029,
+      "grad_norm": 1.9764189720153809,
+      "learning_rate": 8.72539175609639e-05,
+      "loss": 1.7731,
+      "step": 681
+    },
+    {
+      "epoch": 0.5441527446300716,
+      "grad_norm": 1.9319300651550293,
+      "learning_rate": 8.712200904646416e-05,
+      "loss": 1.7558,
+      "step": 684
+    },
+    {
+      "epoch": 0.5465393794749404,
+      "grad_norm": 1.6623660326004028,
+      "learning_rate": 8.698952229677422e-05,
+      "loss": 1.7731,
+      "step": 687
+    },
+    {
+      "epoch": 0.548926014319809,
+      "grad_norm": 2.21167254447937,
+      "learning_rate": 8.685645937558896e-05,
+      "loss": 2.0473,
+      "step": 690
+    },
+    {
+      "epoch": 0.5513126491646778,
+      "grad_norm": 7.886738300323486,
+      "learning_rate": 8.67228223555781e-05,
+      "loss": 1.7793,
+      "step": 693
+    },
+    {
+      "epoch": 0.5536992840095465,
+      "grad_norm": 1.9568718671798706,
+      "learning_rate": 8.658861331835385e-05,
+      "loss": 1.8811,
+      "step": 696
+    },
+    {
+      "epoch": 0.5560859188544153,
+      "grad_norm": 1.9484925270080566,
+      "learning_rate": 8.645383435443852e-05,
+      "loss": 1.7875,
+      "step": 699
+    },
+    {
+      "epoch": 0.5584725536992841,
+      "grad_norm": 1.946845293045044,
+      "learning_rate": 8.631848756323197e-05,
+      "loss": 1.8602,
+      "step": 702
+    },
+    {
+      "epoch": 0.5608591885441527,
+      "grad_norm": 1.8822824954986572,
+      "learning_rate": 8.618257505297886e-05,
+      "loss": 1.873,
+      "step": 705
+    },
+    {
+      "epoch": 0.5632458233890215,
+      "grad_norm": 1.9031224250793457,
+      "learning_rate": 8.604609894073584e-05,
+      "loss": 1.6809,
+      "step": 708
+    },
+    {
+      "epoch": 0.5656324582338902,
+      "grad_norm": 2.2142996788024902,
+      "learning_rate": 8.590906135233854e-05,
+      "loss": 1.8147,
+      "step": 711
+    },
+    {
+      "epoch": 0.568019093078759,
+      "grad_norm": 1.9311624765396118,
+      "learning_rate": 8.577146442236857e-05,
+      "loss": 1.8671,
+      "step": 714
+    },
+    {
+      "epoch": 0.5704057279236276,
+      "grad_norm": 1.874782919883728,
+      "learning_rate": 8.563331029412012e-05,
+      "loss": 1.9027,
+      "step": 717
+    },
+    {
+      "epoch": 0.5727923627684964,
+      "grad_norm": 2.6065828800201416,
+      "learning_rate": 8.549460111956664e-05,
+      "loss": 1.8351,
+      "step": 720
+    },
+    {
+      "epoch": 0.5751789976133651,
+      "grad_norm": 2.043619394302368,
+      "learning_rate": 8.535533905932738e-05,
+      "loss": 1.8218,
+      "step": 723
+    },
+    {
+      "epoch": 0.5775656324582339,
+      "grad_norm": 2.091768264770508,
+      "learning_rate": 8.521552628263362e-05,
+      "loss": 1.8289,
+      "step": 726
+    },
+    {
+      "epoch": 0.5799522673031027,
+      "grad_norm": 2.1494524478912354,
+      "learning_rate": 8.507516496729495e-05,
+      "loss": 1.7669,
+      "step": 729
+    },
+    {
+      "epoch": 0.5823389021479713,
+      "grad_norm": 1.9479185342788696,
+      "learning_rate": 8.493425729966534e-05,
+      "loss": 1.8909,
+      "step": 732
+    },
+    {
+      "epoch": 0.5847255369928401,
+      "grad_norm": 1.799233317375183,
+      "learning_rate": 8.479280547460907e-05,
+      "loss": 1.7795,
+      "step": 735
+    },
+    {
+      "epoch": 0.5871121718377088,
+      "grad_norm": 1.768520474433899,
+      "learning_rate": 8.465081169546659e-05,
+      "loss": 1.9436,
+      "step": 738
+    },
+    {
+      "epoch": 0.5894988066825776,
+      "grad_norm": 1.9335992336273193,
+      "learning_rate": 8.450827817402011e-05,
+      "loss": 1.9402,
+      "step": 741
+    },
+    {
+      "epoch": 0.5918854415274463,
+      "grad_norm": 2.3245134353637695,
+      "learning_rate": 8.436520713045922e-05,
+      "loss": 1.7351,
+      "step": 744
+    },
+    {
+      "epoch": 0.594272076372315,
+      "grad_norm": 1.6517058610916138,
+      "learning_rate": 8.422160079334628e-05,
+      "loss": 1.8117,
+      "step": 747
+    },
+    {
+      "epoch": 0.5966587112171837,
+      "grad_norm": 2.4537134170532227,
+      "learning_rate": 8.40774613995817e-05,
+      "loss": 1.8694,
+      "step": 750
+    },
+    {
+      "epoch": 0.5990453460620525,
+      "grad_norm": 1.9698394536972046,
+      "learning_rate": 8.393279119436912e-05,
+      "loss": 1.8867,
+      "step": 753
+    },
+    {
+      "epoch": 0.6014319809069213,
+      "grad_norm": 1.7775943279266357,
+      "learning_rate": 8.378759243118044e-05,
+      "loss": 2.0282,
+      "step": 756
+    },
+    {
+      "epoch": 0.60381861575179,
+      "grad_norm": 2.286463737487793,
+      "learning_rate": 8.364186737172068e-05,
+      "loss": 1.7754,
+      "step": 759
+    },
+    {
+      "epoch": 0.6062052505966588,
+      "grad_norm": 1.9389877319335938,
+      "learning_rate": 8.349561828589277e-05,
+      "loss": 1.9778,
+      "step": 762
+    },
+    {
+      "epoch": 0.6085918854415274,
+      "grad_norm": 1.778572916984558,
+      "learning_rate": 8.33488474517622e-05,
+      "loss": 1.8265,
+      "step": 765
+    },
+    {
+      "epoch": 0.6109785202863962,
+      "grad_norm": 1.8738365173339844,
+      "learning_rate": 8.320155715552155e-05,
+      "loss": 1.7726,
+      "step": 768
+    },
+    {
+      "epoch": 0.6133651551312649,
+      "grad_norm": 2.0440497398376465,
+      "learning_rate": 8.305374969145488e-05,
+      "loss": 1.9478,
+      "step": 771
+    },
+    {
+      "epoch": 0.6157517899761337,
+      "grad_norm": 1.7448711395263672,
+      "learning_rate": 8.290542736190188e-05,
+      "loss": 1.7166,
+      "step": 774
+    },
+    {
+      "epoch": 0.6181384248210023,
+      "grad_norm": 1.9907166957855225,
+      "learning_rate": 8.275659247722222e-05,
+      "loss": 1.7377,
+      "step": 777
+    },
+    {
+      "epoch": 0.6205250596658711,
+      "grad_norm": 1.9221429824829102,
+      "learning_rate": 8.260724735575933e-05,
+      "loss": 1.8257,
+      "step": 780
+    },
+    {
+      "epoch": 0.6229116945107399,
+      "grad_norm": 2.3536133766174316,
+      "learning_rate": 8.24573943238045e-05,
+      "loss": 1.913,
+      "step": 783
+    },
+    {
+      "epoch": 0.6252983293556086,
+      "grad_norm": 1.7165111303329468,
+      "learning_rate": 8.230703571556048e-05,
+      "loss": 1.7241,
+      "step": 786
+    },
+    {
+      "epoch": 0.6276849642004774,
+      "grad_norm": 1.8131250143051147,
+      "learning_rate": 8.215617387310524e-05,
+      "loss": 1.7696,
+      "step": 789
+    },
+    {
+      "epoch": 0.630071599045346,
+      "grad_norm": 1.9642695188522339,
+      "learning_rate": 8.200481114635536e-05,
+      "loss": 1.8094,
+      "step": 792
+    },
+    {
+      "epoch": 0.6324582338902148,
+      "grad_norm": 1.9827426671981812,
+      "learning_rate": 8.185294989302958e-05,
+      "loss": 1.7453,
+      "step": 795
+    },
+    {
+      "epoch": 0.6348448687350835,
+      "grad_norm": 1.7909404039382935,
+      "learning_rate": 8.170059247861194e-05,
+      "loss": 1.7516,
+      "step": 798
+    },
+    {
+      "epoch": 0.6372315035799523,
+      "grad_norm": 1.903298020362854,
+      "learning_rate": 8.154774127631501e-05,
+      "loss": 1.7161,
+      "step": 801
+    },
+    {
+      "epoch": 0.639618138424821,
+      "grad_norm": 1.9327062368392944,
+      "learning_rate": 8.139439866704293e-05,
+      "loss": 1.7925,
+      "step": 804
+    },
+    {
+      "epoch": 0.6420047732696897,
+      "grad_norm": 1.9829552173614502,
+      "learning_rate": 8.124056703935423e-05,
+      "loss": 1.7784,
+      "step": 807
+    },
+    {
+      "epoch": 0.6443914081145584,
+      "grad_norm": 1.7598705291748047,
+      "learning_rate": 8.108624878942477e-05,
+      "loss": 1.786,
+      "step": 810
+    },
+    {
+      "epoch": 0.6467780429594272,
+      "grad_norm": 1.963447093963623,
+      "learning_rate": 8.093144632101026e-05,
+      "loss": 1.7174,
+      "step": 813
+    },
+    {
+      "epoch": 0.649164677804296,
+      "grad_norm": 1.7478758096694946,
+      "learning_rate": 8.077616204540897e-05,
+      "loss": 1.7849,
+      "step": 816
+    },
+    {
+      "epoch": 0.6515513126491647,
+      "grad_norm": 1.8316524028778076,
+      "learning_rate": 8.062039838142402e-05,
+      "loss": 1.7664,
+      "step": 819
+    },
+    {
+      "epoch": 0.6539379474940334,
+      "grad_norm": 1.9206571578979492,
+      "learning_rate": 8.046415775532585e-05,
+      "loss": 1.7882,
+      "step": 822
+    },
+    {
+      "epoch": 0.6563245823389021,
+      "grad_norm": 2.1145057678222656,
+      "learning_rate": 8.030744260081426e-05,
+      "loss": 1.8059,
+      "step": 825
+    },
+    {
+      "epoch": 0.6587112171837709,
+      "grad_norm": 1.711238145828247,
+      "learning_rate": 8.015025535898073e-05,
+      "loss": 1.8649,
+      "step": 828
+    },
+    {
+      "epoch": 0.6610978520286396,
+      "grad_norm": 1.5813544988632202,
+      "learning_rate": 7.999259847827015e-05,
+      "loss": 1.8448,
+      "step": 831
+    },
+    {
+      "epoch": 0.6634844868735084,
+      "grad_norm": 1.8383519649505615,
+      "learning_rate": 7.983447441444281e-05,
+      "loss": 1.6869,
+      "step": 834
+    },
+    {
+      "epoch": 0.665871121718377,
+      "grad_norm": 1.7440314292907715,
+      "learning_rate": 7.967588563053616e-05,
+      "loss": 1.735,
+      "step": 837
+    },
+    {
+      "epoch": 0.6682577565632458,
+      "grad_norm": 1.9489431381225586,
+      "learning_rate": 7.951683459682641e-05,
+      "loss": 1.7624,
+      "step": 840
+    },
+    {
+      "epoch": 0.6706443914081146,
+      "grad_norm": 1.744482159614563,
+      "learning_rate": 7.935732379079008e-05,
+      "loss": 1.9888,
+      "step": 843
+    },
+    {
+      "epoch": 0.6730310262529833,
+      "grad_norm": 2.0963144302368164,
+      "learning_rate": 7.919735569706533e-05,
+      "loss": 1.8716,
+      "step": 846
+    },
+    {
+      "epoch": 0.6754176610978521,
+      "grad_norm": 1.85282301902771,
+      "learning_rate": 7.903693280741331e-05,
+      "loss": 1.8274,
+      "step": 849
+    },
+    {
+      "epoch": 0.6778042959427207,
+      "grad_norm": 1.7610821723937988,
+      "learning_rate": 7.887605762067945e-05,
+      "loss": 1.7237,
+      "step": 852
+    },
+    {
+      "epoch": 0.6801909307875895,
+      "grad_norm": 1.7411203384399414,
+      "learning_rate": 7.871473264275429e-05,
+      "loss": 1.723,
+      "step": 855
+    },
+    {
+      "epoch": 0.6825775656324582,
+      "grad_norm": 1.8666094541549683,
+      "learning_rate": 7.855296038653475e-05,
+      "loss": 2.0054,
+      "step": 858
+    },
+    {
+      "epoch": 0.684964200477327,
+      "grad_norm": 1.56552255153656,
+      "learning_rate": 7.83907433718847e-05,
+      "loss": 1.7596,
+      "step": 861
+    },
+    {
+      "epoch": 0.6873508353221957,
+      "grad_norm": 1.7352533340454102,
+      "learning_rate": 7.82280841255959e-05,
+      "loss": 1.7279,
+      "step": 864
+    },
+    {
+      "epoch": 0.6897374701670644,
+      "grad_norm": 1.7743481397628784,
+      "learning_rate": 7.80649851813486e-05,
+      "loss": 1.8844,
+      "step": 867
+    },
+    {
+      "epoch": 0.6921241050119332,
+      "grad_norm": 3.154662847518921,
+      "learning_rate": 7.790144907967201e-05,
+      "loss": 1.743,
+      "step": 870
+    },
+    {
+      "epoch": 0.6945107398568019,
+      "grad_norm": 1.835939884185791,
+      "learning_rate": 7.773747836790481e-05,
+      "loss": 1.9497,
+      "step": 873
+    },
+    {
+      "epoch": 0.6968973747016707,
+      "grad_norm": 2.23503041267395,
+      "learning_rate": 7.757307560015538e-05,
+      "loss": 1.7462,
+      "step": 876
+    },
+    {
+      "epoch": 0.6992840095465394,
+      "grad_norm": 1.6978174448013306,
+      "learning_rate": 7.740824333726213e-05,
+      "loss": 1.6822,
+      "step": 879
+    },
+    {
+      "epoch": 0.7016706443914081,
+      "grad_norm": 1.7274194955825806,
+      "learning_rate": 7.724298414675353e-05,
+      "loss": 1.7253,
+      "step": 882
+    },
+    {
+      "epoch": 0.7040572792362768,
+      "grad_norm": 1.969519019126892,
+      "learning_rate": 7.707730060280812e-05,
+      "loss": 1.892,
+      "step": 885
+    },
+    {
+      "epoch": 0.7064439140811456,
+      "grad_norm": 1.7271039485931396,
+      "learning_rate": 7.691119528621444e-05,
+      "loss": 1.7954,
+      "step": 888
+    },
+    {
+      "epoch": 0.7088305489260143,
+      "grad_norm": 1.786372184753418,
+      "learning_rate": 7.674467078433081e-05,
+      "loss": 1.9196,
+      "step": 891
+    },
+    {
+      "epoch": 0.711217183770883,
+      "grad_norm": 1.8364696502685547,
+      "learning_rate": 7.657772969104508e-05,
+      "loss": 1.5962,
+      "step": 894
+    },
+    {
+      "epoch": 0.7136038186157518,
+      "grad_norm": 1.772505760192871,
+      "learning_rate": 7.641037460673412e-05,
+      "loss": 1.576,
+      "step": 897
+    },
+    {
+      "epoch": 0.7159904534606205,
+      "grad_norm": 1.4997633695602417,
+      "learning_rate": 7.624260813822342e-05,
+      "loss": 1.6787,
+      "step": 900
+    },
+    {
+      "epoch": 0.7183770883054893,
+      "grad_norm": 1.8389488458633423,
+      "learning_rate": 7.607443289874642e-05,
+      "loss": 1.832,
+      "step": 903
+    },
+    {
+      "epoch": 0.720763723150358,
+      "grad_norm": 1.6315606832504272,
+      "learning_rate": 7.590585150790389e-05,
+      "loss": 1.8814,
+      "step": 906
+    },
+    {
+      "epoch": 0.7231503579952268,
+      "grad_norm": 1.5416710376739502,
+      "learning_rate": 7.573686659162293e-05,
+      "loss": 1.8169,
+      "step": 909
+    },
+    {
+      "epoch": 0.7255369928400954,
+      "grad_norm": 1.736853003501892,
+      "learning_rate": 7.556748078211635e-05,
+      "loss": 1.9037,
+      "step": 912
+    },
+    {
+      "epoch": 0.7279236276849642,
+      "grad_norm": 1.7387202978134155,
+      "learning_rate": 7.53976967178414e-05,
+      "loss": 1.8332,
+      "step": 915
+    },
+    {
+      "epoch": 0.7303102625298329,
+      "grad_norm": 1.4991620779037476,
+      "learning_rate": 7.522751704345887e-05,
+      "loss": 1.8655,
+      "step": 918
+    },
+    {
+      "epoch": 0.7326968973747017,
+      "grad_norm": 1.6497949361801147,
+      "learning_rate": 7.505694440979178e-05,
+      "loss": 1.7951,
+      "step": 921
+    },
+    {
+      "epoch": 0.7350835322195705,
+      "grad_norm": 1.6602137088775635,
+      "learning_rate": 7.488598147378416e-05,
+      "loss": 1.6238,
+      "step": 924
+    },
+    {
+      "epoch": 0.7374701670644391,
+      "grad_norm": 1.6851720809936523,
+      "learning_rate": 7.471463089845956e-05,
+      "loss": 1.7139,
+      "step": 927
+    },
+    {
+      "epoch": 0.7398568019093079,
+      "grad_norm": 1.911044716835022,
+      "learning_rate": 7.454289535287968e-05,
+      "loss": 1.6739,
+      "step": 930
+    },
+    {
+      "epoch": 0.7422434367541766,
+      "grad_norm": 1.9571938514709473,
+      "learning_rate": 7.437077751210279e-05,
+      "loss": 1.896,
+      "step": 933
+    },
+    {
+      "epoch": 0.7446300715990454,
+      "grad_norm": 5.357728481292725,
+      "learning_rate": 7.419828005714194e-05,
+      "loss": 1.7826,
+      "step": 936
+    },
+    {
+      "epoch": 0.747016706443914,
+      "grad_norm": 1.8948403596878052,
+      "learning_rate": 7.402540567492337e-05,
+      "loss": 1.7544,
+      "step": 939
+    },
+    {
+      "epoch": 0.7494033412887828,
+      "grad_norm": 1.749838948249817,
+      "learning_rate": 7.385215705824449e-05,
+      "loss": 1.9523,
+      "step": 942
+    },
+    {
+      "epoch": 0.7517899761336515,
+      "grad_norm": 2.02341890335083,
+      "learning_rate": 7.367853690573208e-05,
+      "loss": 1.5975,
+      "step": 945
+    },
+    {
+      "epoch": 0.7541766109785203,
+      "grad_norm": 1.999101161956787,
+      "learning_rate": 7.350454792180016e-05,
+      "loss": 1.6717,
+      "step": 948
+    },
+    {
+      "epoch": 0.7565632458233891,
+      "grad_norm": 1.809001088142395,
+      "learning_rate": 7.333019281660789e-05,
+      "loss": 1.9019,
+      "step": 951
+    },
+    {
+      "epoch": 0.7589498806682577,
+      "grad_norm": 1.5597479343414307,
+      "learning_rate": 7.31554743060174e-05,
+      "loss": 1.6719,
+      "step": 954
+    },
+    {
+      "epoch": 0.7613365155131265,
+      "grad_norm": 1.871012806892395,
+      "learning_rate": 7.298039511155138e-05,
+      "loss": 1.7596,
+      "step": 957
+    },
+    {
+      "epoch": 0.7637231503579952,
+      "grad_norm": 1.623950719833374,
+      "learning_rate": 7.280495796035079e-05,
+      "loss": 1.7874,
+      "step": 960
+    },
+    {
+      "epoch": 0.766109785202864,
+      "grad_norm": 1.9036149978637695,
+      "learning_rate": 7.262916558513237e-05,
+      "loss": 1.5974,
+      "step": 963
+    },
+    {
+      "epoch": 0.7684964200477327,
+      "grad_norm": 2.126539468765259,
+      "learning_rate": 7.245302072414601e-05,
+      "loss": 1.747,
+      "step": 966
+    },
+    {
+      "epoch": 0.7708830548926014,
+      "grad_norm": 1.4484821557998657,
+      "learning_rate": 7.227652612113213e-05,
+      "loss": 1.5952,
+      "step": 969
+    },
+    {
+      "epoch": 0.7732696897374701,
+      "grad_norm": 1.6610711812973022,
+      "learning_rate": 7.209968452527896e-05,
+      "loss": 1.7045,
+      "step": 972
+    },
+    {
+      "epoch": 0.7756563245823389,
+      "grad_norm": 1.651370882987976,
+      "learning_rate": 7.192249869117971e-05,
+      "loss": 1.7833,
+      "step": 975
+    },
+    {
+      "epoch": 0.7780429594272077,
+      "grad_norm": 1.8085836172103882,
+      "learning_rate": 7.174497137878966e-05,
+      "loss": 1.6937,
+      "step": 978
+    },
+    {
+      "epoch": 0.7804295942720764,
+      "grad_norm": 1.75294029712677,
+      "learning_rate": 7.156710535338312e-05,
+      "loss": 1.8561,
+      "step": 981
+    },
+    {
+      "epoch": 0.7828162291169452,
+      "grad_norm": 1.712416172027588,
+      "learning_rate": 7.138890338551048e-05,
+      "loss": 1.7812,
+      "step": 984
+    },
+    {
+      "epoch": 0.7852028639618138,
+      "grad_norm": 1.8154973983764648,
+      "learning_rate": 7.121036825095492e-05,
+      "loss": 1.6314,
+      "step": 987
+    },
+    {
+      "epoch": 0.7875894988066826,
+      "grad_norm": 1.8881547451019287,
+      "learning_rate": 7.103150273068921e-05,
+      "loss": 1.6936,
+      "step": 990
+    },
+    {
+      "epoch": 0.7899761336515513,
+      "grad_norm": 1.6982760429382324,
+      "learning_rate": 7.085230961083249e-05,
+      "loss": 1.8004,
+      "step": 993
+    },
+    {
+      "epoch": 0.7923627684964201,
+      "grad_norm": 2.0271944999694824,
+      "learning_rate": 7.067279168260671e-05,
+      "loss": 1.7864,
+      "step": 996
+    },
+    {
+      "epoch": 0.7947494033412887,
+      "grad_norm": 1.6574324369430542,
+      "learning_rate": 7.04929517422933e-05,
+      "loss": 1.7387,
+      "step": 999
+    },
+    {
+      "epoch": 0.7971360381861575,
+      "grad_norm": 1.5255002975463867,
+      "learning_rate": 7.031279259118946e-05,
+      "loss": 1.5828,
+      "step": 1002
+    },
+    {
+      "epoch": 0.7995226730310262,
+      "grad_norm": 1.9319933652877808,
+      "learning_rate": 7.013231703556471e-05,
+      "loss": 1.8482,
+      "step": 1005
+    },
+    {
+      "epoch": 0.801909307875895,
+      "grad_norm": 1.7250359058380127,
+      "learning_rate": 6.995152788661705e-05,
+      "loss": 1.7401,
+      "step": 1008
+    },
+    {
+      "epoch": 0.8042959427207638,
+      "grad_norm": 1.6982582807540894,
+      "learning_rate": 6.977042796042917e-05,
+      "loss": 1.7117,
+      "step": 1011
+    },
+    {
+      "epoch": 0.8066825775656324,
+      "grad_norm": 1.8661295175552368,
+      "learning_rate": 6.958902007792466e-05,
+      "loss": 1.7261,
+      "step": 1014
+    },
+    {
+      "epoch": 0.8090692124105012,
+      "grad_norm": 2.4003262519836426,
+      "learning_rate": 6.940730706482399e-05,
+      "loss": 1.7594,
+      "step": 1017
+    },
+    {
+      "epoch": 0.8114558472553699,
+      "grad_norm": 1.6023939847946167,
+      "learning_rate": 6.922529175160054e-05,
+      "loss": 1.6506,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8138424821002387,
+      "grad_norm": 1.6980756521224976,
+      "learning_rate": 6.904297697343655e-05,
+      "loss": 1.8821,
+      "step": 1023
+    },
+    {
+      "epoch": 0.8162291169451074,
+      "grad_norm": 1.5860844850540161,
+      "learning_rate": 6.886036557017881e-05,
+      "loss": 1.8247,
+      "step": 1026
+    },
+    {
+      "epoch": 0.8186157517899761,
+      "grad_norm": 1.8691662549972534,
+      "learning_rate": 6.867746038629462e-05,
+      "loss": 1.9143,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8210023866348448,
+      "grad_norm": 1.6922295093536377,
+      "learning_rate": 6.849426427082735e-05,
+      "loss": 1.736,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8233890214797136,
+      "grad_norm": 13.19885540008545,
+      "learning_rate": 6.83107800773521e-05,
+      "loss": 1.8041,
+      "step": 1035
+    },
+    {
+      "epoch": 0.8257756563245824,
+      "grad_norm": 1.7809809446334839,
+      "learning_rate": 6.812701066393124e-05,
+      "loss": 1.6853,
+      "step": 1038
+    },
+    {
+      "epoch": 0.8281622911694511,
+      "grad_norm": 1.6743443012237549,
+      "learning_rate": 6.79429588930699e-05,
+      "loss": 1.7091,
+      "step": 1041
+    },
+    {
+      "epoch": 0.8305489260143198,
+      "grad_norm": 1.9790890216827393,
+      "learning_rate": 6.775862763167142e-05,
+      "loss": 1.6965,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8329355608591885,
+      "grad_norm": 1.5133233070373535,
+      "learning_rate": 6.757401975099262e-05,
+      "loss": 1.6387,
+      "step": 1047
+    },
+    {
+      "epoch": 0.8353221957040573,
+      "grad_norm": 3.9208717346191406,
+      "learning_rate": 6.738913812659912e-05,
+      "loss": 1.886,
+      "step": 1050
+    },
+    {
+      "epoch": 0.837708830548926,
+      "grad_norm": 1.884238362312317,
+      "learning_rate": 6.720398563832055e-05,
+      "loss": 1.7375,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8400954653937948,
+      "grad_norm": 1.478226900100708,
+      "learning_rate": 6.701856517020565e-05,
+      "loss": 1.8226,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8424821002386634,
+      "grad_norm": 1.8399808406829834,
+      "learning_rate": 6.683287961047742e-05,
+      "loss": 1.9471,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8448687350835322,
+      "grad_norm": 1.7839453220367432,
+      "learning_rate": 6.664693185148807e-05,
+      "loss": 1.7139,
+      "step": 1062
+    },
+    {
+      "epoch": 0.847255369928401,
+      "grad_norm": 1.7865978479385376,
+      "learning_rate": 6.646072478967397e-05,
+      "loss": 1.8702,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8496420047732697,
+      "grad_norm": 1.572658896446228,
+      "learning_rate": 6.627426132551058e-05,
+      "loss": 1.7129,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8520286396181385,
+      "grad_norm": 1.555977463722229,
+      "learning_rate": 6.608754436346725e-05,
+      "loss": 1.6579,
+      "step": 1071
+    },
+    {
+      "epoch": 0.8544152744630071,
+      "grad_norm": 1.9152740240097046,
+      "learning_rate": 6.590057681196191e-05,
+      "loss": 1.6404,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8568019093078759,
+      "grad_norm": 1.787007451057434,
+      "learning_rate": 6.571336158331589e-05,
+      "loss": 1.9347,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8591885441527446,
+      "grad_norm": 1.757942795753479,
+      "learning_rate": 6.552590159370844e-05,
+      "loss": 1.5946,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8615751789976134,
+      "grad_norm": 1.8404314517974854,
+      "learning_rate": 6.53381997631314e-05,
+      "loss": 1.8198,
+      "step": 1083
+    },
+    {
+      "epoch": 0.863961813842482,
+      "grad_norm": 1.9459943771362305,
+      "learning_rate": 6.515025901534364e-05,
+      "loss": 1.725,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8663484486873508,
+      "grad_norm": 1.6057350635528564,
+      "learning_rate": 6.496208227782556e-05,
+      "loss": 1.7184,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8687350835322196,
+      "grad_norm": 1.80686616897583,
+      "learning_rate": 6.477367248173352e-05,
+      "loss": 1.7365,
+      "step": 1092
+    },
+    {
+      "epoch": 0.8711217183770883,
+      "grad_norm": 1.8622928857803345,
+      "learning_rate": 6.458503256185404e-05,
+      "loss": 1.7449,
+      "step": 1095
+    },
+    {
+      "epoch": 0.8735083532219571,
+      "grad_norm": 1.521681547164917,
+      "learning_rate": 6.439616545655834e-05,
+      "loss": 1.6331,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8758949880668258,
+      "grad_norm": 1.5368080139160156,
+      "learning_rate": 6.420707410775626e-05,
+      "loss": 1.7707,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8782816229116945,
+      "grad_norm": 1.7633475065231323,
+      "learning_rate": 6.401776146085072e-05,
+      "loss": 2.0569,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8806682577565632,
+      "grad_norm": 3.083674430847168,
+      "learning_rate": 6.382823046469167e-05,
+      "loss": 1.7942,
+      "step": 1107
+    },
+    {
+      "epoch": 0.883054892601432,
+      "grad_norm": 1.6440787315368652,
+      "learning_rate": 6.363848407153016e-05,
+      "loss": 1.604,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8854415274463007,
+      "grad_norm": 1.640752911567688,
+      "learning_rate": 6.344852523697247e-05,
+      "loss": 1.8185,
+      "step": 1113
+    },
+    {
+      "epoch": 0.8878281622911695,
+      "grad_norm": 1.959386944770813,
+      "learning_rate": 6.325835691993394e-05,
+      "loss": 1.5247,
+      "step": 1116
+    },
+    {
+      "epoch": 0.8902147971360382,
+      "grad_norm": 1.8059946298599243,
+      "learning_rate": 6.306798208259297e-05,
+      "loss": 1.7049,
+      "step": 1119
+    },
+    {
+      "epoch": 0.8926014319809069,
+      "grad_norm": 1.619410753250122,
+      "learning_rate": 6.287740369034485e-05,
+      "loss": 1.5096,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8949880668257757,
+      "grad_norm": 1.6625523567199707,
+      "learning_rate": 6.26866247117555e-05,
+      "loss": 1.5899,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8973747016706444,
+      "grad_norm": 1.6419795751571655,
+      "learning_rate": 6.249564811851543e-05,
+      "loss": 1.7902,
+      "step": 1128
+    },
+    {
+      "epoch": 0.8997613365155132,
+      "grad_norm": 1.6195034980773926,
+      "learning_rate": 6.230447688539316e-05,
+      "loss": 1.592,
+      "step": 1131
+    },
+    {
+      "epoch": 0.9021479713603818,
+      "grad_norm": 1.6635634899139404,
+      "learning_rate": 6.211311399018916e-05,
+      "loss": 1.7621,
+      "step": 1134
+    },
+    {
+      "epoch": 0.9045346062052506,
+      "grad_norm": 1.5676401853561401,
+      "learning_rate": 6.192156241368929e-05,
+      "loss": 1.7352,
+      "step": 1137
+    },
+    {
+      "epoch": 0.9069212410501193,
+      "grad_norm": 1.879977822303772,
+      "learning_rate": 6.172982513961845e-05,
+      "loss": 1.6849,
+      "step": 1140
+    },
+    {
+      "epoch": 0.9093078758949881,
+      "grad_norm": 2.4494450092315674,
+      "learning_rate": 6.153790515459404e-05,
+      "loss": 1.6082,
+      "step": 1143
+    },
+    {
+      "epoch": 0.9116945107398569,
+      "grad_norm": 1.6648575067520142,
+      "learning_rate": 6.13458054480795e-05,
+      "loss": 1.8233,
+      "step": 1146
+    },
+    {
+      "epoch": 0.9140811455847255,
+      "grad_norm": 1.6422799825668335,
+      "learning_rate": 6.115352901233779e-05,
+      "loss": 1.7669,
+      "step": 1149
+    },
+    {
+      "epoch": 0.9164677804295943,
+      "grad_norm": 1.483912706375122,
+      "learning_rate": 6.096107884238458e-05,
+      "loss": 1.6037,
+      "step": 1152
+    },
+    {
+      "epoch": 0.918854415274463,
+      "grad_norm": 1.664514422416687,
+      "learning_rate": 6.0768457935941817e-05,
+      "loss": 1.7514,
+      "step": 1155
+    },
+    {
+      "epoch": 0.9212410501193318,
+      "grad_norm": 2.0139923095703125,
+      "learning_rate": 6.0575669293390954e-05,
+      "loss": 1.7803,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9236276849642004,
+      "grad_norm": 1.7868553400039673,
+      "learning_rate": 6.038271591772615e-05,
+      "loss": 1.7884,
+      "step": 1161
+    },
+    {
+      "epoch": 0.9260143198090692,
+      "grad_norm": 2.1342079639434814,
+      "learning_rate": 6.0189600814507604e-05,
+      "loss": 1.8653,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9284009546539379,
+      "grad_norm": 1.3909924030303955,
+      "learning_rate": 5.9996326991814654e-05,
+      "loss": 1.5607,
+      "step": 1167
+    },
+    {
+      "epoch": 0.9307875894988067,
+      "grad_norm": 2.2413270473480225,
+      "learning_rate": 5.980289746019892e-05,
+      "loss": 1.8347,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9331742243436754,
+      "grad_norm": 1.802080512046814,
+      "learning_rate": 5.9609315232637483e-05,
+      "loss": 1.5535,
+      "step": 1173
+    },
+    {
+      "epoch": 0.9355608591885441,
+      "grad_norm": 1.6270626783370972,
+      "learning_rate": 5.941558332448589e-05,
+      "loss": 1.5801,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9379474940334129,
+      "grad_norm": 1.567086935043335,
+      "learning_rate": 5.922170475343125e-05,
+      "loss": 1.6446,
+      "step": 1179
+    },
+    {
+      "epoch": 0.9403341288782816,
+      "grad_norm": 1.7809765338897705,
+      "learning_rate": 5.9027682539445104e-05,
+      "loss": 1.6499,
+      "step": 1182
+    },
+    {
+      "epoch": 0.9427207637231504,
+      "grad_norm": 1.5927331447601318,
+      "learning_rate": 5.883351970473654e-05,
+      "loss": 1.8008,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9451073985680191,
+      "grad_norm": 1.7844749689102173,
+      "learning_rate": 5.863921927370498e-05,
+      "loss": 1.7045,
+      "step": 1188
+    },
+    {
+      "epoch": 0.9474940334128878,
+      "grad_norm": 1.6692689657211304,
+      "learning_rate": 5.8444784272893175e-05,
+      "loss": 1.6428,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9498806682577565,
+      "grad_norm": 1.7680476903915405,
+      "learning_rate": 5.8250217730939973e-05,
+      "loss": 1.6511,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9522673031026253,
+      "grad_norm": 1.5252869129180908,
+      "learning_rate": 5.8055522678533225e-05,
+      "loss": 1.6309,
+      "step": 1197
+    },
+    {
+      "epoch": 0.954653937947494,
+      "grad_norm": 1.713062047958374,
+      "learning_rate": 5.786070214836254e-05,
+      "loss": 1.5895,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9570405727923628,
+      "grad_norm": 1.681307315826416,
+      "learning_rate": 5.7665759175072034e-05,
+      "loss": 1.8348,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9594272076372315,
+      "grad_norm": 11.129051208496094,
+      "learning_rate": 5.747069679521305e-05,
+      "loss": 1.7288,
+      "step": 1206
+    },
+    {
+      "epoch": 0.9618138424821002,
+      "grad_norm": 2.107663869857788,
+      "learning_rate": 5.727551804719693e-05,
+      "loss": 1.6863,
+      "step": 1209
+    },
+    {
+      "epoch": 0.964200477326969,
+      "grad_norm": 1.520448088645935,
+      "learning_rate": 5.708022597124758e-05,
+      "loss": 1.6418,
+      "step": 1212
+    },
+    {
+      "epoch": 0.9665871121718377,
+      "grad_norm": 1.7622371912002563,
+      "learning_rate": 5.688482360935423e-05,
+      "loss": 1.8378,
+      "step": 1215
+    },
+    {
+      "epoch": 0.9689737470167065,
+      "grad_norm": 1.4517844915390015,
+      "learning_rate": 5.668931400522396e-05,
+      "loss": 1.7499,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9713603818615751,
+      "grad_norm": 1.7153959274291992,
+      "learning_rate": 5.649370020423431e-05,
+      "loss": 1.6391,
+      "step": 1221
+    },
+    {
+      "epoch": 0.9737470167064439,
+      "grad_norm": 1.6043071746826172,
+      "learning_rate": 5.629798525338589e-05,
+      "loss": 1.6558,
+      "step": 1224
+    },
+    {
+      "epoch": 0.9761336515513126,
+      "grad_norm": 1.5428578853607178,
+      "learning_rate": 5.6102172201254835e-05,
+      "loss": 1.7245,
+      "step": 1227
+    },
+    {
+      "epoch": 0.9785202863961814,
+      "grad_norm": 3.9565799236297607,
+      "learning_rate": 5.5906264097945407e-05,
+      "loss": 1.7375,
+      "step": 1230
+    },
+    {
+      "epoch": 0.9809069212410502,
+      "grad_norm": 1.5014811754226685,
+      "learning_rate": 5.5710263995042434e-05,
+      "loss": 1.8023,
+      "step": 1233
+    },
+    {
+      "epoch": 0.9832935560859188,
+      "grad_norm": 1.6702003479003906,
+      "learning_rate": 5.551417494556376e-05,
+      "loss": 1.7092,
+      "step": 1236
+    },
+    {
+      "epoch": 0.9856801909307876,
+      "grad_norm": 1.5324056148529053,
+      "learning_rate": 5.531800000391275e-05,
+      "loss": 1.7092,
+      "step": 1239
+    },
+    {
+      "epoch": 0.9880668257756563,
+      "grad_norm": 1.6837244033813477,
+      "learning_rate": 5.5121742225830665e-05,
+      "loss": 1.8185,
+      "step": 1242
+    },
+    {
+      "epoch": 0.9904534606205251,
+      "grad_norm": 2.0750982761383057,
+      "learning_rate": 5.4925404668349076e-05,
+      "loss": 1.7451,
+      "step": 1245
+    },
+    {
+      "epoch": 0.9928400954653938,
+      "grad_norm": 1.609738826751709,
+      "learning_rate": 5.472899038974225e-05,
+      "loss": 1.6469,
+      "step": 1248
+    },
+    {
+      "epoch": 0.9952267303102625,
+      "grad_norm": 3.454913377761841,
+      "learning_rate": 5.45325024494795e-05,
+      "loss": 1.7291,
+      "step": 1251
+    },
+    {
+      "epoch": 0.9976133651551312,
+      "grad_norm": 1.7975199222564697,
+      "learning_rate": 5.433594390817756e-05,
+      "loss": 1.8493,
+      "step": 1254
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.325455904006958,
+      "learning_rate": 5.413931782755283e-05,
+      "loss": 1.8083,
+      "step": 1257
+    },
+    {
+      "epoch": 1.0023866348448687,
+      "grad_norm": 1.4781057834625244,
+      "learning_rate": 5.3942627270373826e-05,
+      "loss": 1.4756,
+      "step": 1260
+    },
+    {
+      "epoch": 1.0047732696897376,
+      "grad_norm": 1.787947177886963,
+      "learning_rate": 5.374587530041335e-05,
+      "loss": 1.5096,
+      "step": 1263
+    },
+    {
+      "epoch": 1.0071599045346062,
+      "grad_norm": 1.7651340961456299,
+      "learning_rate": 5.35490649824008e-05,
+      "loss": 1.4784,
+      "step": 1266
+    },
+    {
+      "epoch": 1.009546539379475,
+      "grad_norm": 1.4127767086029053,
+      "learning_rate": 5.335219938197445e-05,
+      "loss": 1.4218,
+      "step": 1269
+    },
+    {
+      "epoch": 1.0119331742243436,
+      "grad_norm": 1.8526569604873657,
+      "learning_rate": 5.315528156563367e-05,
+      "loss": 1.4832,
+      "step": 1272
+    },
+    {
+      "epoch": 1.0143198090692125,
+      "grad_norm": 1.7442090511322021,
+      "learning_rate": 5.295831460069124e-05,
+      "loss": 1.3938,
+      "step": 1275
+    },
+    {
+      "epoch": 1.0167064439140812,
+      "grad_norm": 1.656014323234558,
+      "learning_rate": 5.276130155522541e-05,
+      "loss": 1.5448,
+      "step": 1278
+    },
+    {
+      "epoch": 1.0190930787589498,
+      "grad_norm": 1.4529969692230225,
+      "learning_rate": 5.256424549803228e-05,
+      "loss": 1.5037,
+      "step": 1281
+    },
+    {
+      "epoch": 1.0214797136038185,
+      "grad_norm": 1.4293503761291504,
+      "learning_rate": 5.236714949857791e-05,
+      "loss": 1.3965,
+      "step": 1284
+    },
+    {
+      "epoch": 1.0238663484486874,
+      "grad_norm": 1.8054838180541992,
+      "learning_rate": 5.2170016626950505e-05,
+      "loss": 1.4733,
+      "step": 1287
+    },
+    {
+      "epoch": 1.026252983293556,
+      "grad_norm": 1.5657786130905151,
+      "learning_rate": 5.1972849953812644e-05,
+      "loss": 1.4205,
+      "step": 1290
+    },
+    {
+      "epoch": 1.0286396181384247,
+      "grad_norm": 1.5996116399765015,
+      "learning_rate": 5.1775652550353405e-05,
+      "loss": 1.5207,
+      "step": 1293
+    },
+    {
+      "epoch": 1.0310262529832936,
+      "grad_norm": 1.7841379642486572,
+      "learning_rate": 5.157842748824053e-05,
+      "loss": 1.4094,
+      "step": 1296
+    },
+    {
+      "epoch": 1.0334128878281623,
+      "grad_norm": 1.8023128509521484,
+      "learning_rate": 5.138117783957261e-05,
+      "loss": 1.5327,
+      "step": 1299
+    },
+    {
+      "epoch": 1.035799522673031,
+      "grad_norm": 1.6026707887649536,
+      "learning_rate": 5.1183906676831197e-05,
+      "loss": 1.5903,
+      "step": 1302
+    },
+    {
+      "epoch": 1.0381861575178997,
+      "grad_norm": 1.5639821290969849,
+      "learning_rate": 5.098661707283298e-05,
+      "loss": 1.4613,
+      "step": 1305
+    },
+    {
+      "epoch": 1.0405727923627686,
+      "grad_norm": 1.6041333675384521,
+      "learning_rate": 5.078931210068185e-05,
+      "loss": 1.4202,
+      "step": 1308
+    },
+    {
+      "epoch": 1.0429594272076372,
+      "grad_norm": 1.774864673614502,
+      "learning_rate": 5.059199483372114e-05,
+      "loss": 1.4175,
+      "step": 1311
+    },
+    {
+      "epoch": 1.045346062052506,
+      "grad_norm": 1.5444409847259521,
+      "learning_rate": 5.039466834548568e-05,
+      "loss": 1.5621,
+      "step": 1314
+    },
+    {
+      "epoch": 1.0477326968973748,
+      "grad_norm": 1.456629753112793,
+      "learning_rate": 5.0197335709653883e-05,
+      "loss": 1.2813,
+      "step": 1317
+    },
+    {
+      "epoch": 1.0501193317422435,
+      "grad_norm": 1.5410020351409912,
+      "learning_rate": 5e-05,
+      "loss": 1.5317,
+      "step": 1320
+    },
+    {
+      "epoch": 1.0525059665871122,
+      "grad_norm": 1.7866257429122925,
+      "learning_rate": 4.980266429034613e-05,
+      "loss": 1.4627,
+      "step": 1323
+    },
+    {
+      "epoch": 1.0548926014319808,
+      "grad_norm": 1.5053359270095825,
+      "learning_rate": 4.960533165451435e-05,
+      "loss": 1.3497,
+      "step": 1326
+    },
+    {
+      "epoch": 1.0572792362768497,
+      "grad_norm": 1.4978351593017578,
+      "learning_rate": 4.9408005166278855e-05,
+      "loss": 1.4533,
+      "step": 1329
+    },
+    {
+      "epoch": 1.0596658711217184,
+      "grad_norm": 1.6775259971618652,
+      "learning_rate": 4.921068789931816e-05,
+      "loss": 1.4523,
+      "step": 1332
+    },
+    {
+      "epoch": 1.062052505966587,
+      "grad_norm": 1.4441739320755005,
+      "learning_rate": 4.901338292716704e-05,
+      "loss": 1.302,
+      "step": 1335
+    },
+    {
+      "epoch": 1.064439140811456,
+      "grad_norm": 1.6126224994659424,
+      "learning_rate": 4.8816093323168815e-05,
+      "loss": 1.4091,
+      "step": 1338
+    },
+    {
+      "epoch": 1.0668257756563246,
+      "grad_norm": 1.5988109111785889,
+      "learning_rate": 4.8618822160427406e-05,
+      "loss": 1.4568,
+      "step": 1341
+    },
+    {
+      "epoch": 1.0692124105011933,
+      "grad_norm": 1.4735040664672852,
+      "learning_rate": 4.842157251175947e-05,
+      "loss": 1.4034,
+      "step": 1344
+    },
+    {
+      "epoch": 1.071599045346062,
+      "grad_norm": 1.6535234451293945,
+      "learning_rate": 4.822434744964661e-05,
+      "loss": 1.4543,
+      "step": 1347
+    },
+    {
+      "epoch": 1.0739856801909309,
+      "grad_norm": 1.4830214977264404,
+      "learning_rate": 4.802715004618737e-05,
+      "loss": 1.4614,
+      "step": 1350
+    },
+    {
+      "epoch": 1.0763723150357996,
+      "grad_norm": 1.60805344581604,
+      "learning_rate": 4.7829983373049507e-05,
+      "loss": 1.4318,
+      "step": 1353
+    },
+    {
+      "epoch": 1.0787589498806682,
+      "grad_norm": 1.2208465337753296,
+      "learning_rate": 4.763285050142211e-05,
+      "loss": 1.3005,
+      "step": 1356
+    },
+    {
+      "epoch": 1.081145584725537,
+      "grad_norm": 1.3961493968963623,
+      "learning_rate": 4.743575450196773e-05,
+      "loss": 1.4343,
+      "step": 1359
+    },
+    {
+      "epoch": 1.0835322195704058,
+      "grad_norm": 1.7156232595443726,
+      "learning_rate": 4.7238698444774595e-05,
+      "loss": 1.3852,
+      "step": 1362
+    },
+    {
+      "epoch": 1.0859188544152745,
+      "grad_norm": 1.3676952123641968,
+      "learning_rate": 4.704168539930878e-05,
+      "loss": 1.3402,
+      "step": 1365
+    },
+    {
+      "epoch": 1.0883054892601431,
+      "grad_norm": 1.734222173690796,
+      "learning_rate": 4.6844718434366334e-05,
+      "loss": 1.5666,
+      "step": 1368
+    },
+    {
+      "epoch": 1.0906921241050118,
+      "grad_norm": 1.791410207748413,
+      "learning_rate": 4.664780061802557e-05,
+      "loss": 1.4359,
+      "step": 1371
+    },
+    {
+      "epoch": 1.0930787589498807,
+      "grad_norm": 1.553160309791565,
+      "learning_rate": 4.64509350175992e-05,
+      "loss": 1.4339,
+      "step": 1374
+    },
+    {
+      "epoch": 1.0954653937947494,
+      "grad_norm": 1.6250778436660767,
+      "learning_rate": 4.6254124699586656e-05,
+      "loss": 1.4916,
+      "step": 1377
+    },
+    {
+      "epoch": 1.097852028639618,
+      "grad_norm": 1.6796406507492065,
+      "learning_rate": 4.605737272962618e-05,
+      "loss": 1.4177,
+      "step": 1380
+    },
+    {
+      "epoch": 1.100238663484487,
+      "grad_norm": 1.5058690309524536,
+      "learning_rate": 4.5860682172447184e-05,
+      "loss": 1.455,
+      "step": 1383
+    },
+    {
+      "epoch": 1.1026252983293556,
+      "grad_norm": 3.526346206665039,
+      "learning_rate": 4.566405609182247e-05,
+      "loss": 1.5,
+      "step": 1386
+    },
+    {
+      "epoch": 1.1050119331742243,
+      "grad_norm": 1.591094970703125,
+      "learning_rate": 4.546749755052051e-05,
+      "loss": 1.5146,
+      "step": 1389
+    },
+    {
+      "epoch": 1.107398568019093,
+      "grad_norm": 1.7309263944625854,
+      "learning_rate": 4.527100961025776e-05,
+      "loss": 1.4662,
+      "step": 1392
+    },
+    {
+      "epoch": 1.1097852028639619,
+      "grad_norm": 1.7857303619384766,
+      "learning_rate": 4.507459533165093e-05,
+      "loss": 1.4219,
+      "step": 1395
+    },
+    {
+      "epoch": 1.1121718377088305,
+      "grad_norm": 1.5359363555908203,
+      "learning_rate": 4.4878257774169346e-05,
+      "loss": 1.3648,
+      "step": 1398
+    },
+    {
+      "epoch": 1.1145584725536992,
+      "grad_norm": 1.6269012689590454,
+      "learning_rate": 4.4681999996087274e-05,
+      "loss": 1.4435,
+      "step": 1401
+    },
+    {
+      "epoch": 1.1169451073985681,
+      "grad_norm": 1.8540282249450684,
+      "learning_rate": 4.448582505443625e-05,
+      "loss": 1.515,
+      "step": 1404
+    },
+    {
+      "epoch": 1.1193317422434368,
+      "grad_norm": 1.5260019302368164,
+      "learning_rate": 4.4289736004957585e-05,
+      "loss": 1.4334,
+      "step": 1407
+    },
+    {
+      "epoch": 1.1217183770883055,
+      "grad_norm": 1.5051707029342651,
+      "learning_rate": 4.4093735902054605e-05,
+      "loss": 1.4261,
+      "step": 1410
+    },
+    {
+      "epoch": 1.1241050119331741,
+      "grad_norm": 1.4636119604110718,
+      "learning_rate": 4.3897827798745183e-05,
+      "loss": 1.4085,
+      "step": 1413
+    },
+    {
+      "epoch": 1.126491646778043,
+      "grad_norm": 1.660888910293579,
+      "learning_rate": 4.3702014746614136e-05,
+      "loss": 1.4554,
+      "step": 1416
+    },
+    {
+      "epoch": 1.1288782816229117,
+      "grad_norm": 1.8002493381500244,
+      "learning_rate": 4.350629979576569e-05,
+      "loss": 1.4019,
+      "step": 1419
+    },
+    {
+      "epoch": 1.1312649164677804,
+      "grad_norm": 1.442589282989502,
+      "learning_rate": 4.331068599477605e-05,
+      "loss": 1.3116,
+      "step": 1422
+    },
+    {
+      "epoch": 1.1336515513126493,
+      "grad_norm": 1.5035159587860107,
+      "learning_rate": 4.311517639064578e-05,
+      "loss": 1.2794,
+      "step": 1425
+    },
+    {
+      "epoch": 1.136038186157518,
+      "grad_norm": 1.8796526193618774,
+      "learning_rate": 4.2919774028752436e-05,
+      "loss": 1.3611,
+      "step": 1428
+    },
+    {
+      "epoch": 1.1384248210023866,
+      "grad_norm": 1.4292360544204712,
+      "learning_rate": 4.27244819528031e-05,
+      "loss": 1.4402,
+      "step": 1431
+    },
+    {
+      "epoch": 1.1408114558472553,
+      "grad_norm": 1.6761499643325806,
+      "learning_rate": 4.2529303204786953e-05,
+      "loss": 1.4959,
+      "step": 1434
+    },
+    {
+      "epoch": 1.1431980906921242,
+      "grad_norm": 1.7346595525741577,
+      "learning_rate": 4.233424082492797e-05,
+      "loss": 1.4828,
+      "step": 1437
+    },
+    {
+      "epoch": 1.1455847255369929,
+      "grad_norm": 1.394885778427124,
+      "learning_rate": 4.213929785163747e-05,
+      "loss": 1.4321,
+      "step": 1440
+    },
+    {
+      "epoch": 1.1479713603818615,
+      "grad_norm": 1.510941743850708,
+      "learning_rate": 4.1944477321466786e-05,
+      "loss": 1.4397,
+      "step": 1443
+    },
+    {
+      "epoch": 1.1503579952267304,
+      "grad_norm": 1.5493265390396118,
+      "learning_rate": 4.1749782269060045e-05,
+      "loss": 1.457,
+      "step": 1446
+    },
+    {
+      "epoch": 1.152744630071599,
+      "grad_norm": 1.6018649339675903,
+      "learning_rate": 4.1555215727106844e-05,
+      "loss": 1.4258,
+      "step": 1449
+    },
+    {
+      "epoch": 1.1551312649164678,
+      "grad_norm": 1.5491352081298828,
+      "learning_rate": 4.136078072629503e-05,
+      "loss": 1.422,
+      "step": 1452
+    },
+    {
+      "epoch": 1.1575178997613365,
+      "grad_norm": 1.7042955160140991,
+      "learning_rate": 4.116648029526347e-05,
+      "loss": 1.406,
+      "step": 1455
+    },
+    {
+      "epoch": 1.1599045346062051,
+      "grad_norm": 1.5230995416641235,
+      "learning_rate": 4.097231746055491e-05,
+      "loss": 1.5026,
+      "step": 1458
+    },
+    {
+      "epoch": 1.162291169451074,
+      "grad_norm": 1.8555903434753418,
+      "learning_rate": 4.077829524656877e-05,
+      "loss": 1.3636,
+      "step": 1461
+    },
+    {
+      "epoch": 1.1646778042959427,
+      "grad_norm": 1.4330073595046997,
+      "learning_rate": 4.05844166755141e-05,
+      "loss": 1.4293,
+      "step": 1464
+    },
+    {
+      "epoch": 1.1670644391408114,
+      "grad_norm": 1.5035028457641602,
+      "learning_rate": 4.039068476736253e-05,
+      "loss": 1.406,
+      "step": 1467
+    },
+    {
+      "epoch": 1.1694510739856803,
+      "grad_norm": 1.7410966157913208,
+      "learning_rate": 4.01971025398011e-05,
+      "loss": 1.3049,
+      "step": 1470
+    },
+    {
+      "epoch": 1.171837708830549,
+      "grad_norm": 1.6594077348709106,
+      "learning_rate": 4.000367300818537e-05,
+      "loss": 1.6195,
+      "step": 1473
+    },
+    {
+      "epoch": 1.1742243436754176,
+      "grad_norm": 1.6386613845825195,
+      "learning_rate": 3.98103991854924e-05,
+      "loss": 1.4577,
+      "step": 1476
+    },
+    {
+      "epoch": 1.1766109785202863,
+      "grad_norm": 1.6970839500427246,
+      "learning_rate": 3.961728408227384e-05,
+      "loss": 1.5535,
+      "step": 1479
+    },
+    {
+      "epoch": 1.1789976133651552,
+      "grad_norm": 1.6995643377304077,
+      "learning_rate": 3.942433070660905e-05,
+      "loss": 1.4154,
+      "step": 1482
+    },
+    {
+      "epoch": 1.1813842482100239,
+      "grad_norm": 1.4957417249679565,
+      "learning_rate": 3.923154206405819e-05,
+      "loss": 1.4574,
+      "step": 1485
+    },
+    {
+      "epoch": 1.1837708830548925,
+      "grad_norm": 1.7634819746017456,
+      "learning_rate": 3.9038921157615444e-05,
+      "loss": 1.3863,
+      "step": 1488
+    },
+    {
+      "epoch": 1.1861575178997614,
+      "grad_norm": 1.64838707447052,
+      "learning_rate": 3.884647098766224e-05,
+      "loss": 1.4176,
+      "step": 1491
+    },
+    {
+      "epoch": 1.18854415274463,
+      "grad_norm": 1.379270315170288,
+      "learning_rate": 3.8654194551920485e-05,
+      "loss": 1.4031,
+      "step": 1494
+    },
+    {
+      "epoch": 1.1909307875894988,
+      "grad_norm": 1.4653505086898804,
+      "learning_rate": 3.846209484540597e-05,
+      "loss": 1.407,
+      "step": 1497
+    },
+    {
+      "epoch": 1.1933174224343674,
+      "grad_norm": 1.425595760345459,
+      "learning_rate": 3.827017486038157e-05,
+      "loss": 1.3342,
+      "step": 1500
+    },
+    {
+      "epoch": 1.1957040572792363,
+      "grad_norm": 1.3729442358016968,
+      "learning_rate": 3.8078437586310716e-05,
+      "loss": 1.5717,
+      "step": 1503
+    },
+    {
+      "epoch": 1.198090692124105,
+      "grad_norm": 1.317692756652832,
+      "learning_rate": 3.788688600981085e-05,
+      "loss": 1.4324,
+      "step": 1506
+    },
+    {
+      "epoch": 1.2004773269689737,
+      "grad_norm": 1.3449257612228394,
+      "learning_rate": 3.769552311460684e-05,
+      "loss": 1.454,
+      "step": 1509
+    },
+    {
+      "epoch": 1.2028639618138426,
+      "grad_norm": 1.6187500953674316,
+      "learning_rate": 3.750435188148459e-05,
+      "loss": 1.5128,
+      "step": 1512
+    },
+    {
+      "epoch": 1.2052505966587113,
+      "grad_norm": 2.184224843978882,
+      "learning_rate": 3.73133752882445e-05,
+      "loss": 1.2855,
+      "step": 1515
+    },
+    {
+      "epoch": 1.20763723150358,
+      "grad_norm": 1.5003141164779663,
+      "learning_rate": 3.712259630965518e-05,
+      "loss": 1.4485,
+      "step": 1518
+    },
+    {
+      "epoch": 1.2100238663484486,
+      "grad_norm": 1.4491519927978516,
+      "learning_rate": 3.6932017917407045e-05,
+      "loss": 1.4343,
+      "step": 1521
+    },
+    {
+      "epoch": 1.2124105011933175,
+      "grad_norm": 1.332411289215088,
+      "learning_rate": 3.6741643080066065e-05,
+      "loss": 1.3984,
+      "step": 1524
+    },
+    {
+      "epoch": 1.2147971360381862,
+      "grad_norm": 1.6156560182571411,
+      "learning_rate": 3.655147476302754e-05,
+      "loss": 1.4771,
+      "step": 1527
+    },
+    {
+      "epoch": 1.2171837708830548,
+      "grad_norm": 1.6296911239624023,
+      "learning_rate": 3.636151592846985e-05,
+      "loss": 1.3419,
+      "step": 1530
+    },
+    {
+      "epoch": 1.2195704057279237,
+      "grad_norm": 1.556209921836853,
+      "learning_rate": 3.617176953530835e-05,
+      "loss": 1.3619,
+      "step": 1533
+    },
+    {
+      "epoch": 1.2219570405727924,
+      "grad_norm": 1.383895754814148,
+      "learning_rate": 3.5982238539149285e-05,
+      "loss": 1.2554,
+      "step": 1536
+    },
+    {
+      "epoch": 1.224343675417661,
+      "grad_norm": 2.2628183364868164,
+      "learning_rate": 3.579292589224375e-05,
+      "loss": 1.4491,
+      "step": 1539
+    },
+    {
+      "epoch": 1.2267303102625298,
+      "grad_norm": 1.618979573249817,
+      "learning_rate": 3.560383454344168e-05,
+      "loss": 1.5344,
+      "step": 1542
+    },
+    {
+      "epoch": 1.2291169451073987,
+      "grad_norm": 1.386203646659851,
+      "learning_rate": 3.541496743814596e-05,
+      "loss": 1.4364,
+      "step": 1545
+    },
+    {
+      "epoch": 1.2315035799522673,
+      "grad_norm": 1.5511690378189087,
+      "learning_rate": 3.522632751826651e-05,
+      "loss": 1.3568,
+      "step": 1548
+    },
+    {
+      "epoch": 1.233890214797136,
+      "grad_norm": 1.6482512950897217,
+      "learning_rate": 3.503791772217445e-05,
+      "loss": 1.4858,
+      "step": 1551
+    },
+    {
+      "epoch": 1.2362768496420047,
+      "grad_norm": 1.5876716375350952,
+      "learning_rate": 3.484974098465636e-05,
+      "loss": 1.3507,
+      "step": 1554
+    },
+    {
+      "epoch": 1.2386634844868736,
+      "grad_norm": 1.7436355352401733,
+      "learning_rate": 3.4661800236868604e-05,
+      "loss": 1.4313,
+      "step": 1557
+    },
+    {
+      "epoch": 1.2410501193317423,
+      "grad_norm": 1.5434948205947876,
+      "learning_rate": 3.447409840629156e-05,
+      "loss": 1.3697,
+      "step": 1560
+    },
+    {
+      "epoch": 1.243436754176611,
+      "grad_norm": 1.7537883520126343,
+      "learning_rate": 3.428663841668412e-05,
+      "loss": 1.5541,
+      "step": 1563
+    },
+    {
+      "epoch": 1.2458233890214796,
+      "grad_norm": 1.4465534687042236,
+      "learning_rate": 3.409942318803809e-05,
+      "loss": 1.4101,
+      "step": 1566
+    },
+    {
+      "epoch": 1.2482100238663485,
+      "grad_norm": 1.4425692558288574,
+      "learning_rate": 3.391245563653276e-05,
+      "loss": 1.5231,
+      "step": 1569
+    },
+    {
+      "epoch": 1.2505966587112172,
+      "grad_norm": 1.571582317352295,
+      "learning_rate": 3.3725738674489414e-05,
+      "loss": 1.4215,
+      "step": 1572
+    },
+    {
+      "epoch": 1.2529832935560858,
+      "grad_norm": 1.4353744983673096,
+      "learning_rate": 3.3539275210326044e-05,
+      "loss": 1.4282,
+      "step": 1575
+    },
+    {
+      "epoch": 1.2553699284009547,
+      "grad_norm": 1.4851089715957642,
+      "learning_rate": 3.335306814851196e-05,
+      "loss": 1.4177,
+      "step": 1578
+    },
+    {
+      "epoch": 1.2577565632458234,
+      "grad_norm": 1.3933429718017578,
+      "learning_rate": 3.31671203895226e-05,
+      "loss": 1.3531,
+      "step": 1581
+    },
+    {
+      "epoch": 1.260143198090692,
+      "grad_norm": 3.346571207046509,
+      "learning_rate": 3.298143482979436e-05,
+      "loss": 1.3556,
+      "step": 1584
+    },
+    {
+      "epoch": 1.2625298329355608,
+      "grad_norm": 1.9931849241256714,
+      "learning_rate": 3.2796014361679464e-05,
+      "loss": 1.4935,
+      "step": 1587
+    },
+    {
+      "epoch": 1.2649164677804297,
+      "grad_norm": 1.5960787534713745,
+      "learning_rate": 3.261086187340088e-05,
+      "loss": 1.3861,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2673031026252983,
+      "grad_norm": 1.4517109394073486,
+      "learning_rate": 3.242598024900738e-05,
+      "loss": 1.4512,
+      "step": 1593
+    },
+    {
+      "epoch": 1.269689737470167,
+      "grad_norm": 1.7831088304519653,
+      "learning_rate": 3.224137236832859e-05,
+      "loss": 1.336,
+      "step": 1596
+    },
+    {
+      "epoch": 1.272076372315036,
+      "grad_norm": 1.3997858762741089,
+      "learning_rate": 3.2057041106930104e-05,
+      "loss": 1.3483,
+      "step": 1599
+    },
+    {
+      "epoch": 1.2744630071599046,
+      "grad_norm": 1.4272023439407349,
+      "learning_rate": 3.187298933606878e-05,
+      "loss": 1.4455,
+      "step": 1602
+    },
+    {
+      "epoch": 1.2768496420047732,
+      "grad_norm": 1.44106924533844,
+      "learning_rate": 3.1689219922647924e-05,
+      "loss": 1.3062,
+      "step": 1605
+    },
+    {
+      "epoch": 1.279236276849642,
+      "grad_norm": 1.472388744354248,
+      "learning_rate": 3.150573572917267e-05,
+      "loss": 1.5172,
+      "step": 1608
+    },
+    {
+      "epoch": 1.2816229116945108,
+      "grad_norm": 1.4416998624801636,
+      "learning_rate": 3.13225396137054e-05,
+      "loss": 1.3624,
+      "step": 1611
+    },
+    {
+      "epoch": 1.2840095465393795,
+      "grad_norm": 1.5669586658477783,
+      "learning_rate": 3.11396344298212e-05,
+      "loss": 1.4928,
+      "step": 1614
+    },
+    {
+      "epoch": 1.2863961813842482,
+      "grad_norm": 1.4020856618881226,
+      "learning_rate": 3.095702302656347e-05,
+      "loss": 1.4451,
+      "step": 1617
+    },
+    {
+      "epoch": 1.288782816229117,
+      "grad_norm": 1.58262038230896,
+      "learning_rate": 3.077470824839947e-05,
+      "loss": 1.4129,
+      "step": 1620
+    },
+    {
+      "epoch": 1.2911694510739857,
+      "grad_norm": 1.3566120862960815,
+      "learning_rate": 3.059269293517603e-05,
+      "loss": 1.4103,
+      "step": 1623
+    },
+    {
+      "epoch": 1.2935560859188544,
+      "grad_norm": 1.4482066631317139,
+      "learning_rate": 3.0410979922075343e-05,
+      "loss": 1.3193,
+      "step": 1626
+    },
+    {
+      "epoch": 1.295942720763723,
+      "grad_norm": 1.5326952934265137,
+      "learning_rate": 3.022957203957083e-05,
+      "loss": 1.508,
+      "step": 1629
+    },
+    {
+      "epoch": 1.2983293556085918,
+      "grad_norm": 1.4580435752868652,
+      "learning_rate": 3.004847211338295e-05,
+      "loss": 1.3046,
+      "step": 1632
+    },
+    {
+      "epoch": 1.3007159904534606,
+      "grad_norm": 1.4401758909225464,
+      "learning_rate": 2.9867682964435294e-05,
+      "loss": 1.403,
+      "step": 1635
+    },
+    {
+      "epoch": 1.3031026252983293,
+      "grad_norm": 1.3278849124908447,
+      "learning_rate": 2.9687207408810557e-05,
+      "loss": 1.4574,
+      "step": 1638
+    },
+    {
+      "epoch": 1.3054892601431982,
+      "grad_norm": 1.5741738080978394,
+      "learning_rate": 2.9507048257706727e-05,
+      "loss": 1.5322,
+      "step": 1641
+    },
+    {
+      "epoch": 1.307875894988067,
+      "grad_norm": 1.517532229423523,
+      "learning_rate": 2.9327208317393303e-05,
+      "loss": 1.5201,
+      "step": 1644
+    },
+    {
+      "epoch": 1.3102625298329356,
+      "grad_norm": 1.5097103118896484,
+      "learning_rate": 2.9147690389167514e-05,
+      "loss": 1.3938,
+      "step": 1647
+    },
+    {
+      "epoch": 1.3126491646778042,
+      "grad_norm": 1.369953989982605,
+      "learning_rate": 2.8968497269310803e-05,
+      "loss": 1.2942,
+      "step": 1650
+    },
+    {
+      "epoch": 1.315035799522673,
+      "grad_norm": 1.4780120849609375,
+      "learning_rate": 2.8789631749045097e-05,
+      "loss": 1.3554,
+      "step": 1653
+    },
+    {
+      "epoch": 1.3174224343675418,
+      "grad_norm": 2.5901074409484863,
+      "learning_rate": 2.8611096614489518e-05,
+      "loss": 1.3647,
+      "step": 1656
+    },
+    {
+      "epoch": 1.3198090692124105,
+      "grad_norm": 1.4900747537612915,
+      "learning_rate": 2.8432894646616885e-05,
+      "loss": 1.3546,
+      "step": 1659
+    },
+    {
+      "epoch": 1.3221957040572792,
+      "grad_norm": 1.3987863063812256,
+      "learning_rate": 2.8255028621210355e-05,
+      "loss": 1.4177,
+      "step": 1662
+    },
+    {
+      "epoch": 1.324582338902148,
+      "grad_norm": 11.975431442260742,
+      "learning_rate": 2.8077501308820308e-05,
+      "loss": 1.4814,
+      "step": 1665
+    },
+    {
+      "epoch": 1.3269689737470167,
+      "grad_norm": 1.503467082977295,
+      "learning_rate": 2.790031547472105e-05,
+      "loss": 1.4222,
+      "step": 1668
+    },
+    {
+      "epoch": 1.3293556085918854,
+      "grad_norm": 1.5251504182815552,
+      "learning_rate": 2.7723473878867877e-05,
+      "loss": 1.4811,
+      "step": 1671
+    },
+    {
+      "epoch": 1.331742243436754,
+      "grad_norm": 1.567043423652649,
+      "learning_rate": 2.754697927585399e-05,
+      "loss": 1.3752,
+      "step": 1674
+    },
+    {
+      "epoch": 1.334128878281623,
+      "grad_norm": 1.3771165609359741,
+      "learning_rate": 2.737083441486763e-05,
+      "loss": 1.3422,
+      "step": 1677
+    },
+    {
+      "epoch": 1.3365155131264916,
+      "grad_norm": 1.4927712678909302,
+      "learning_rate": 2.71950420396492e-05,
+      "loss": 1.3736,
+      "step": 1680
+    },
+    {
+      "epoch": 1.3389021479713603,
+      "grad_norm": 1.3801411390304565,
+      "learning_rate": 2.7019604888448642e-05,
+      "loss": 1.3837,
+      "step": 1683
+    },
+    {
+      "epoch": 1.3412887828162292,
+      "grad_norm": 1.4260993003845215,
+      "learning_rate": 2.6844525693982613e-05,
+      "loss": 1.3627,
+      "step": 1686
+    },
+    {
+      "epoch": 1.3436754176610979,
+      "grad_norm": 1.3967187404632568,
+      "learning_rate": 2.666980718339211e-05,
+      "loss": 1.4247,
+      "step": 1689
+    },
+    {
+      "epoch": 1.3460620525059666,
+      "grad_norm": 1.6355633735656738,
+      "learning_rate": 2.6495452078199863e-05,
+      "loss": 1.3911,
+      "step": 1692
+    },
+    {
+      "epoch": 1.3484486873508352,
+      "grad_norm": 1.4555933475494385,
+      "learning_rate": 2.6321463094267934e-05,
+      "loss": 1.3659,
+      "step": 1695
+    },
+    {
+      "epoch": 1.3508353221957041,
+      "grad_norm": 1.8073235750198364,
+      "learning_rate": 2.614784294175554e-05,
+      "loss": 1.4982,
+      "step": 1698
+    },
+    {
+      "epoch": 1.3532219570405728,
+      "grad_norm": 1.790961503982544,
+      "learning_rate": 2.597459432507664e-05,
+      "loss": 1.427,
+      "step": 1701
+    },
+    {
+      "epoch": 1.3556085918854415,
+      "grad_norm": 1.4583306312561035,
+      "learning_rate": 2.5801719942858065e-05,
+      "loss": 1.4358,
+      "step": 1704
+    },
+    {
+      "epoch": 1.3579952267303104,
+      "grad_norm": 1.3691617250442505,
+      "learning_rate": 2.562922248789722e-05,
+      "loss": 1.4026,
+      "step": 1707
+    },
+    {
+      "epoch": 1.360381861575179,
+      "grad_norm": 1.4829280376434326,
+      "learning_rate": 2.5457104647120322e-05,
+      "loss": 1.5299,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3627684964200477,
+      "grad_norm": 1.4290732145309448,
+      "learning_rate": 2.5285369101540445e-05,
+      "loss": 1.4366,
+      "step": 1713
+    },
+    {
+      "epoch": 1.3651551312649164,
+      "grad_norm": 1.5401760339736938,
+      "learning_rate": 2.5114018526215844e-05,
+      "loss": 1.4262,
+      "step": 1716
+    },
+    {
+      "epoch": 1.3675417661097853,
+      "grad_norm": 1.3975614309310913,
+      "learning_rate": 2.494305559020822e-05,
+      "loss": 1.4702,
+      "step": 1719
+    },
+    {
+      "epoch": 1.369928400954654,
+      "grad_norm": 1.3241435289382935,
+      "learning_rate": 2.4772482956541132e-05,
+      "loss": 1.3565,
+      "step": 1722
+    },
+    {
+      "epoch": 1.3723150357995226,
+      "grad_norm": 1.4867254495620728,
+      "learning_rate": 2.4602303282158616e-05,
+      "loss": 1.3253,
+      "step": 1725
+    },
+    {
+      "epoch": 1.3747016706443915,
+      "grad_norm": 1.5833885669708252,
+      "learning_rate": 2.4432519217883676e-05,
+      "loss": 1.421,
+      "step": 1728
+    },
+    {
+      "epoch": 1.3770883054892602,
+      "grad_norm": 1.5261849164962769,
+      "learning_rate": 2.4263133408377076e-05,
+      "loss": 1.4693,
+      "step": 1731
+    },
+    {
+      "epoch": 1.3794749403341289,
+      "grad_norm": 1.4552416801452637,
+      "learning_rate": 2.4094148492096125e-05,
+      "loss": 1.5436,
+      "step": 1734
+    },
+    {
+      "epoch": 1.3818615751789975,
+      "grad_norm": 1.5196255445480347,
+      "learning_rate": 2.3925567101253576e-05,
+      "loss": 1.5093,
+      "step": 1737
+    },
+    {
+      "epoch": 1.3842482100238662,
+      "grad_norm": 1.3654688596725464,
+      "learning_rate": 2.3757391861776585e-05,
+      "loss": 1.3681,
+      "step": 1740
+    },
+    {
+      "epoch": 1.3866348448687351,
+      "grad_norm": 1.4244242906570435,
+      "learning_rate": 2.3589625393265895e-05,
+      "loss": 1.3754,
+      "step": 1743
+    },
+    {
+      "epoch": 1.3890214797136038,
+      "grad_norm": 1.6873931884765625,
+      "learning_rate": 2.3422270308954934e-05,
+      "loss": 1.4434,
+      "step": 1746
+    },
+    {
+      "epoch": 1.3914081145584727,
+      "grad_norm": 1.2441445589065552,
+      "learning_rate": 2.3255329215669185e-05,
+      "loss": 1.3577,
+      "step": 1749
+    },
+    {
+      "epoch": 1.3937947494033414,
+      "grad_norm": 1.4804401397705078,
+      "learning_rate": 2.3088804713785584e-05,
+      "loss": 1.4486,
+      "step": 1752
+    },
+    {
+      "epoch": 1.39618138424821,
+      "grad_norm": 1.3242276906967163,
+      "learning_rate": 2.2922699397191893e-05,
+      "loss": 1.3782,
+      "step": 1755
+    },
+    {
+      "epoch": 1.3985680190930787,
+      "grad_norm": 1.4735126495361328,
+      "learning_rate": 2.2757015853246493e-05,
+      "loss": 1.3395,
+      "step": 1758
+    },
+    {
+      "epoch": 1.4009546539379474,
+      "grad_norm": 1.2929046154022217,
+      "learning_rate": 2.2591756662737862e-05,
+      "loss": 1.4035,
+      "step": 1761
+    },
+    {
+      "epoch": 1.4033412887828163,
+      "grad_norm": 1.3094271421432495,
+      "learning_rate": 2.242692439984463e-05,
+      "loss": 1.3874,
+      "step": 1764
+    },
+    {
+      "epoch": 1.405727923627685,
+      "grad_norm": 1.598767876625061,
+      "learning_rate": 2.2262521632095203e-05,
+      "loss": 1.4688,
+      "step": 1767
+    },
+    {
+      "epoch": 1.4081145584725536,
+      "grad_norm": 1.4594007730484009,
+      "learning_rate": 2.2098550920327998e-05,
+      "loss": 1.2801,
+      "step": 1770
+    },
+    {
+      "epoch": 1.4105011933174225,
+      "grad_norm": 1.6119073629379272,
+      "learning_rate": 2.1935014818651405e-05,
+      "loss": 1.3331,
+      "step": 1773
+    },
+    {
+      "epoch": 1.4128878281622912,
+      "grad_norm": 1.351380705833435,
+      "learning_rate": 2.177191587440409e-05,
+      "loss": 1.3735,
+      "step": 1776
+    },
+    {
+      "epoch": 1.4152744630071599,
+      "grad_norm": 1.4811229705810547,
+      "learning_rate": 2.1609256628115316e-05,
+      "loss": 1.4969,
+      "step": 1779
+    },
+    {
+      "epoch": 1.4176610978520285,
+      "grad_norm": 1.5260910987854004,
+      "learning_rate": 2.1447039613465265e-05,
+      "loss": 1.4194,
+      "step": 1782
+    },
+    {
+      "epoch": 1.4200477326968974,
+      "grad_norm": 1.8757227659225464,
+      "learning_rate": 2.128526735724572e-05,
+      "loss": 1.4043,
+      "step": 1785
+    },
+    {
+      "epoch": 1.422434367541766,
+      "grad_norm": 1.8287010192871094,
+      "learning_rate": 2.1123942379320576e-05,
+      "loss": 1.3959,
+      "step": 1788
+    },
+    {
+      "epoch": 1.4248210023866348,
+      "grad_norm": 1.6383695602416992,
+      "learning_rate": 2.096306719258669e-05,
+      "loss": 1.3586,
+      "step": 1791
+    },
+    {
+      "epoch": 1.4272076372315037,
+      "grad_norm": 1.5422853231430054,
+      "learning_rate": 2.0802644302934683e-05,
+      "loss": 1.4287,
+      "step": 1794
+    },
+    {
+      "epoch": 1.4295942720763724,
+      "grad_norm": 1.5485094785690308,
+      "learning_rate": 2.0642676209209934e-05,
+      "loss": 1.5216,
+      "step": 1797
+    },
+    {
+      "epoch": 1.431980906921241,
+      "grad_norm": 1.9971394538879395,
+      "learning_rate": 2.0483165403173583e-05,
+      "loss": 1.4238,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4343675417661097,
+      "grad_norm": 1.4561408758163452,
+      "learning_rate": 2.0324114369463855e-05,
+      "loss": 1.3847,
+      "step": 1803
+    },
+    {
+      "epoch": 1.4367541766109786,
+      "grad_norm": 1.280886173248291,
+      "learning_rate": 2.0165525585557204e-05,
+      "loss": 1.3258,
+      "step": 1806
+    },
+    {
+      "epoch": 1.4391408114558473,
+      "grad_norm": 1.3426673412322998,
+      "learning_rate": 2.0007401521729863e-05,
+      "loss": 1.3456,
+      "step": 1809
+    },
+    {
+      "epoch": 1.441527446300716,
+      "grad_norm": 1.6823667287826538,
+      "learning_rate": 1.984974464101928e-05,
+      "loss": 1.406,
+      "step": 1812
+    },
+    {
+      "epoch": 1.4439140811455848,
+      "grad_norm": 1.64955735206604,
+      "learning_rate": 1.9692557399185734e-05,
+      "loss": 1.5698,
+      "step": 1815
+    },
+    {
+      "epoch": 1.4463007159904535,
+      "grad_norm": 1.390384316444397,
+      "learning_rate": 1.953584224467418e-05,
+      "loss": 1.3851,
+      "step": 1818
+    },
+    {
+      "epoch": 1.4486873508353222,
+      "grad_norm": 1.543288230895996,
+      "learning_rate": 1.9379601618575977e-05,
+      "loss": 1.3457,
+      "step": 1821
+    },
+    {
+      "epoch": 1.4510739856801909,
+      "grad_norm": 1.235751986503601,
+      "learning_rate": 1.9223837954591046e-05,
+      "loss": 1.398,
+      "step": 1824
+    },
+    {
+      "epoch": 1.4534606205250595,
+      "grad_norm": 1.6137385368347168,
+      "learning_rate": 1.9068553678989736e-05,
+      "loss": 1.513,
+      "step": 1827
+    },
+    {
+      "epoch": 1.4558472553699284,
+      "grad_norm": 1.5480167865753174,
+      "learning_rate": 1.8913751210575248e-05,
+      "loss": 1.4052,
+      "step": 1830
+    },
+    {
+      "epoch": 1.458233890214797,
+      "grad_norm": 1.3694156408309937,
+      "learning_rate": 1.8759432960645774e-05,
+      "loss": 1.4116,
+      "step": 1833
+    },
+    {
+      "epoch": 1.460620525059666,
+      "grad_norm": 1.3757750988006592,
+      "learning_rate": 1.8605601332957077e-05,
+      "loss": 1.3357,
+      "step": 1836
+    },
+    {
+      "epoch": 1.4630071599045347,
+      "grad_norm": 1.315748691558838,
+      "learning_rate": 1.8452258723684995e-05,
+      "loss": 1.3027,
+      "step": 1839
+    },
+    {
+      "epoch": 1.4653937947494033,
+      "grad_norm": 1.4307796955108643,
+      "learning_rate": 1.8299407521388067e-05,
+      "loss": 1.323,
+      "step": 1842
+    },
+    {
+      "epoch": 1.467780429594272,
+      "grad_norm": 1.3611763715744019,
+      "learning_rate": 1.8147050106970437e-05,
+      "loss": 1.4422,
+      "step": 1845
+    },
+    {
+      "epoch": 1.4701670644391407,
+      "grad_norm": 1.3486278057098389,
+      "learning_rate": 1.7995188853644646e-05,
+      "loss": 1.4126,
+      "step": 1848
+    },
+    {
+      "epoch": 1.4725536992840096,
+      "grad_norm": 1.381899118423462,
+      "learning_rate": 1.784382612689477e-05,
+      "loss": 1.3985,
+      "step": 1851
+    },
+    {
+      "epoch": 1.4749403341288783,
+      "grad_norm": 1.3606630563735962,
+      "learning_rate": 1.7692964284439505e-05,
+      "loss": 1.3024,
+      "step": 1854
+    },
+    {
+      "epoch": 1.477326968973747,
+      "grad_norm": 1.542244791984558,
+      "learning_rate": 1.7542605676195506e-05,
+      "loss": 1.3591,
+      "step": 1857
+    },
+    {
+      "epoch": 1.4797136038186158,
+      "grad_norm": 1.285721778869629,
+      "learning_rate": 1.739275264424067e-05,
+      "loss": 1.4591,
+      "step": 1860
+    },
+    {
+      "epoch": 1.4821002386634845,
+      "grad_norm": 1.584008812904358,
+      "learning_rate": 1.7243407522777806e-05,
+      "loss": 1.4387,
+      "step": 1863
+    },
+    {
+      "epoch": 1.4844868735083532,
+      "grad_norm": 1.2456049919128418,
+      "learning_rate": 1.7094572638098123e-05,
+      "loss": 1.4347,
+      "step": 1866
+    },
+    {
+      "epoch": 1.4868735083532219,
+      "grad_norm": 1.5531784296035767,
+      "learning_rate": 1.6946250308545125e-05,
+      "loss": 1.3157,
+      "step": 1869
+    },
+    {
+      "epoch": 1.4892601431980907,
+      "grad_norm": 1.3614214658737183,
+      "learning_rate": 1.6798442844478445e-05,
+      "loss": 1.3473,
+      "step": 1872
+    },
+    {
+      "epoch": 1.4916467780429594,
+      "grad_norm": 1.2862516641616821,
+      "learning_rate": 1.6651152548237802e-05,
+      "loss": 1.3371,
+      "step": 1875
+    },
+    {
+      "epoch": 1.494033412887828,
+      "grad_norm": 1.3549131155014038,
+      "learning_rate": 1.6504381714107252e-05,
+      "loss": 1.4438,
+      "step": 1878
+    },
+    {
+      "epoch": 1.496420047732697,
+      "grad_norm": 1.3501994609832764,
+      "learning_rate": 1.6358132628279322e-05,
+      "loss": 1.3332,
+      "step": 1881
+    },
+    {
+      "epoch": 1.4988066825775657,
+      "grad_norm": 1.3975999355316162,
+      "learning_rate": 1.6212407568819565e-05,
+      "loss": 1.3167,
+      "step": 1884
+    },
+    {
+      "epoch": 1.5011933174224343,
+      "grad_norm": 1.443943738937378,
+      "learning_rate": 1.6067208805630877e-05,
+      "loss": 1.3842,
+      "step": 1887
+    },
+    {
+      "epoch": 1.503579952267303,
+      "grad_norm": 2.483057975769043,
+      "learning_rate": 1.5922538600418318e-05,
+      "loss": 1.351,
+      "step": 1890
+    },
+    {
+      "epoch": 1.5059665871121717,
+      "grad_norm": 1.4036768674850464,
+      "learning_rate": 1.5778399206653734e-05,
+      "loss": 1.3499,
+      "step": 1893
+    },
+    {
+      "epoch": 1.5083532219570406,
+      "grad_norm": 1.3263362646102905,
+      "learning_rate": 1.563479286954078e-05,
+      "loss": 1.3777,
+      "step": 1896
+    },
+    {
+      "epoch": 1.5107398568019093,
+      "grad_norm": 1.3591517210006714,
+      "learning_rate": 1.54917218259799e-05,
+      "loss": 1.293,
+      "step": 1899
+    },
+    {
+      "epoch": 1.5131264916467781,
+      "grad_norm": 1.310631513595581,
+      "learning_rate": 1.5349188304533413e-05,
+      "loss": 1.4139,
+      "step": 1902
+    },
+    {
+      "epoch": 1.5155131264916468,
+      "grad_norm": 1.3131390810012817,
+      "learning_rate": 1.5207194525390938e-05,
+      "loss": 1.4123,
+      "step": 1905
+    },
+    {
+      "epoch": 1.5178997613365155,
+      "grad_norm": 1.4222484827041626,
+      "learning_rate": 1.5065742700334678e-05,
+      "loss": 1.3474,
+      "step": 1908
+    },
+    {
+      "epoch": 1.5202863961813842,
+      "grad_norm": 1.4119162559509277,
+      "learning_rate": 1.4924835032705064e-05,
+      "loss": 1.3558,
+      "step": 1911
+    },
+    {
+      "epoch": 1.5226730310262528,
+      "grad_norm": 1.356094479560852,
+      "learning_rate": 1.4784473717366387e-05,
+      "loss": 1.4808,
+      "step": 1914
+    },
+    {
+      "epoch": 1.5250596658711217,
+      "grad_norm": 1.249845266342163,
+      "learning_rate": 1.4644660940672627e-05,
+      "loss": 1.3971,
+      "step": 1917
+    },
+    {
+      "epoch": 1.5274463007159904,
+      "grad_norm": 1.2531503438949585,
+      "learning_rate": 1.4505398880433369e-05,
+      "loss": 1.3037,
+      "step": 1920
+    },
+    {
+      "epoch": 1.5298329355608593,
+      "grad_norm": 1.3742297887802124,
+      "learning_rate": 1.4366689705879898e-05,
+      "loss": 1.3794,
+      "step": 1923
+    },
+    {
+      "epoch": 1.532219570405728,
+      "grad_norm": 1.2931220531463623,
+      "learning_rate": 1.4228535577631442e-05,
+      "loss": 1.2969,
+      "step": 1926
+    },
+    {
+      "epoch": 1.5346062052505967,
+      "grad_norm": 1.314113974571228,
+      "learning_rate": 1.4090938647661461e-05,
+      "loss": 1.4102,
+      "step": 1929
+    },
+    {
+      "epoch": 1.5369928400954653,
+      "grad_norm": 1.6142003536224365,
+      "learning_rate": 1.3953901059264191e-05,
+      "loss": 1.5013,
+      "step": 1932
+    },
+    {
+      "epoch": 1.539379474940334,
+      "grad_norm": 1.374854564666748,
+      "learning_rate": 1.3817424947021151e-05,
+      "loss": 1.3812,
+      "step": 1935
+    },
+    {
+      "epoch": 1.541766109785203,
+      "grad_norm": 1.574981689453125,
+      "learning_rate": 1.3681512436768045e-05,
+      "loss": 1.5476,
+      "step": 1938
+    },
+    {
+      "epoch": 1.5441527446300716,
+      "grad_norm": 1.3789339065551758,
+      "learning_rate": 1.3546165645561487e-05,
+      "loss": 1.3535,
+      "step": 1941
+    },
+    {
+      "epoch": 1.5465393794749405,
+      "grad_norm": 1.4727178812026978,
+      "learning_rate": 1.3411386681646164e-05,
+      "loss": 1.3763,
+      "step": 1944
+    },
+    {
+      "epoch": 1.5489260143198091,
+      "grad_norm": 1.3274617195129395,
+      "learning_rate": 1.3277177644421924e-05,
+      "loss": 1.3276,
+      "step": 1947
+    },
+    {
+      "epoch": 1.5513126491646778,
+      "grad_norm": 1.5898061990737915,
+      "learning_rate": 1.314354062441106e-05,
+      "loss": 1.4907,
+      "step": 1950
+    },
+    {
+      "epoch": 1.5536992840095465,
+      "grad_norm": 1.401552677154541,
+      "learning_rate": 1.301047770322581e-05,
+      "loss": 1.3592,
+      "step": 1953
+    },
+    {
+      "epoch": 1.5560859188544152,
+      "grad_norm": 1.2774077653884888,
+      "learning_rate": 1.287799095353584e-05,
+      "loss": 1.2775,
+      "step": 1956
+    },
+    {
+      "epoch": 1.558472553699284,
+      "grad_norm": 1.458093523979187,
+      "learning_rate": 1.2746082439036117e-05,
+      "loss": 1.3795,
+      "step": 1959
+    },
+    {
+      "epoch": 1.5608591885441527,
+      "grad_norm": 1.4670827388763428,
+      "learning_rate": 1.2614754214414548e-05,
+      "loss": 1.3595,
+      "step": 1962
+    },
+    {
+      "epoch": 1.5632458233890216,
+      "grad_norm": 1.629981279373169,
+      "learning_rate": 1.2484008325320174e-05,
+      "loss": 1.3965,
+      "step": 1965
+    },
+    {
+      "epoch": 1.5656324582338903,
+      "grad_norm": 1.4332212209701538,
+      "learning_rate": 1.2353846808331154e-05,
+      "loss": 1.3406,
+      "step": 1968
+    },
+    {
+      "epoch": 1.568019093078759,
+      "grad_norm": 1.4571059942245483,
+      "learning_rate": 1.2224271690923155e-05,
+      "loss": 1.3083,
+      "step": 1971
+    },
+    {
+      "epoch": 1.5704057279236276,
+      "grad_norm": 1.2997206449508667,
+      "learning_rate": 1.2095284991437733e-05,
+      "loss": 1.3003,
+      "step": 1974
+    },
+    {
+      "epoch": 1.5727923627684963,
+      "grad_norm": 1.315752625465393,
+      "learning_rate": 1.1966888719050829e-05,
+      "loss": 1.4088,
+      "step": 1977
+    },
+    {
+      "epoch": 1.575178997613365,
+      "grad_norm": 1.3280858993530273,
+      "learning_rate": 1.1839084873741584e-05,
+      "loss": 1.3781,
+      "step": 1980
+    },
+    {
+      "epoch": 1.577565632458234,
+      "grad_norm": 1.4649864435195923,
+      "learning_rate": 1.1711875446261094e-05,
+      "loss": 1.4019,
+      "step": 1983
+    },
+    {
+      "epoch": 1.5799522673031028,
+      "grad_norm": 1.6478016376495361,
+      "learning_rate": 1.1585262418101467e-05,
+      "loss": 1.4548,
+      "step": 1986
+    },
+    {
+      "epoch": 1.5823389021479715,
+      "grad_norm": 1.4969959259033203,
+      "learning_rate": 1.1459247761464909e-05,
+      "loss": 1.3712,
+      "step": 1989
+    },
+    {
+      "epoch": 1.5847255369928401,
+      "grad_norm": 1.4021399021148682,
+      "learning_rate": 1.1333833439233055e-05,
+      "loss": 1.3716,
+      "step": 1992
+    },
+    {
+      "epoch": 1.5871121718377088,
+      "grad_norm": 1.4327634572982788,
+      "learning_rate": 1.1209021404936304e-05,
+      "loss": 1.3505,
+      "step": 1995
+    },
+    {
+      "epoch": 1.5894988066825775,
+      "grad_norm": 1.3530958890914917,
+      "learning_rate": 1.1084813602723515e-05,
+      "loss": 1.4087,
+      "step": 1998
+    },
+    {
+      "epoch": 1.5918854415274462,
+      "grad_norm": 1.4223620891571045,
+      "learning_rate": 1.0961211967331597e-05,
+      "loss": 1.4233,
+      "step": 2001
+    },
+    {
+      "epoch": 1.594272076372315,
+      "grad_norm": 1.3832675218582153,
+      "learning_rate": 1.083821842405548e-05,
+      "loss": 1.3184,
+      "step": 2004
+    },
+    {
+      "epoch": 1.5966587112171837,
+      "grad_norm": 4.159378528594971,
+      "learning_rate": 1.0715834888718074e-05,
+      "loss": 1.2636,
+      "step": 2007
+    },
+    {
+      "epoch": 1.5990453460620526,
+      "grad_norm": 1.4075279235839844,
+      "learning_rate": 1.0594063267640386e-05,
+      "loss": 1.2941,
+      "step": 2010
+    },
+    {
+      "epoch": 1.6014319809069213,
+      "grad_norm": 1.3836835622787476,
+      "learning_rate": 1.0472905457611936e-05,
+      "loss": 1.3858,
+      "step": 2013
+    },
+    {
+      "epoch": 1.60381861575179,
+      "grad_norm": 1.3855738639831543,
+      "learning_rate": 1.0352363345861065e-05,
+      "loss": 1.3013,
+      "step": 2016
+    },
+    {
+      "epoch": 1.6062052505966586,
+      "grad_norm": 1.3453712463378906,
+      "learning_rate": 1.023243881002573e-05,
+      "loss": 1.5934,
+      "step": 2019
+    },
+    {
+      "epoch": 1.6085918854415273,
+      "grad_norm": 14.255722999572754,
+      "learning_rate": 1.0113133718124035e-05,
+      "loss": 1.4809,
+      "step": 2022
+    },
+    {
+      "epoch": 1.6109785202863962,
+      "grad_norm": 1.4346468448638916,
+      "learning_rate": 9.994449928525324e-06,
+      "loss": 1.4255,
+      "step": 2025
+    },
+    {
+      "epoch": 1.6133651551312649,
+      "grad_norm": 1.797566533088684,
+      "learning_rate": 9.876389289921106e-06,
+      "loss": 1.4909,
+      "step": 2028
+    },
+    {
+      "epoch": 1.6157517899761338,
+      "grad_norm": 1.2779227495193481,
+      "learning_rate": 9.758953641296331e-06,
+      "loss": 1.3708,
+      "step": 2031
+    },
+    {
+      "epoch": 1.6181384248210025,
+      "grad_norm": 1.439191460609436,
+      "learning_rate": 9.642144811900739e-06,
+      "loss": 1.3745,
+      "step": 2034
+    },
+    {
+      "epoch": 1.6205250596658711,
+      "grad_norm": 1.3853205442428589,
+      "learning_rate": 9.5259646212203e-06,
+      "loss": 1.3369,
+      "step": 2037
+    },
+    {
+      "epoch": 1.6229116945107398,
+      "grad_norm": 1.397948145866394,
+      "learning_rate": 9.410414878948975e-06,
+      "loss": 1.365,
+      "step": 2040
+    },
+    {
+      "epoch": 1.6252983293556085,
+      "grad_norm": 1.2659891843795776,
+      "learning_rate": 9.295497384960416e-06,
+      "loss": 1.3817,
+      "step": 2043
+    },
+    {
+      "epoch": 1.6276849642004774,
+      "grad_norm": 1.4338490962982178,
+      "learning_rate": 9.181213929280046e-06,
+      "loss": 1.4503,
+      "step": 2046
+    },
+    {
+      "epoch": 1.630071599045346,
+      "grad_norm": 1.3001736402511597,
+      "learning_rate": 9.067566292057084e-06,
+      "loss": 1.3814,
+      "step": 2049
+    },
+    {
+      "epoch": 1.632458233890215,
+      "grad_norm": 1.4515080451965332,
+      "learning_rate": 8.954556243536877e-06,
+      "loss": 1.2801,
+      "step": 2052
+    },
+    {
+      "epoch": 1.6348448687350836,
+      "grad_norm": 1.4361463785171509,
+      "learning_rate": 8.842185544033255e-06,
+      "loss": 1.4327,
+      "step": 2055
+    },
+    {
+      "epoch": 1.6372315035799523,
+      "grad_norm": 1.36533522605896,
+      "learning_rate": 8.7304559439012e-06,
+      "loss": 1.4052,
+      "step": 2058
+    },
+    {
+      "epoch": 1.639618138424821,
+      "grad_norm": 1.3295538425445557,
+      "learning_rate": 8.619369183509501e-06,
+      "loss": 1.3336,
+      "step": 2061
+    },
+    {
+      "epoch": 1.6420047732696896,
+      "grad_norm": 1.3914989233016968,
+      "learning_rate": 8.508926993213712e-06,
+      "loss": 1.4033,
+      "step": 2064
+    },
+    {
+      "epoch": 1.6443914081145583,
+      "grad_norm": 1.3706380128860474,
+      "learning_rate": 8.39913109332916e-06,
+      "loss": 1.3468,
+      "step": 2067
+    },
+    {
+      "epoch": 1.6467780429594272,
+      "grad_norm": 1.386931300163269,
+      "learning_rate": 8.28998319410413e-06,
+      "loss": 1.3398,
+      "step": 2070
+    },
+    {
+      "epoch": 1.649164677804296,
+      "grad_norm": 1.307201623916626,
+      "learning_rate": 8.181484995693295e-06,
+      "loss": 1.3215,
+      "step": 2073
+    },
+    {
+      "epoch": 1.6515513126491648,
+      "grad_norm": 1.7024941444396973,
+      "learning_rate": 8.073638188131128e-06,
+      "loss": 1.3774,
+      "step": 2076
+    },
+    {
+      "epoch": 1.6539379474940334,
+      "grad_norm": 1.461051106452942,
+      "learning_rate": 7.966444451305726e-06,
+      "loss": 1.4113,
+      "step": 2079
+    },
+    {
+      "epoch": 1.6563245823389021,
+      "grad_norm": 1.3585187196731567,
+      "learning_rate": 7.859905454932471e-06,
+      "loss": 1.2784,
+      "step": 2082
+    },
+    {
+      "epoch": 1.6587112171837708,
+      "grad_norm": 1.2371797561645508,
+      "learning_rate": 7.75402285852816e-06,
+      "loss": 1.3123,
+      "step": 2085
+    },
+    {
+      "epoch": 1.6610978520286395,
+      "grad_norm": 1.3656738996505737,
+      "learning_rate": 7.648798311385058e-06,
+      "loss": 1.3187,
+      "step": 2088
+    },
+    {
+      "epoch": 1.6634844868735084,
+      "grad_norm": 1.3773971796035767,
+      "learning_rate": 7.5442334525452964e-06,
+      "loss": 1.2809,
+      "step": 2091
+    },
+    {
+      "epoch": 1.665871121718377,
+      "grad_norm": 1.370139718055725,
+      "learning_rate": 7.440329910775273e-06,
+      "loss": 1.3349,
+      "step": 2094
+    },
+    {
+      "epoch": 1.668257756563246,
+      "grad_norm": 1.5281294584274292,
+      "learning_rate": 7.337089304540301e-06,
+      "loss": 1.3332,
+      "step": 2097
+    },
+    {
+      "epoch": 1.6706443914081146,
+      "grad_norm": 1.3767362833023071,
+      "learning_rate": 7.234513241979418e-06,
+      "loss": 1.3754,
+      "step": 2100
+    },
+    {
+      "epoch": 1.6730310262529833,
+      "grad_norm": 1.2722810506820679,
+      "learning_rate": 7.132603320880294e-06,
+      "loss": 1.3081,
+      "step": 2103
+    },
+    {
+      "epoch": 1.675417661097852,
+      "grad_norm": 1.3534489870071411,
+      "learning_rate": 7.031361128654401e-06,
+      "loss": 1.4329,
+      "step": 2106
+    },
+    {
+      "epoch": 1.6778042959427206,
+      "grad_norm": 1.2815319299697876,
+      "learning_rate": 6.930788242312253e-06,
+      "loss": 1.3198,
+      "step": 2109
+    },
+    {
+      "epoch": 1.6801909307875895,
+      "grad_norm": 1.3864092826843262,
+      "learning_rate": 6.830886228438837e-06,
+      "loss": 1.4143,
+      "step": 2112
+    },
+    {
+      "epoch": 1.6825775656324582,
+      "grad_norm": 1.338248372077942,
+      "learning_rate": 6.731656643169204e-06,
+      "loss": 1.3364,
+      "step": 2115
+    },
+    {
+      "epoch": 1.684964200477327,
+      "grad_norm": 1.4621409177780151,
+      "learning_rate": 6.633101032164274e-06,
+      "loss": 1.4438,
+      "step": 2118
+    },
+    {
+      "epoch": 1.6873508353221958,
+      "grad_norm": 1.3858606815338135,
+      "learning_rate": 6.535220930586705e-06,
+      "loss": 1.5138,
+      "step": 2121
+    },
+    {
+      "epoch": 1.6897374701670644,
+      "grad_norm": 1.2535791397094727,
+      "learning_rate": 6.4380178630770225e-06,
+      "loss": 1.3784,
+      "step": 2124
+    },
+    {
+      "epoch": 1.692124105011933,
+      "grad_norm": 1.3751507997512817,
+      "learning_rate": 6.341493343729854e-06,
+      "loss": 1.4508,
+      "step": 2127
+    },
+    {
+      "epoch": 1.6945107398568018,
+      "grad_norm": 1.2520653009414673,
+      "learning_rate": 6.2456488760703205e-06,
+      "loss": 1.4547,
+      "step": 2130
+    },
+    {
+      "epoch": 1.6968973747016707,
+      "grad_norm": 1.4060708284378052,
+      "learning_rate": 6.150485953030677e-06,
+      "loss": 1.223,
+      "step": 2133
+    },
+    {
+      "epoch": 1.6992840095465394,
+      "grad_norm": 1.4133458137512207,
+      "learning_rate": 6.056006056926977e-06,
+      "loss": 1.4631,
+      "step": 2136
+    },
+    {
+      "epoch": 1.7016706443914082,
+      "grad_norm": 1.3407944440841675,
+      "learning_rate": 5.962210659436091e-06,
+      "loss": 1.3376,
+      "step": 2139
+    },
+    {
+      "epoch": 1.704057279236277,
+      "grad_norm": 1.3000420331954956,
+      "learning_rate": 5.869101221572654e-06,
+      "loss": 1.2783,
+      "step": 2142
+    },
+    {
+      "epoch": 1.7064439140811456,
+      "grad_norm": 1.3916162252426147,
+      "learning_rate": 5.776679193666412e-06,
+      "loss": 1.4991,
+      "step": 2145
+    },
+    {
+      "epoch": 1.7088305489260143,
+      "grad_norm": 1.5495882034301758,
+      "learning_rate": 5.6849460153395706e-06,
+      "loss": 1.3435,
+      "step": 2148
+    },
+    {
+      "epoch": 1.711217183770883,
+      "grad_norm": 1.365742564201355,
+      "learning_rate": 5.5939031154844e-06,
+      "loss": 1.3268,
+      "step": 2151
+    },
+    {
+      "epoch": 1.7136038186157518,
+      "grad_norm": 1.3886197805404663,
+      "learning_rate": 5.5035519122409895e-06,
+      "loss": 1.3701,
+      "step": 2154
+    },
+    {
+      "epoch": 1.7159904534606205,
+      "grad_norm": 1.5498794317245483,
+      "learning_rate": 5.413893812975096e-06,
+      "loss": 1.4669,
+      "step": 2157
+    },
+    {
+      "epoch": 1.7183770883054894,
+      "grad_norm": 1.51931893825531,
+      "learning_rate": 5.324930214256302e-06,
+      "loss": 1.4072,
+      "step": 2160
+    },
+    {
+      "epoch": 1.720763723150358,
+      "grad_norm": 1.441819667816162,
+      "learning_rate": 5.236662501836192e-06,
+      "loss": 1.3644,
+      "step": 2163
+    },
+    {
+      "epoch": 1.7231503579952268,
+      "grad_norm": 1.5425219535827637,
+      "learning_rate": 5.149092050626825e-06,
+      "loss": 1.5182,
+      "step": 2166
+    },
+    {
+      "epoch": 1.7255369928400954,
+      "grad_norm": 1.4178110361099243,
+      "learning_rate": 5.062220224679276e-06,
+      "loss": 1.3155,
+      "step": 2169
+    },
+    {
+      "epoch": 1.727923627684964,
+      "grad_norm": 1.401777982711792,
+      "learning_rate": 4.9760483771624236e-06,
+      "loss": 1.3718,
+      "step": 2172
+    },
+    {
+      "epoch": 1.7303102625298328,
+      "grad_norm": 1.3606517314910889,
+      "learning_rate": 4.89057785034181e-06,
+      "loss": 1.3611,
+      "step": 2175
+    },
+    {
+      "epoch": 1.7326968973747017,
+      "grad_norm": 1.4071412086486816,
+      "learning_rate": 4.805809975558828e-06,
+      "loss": 1.3619,
+      "step": 2178
+    },
+    {
+      "epoch": 1.7350835322195706,
+      "grad_norm": 1.3548985719680786,
+      "learning_rate": 4.721746073209893e-06,
+      "loss": 1.3098,
+      "step": 2181
+    },
+    {
+      "epoch": 1.7374701670644392,
+      "grad_norm": 1.4642560482025146,
+      "learning_rate": 4.6383874527259345e-06,
+      "loss": 1.4429,
+      "step": 2184
+    },
+    {
+      "epoch": 1.739856801909308,
+      "grad_norm": 1.3343088626861572,
+      "learning_rate": 4.555735412551975e-06,
+      "loss": 1.3698,
+      "step": 2187
+    },
+    {
+      "epoch": 1.7422434367541766,
+      "grad_norm": 1.6669230461120605,
+      "learning_rate": 4.47379124012689e-06,
+      "loss": 1.3676,
+      "step": 2190
+    },
+    {
+      "epoch": 1.7446300715990453,
+      "grad_norm": 1.513681173324585,
+      "learning_rate": 4.3925562118634135e-06,
+      "loss": 1.4566,
+      "step": 2193
+    },
+    {
+      "epoch": 1.747016706443914,
+      "grad_norm": 1.3339098691940308,
+      "learning_rate": 4.312031593128163e-06,
+      "loss": 1.513,
+      "step": 2196
+    },
+    {
+      "epoch": 1.7494033412887828,
+      "grad_norm": 1.3722683191299438,
+      "learning_rate": 4.232218638222029e-06,
+      "loss": 1.3236,
+      "step": 2199
+    },
+    {
+      "epoch": 1.7517899761336515,
+      "grad_norm": 1.3475229740142822,
+      "learning_rate": 4.153118590360561e-06,
+      "loss": 1.3351,
+      "step": 2202
+    },
+    {
+      "epoch": 1.7541766109785204,
+      "grad_norm": 1.2870583534240723,
+      "learning_rate": 4.074732681654647e-06,
+      "loss": 1.291,
+      "step": 2205
+    },
+    {
+      "epoch": 1.756563245823389,
+      "grad_norm": 1.3666460514068604,
+      "learning_rate": 3.997062133091284e-06,
+      "loss": 1.274,
+      "step": 2208
+    },
+    {
+      "epoch": 1.7589498806682577,
+      "grad_norm": 1.243913173675537,
+      "learning_rate": 3.920108154514585e-06,
+      "loss": 1.2515,
+      "step": 2211
+    },
+    {
+      "epoch": 1.7613365155131264,
+      "grad_norm": 1.3077210187911987,
+      "learning_rate": 3.843871944606969e-06,
+      "loss": 1.2814,
+      "step": 2214
+    },
+    {
+      "epoch": 1.763723150357995,
+      "grad_norm": 1.4350789785385132,
+      "learning_rate": 3.7683546908703903e-06,
+      "loss": 1.4259,
+      "step": 2217
+    },
+    {
+      "epoch": 1.766109785202864,
+      "grad_norm": 1.3633619546890259,
+      "learning_rate": 3.693557569607947e-06,
+      "loss": 1.3094,
+      "step": 2220
+    },
+    {
+      "epoch": 1.7684964200477327,
+      "grad_norm": 1.5014336109161377,
+      "learning_rate": 3.6194817459054676e-06,
+      "loss": 1.4016,
+      "step": 2223
+    },
+    {
+      "epoch": 1.7708830548926016,
+      "grad_norm": 1.2430241107940674,
+      "learning_rate": 3.5461283736134722e-06,
+      "loss": 1.3035,
+      "step": 2226
+    },
+    {
+      "epoch": 1.7732696897374702,
+      "grad_norm": 1.3036683797836304,
+      "learning_rate": 3.4734985953290778e-06,
+      "loss": 1.3666,
+      "step": 2229
+    },
+    {
+      "epoch": 1.775656324582339,
+      "grad_norm": 1.1916844844818115,
+      "learning_rate": 3.401593542378262e-06,
+      "loss": 1.364,
+      "step": 2232
+    },
+    {
+      "epoch": 1.7780429594272076,
+      "grad_norm": 1.4589468240737915,
+      "learning_rate": 3.330414334798265e-06,
+      "loss": 1.2668,
+      "step": 2235
+    },
+    {
+      "epoch": 1.7804295942720763,
+      "grad_norm": 1.426949143409729,
+      "learning_rate": 3.2599620813200837e-06,
+      "loss": 1.3789,
+      "step": 2238
+    },
+    {
+      "epoch": 1.7828162291169452,
+      "grad_norm": 1.4803996086120605,
+      "learning_rate": 3.1902378793512657e-06,
+      "loss": 1.4121,
+      "step": 2241
+    },
+    {
+      "epoch": 1.7852028639618138,
+      "grad_norm": 1.3442944288253784,
+      "learning_rate": 3.121242814958747e-06,
+      "loss": 1.3518,
+      "step": 2244
+    },
+    {
+      "epoch": 1.7875894988066827,
+      "grad_norm": 1.3079568147659302,
+      "learning_rate": 3.0529779628519992e-06,
+      "loss": 1.368,
+      "step": 2247
+    },
+    {
+      "epoch": 1.7899761336515514,
+      "grad_norm": 1.3837271928787231,
+      "learning_rate": 2.9854443863662262e-06,
+      "loss": 1.403,
+      "step": 2250
+    },
+    {
+      "epoch": 1.79236276849642,
+      "grad_norm": 2.337009906768799,
+      "learning_rate": 2.918643137445859e-06,
+      "loss": 1.3971,
+      "step": 2253
+    },
+    {
+      "epoch": 1.7947494033412887,
+      "grad_norm": 1.3745216131210327,
+      "learning_rate": 2.8525752566281482e-06,
+      "loss": 1.3739,
+      "step": 2256
+    },
+    {
+      "epoch": 1.7971360381861574,
+      "grad_norm": 1.4141366481781006,
+      "learning_rate": 2.787241773026933e-06,
+      "loss": 1.3846,
+      "step": 2259
+    },
+    {
+      "epoch": 1.799522673031026,
+      "grad_norm": 1.413260579109192,
+      "learning_rate": 2.722643704316652e-06,
+      "loss": 1.4082,
+      "step": 2262
+    },
+    {
+      "epoch": 1.801909307875895,
+      "grad_norm": 1.37075936794281,
+      "learning_rate": 2.658782056716441e-06,
+      "loss": 1.3568,
+      "step": 2265
+    },
+    {
+      "epoch": 1.8042959427207639,
+      "grad_norm": 1.3348731994628906,
+      "learning_rate": 2.5956578249745236e-06,
+      "loss": 1.3443,
+      "step": 2268
+    },
+    {
+      "epoch": 1.8066825775656326,
+      "grad_norm": 1.3601338863372803,
+      "learning_rate": 2.533271992352659e-06,
+      "loss": 1.3713,
+      "step": 2271
+    },
+    {
+      "epoch": 1.8090692124105012,
+      "grad_norm": 1.3218231201171875,
+      "learning_rate": 2.4716255306108605e-06,
+      "loss": 1.347,
+      "step": 2274
+    },
+    {
+      "epoch": 1.81145584725537,
+      "grad_norm": 1.3518648147583008,
+      "learning_rate": 2.4107193999922286e-06,
+      "loss": 1.2742,
+      "step": 2277
+    },
+    {
+      "epoch": 1.8138424821002386,
+      "grad_norm": 1.2940058708190918,
+      "learning_rate": 2.3505545492080395e-06,
+      "loss": 1.246,
+      "step": 2280
+    },
+    {
+      "epoch": 1.8162291169451072,
+      "grad_norm": 1.4600387811660767,
+      "learning_rate": 2.291131915422917e-06,
+      "loss": 1.3453,
+      "step": 2283
+    },
+    {
+      "epoch": 1.8186157517899761,
+      "grad_norm": 1.476802945137024,
+      "learning_rate": 2.2324524242402613e-06,
+      "loss": 1.3603,
+      "step": 2286
+    },
+    {
+      "epoch": 1.8210023866348448,
+      "grad_norm": 2.0956149101257324,
+      "learning_rate": 2.1745169896878414e-06,
+      "loss": 1.3493,
+      "step": 2289
+    },
+    {
+      "epoch": 1.8233890214797137,
+      "grad_norm": 1.4859145879745483,
+      "learning_rate": 2.117326514203527e-06,
+      "loss": 1.4109,
+      "step": 2292
+    },
+    {
+      "epoch": 1.8257756563245824,
+      "grad_norm": 1.400531530380249,
+      "learning_rate": 2.0608818886212576e-06,
+      "loss": 1.3619,
+      "step": 2295
+    },
+    {
+      "epoch": 1.828162291169451,
+      "grad_norm": 1.3414305448532104,
+      "learning_rate": 2.0051839921571448e-06,
+      "loss": 1.404,
+      "step": 2298
+    },
+    {
+      "epoch": 1.8305489260143197,
+      "grad_norm": 1.4340054988861084,
+      "learning_rate": 1.9502336923958255e-06,
+      "loss": 1.2401,
+      "step": 2301
+    },
+    {
+      "epoch": 1.8329355608591884,
+      "grad_norm": 1.320109486579895,
+      "learning_rate": 1.8960318452768577e-06,
+      "loss": 1.326,
+      "step": 2304
+    },
+    {
+      "epoch": 1.8353221957040573,
+      "grad_norm": 1.440313458442688,
+      "learning_rate": 1.8425792950814868e-06,
+      "loss": 1.4369,
+      "step": 2307
+    },
+    {
+      "epoch": 1.837708830548926,
+      "grad_norm": 1.357653260231018,
+      "learning_rate": 1.7898768744194162e-06,
+      "loss": 1.3528,
+      "step": 2310
+    },
+    {
+      "epoch": 1.8400954653937949,
+      "grad_norm": 1.28446364402771,
+      "learning_rate": 1.7379254042158955e-06,
+      "loss": 1.3513,
+      "step": 2313
+    },
+    {
+      "epoch": 1.8424821002386635,
+      "grad_norm": 1.5899930000305176,
+      "learning_rate": 1.6867256936989096e-06,
+      "loss": 1.4262,
+      "step": 2316
+    },
+    {
+      "epoch": 1.8448687350835322,
+      "grad_norm": 1.353194236755371,
+      "learning_rate": 1.6362785403865488e-06,
+      "loss": 1.345,
+      "step": 2319
+    },
+    {
+      "epoch": 1.847255369928401,
+      "grad_norm": 1.2713377475738525,
+      "learning_rate": 1.5865847300746417e-06,
+      "loss": 1.284,
+      "step": 2322
+    },
+    {
+      "epoch": 1.8496420047732696,
+      "grad_norm": 1.2895587682724,
+      "learning_rate": 1.5376450368244589e-06,
+      "loss": 1.3386,
+      "step": 2325
+    },
+    {
+      "epoch": 1.8520286396181385,
+      "grad_norm": 1.3775376081466675,
+      "learning_rate": 1.4894602229506892e-06,
+      "loss": 1.3587,
+      "step": 2328
+    },
+    {
+      "epoch": 1.8544152744630071,
+      "grad_norm": 1.153599739074707,
+      "learning_rate": 1.4420310390095615e-06,
+      "loss": 1.25,
+      "step": 2331
+    },
+    {
+      "epoch": 1.856801909307876,
+      "grad_norm": 1.2969733476638794,
+      "learning_rate": 1.3953582237871521e-06,
+      "loss": 1.3874,
+      "step": 2334
+    },
+    {
+      "epoch": 1.8591885441527447,
+      "grad_norm": 1.3307387828826904,
+      "learning_rate": 1.3494425042878622e-06,
+      "loss": 1.389,
+      "step": 2337
+    },
+    {
+      "epoch": 1.8615751789976134,
+      "grad_norm": 1.3975698947906494,
+      "learning_rate": 1.3042845957231153e-06,
+      "loss": 1.3198,
+      "step": 2340
+    },
+    {
+      "epoch": 1.863961813842482,
+      "grad_norm": 1.4592026472091675,
+      "learning_rate": 1.2598852015001994e-06,
+      "loss": 1.354,
+      "step": 2343
+    },
+    {
+      "epoch": 1.8663484486873507,
+      "grad_norm": 1.3790894746780396,
+      "learning_rate": 1.2162450132113201e-06,
+      "loss": 1.46,
+      "step": 2346
+    },
+    {
+      "epoch": 1.8687350835322196,
+      "grad_norm": 1.3110777139663696,
+      "learning_rate": 1.1733647106228375e-06,
+      "loss": 1.3819,
+      "step": 2349
+    },
+    {
+      "epoch": 1.8711217183770883,
+      "grad_norm": 1.447709083557129,
+      "learning_rate": 1.1312449616646403e-06,
+      "loss": 1.293,
+      "step": 2352
+    },
+    {
+      "epoch": 1.8735083532219572,
+      "grad_norm": 1.2239938974380493,
+      "learning_rate": 1.0898864224197946e-06,
+      "loss": 1.2388,
+      "step": 2355
+    },
+    {
+      "epoch": 1.8758949880668259,
+      "grad_norm": 1.3444265127182007,
+      "learning_rate": 1.049289737114273e-06,
+      "loss": 1.3817,
+      "step": 2358
+    },
+    {
+      "epoch": 1.8782816229116945,
+      "grad_norm": 1.226277232170105,
+      "learning_rate": 1.009455538106968e-06,
+      "loss": 1.2544,
+      "step": 2361
+    },
+    {
+      "epoch": 1.8806682577565632,
+      "grad_norm": 1.3254185914993286,
+      "learning_rate": 9.703844458797962e-07,
+      "loss": 1.4172,
+      "step": 2364
+    },
+    {
+      "epoch": 1.8830548926014319,
+      "grad_norm": 1.290475606918335,
+      "learning_rate": 9.320770690280645e-07,
+      "loss": 1.3941,
+      "step": 2367
+    },
+    {
+      "epoch": 1.8854415274463006,
+      "grad_norm": 1.4123435020446777,
+      "learning_rate": 8.945340042509797e-07,
+      "loss": 1.3697,
+      "step": 2370
+    },
+    {
+      "epoch": 1.8878281622911695,
+      "grad_norm": 1.3699328899383545,
+      "learning_rate": 8.577558363423554e-07,
+      "loss": 1.3808,
+      "step": 2373
+    },
+    {
+      "epoch": 1.8902147971360383,
+      "grad_norm": 1.2443058490753174,
+      "learning_rate": 8.217431381815077e-07,
+      "loss": 1.3819,
+      "step": 2376
+    },
+    {
+      "epoch": 1.892601431980907,
+      "grad_norm": 1.4468942880630493,
+      "learning_rate": 7.864964707243072e-07,
+      "loss": 1.3101,
+      "step": 2379
+    },
+    {
+      "epoch": 1.8949880668257757,
+      "grad_norm": 1.3872753381729126,
+      "learning_rate": 7.520163829944804e-07,
+      "loss": 1.2876,
+      "step": 2382
+    },
+    {
+      "epoch": 1.8973747016706444,
+      "grad_norm": 1.4598524570465088,
+      "learning_rate": 7.183034120750221e-07,
+      "loss": 1.3302,
+      "step": 2385
+    },
+    {
+      "epoch": 1.899761336515513,
+      "grad_norm": 1.4307209253311157,
+      "learning_rate": 6.85358083099863e-07,
+      "loss": 1.3474,
+      "step": 2388
+    },
+    {
+      "epoch": 1.9021479713603817,
+      "grad_norm": 1.347608208656311,
+      "learning_rate": 6.531809092456598e-07,
+      "loss": 1.2955,
+      "step": 2391
+    },
+    {
+      "epoch": 1.9045346062052506,
+      "grad_norm": 1.2689610719680786,
+      "learning_rate": 6.217723917238128e-07,
+      "loss": 1.4583,
+      "step": 2394
+    },
+    {
+      "epoch": 1.9069212410501193,
+      "grad_norm": 1.3200253248214722,
+      "learning_rate": 5.911330197726661e-07,
+      "loss": 1.2874,
+      "step": 2397
+    },
+    {
+      "epoch": 1.9093078758949882,
+      "grad_norm": 1.2589712142944336,
+      "learning_rate": 5.612632706498755e-07,
+      "loss": 1.3658,
+      "step": 2400
+    },
+    {
+      "epoch": 1.9116945107398569,
+      "grad_norm": 1.2950568199157715,
+      "learning_rate": 5.321636096249749e-07,
+      "loss": 1.3954,
+      "step": 2403
+    },
+    {
+      "epoch": 1.9140811455847255,
+      "grad_norm": 1.563720703125,
+      "learning_rate": 5.038344899721436e-07,
+      "loss": 1.326,
+      "step": 2406
+    },
+    {
+      "epoch": 1.9164677804295942,
+      "grad_norm": 1.2961969375610352,
+      "learning_rate": 4.762763529631342e-07,
+      "loss": 1.3369,
+      "step": 2409
+    },
+    {
+      "epoch": 1.9188544152744629,
+      "grad_norm": 1.4480239152908325,
+      "learning_rate": 4.4948962786039437e-07,
+      "loss": 1.3879,
+      "step": 2412
+    },
+    {
+      "epoch": 1.9212410501193318,
+      "grad_norm": 1.3828881978988647,
+      "learning_rate": 4.234747319103949e-07,
+      "loss": 1.2412,
+      "step": 2415
+    },
+    {
+      "epoch": 1.9236276849642004,
+      "grad_norm": 1.4093812704086304,
+      "learning_rate": 3.9823207033710676e-07,
+      "loss": 1.3996,
+      "step": 2418
+    },
+    {
+      "epoch": 1.9260143198090693,
+      "grad_norm": 1.438379168510437,
+      "learning_rate": 3.737620363357286e-07,
+      "loss": 1.2387,
+      "step": 2421
+    },
+    {
+      "epoch": 1.928400954653938,
+      "grad_norm": 1.32201087474823,
+      "learning_rate": 3.5006501106651937e-07,
+      "loss": 1.3328,
+      "step": 2424
+    },
+    {
+      "epoch": 1.9307875894988067,
+      "grad_norm": 1.276695966720581,
+      "learning_rate": 3.2714136364888073e-07,
+      "loss": 1.3076,
+      "step": 2427
+    },
+    {
+      "epoch": 1.9331742243436754,
+      "grad_norm": 1.8650705814361572,
+      "learning_rate": 3.0499145115561176e-07,
+      "loss": 1.4659,
+      "step": 2430
+    },
+    {
+      "epoch": 1.935560859188544,
+      "grad_norm": 1.3024852275848389,
+      "learning_rate": 2.836156186073413e-07,
+      "loss": 1.3261,
+      "step": 2433
+    },
+    {
+      "epoch": 1.937947494033413,
+      "grad_norm": 1.2863266468048096,
+      "learning_rate": 2.630141989671542e-07,
+      "loss": 1.3533,
+      "step": 2436
+    },
+    {
+      "epoch": 1.9403341288782816,
+      "grad_norm": 1.443809151649475,
+      "learning_rate": 2.431875131354011e-07,
+      "loss": 1.444,
+      "step": 2439
+    },
+    {
+      "epoch": 1.9427207637231505,
+      "grad_norm": 1.2889124155044556,
+      "learning_rate": 2.2413586994470825e-07,
+      "loss": 1.3069,
+      "step": 2442
+    },
+    {
+      "epoch": 1.9451073985680192,
+      "grad_norm": 1.3247694969177246,
+      "learning_rate": 2.0585956615515323e-07,
+      "loss": 1.3889,
+      "step": 2445
+    },
+    {
+      "epoch": 1.9474940334128878,
+      "grad_norm": 1.3477109670639038,
+      "learning_rate": 1.8835888644966325e-07,
+      "loss": 1.3446,
+      "step": 2448
+    },
+    {
+      "epoch": 1.9498806682577565,
+      "grad_norm": 1.4671467542648315,
+      "learning_rate": 1.7163410342956875e-07,
+      "loss": 1.3727,
+      "step": 2451
+    },
+    {
+      "epoch": 1.9522673031026252,
+      "grad_norm": 1.4846844673156738,
+      "learning_rate": 1.5568547761034004e-07,
+      "loss": 1.3132,
+      "step": 2454
+    },
+    {
+      "epoch": 1.9546539379474939,
+      "grad_norm": 1.4392292499542236,
+      "learning_rate": 1.4051325741756828e-07,
+      "loss": 1.4372,
+      "step": 2457
+    },
+    {
+      "epoch": 1.9570405727923628,
+      "grad_norm": 1.2357834577560425,
+      "learning_rate": 1.2611767918306316e-07,
+      "loss": 1.357,
+      "step": 2460
+    },
+    {
+      "epoch": 1.9594272076372317,
+      "grad_norm": 1.7262612581253052,
+      "learning_rate": 1.1249896714117802e-07,
+      "loss": 1.3613,
+      "step": 2463
+    },
+    {
+      "epoch": 1.9618138424821003,
+      "grad_norm": 1.2283878326416016,
+      "learning_rate": 9.965733342532924e-08,
+      "loss": 1.3145,
+      "step": 2466
+    },
+    {
+      "epoch": 1.964200477326969,
+      "grad_norm": 1.2315729856491089,
+      "learning_rate": 8.759297806469335e-08,
+      "loss": 1.254,
+      "step": 2469
+    },
+    {
+      "epoch": 1.9665871121718377,
+      "grad_norm": 1.3991883993148804,
+      "learning_rate": 7.630608898105962e-08,
+      "loss": 1.3403,
+      "step": 2472
+    },
+    {
+      "epoch": 1.9689737470167064,
+      "grad_norm": 1.2701016664505005,
+      "learning_rate": 6.579684198594338e-08,
+      "loss": 1.2927,
+      "step": 2475
+    },
+    {
+      "epoch": 1.971360381861575,
+      "grad_norm": 1.2588876485824585,
+      "learning_rate": 5.606540077782163e-08,
+      "loss": 1.2546,
+      "step": 2478
+    },
+    {
+      "epoch": 1.973747016706444,
+      "grad_norm": 1.5006442070007324,
+      "learning_rate": 4.711191693959616e-08,
+      "loss": 1.3456,
+      "step": 2481
+    },
+    {
+      "epoch": 1.9761336515513126,
+      "grad_norm": 1.3775635957717896,
+      "learning_rate": 3.893652993621766e-08,
+      "loss": 1.3456,
+      "step": 2484
+    },
+    {
+      "epoch": 1.9785202863961815,
+      "grad_norm": 1.5189213752746582,
+      "learning_rate": 3.1539367112543014e-08,
+      "loss": 1.4398,
+      "step": 2487
+    },
+    {
+      "epoch": 1.9809069212410502,
+      "grad_norm": 1.4015928506851196,
+      "learning_rate": 2.4920543691309138e-08,
+      "loss": 1.3546,
+      "step": 2490
+    },
+    {
+      "epoch": 1.9832935560859188,
+      "grad_norm": 1.2725491523742676,
+      "learning_rate": 1.9080162771378808e-08,
+      "loss": 1.3043,
+      "step": 2493
+    },
+    {
+      "epoch": 1.9856801909307875,
+      "grad_norm": 1.3217906951904297,
+      "learning_rate": 1.4018315326103094e-08,
+      "loss": 1.3286,
+      "step": 2496
+    },
+    {
+      "epoch": 1.9880668257756562,
+      "grad_norm": 1.2626640796661377,
+      "learning_rate": 9.735080201922487e-09,
+      "loss": 1.281,
+      "step": 2499
+    },
+    {
+      "epoch": 1.990453460620525,
+      "grad_norm": 1.3296139240264893,
+      "learning_rate": 6.2305241171345395e-09,
+      "loss": 1.3501,
+      "step": 2502
+    },
+    {
+      "epoch": 1.9928400954653938,
+      "grad_norm": 1.2684460878372192,
+      "learning_rate": 3.5047016608613647e-09,
+      "loss": 1.2971,
+      "step": 2505
+    },
+    {
+      "epoch": 1.9952267303102627,
+      "grad_norm": 1.4350745677947998,
+      "learning_rate": 1.5576552921836574e-09,
+      "loss": 1.3621,
+      "step": 2508
+    },
+    {
+      "epoch": 1.9976133651551313,
+      "grad_norm": 1.4221293926239014,
+      "learning_rate": 3.89415339491217e-10,
+      "loss": 1.3512,
+      "step": 2511
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6571133136749268,
+      "learning_rate": 0.0,
+      "loss": 1.3492,
+      "step": 2514
+    }
+  ],
+  "logging_steps": 3,
+  "max_steps": 2514,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1257,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.254235526619464e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}