diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -3,6155 +3,3089 @@
   "best_model_checkpoint": null,
   "epoch": 2.0,
   "eval_steps": 500,
-  "global_step": 43800,
+  "global_step": 21900,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.00228310502283105,
-      "grad_norm": 33.85030746459961,
-      "learning_rate": 2.9965753424657534e-05,
-      "loss": 4.7256,
+      "epoch": 0.0045662100456621,
+      "grad_norm": 22.38092803955078,
+      "learning_rate": 2.993150684931507e-05,
+      "loss": 5.0082,
       "step": 50
     },
     {
-      "epoch": 0.0045662100456621,
-      "grad_norm": 28.36233139038086,
-      "learning_rate": 2.993150684931507e-05,
-      "loss": 4.0093,
+      "epoch": 0.0091324200913242,
+      "grad_norm": 19.525426864624023,
+      "learning_rate": 2.9863013698630136e-05,
+      "loss": 4.0853,
       "step": 100
     },
     {
-      "epoch": 0.00684931506849315,
-      "grad_norm": 40.79988098144531,
-      "learning_rate": 2.9897260273972603e-05,
-      "loss": 3.3789,
+      "epoch": 0.0136986301369863,
+      "grad_norm": 29.21298599243164,
+      "learning_rate": 2.9794520547945206e-05,
+      "loss": 3.8879,
       "step": 150
     },
     {
-      "epoch": 0.0091324200913242,
-      "grad_norm": 38.37752914428711,
-      "learning_rate": 2.9863013698630136e-05,
-      "loss": 2.5531,
+      "epoch": 0.0182648401826484,
+      "grad_norm": 29.60974884033203,
+      "learning_rate": 2.9726027397260275e-05,
+      "loss": 3.5946,
       "step": 200
     },
     {
-      "epoch": 0.01141552511415525,
-      "grad_norm": 46.2108268737793,
-      "learning_rate": 2.9828767123287673e-05,
-      "loss": 2.4412,
+      "epoch": 0.0228310502283105,
+      "grad_norm": 20.75732421875,
+      "learning_rate": 2.9657534246575345e-05,
+      "loss": 3.1165,
       "step": 250
     },
     {
-      "epoch": 0.0136986301369863,
-      "grad_norm": 22.16556739807129,
-      "learning_rate": 2.9794520547945206e-05,
-      "loss": 2.1968,
+      "epoch": 0.0273972602739726,
+      "grad_norm": 31.49699592590332,
+      "learning_rate": 2.958904109589041e-05,
+      "loss": 2.7346,
       "step": 300
     },
     {
-      "epoch": 0.01598173515981735,
-      "grad_norm": 38.14749526977539,
-      "learning_rate": 2.9760273972602742e-05,
-      "loss": 2.0913,
+      "epoch": 0.0319634703196347,
+      "grad_norm": 43.215335845947266,
+      "learning_rate": 2.952054794520548e-05,
+      "loss": 2.3509,
       "step": 350
     },
     {
-      "epoch": 0.0182648401826484,
-      "grad_norm": 38.938079833984375,
-      "learning_rate": 2.9726027397260275e-05,
-      "loss": 2.0564,
+      "epoch": 0.0365296803652968,
+      "grad_norm": 29.553388595581055,
+      "learning_rate": 2.945205479452055e-05,
+      "loss": 2.3374,
       "step": 400
     },
     {
-      "epoch": 0.02054794520547945,
-      "grad_norm": 51.680259704589844,
-      "learning_rate": 2.969178082191781e-05,
-      "loss": 1.9041,
+      "epoch": 0.0410958904109589,
+      "grad_norm": 30.272083282470703,
+      "learning_rate": 2.938356164383562e-05,
+      "loss": 2.0551,
       "step": 450
     },
     {
-      "epoch": 0.0228310502283105,
-      "grad_norm": 23.21047019958496,
-      "learning_rate": 2.9657534246575345e-05,
-      "loss": 1.7974,
+      "epoch": 0.045662100456621,
+      "grad_norm": 23.157560348510742,
+      "learning_rate": 2.9315068493150685e-05,
+      "loss": 1.9659,
       "step": 500
     },
     {
-      "epoch": 0.02511415525114155,
-      "grad_norm": 27.936336517333984,
-      "learning_rate": 2.9623287671232878e-05,
-      "loss": 1.6824,
+      "epoch": 0.0502283105022831,
+      "grad_norm": 39.72032165527344,
+      "learning_rate": 2.9246575342465755e-05,
+      "loss": 1.7552,
       "step": 550
     },
     {
-      "epoch": 0.0273972602739726,
-      "grad_norm": 121.84676361083984,
-      "learning_rate": 2.958904109589041e-05,
-      "loss": 2.0853,
+      "epoch": 0.0547945205479452,
+      "grad_norm": 24.441036224365234,
+      "learning_rate": 2.9178082191780824e-05,
+      "loss": 1.7992,
       "step": 600
     },
     {
-      "epoch": 0.02968036529680365,
-      "grad_norm": 51.72910690307617,
-      "learning_rate": 2.9554794520547947e-05,
-      "loss": 1.8569,
+      "epoch": 0.0593607305936073,
+      "grad_norm": 32.07455825805664,
+      "learning_rate": 2.910958904109589e-05,
+      "loss": 1.6968,
       "step": 650
     },
     {
-      "epoch": 0.0319634703196347,
-      "grad_norm": 27.697223663330078,
-      "learning_rate": 2.952054794520548e-05,
-      "loss": 1.7805,
+      "epoch": 0.0639269406392694,
+      "grad_norm": 25.034423828125,
+      "learning_rate": 2.9041095890410956e-05,
+      "loss": 1.788,
       "step": 700
     },
     {
-      "epoch": 0.03424657534246575,
-      "grad_norm": 59.87869644165039,
-      "learning_rate": 2.9486301369863017e-05,
-      "loss": 1.8421,
+      "epoch": 0.0684931506849315,
+      "grad_norm": 33.06129455566406,
+      "learning_rate": 2.8972602739726026e-05,
+      "loss": 1.8157,
       "step": 750
     },
     {
-      "epoch": 0.0365296803652968,
-      "grad_norm": 20.744386672973633,
-      "learning_rate": 2.945205479452055e-05,
-      "loss": 1.8149,
+      "epoch": 0.0730593607305936,
+      "grad_norm": 32.716251373291016,
+      "learning_rate": 2.8904109589041095e-05,
+      "loss": 1.7502,
       "step": 800
     },
     {
-      "epoch": 0.03881278538812785,
-      "grad_norm": 22.272050857543945,
-      "learning_rate": 2.9417808219178083e-05,
-      "loss": 1.6892,
+      "epoch": 0.0776255707762557,
+      "grad_norm": 27.18515968322754,
+      "learning_rate": 2.8835616438356165e-05,
+      "loss": 1.5629,
       "step": 850
     },
     {
-      "epoch": 0.0410958904109589,
-      "grad_norm": 34.34733963012695,
-      "learning_rate": 2.938356164383562e-05,
-      "loss": 1.6013,
+      "epoch": 0.0821917808219178,
+      "grad_norm": 19.844303131103516,
+      "learning_rate": 2.876712328767123e-05,
+      "loss": 1.6421,
       "step": 900
     },
     {
-      "epoch": 0.04337899543378995,
-      "grad_norm": 19.482553482055664,
-      "learning_rate": 2.9349315068493152e-05,
-      "loss": 1.6772,
+      "epoch": 0.0867579908675799,
+      "grad_norm": 160.48770141601562,
+      "learning_rate": 2.86986301369863e-05,
+      "loss": 1.6035,
       "step": 950
     },
     {
-      "epoch": 0.045662100456621,
-      "grad_norm": 19.61932945251465,
-      "learning_rate": 2.9315068493150685e-05,
-      "loss": 1.7372,
+      "epoch": 0.091324200913242,
+      "grad_norm": 35.13261032104492,
+      "learning_rate": 2.863013698630137e-05,
+      "loss": 1.4939,
       "step": 1000
     },
     {
-      "epoch": 0.04794520547945205,
-      "grad_norm": 47.28583526611328,
-      "learning_rate": 2.9280821917808222e-05,
-      "loss": 1.6459,
+      "epoch": 0.0958904109589041,
+      "grad_norm": 24.28845977783203,
+      "learning_rate": 2.856164383561644e-05,
+      "loss": 1.5611,
       "step": 1050
     },
     {
-      "epoch": 0.0502283105022831,
-      "grad_norm": 74.12429809570312,
-      "learning_rate": 2.9246575342465755e-05,
-      "loss": 1.4242,
+      "epoch": 0.1004566210045662,
+      "grad_norm": 30.07063865661621,
+      "learning_rate": 2.8493150684931505e-05,
+      "loss": 1.4715,
       "step": 1100
     },
     {
-      "epoch": 0.05251141552511415,
-      "grad_norm": 28.49104118347168,
-      "learning_rate": 2.921232876712329e-05,
-      "loss": 1.4958,
+      "epoch": 0.1050228310502283,
+      "grad_norm": 23.672300338745117,
+      "learning_rate": 2.8424657534246575e-05,
+      "loss": 1.534,
       "step": 1150
     },
     {
-      "epoch": 0.0547945205479452,
-      "grad_norm": 69.21525573730469,
-      "learning_rate": 2.9178082191780824e-05,
-      "loss": 1.4848,
+      "epoch": 0.1095890410958904,
+      "grad_norm": 27.45283317565918,
+      "learning_rate": 2.8356164383561644e-05,
+      "loss": 1.5268,
       "step": 1200
     },
     {
-      "epoch": 0.05707762557077625,
-      "grad_norm": 28.827421188354492,
-      "learning_rate": 2.9143835616438357e-05,
-      "loss": 1.451,
+      "epoch": 0.1141552511415525,
+      "grad_norm": 26.83701515197754,
+      "learning_rate": 2.8287671232876714e-05,
+      "loss": 1.4089,
       "step": 1250
     },
     {
-      "epoch": 0.0593607305936073,
-      "grad_norm": 69.60295867919922,
-      "learning_rate": 2.910958904109589e-05,
-      "loss": 1.6166,
+      "epoch": 0.1187214611872146,
+      "grad_norm": 19.307844161987305,
+      "learning_rate": 2.821917808219178e-05,
+      "loss": 1.4929,
       "step": 1300
     },
     {
-      "epoch": 0.06164383561643835,
-      "grad_norm": 43.529075622558594,
-      "learning_rate": 2.9075342465753423e-05,
-      "loss": 1.5593,
+      "epoch": 0.1232876712328767,
+      "grad_norm": 14.790838241577148,
+      "learning_rate": 2.815068493150685e-05,
+      "loss": 1.4687,
       "step": 1350
     },
     {
-      "epoch": 0.0639269406392694,
-      "grad_norm": 21.382102966308594,
-      "learning_rate": 2.9041095890410956e-05,
-      "loss": 1.5083,
+      "epoch": 0.1278538812785388,
+      "grad_norm": 16.656993865966797,
+      "learning_rate": 2.808219178082192e-05,
+      "loss": 1.3456,
       "step": 1400
     },
     {
-      "epoch": 0.06621004566210045,
-      "grad_norm": 38.85325622558594,
-      "learning_rate": 2.9006849315068493e-05,
-      "loss": 1.6731,
+      "epoch": 0.1324200913242009,
+      "grad_norm": 14.602429389953613,
+      "learning_rate": 2.801369863013699e-05,
+      "loss": 1.2725,
       "step": 1450
     },
     {
-      "epoch": 0.0684931506849315,
-      "grad_norm": 31.006072998046875,
-      "learning_rate": 2.8972602739726026e-05,
-      "loss": 1.4695,
+      "epoch": 0.136986301369863,
+      "grad_norm": 23.853199005126953,
+      "learning_rate": 2.7945205479452054e-05,
+      "loss": 1.2684,
       "step": 1500
     },
     {
-      "epoch": 0.07077625570776255,
-      "grad_norm": 26.75164222717285,
-      "learning_rate": 2.8938356164383562e-05,
-      "loss": 1.574,
+      "epoch": 0.1415525114155251,
+      "grad_norm": 40.28663635253906,
+      "learning_rate": 2.7876712328767124e-05,
+      "loss": 1.2728,
       "step": 1550
     },
     {
-      "epoch": 0.0730593607305936,
-      "grad_norm": 38.76344299316406,
-      "learning_rate": 2.8904109589041095e-05,
-      "loss": 1.5674,
+      "epoch": 0.1461187214611872,
+      "grad_norm": 11.038789749145508,
+      "learning_rate": 2.7808219178082193e-05,
+      "loss": 1.3578,
       "step": 1600
     },
     {
-      "epoch": 0.07534246575342465,
-      "grad_norm": 34.11776351928711,
-      "learning_rate": 2.886986301369863e-05,
-      "loss": 1.3804,
+      "epoch": 0.1506849315068493,
+      "grad_norm": 22.868593215942383,
+      "learning_rate": 2.7739726027397263e-05,
+      "loss": 1.3941,
       "step": 1650
     },
     {
-      "epoch": 0.0776255707762557,
-      "grad_norm": 45.00776290893555,
-      "learning_rate": 2.8835616438356165e-05,
-      "loss": 1.5021,
+      "epoch": 0.1552511415525114,
+      "grad_norm": 23.89753532409668,
+      "learning_rate": 2.767123287671233e-05,
+      "loss": 1.3071,
       "step": 1700
     },
     {
-      "epoch": 0.07990867579908675,
-      "grad_norm": 85.93338775634766,
-      "learning_rate": 2.8801369863013698e-05,
-      "loss": 1.4673,
+      "epoch": 0.1598173515981735,
+      "grad_norm": 40.02177429199219,
+      "learning_rate": 2.76027397260274e-05,
+      "loss": 1.3711,
       "step": 1750
     },
     {
-      "epoch": 0.0821917808219178,
-      "grad_norm": 20.37348747253418,
-      "learning_rate": 2.876712328767123e-05,
-      "loss": 1.5315,
+      "epoch": 0.1643835616438356,
+      "grad_norm": 40.53044128417969,
+      "learning_rate": 2.7534246575342468e-05,
+      "loss": 1.1694,
       "step": 1800
     },
     {
-      "epoch": 0.08447488584474885,
-      "grad_norm": 18.73834800720215,
-      "learning_rate": 2.8732876712328767e-05,
-      "loss": 1.4074,
+      "epoch": 0.1689497716894977,
+      "grad_norm": 38.70663070678711,
+      "learning_rate": 2.7465753424657537e-05,
+      "loss": 1.2284,
       "step": 1850
     },
     {
-      "epoch": 0.0867579908675799,
-      "grad_norm": 38.19795608520508,
-      "learning_rate": 2.86986301369863e-05,
-      "loss": 1.4442,
+      "epoch": 0.1735159817351598,
+      "grad_norm": 114.71417236328125,
+      "learning_rate": 2.7397260273972603e-05,
+      "loss": 1.1634,
       "step": 1900
     },
     {
-      "epoch": 0.08904109589041095,
-      "grad_norm": 18.31586265563965,
-      "learning_rate": 2.8664383561643837e-05,
-      "loss": 1.4918,
+      "epoch": 0.1780821917808219,
+      "grad_norm": 34.596923828125,
+      "learning_rate": 2.732876712328767e-05,
+      "loss": 1.2858,
       "step": 1950
     },
     {
-      "epoch": 0.091324200913242,
-      "grad_norm": 44.28574752807617,
-      "learning_rate": 2.863013698630137e-05,
-      "loss": 1.3071,
+      "epoch": 0.182648401826484,
+      "grad_norm": 23.52661895751953,
+      "learning_rate": 2.726027397260274e-05,
+      "loss": 1.2273,
       "step": 2000
     },
     {
-      "epoch": 0.09360730593607305,
-      "grad_norm": 51.20829391479492,
-      "learning_rate": 2.8595890410958903e-05,
-      "loss": 1.6107,
+      "epoch": 0.1872146118721461,
+      "grad_norm": 22.8360652923584,
+      "learning_rate": 2.719178082191781e-05,
+      "loss": 1.2622,
       "step": 2050
     },
     {
-      "epoch": 0.0958904109589041,
-      "grad_norm": 45.16651153564453,
-      "learning_rate": 2.856164383561644e-05,
-      "loss": 1.1947,
+      "epoch": 0.1917808219178082,
+      "grad_norm": 17.08322525024414,
+      "learning_rate": 2.7123287671232875e-05,
+      "loss": 1.3012,
       "step": 2100
     },
     {
-      "epoch": 0.09817351598173515,
-      "grad_norm": 3.6301872730255127,
-      "learning_rate": 2.8527397260273972e-05,
-      "loss": 1.4133,
+      "epoch": 0.1963470319634703,
+      "grad_norm": 12.923150062561035,
+      "learning_rate": 2.7054794520547944e-05,
+      "loss": 1.3847,
       "step": 2150
     },
     {
-      "epoch": 0.1004566210045662,
-      "grad_norm": 19.277009963989258,
-      "learning_rate": 2.8493150684931505e-05,
-      "loss": 1.2874,
+      "epoch": 0.2009132420091324,
+      "grad_norm": 73.03246307373047,
+      "learning_rate": 2.6986301369863014e-05,
+      "loss": 1.3173,
       "step": 2200
     },
     {
-      "epoch": 0.10273972602739725,
-      "grad_norm": 36.330936431884766,
-      "learning_rate": 2.8458904109589042e-05,
-      "loss": 1.4955,
+      "epoch": 0.2054794520547945,
+      "grad_norm": 16.656768798828125,
+      "learning_rate": 2.6917808219178083e-05,
+      "loss": 1.1779,
       "step": 2250
     },
     {
-      "epoch": 0.1050228310502283,
-      "grad_norm": 15.057435035705566,
-      "learning_rate": 2.8424657534246575e-05,
-      "loss": 1.3309,
+      "epoch": 0.2100456621004566,
+      "grad_norm": 22.919973373413086,
+      "learning_rate": 2.684931506849315e-05,
+      "loss": 1.2497,
       "step": 2300
     },
     {
-      "epoch": 0.10730593607305935,
-      "grad_norm": 31.65122413635254,
-      "learning_rate": 2.839041095890411e-05,
-      "loss": 1.3836,
+      "epoch": 0.2146118721461187,
+      "grad_norm": 20.0987491607666,
+      "learning_rate": 2.678082191780822e-05,
+      "loss": 1.1787,
       "step": 2350
     },
     {
-      "epoch": 0.1095890410958904,
-      "grad_norm": 26.276851654052734,
-      "learning_rate": 2.8356164383561644e-05,
-      "loss": 1.4872,
+      "epoch": 0.2191780821917808,
+      "grad_norm": 84.04356384277344,
+      "learning_rate": 2.6712328767123288e-05,
+      "loss": 1.1989,
       "step": 2400
     },
     {
-      "epoch": 0.11187214611872145,
-      "grad_norm": 80.19303131103516,
-      "learning_rate": 2.8321917808219177e-05,
-      "loss": 1.4708,
+      "epoch": 0.2237442922374429,
+      "grad_norm": 12.585476875305176,
+      "learning_rate": 2.6643835616438358e-05,
+      "loss": 1.2324,
       "step": 2450
     },
     {
-      "epoch": 0.1141552511415525,
-      "grad_norm": 23.674320220947266,
-      "learning_rate": 2.8287671232876714e-05,
-      "loss": 1.1475,
+      "epoch": 0.228310502283105,
+      "grad_norm": 18.206226348876953,
+      "learning_rate": 2.6575342465753424e-05,
+      "loss": 1.2541,
       "step": 2500
     },
     {
-      "epoch": 0.11643835616438356,
-      "grad_norm": 29.5810604095459,
-      "learning_rate": 2.8253424657534247e-05,
-      "loss": 1.5903,
+      "epoch": 0.2328767123287671,
+      "grad_norm": 24.236295700073242,
+      "learning_rate": 2.6506849315068493e-05,
+      "loss": 1.1755,
       "step": 2550
     },
     {
-      "epoch": 0.1187214611872146,
-      "grad_norm": 32.49135208129883,
-      "learning_rate": 2.821917808219178e-05,
-      "loss": 1.3248,
+      "epoch": 0.2374429223744292,
+      "grad_norm": 21.91484260559082,
+      "learning_rate": 2.6438356164383563e-05,
+      "loss": 1.1436,
       "step": 2600
     },
     {
-      "epoch": 0.12100456621004566,
-      "grad_norm": 33.26783752441406,
-      "learning_rate": 2.8184931506849316e-05,
-      "loss": 1.5587,
+      "epoch": 0.2420091324200913,
+      "grad_norm": 25.741945266723633,
+      "learning_rate": 2.6369863013698632e-05,
+      "loss": 1.146,
       "step": 2650
     },
     {
-      "epoch": 0.1232876712328767,
-      "grad_norm": 34.50761795043945,
-      "learning_rate": 2.815068493150685e-05,
-      "loss": 1.3713,
+      "epoch": 0.2465753424657534,
+      "grad_norm": 21.776227951049805,
+      "learning_rate": 2.6301369863013698e-05,
+      "loss": 1.1521,
       "step": 2700
     },
     {
-      "epoch": 0.12557077625570776,
-      "grad_norm": 21.493181228637695,
-      "learning_rate": 2.8116438356164386e-05,
-      "loss": 1.2665,
+      "epoch": 0.2511415525114155,
+      "grad_norm": 30.281368255615234,
+      "learning_rate": 2.6232876712328768e-05,
+      "loss": 1.0747,
       "step": 2750
     },
     {
-      "epoch": 0.1278538812785388,
-      "grad_norm": 10.42421817779541,
-      "learning_rate": 2.808219178082192e-05,
-      "loss": 1.3406,
+      "epoch": 0.2557077625570776,
+      "grad_norm": 16.21628189086914,
+      "learning_rate": 2.6164383561643837e-05,
+      "loss": 1.2503,
       "step": 2800
     },
     {
-      "epoch": 0.13013698630136986,
-      "grad_norm": 18.451990127563477,
-      "learning_rate": 2.8047945205479452e-05,
-      "loss": 1.2502,
+      "epoch": 0.2602739726027397,
+      "grad_norm": 17.022933959960938,
+      "learning_rate": 2.6095890410958907e-05,
+      "loss": 1.1719,
       "step": 2850
     },
     {
-      "epoch": 0.1324200913242009,
-      "grad_norm": 27.949146270751953,
-      "learning_rate": 2.801369863013699e-05,
-      "loss": 1.2492,
+      "epoch": 0.2648401826484018,
+      "grad_norm": 17.665477752685547,
+      "learning_rate": 2.6027397260273973e-05,
+      "loss": 1.091,
       "step": 2900
     },
     {
-      "epoch": 0.13470319634703196,
-      "grad_norm": 19.943056106567383,
-      "learning_rate": 2.797945205479452e-05,
-      "loss": 1.4071,
+      "epoch": 0.2694063926940639,
+      "grad_norm": 12.710230827331543,
+      "learning_rate": 2.5958904109589042e-05,
+      "loss": 1.1828,
       "step": 2950
     },
     {
-      "epoch": 0.136986301369863,
-      "grad_norm": 30.464204788208008,
-      "learning_rate": 2.7945205479452054e-05,
-      "loss": 1.2135,
+      "epoch": 0.273972602739726,
+      "grad_norm": 23.11914825439453,
+      "learning_rate": 2.589041095890411e-05,
+      "loss": 1.0025,
       "step": 3000
     },
     {
-      "epoch": 0.13926940639269406,
-      "grad_norm": 20.832355499267578,
-      "learning_rate": 2.791095890410959e-05,
-      "loss": 1.2872,
+      "epoch": 0.2785388127853881,
+      "grad_norm": 17.97935676574707,
+      "learning_rate": 2.582191780821918e-05,
+      "loss": 1.2145,
       "step": 3050
     },
     {
-      "epoch": 0.1415525114155251,
-      "grad_norm": 13.37592601776123,
-      "learning_rate": 2.7876712328767124e-05,
-      "loss": 1.1165,
+      "epoch": 0.2831050228310502,
+      "grad_norm": 28.123554229736328,
+      "learning_rate": 2.5753424657534247e-05,
+      "loss": 1.2629,
       "step": 3100
     },
     {
-      "epoch": 0.14383561643835616,
-      "grad_norm": 36.62568664550781,
-      "learning_rate": 2.784246575342466e-05,
-      "loss": 1.2849,
+      "epoch": 0.2876712328767123,
+      "grad_norm": 17.924665451049805,
+      "learning_rate": 2.5684931506849317e-05,
+      "loss": 1.0439,
       "step": 3150
     },
     {
-      "epoch": 0.1461187214611872,
-      "grad_norm": 11.884871482849121,
-      "learning_rate": 2.7808219178082193e-05,
-      "loss": 1.241,
+      "epoch": 0.2922374429223744,
+      "grad_norm": 24.03557777404785,
+      "learning_rate": 2.5616438356164386e-05,
+      "loss": 1.127,
       "step": 3200
     },
     {
-      "epoch": 0.14840182648401826,
-      "grad_norm": 22.552181243896484,
-      "learning_rate": 2.7773972602739726e-05,
-      "loss": 1.1695,
+      "epoch": 0.2968036529680365,
+      "grad_norm": 30.70620346069336,
+      "learning_rate": 2.5547945205479452e-05,
+      "loss": 1.1464,
       "step": 3250
     },
     {
-      "epoch": 0.1506849315068493,
-      "grad_norm": 29.05255126953125,
-      "learning_rate": 2.7739726027397263e-05,
-      "loss": 1.3266,
+      "epoch": 0.3013698630136986,
+      "grad_norm": 21.80393409729004,
+      "learning_rate": 2.5479452054794518e-05,
+      "loss": 0.9603,
       "step": 3300
     },
     {
-      "epoch": 0.15296803652968036,
-      "grad_norm": 27.039731979370117,
-      "learning_rate": 2.7705479452054796e-05,
-      "loss": 1.4528,
+      "epoch": 0.3059360730593607,
+      "grad_norm": 23.686079025268555,
+      "learning_rate": 2.5410958904109588e-05,
+      "loss": 1.0745,
       "step": 3350
     },
     {
-      "epoch": 0.1552511415525114,
-      "grad_norm": 12.077860832214355,
-      "learning_rate": 2.767123287671233e-05,
-      "loss": 1.0587,
+      "epoch": 0.3105022831050228,
+      "grad_norm": 10.258773803710938,
+      "learning_rate": 2.5342465753424657e-05,
+      "loss": 1.0776,
       "step": 3400
     },
     {
-      "epoch": 0.15753424657534246,
-      "grad_norm": 33.64236068725586,
-      "learning_rate": 2.7636986301369865e-05,
-      "loss": 1.3053,
+      "epoch": 0.3150684931506849,
+      "grad_norm": 28.77837562561035,
+      "learning_rate": 2.5273972602739727e-05,
+      "loss": 1.1781,
       "step": 3450
     },
     {
-      "epoch": 0.1598173515981735,
-      "grad_norm": 7.781187534332275,
-      "learning_rate": 2.76027397260274e-05,
-      "loss": 1.5692,
+      "epoch": 0.319634703196347,
+      "grad_norm": 39.38608932495117,
+      "learning_rate": 2.5205479452054793e-05,
+      "loss": 0.9749,
       "step": 3500
     },
     {
-      "epoch": 0.16210045662100456,
-      "grad_norm": 58.28102111816406,
-      "learning_rate": 2.7568493150684935e-05,
-      "loss": 1.0579,
+      "epoch": 0.3242009132420091,
+      "grad_norm": 29.596742630004883,
+      "learning_rate": 2.5136986301369862e-05,
+      "loss": 1.1099,
       "step": 3550
     },
     {
-      "epoch": 0.1643835616438356,
-      "grad_norm": 46.713260650634766,
-      "learning_rate": 2.7534246575342468e-05,
-      "loss": 1.1327,
+      "epoch": 0.3287671232876712,
+      "grad_norm": 15.371848106384277,
+      "learning_rate": 2.5068493150684932e-05,
+      "loss": 1.1363,
       "step": 3600
     },
     {
-      "epoch": 0.16666666666666666,
-      "grad_norm": 38.657493591308594,
-      "learning_rate": 2.75e-05,
-      "loss": 1.299,
+      "epoch": 0.3333333333333333,
+      "grad_norm": 27.008398056030273,
+      "learning_rate": 2.5e-05,
+      "loss": 1.2275,
       "step": 3650
     },
     {
-      "epoch": 0.1689497716894977,
-      "grad_norm": 7.117087364196777,
-      "learning_rate": 2.7465753424657537e-05,
-      "loss": 1.0367,
+      "epoch": 0.3378995433789954,
+      "grad_norm": 54.85331726074219,
+      "learning_rate": 2.4931506849315067e-05,
+      "loss": 1.2297,
       "step": 3700
     },
     {
-      "epoch": 0.17123287671232876,
-      "grad_norm": 33.10891342163086,
-      "learning_rate": 2.743150684931507e-05,
-      "loss": 1.1318,
+      "epoch": 0.3424657534246575,
+      "grad_norm": 10.650010108947754,
+      "learning_rate": 2.4863013698630137e-05,
+      "loss": 1.1716,
       "step": 3750
     },
     {
-      "epoch": 0.1735159817351598,
-      "grad_norm": 38.68271255493164,
-      "learning_rate": 2.7397260273972603e-05,
-      "loss": 1.1906,
+      "epoch": 0.3470319634703196,
+      "grad_norm": 21.537841796875,
+      "learning_rate": 2.4794520547945206e-05,
+      "loss": 1.2375,
       "step": 3800
     },
     {
-      "epoch": 0.17579908675799086,
-      "grad_norm": 45.01628494262695,
-      "learning_rate": 2.736301369863014e-05,
-      "loss": 1.1285,
+      "epoch": 0.3515981735159817,
+      "grad_norm": 31.040218353271484,
+      "learning_rate": 2.4726027397260276e-05,
+      "loss": 1.0511,
       "step": 3850
     },
     {
-      "epoch": 0.1780821917808219,
-      "grad_norm": 20.761018753051758,
-      "learning_rate": 2.732876712328767e-05,
-      "loss": 1.4117,
+      "epoch": 0.3561643835616438,
+      "grad_norm": 13.31843090057373,
+      "learning_rate": 2.4657534246575342e-05,
+      "loss": 1.1581,
       "step": 3900
     },
     {
-      "epoch": 0.18036529680365296,
-      "grad_norm": 46.919952392578125,
-      "learning_rate": 2.7294520547945206e-05,
-      "loss": 1.0591,
+      "epoch": 0.3607305936073059,
+      "grad_norm": 22.53802490234375,
+      "learning_rate": 2.458904109589041e-05,
+      "loss": 1.0627,
       "step": 3950
     },
     {
-      "epoch": 0.182648401826484,
-      "grad_norm": 78.8492431640625,
-      "learning_rate": 2.726027397260274e-05,
-      "loss": 1.3741,
+      "epoch": 0.365296803652968,
+      "grad_norm": 20.266468048095703,
+      "learning_rate": 2.452054794520548e-05,
+      "loss": 1.1103,
       "step": 4000
     },
     {
-      "epoch": 0.18493150684931506,
-      "grad_norm": 25.57372283935547,
-      "learning_rate": 2.7226027397260272e-05,
-      "loss": 1.1287,
+      "epoch": 0.3698630136986301,
+      "grad_norm": 9.198668479919434,
+      "learning_rate": 2.445205479452055e-05,
+      "loss": 1.136,
       "step": 4050
     },
     {
-      "epoch": 0.1872146118721461,
-      "grad_norm": 16.47197723388672,
-      "learning_rate": 2.719178082191781e-05,
-      "loss": 1.3216,
+      "epoch": 0.3744292237442922,
+      "grad_norm": 13.770018577575684,
+      "learning_rate": 2.4383561643835616e-05,
+      "loss": 1.0984,
       "step": 4100
     },
     {
-      "epoch": 0.18949771689497716,
-      "grad_norm": 36.24203109741211,
-      "learning_rate": 2.715753424657534e-05,
-      "loss": 1.3095,
+      "epoch": 0.3789954337899543,
+      "grad_norm": 25.558441162109375,
+      "learning_rate": 2.4315068493150686e-05,
+      "loss": 1.062,
       "step": 4150
     },
     {
-      "epoch": 0.1917808219178082,
-      "grad_norm": 8.002535820007324,
-      "learning_rate": 2.7123287671232875e-05,
-      "loss": 1.1426,
+      "epoch": 0.3835616438356164,
+      "grad_norm": 25.940107345581055,
+      "learning_rate": 2.4246575342465755e-05,
+      "loss": 1.2237,
       "step": 4200
     },
     {
-      "epoch": 0.19406392694063926,
-      "grad_norm": 9.112885475158691,
-      "learning_rate": 2.708904109589041e-05,
-      "loss": 1.4292,
+      "epoch": 0.3881278538812785,
+      "grad_norm": 11.530945777893066,
+      "learning_rate": 2.4178082191780825e-05,
+      "loss": 1.0717,
       "step": 4250
     },
     {
-      "epoch": 0.1963470319634703,
-      "grad_norm": 7.080036640167236,
-      "learning_rate": 2.7054794520547944e-05,
-      "loss": 1.2665,
+      "epoch": 0.3926940639269406,
+      "grad_norm": 13.686588287353516,
+      "learning_rate": 2.410958904109589e-05,
+      "loss": 1.0011,
       "step": 4300
     },
     {
-      "epoch": 0.19863013698630136,
-      "grad_norm": 8.403115272521973,
-      "learning_rate": 2.702054794520548e-05,
-      "loss": 1.3686,
+      "epoch": 0.3972602739726027,
+      "grad_norm": 24.325557708740234,
+      "learning_rate": 2.404109589041096e-05,
+      "loss": 1.2033,
       "step": 4350
     },
     {
-      "epoch": 0.2009132420091324,
-      "grad_norm": 29.843015670776367,
-      "learning_rate": 2.6986301369863014e-05,
-      "loss": 1.0842,
+      "epoch": 0.4018264840182648,
+      "grad_norm": 24.21858787536621,
+      "learning_rate": 2.397260273972603e-05,
+      "loss": 1.01,
       "step": 4400
     },
     {
-      "epoch": 0.20319634703196346,
-      "grad_norm": 16.962310791015625,
-      "learning_rate": 2.6952054794520547e-05,
-      "loss": 1.1983,
+      "epoch": 0.4063926940639269,
+      "grad_norm": 14.875985145568848,
+      "learning_rate": 2.39041095890411e-05,
+      "loss": 1.0378,
       "step": 4450
     },
     {
-      "epoch": 0.2054794520547945,
-      "grad_norm": 19.98363494873047,
-      "learning_rate": 2.6917808219178083e-05,
-      "loss": 1.136,
+      "epoch": 0.410958904109589,
+      "grad_norm": 38.33986282348633,
+      "learning_rate": 2.3835616438356165e-05,
+      "loss": 1.0722,
       "step": 4500
     },
     {
-      "epoch": 0.20776255707762556,
-      "grad_norm": 23.507699966430664,
-      "learning_rate": 2.6883561643835616e-05,
-      "loss": 1.4167,
+      "epoch": 0.4155251141552511,
+      "grad_norm": 10.655511856079102,
+      "learning_rate": 2.376712328767123e-05,
+      "loss": 1.1182,
       "step": 4550
     },
     {
-      "epoch": 0.2100456621004566,
-      "grad_norm": 85.11144256591797,
-      "learning_rate": 2.684931506849315e-05,
-      "loss": 1.0667,
+      "epoch": 0.4200913242009132,
+      "grad_norm": 26.336122512817383,
+      "learning_rate": 2.36986301369863e-05,
+      "loss": 1.0772,
       "step": 4600
     },
     {
-      "epoch": 0.21232876712328766,
-      "grad_norm": 25.13855743408203,
-      "learning_rate": 2.6815068493150686e-05,
-      "loss": 1.1755,
+      "epoch": 0.4246575342465753,
+      "grad_norm": 26.254093170166016,
+      "learning_rate": 2.363013698630137e-05,
+      "loss": 0.9908,
       "step": 4650
     },
     {
-      "epoch": 0.2146118721461187,
-      "grad_norm": 21.816020965576172,
-      "learning_rate": 2.678082191780822e-05,
-      "loss": 1.1357,
+      "epoch": 0.4292237442922374,
+      "grad_norm": 7.081357002258301,
+      "learning_rate": 2.3561643835616436e-05,
+      "loss": 1.0079,
       "step": 4700
     },
     {
-      "epoch": 0.21689497716894976,
-      "grad_norm": 27.655017852783203,
-      "learning_rate": 2.6746575342465755e-05,
-      "loss": 1.2578,
+      "epoch": 0.4337899543378995,
+      "grad_norm": 20.019088745117188,
+      "learning_rate": 2.3493150684931506e-05,
+      "loss": 1.1188,
       "step": 4750
     },
     {
-      "epoch": 0.2191780821917808,
-      "grad_norm": 24.329933166503906,
-      "learning_rate": 2.6712328767123288e-05,
-      "loss": 1.1243,
+      "epoch": 0.4383561643835616,
+      "grad_norm": 30.188098907470703,
+      "learning_rate": 2.3424657534246575e-05,
+      "loss": 1.0616,
       "step": 4800
     },
     {
-      "epoch": 0.22146118721461186,
-      "grad_norm": 9.320807456970215,
-      "learning_rate": 2.667808219178082e-05,
-      "loss": 1.2211,
+      "epoch": 0.4429223744292237,
+      "grad_norm": 6.621674537658691,
+      "learning_rate": 2.3356164383561645e-05,
+      "loss": 1.0935,
       "step": 4850
     },
     {
-      "epoch": 0.2237442922374429,
-      "grad_norm": 14.245413780212402,
-      "learning_rate": 2.6643835616438358e-05,
-      "loss": 1.2804,
+      "epoch": 0.4474885844748858,
+      "grad_norm": 21.145673751831055,
+      "learning_rate": 2.328767123287671e-05,
+      "loss": 1.0178,
       "step": 4900
     },
     {
-      "epoch": 0.22602739726027396,
-      "grad_norm": 46.82653045654297,
-      "learning_rate": 2.660958904109589e-05,
-      "loss": 1.2949,
+      "epoch": 0.4520547945205479,
+      "grad_norm": 26.22977066040039,
+      "learning_rate": 2.321917808219178e-05,
+      "loss": 1.1739,
       "step": 4950
     },
     {
-      "epoch": 0.228310502283105,
-      "grad_norm": 43.06188201904297,
-      "learning_rate": 2.6575342465753424e-05,
-      "loss": 1.2012,
+      "epoch": 0.45662100456621,
+      "grad_norm": 22.71933364868164,
+      "learning_rate": 2.315068493150685e-05,
+      "loss": 1.122,
       "step": 5000
     },
     {
-      "epoch": 0.23059360730593606,
-      "grad_norm": 50.084381103515625,
-      "learning_rate": 2.654109589041096e-05,
-      "loss": 1.2965,
+      "epoch": 0.4611872146118721,
+      "grad_norm": 16.020483016967773,
+      "learning_rate": 2.308219178082192e-05,
+      "loss": 1.0494,
       "step": 5050
     },
     {
-      "epoch": 0.2328767123287671,
-      "grad_norm": 32.71940994262695,
-      "learning_rate": 2.6506849315068493e-05,
-      "loss": 1.0847,
+      "epoch": 0.4657534246575342,
+      "grad_norm": 10.217668533325195,
+      "learning_rate": 2.3013698630136985e-05,
+      "loss": 1.1577,
       "step": 5100
     },
     {
-      "epoch": 0.23515981735159816,
-      "grad_norm": 53.56361389160156,
-      "learning_rate": 2.647260273972603e-05,
-      "loss": 1.2125,
+      "epoch": 0.4703196347031963,
+      "grad_norm": 13.561128616333008,
+      "learning_rate": 2.2945205479452055e-05,
+      "loss": 1.1065,
       "step": 5150
     },
     {
-      "epoch": 0.2374429223744292,
-      "grad_norm": 57.06380081176758,
-      "learning_rate": 2.6438356164383563e-05,
-      "loss": 1.06,
+      "epoch": 0.4748858447488584,
+      "grad_norm": 24.07544708251953,
+      "learning_rate": 2.2876712328767124e-05,
+      "loss": 1.1721,
       "step": 5200
     },
     {
-      "epoch": 0.23972602739726026,
-      "grad_norm": 41.62690734863281,
-      "learning_rate": 2.6404109589041096e-05,
-      "loss": 1.1048,
+      "epoch": 0.4794520547945205,
+      "grad_norm": 27.471532821655273,
+      "learning_rate": 2.2808219178082194e-05,
+      "loss": 1.0425,
       "step": 5250
     },
     {
-      "epoch": 0.2420091324200913,
-      "grad_norm": 44.06789016723633,
-      "learning_rate": 2.6369863013698632e-05,
-      "loss": 1.2219,
+      "epoch": 0.4840182648401826,
+      "grad_norm": 39.17247772216797,
+      "learning_rate": 2.273972602739726e-05,
+      "loss": 1.0182,
       "step": 5300
     },
     {
-      "epoch": 0.24429223744292236,
-      "grad_norm": 14.321037292480469,
-      "learning_rate": 2.6335616438356165e-05,
-      "loss": 1.2137,
+      "epoch": 0.4885844748858447,
+      "grad_norm": 11.253133773803711,
+      "learning_rate": 2.267123287671233e-05,
+      "loss": 1.0516,
       "step": 5350
     },
     {
-      "epoch": 0.2465753424657534,
-      "grad_norm": 31.366552352905273,
-      "learning_rate": 2.6301369863013698e-05,
-      "loss": 1.2281,
+      "epoch": 0.4931506849315068,
+      "grad_norm": 19.816268920898438,
+      "learning_rate": 2.26027397260274e-05,
+      "loss": 0.9359,
       "step": 5400
     },
     {
-      "epoch": 0.24885844748858446,
-      "grad_norm": 19.642141342163086,
-      "learning_rate": 2.6267123287671235e-05,
-      "loss": 1.0646,
+      "epoch": 0.4977168949771689,
+      "grad_norm": 21.89589500427246,
+      "learning_rate": 2.253424657534247e-05,
+      "loss": 1.0047,
       "step": 5450
     },
     {
-      "epoch": 0.2511415525114155,
-      "grad_norm": 24.77635955810547,
-      "learning_rate": 2.6232876712328768e-05,
-      "loss": 1.0827,
+      "epoch": 0.502283105022831,
+      "grad_norm": 18.662717819213867,
+      "learning_rate": 2.2465753424657534e-05,
+      "loss": 1.1092,
       "step": 5500
     },
     {
-      "epoch": 0.2534246575342466,
-      "grad_norm": 21.257320404052734,
-      "learning_rate": 2.6198630136986304e-05,
-      "loss": 1.1798,
+      "epoch": 0.5068493150684932,
+      "grad_norm": 13.65588092803955,
+      "learning_rate": 2.2397260273972604e-05,
+      "loss": 1.2777,
       "step": 5550
     },
     {
-      "epoch": 0.2557077625570776,
-      "grad_norm": 7.442146301269531,
-      "learning_rate": 2.6164383561643837e-05,
-      "loss": 1.2409,
+      "epoch": 0.5114155251141552,
+      "grad_norm": 22.897354125976562,
+      "learning_rate": 2.2328767123287673e-05,
+      "loss": 1.0557,
       "step": 5600
     },
     {
-      "epoch": 0.2579908675799087,
-      "grad_norm": 40.682579040527344,
-      "learning_rate": 2.613013698630137e-05,
-      "loss": 1.2001,
+      "epoch": 0.5159817351598174,
+      "grad_norm": 22.30970573425293,
+      "learning_rate": 2.2260273972602743e-05,
+      "loss": 1.0701,
       "step": 5650
     },
     {
-      "epoch": 0.2602739726027397,
-      "grad_norm": 22.984914779663086,
-      "learning_rate": 2.6095890410958907e-05,
-      "loss": 1.0922,
+      "epoch": 0.5205479452054794,
+      "grad_norm": 55.639892578125,
+      "learning_rate": 2.219178082191781e-05,
+      "loss": 1.1001,
       "step": 5700
     },
     {
-      "epoch": 0.2625570776255708,
-      "grad_norm": 14.088912010192871,
-      "learning_rate": 2.606164383561644e-05,
-      "loss": 1.0361,
+      "epoch": 0.5251141552511416,
+      "grad_norm": 11.327408790588379,
+      "learning_rate": 2.212328767123288e-05,
+      "loss": 1.1506,
       "step": 5750
     },
     {
-      "epoch": 0.2648401826484018,
-      "grad_norm": 26.859743118286133,
-      "learning_rate": 2.6027397260273973e-05,
-      "loss": 1.3059,
+      "epoch": 0.5296803652968036,
+      "grad_norm": 31.288894653320312,
+      "learning_rate": 2.2054794520547945e-05,
+      "loss": 1.0774,
       "step": 5800
     },
     {
-      "epoch": 0.2671232876712329,
-      "grad_norm": 27.274446487426758,
-      "learning_rate": 2.599315068493151e-05,
-      "loss": 1.2146,
+      "epoch": 0.5342465753424658,
+      "grad_norm": 11.587857246398926,
+      "learning_rate": 2.1986301369863014e-05,
+      "loss": 1.0175,
       "step": 5850
     },
     {
-      "epoch": 0.2694063926940639,
-      "grad_norm": 15.877359390258789,
-      "learning_rate": 2.5958904109589042e-05,
-      "loss": 1.047,
+      "epoch": 0.5388127853881278,
+      "grad_norm": 27.174907684326172,
+      "learning_rate": 2.191780821917808e-05,
+      "loss": 1.1566,
       "step": 5900
     },
     {
-      "epoch": 0.271689497716895,
-      "grad_norm": 27.11214256286621,
-      "learning_rate": 2.592465753424658e-05,
-      "loss": 1.1303,
+      "epoch": 0.54337899543379,
+      "grad_norm": 28.325786590576172,
+      "learning_rate": 2.184931506849315e-05,
+      "loss": 1.1089,
       "step": 5950
     },
     {
-      "epoch": 0.273972602739726,
-      "grad_norm": 25.463417053222656,
-      "learning_rate": 2.589041095890411e-05,
-      "loss": 0.9916,
+      "epoch": 0.547945205479452,
+      "grad_norm": 21.70798683166504,
+      "learning_rate": 2.178082191780822e-05,
+      "loss": 1.0575,
       "step": 6000
     },
     {
-      "epoch": 0.2762557077625571,
-      "grad_norm": 5.419765472412109,
-      "learning_rate": 2.5856164383561645e-05,
-      "loss": 1.3228,
+      "epoch": 0.5525114155251142,
+      "grad_norm": 18.79369354248047,
+      "learning_rate": 2.171232876712329e-05,
+      "loss": 1.1327,
       "step": 6050
     },
     {
-      "epoch": 0.2785388127853881,
-      "grad_norm": 26.771987915039062,
-      "learning_rate": 2.582191780821918e-05,
-      "loss": 1.1163,
+      "epoch": 0.5570776255707762,
+      "grad_norm": 31.291170120239258,
+      "learning_rate": 2.1643835616438355e-05,
+      "loss": 1.1412,
       "step": 6100
     },
     {
-      "epoch": 0.2808219178082192,
-      "grad_norm": 40.04314422607422,
-      "learning_rate": 2.5787671232876714e-05,
-      "loss": 1.2227,
+      "epoch": 0.5616438356164384,
+      "grad_norm": 15.10459041595459,
+      "learning_rate": 2.1575342465753424e-05,
+      "loss": 1.1289,
       "step": 6150
     },
     {
-      "epoch": 0.2831050228310502,
-      "grad_norm": 31.262903213500977,
-      "learning_rate": 2.5753424657534247e-05,
-      "loss": 1.0956,
+      "epoch": 0.5662100456621004,
+      "grad_norm": 23.595136642456055,
+      "learning_rate": 2.1506849315068494e-05,
+      "loss": 0.9419,
       "step": 6200
     },
     {
-      "epoch": 0.2853881278538813,
-      "grad_norm": 8.211708068847656,
-      "learning_rate": 2.5719178082191784e-05,
-      "loss": 0.8686,
+      "epoch": 0.5707762557077626,
+      "grad_norm": 9.336952209472656,
+      "learning_rate": 2.1438356164383563e-05,
+      "loss": 0.9357,
       "step": 6250
     },
     {
-      "epoch": 0.2876712328767123,
-      "grad_norm": 7.24912166595459,
-      "learning_rate": 2.5684931506849317e-05,
-      "loss": 1.2297,
+      "epoch": 0.5753424657534246,
+      "grad_norm": 34.959354400634766,
+      "learning_rate": 2.136986301369863e-05,
+      "loss": 1.1525,
       "step": 6300
     },
     {
-      "epoch": 0.2899543378995434,
-      "grad_norm": 11.469908714294434,
-      "learning_rate": 2.5650684931506853e-05,
-      "loss": 1.1037,
+      "epoch": 0.5799086757990868,
+      "grad_norm": 25.974088668823242,
+      "learning_rate": 2.13013698630137e-05,
+      "loss": 1.0915,
       "step": 6350
     },
     {
-      "epoch": 0.2922374429223744,
-      "grad_norm": 11.30005168914795,
-      "learning_rate": 2.5616438356164386e-05,
-      "loss": 1.1043,
+      "epoch": 0.5844748858447488,
+      "grad_norm": 26.807876586914062,
+      "learning_rate": 2.1232876712328768e-05,
+      "loss": 1.0783,
       "step": 6400
     },
     {
-      "epoch": 0.2945205479452055,
-      "grad_norm": 9.771041870117188,
-      "learning_rate": 2.558219178082192e-05,
-      "loss": 1.09,
+      "epoch": 0.589041095890411,
+      "grad_norm": 11.836869239807129,
+      "learning_rate": 2.1164383561643838e-05,
+      "loss": 1.1106,
       "step": 6450
     },
     {
-      "epoch": 0.2968036529680365,
-      "grad_norm": 22.025493621826172,
-      "learning_rate": 2.5547945205479452e-05,
-      "loss": 1.1252,
+      "epoch": 0.593607305936073,
+      "grad_norm": 20.5205078125,
+      "learning_rate": 2.1095890410958904e-05,
+      "loss": 1.0021,
       "step": 6500
     },
     {
-      "epoch": 0.2990867579908676,
-      "grad_norm": 22.842763900756836,
-      "learning_rate": 2.5513698630136985e-05,
-      "loss": 1.0219,
+      "epoch": 0.5981735159817352,
+      "grad_norm": 63.918914794921875,
+      "learning_rate": 2.1027397260273973e-05,
+      "loss": 0.9942,
       "step": 6550
     },
     {
-      "epoch": 0.3013698630136986,
-      "grad_norm": 49.29541015625,
-      "learning_rate": 2.5479452054794518e-05,
-      "loss": 0.8537,
+      "epoch": 0.6027397260273972,
+      "grad_norm": 15.824874877929688,
+      "learning_rate": 2.0958904109589043e-05,
+      "loss": 1.0382,
       "step": 6600
     },
     {
-      "epoch": 0.3036529680365297,
-      "grad_norm": 32.22005844116211,
-      "learning_rate": 2.5445205479452055e-05,
-      "loss": 1.1211,
+      "epoch": 0.6073059360730594,
+      "grad_norm": 23.539623260498047,
+      "learning_rate": 2.0890410958904112e-05,
+      "loss": 1.1323,
       "step": 6650
     },
     {
-      "epoch": 0.3059360730593607,
-      "grad_norm": 49.50102615356445,
-      "learning_rate": 2.5410958904109588e-05,
-      "loss": 1.0462,
+      "epoch": 0.6118721461187214,
+      "grad_norm": 22.72905921936035,
+      "learning_rate": 2.0821917808219178e-05,
+      "loss": 1.096,
       "step": 6700
     },
     {
-      "epoch": 0.3082191780821918,
-      "grad_norm": 69.23637390136719,
-      "learning_rate": 2.5376712328767124e-05,
-      "loss": 0.9964,
+      "epoch": 0.6164383561643836,
+      "grad_norm": 21.393850326538086,
+      "learning_rate": 2.0753424657534248e-05,
+      "loss": 1.0491,
       "step": 6750
     },
     {
-      "epoch": 0.3105022831050228,
-      "grad_norm": 15.762770652770996,
-      "learning_rate": 2.5342465753424657e-05,
-      "loss": 1.3032,
+      "epoch": 0.6210045662100456,
+      "grad_norm": 10.425342559814453,
+      "learning_rate": 2.0684931506849317e-05,
+      "loss": 0.9894,
       "step": 6800
     },
     {
-      "epoch": 0.3127853881278539,
-      "grad_norm": 36.22660827636719,
-      "learning_rate": 2.530821917808219e-05,
-      "loss": 1.1776,
+      "epoch": 0.6255707762557078,
+      "grad_norm": 31.147842407226562,
+      "learning_rate": 2.0616438356164387e-05,
+      "loss": 1.0229,
       "step": 6850
     },
     {
-      "epoch": 0.3150684931506849,
-      "grad_norm": 6.89371395111084,
-      "learning_rate": 2.5273972602739727e-05,
-      "loss": 1.1224,
+      "epoch": 0.6301369863013698,
+      "grad_norm": 13.265896797180176,
+      "learning_rate": 2.0547945205479453e-05,
+      "loss": 1.0372,
       "step": 6900
     },
     {
-      "epoch": 0.317351598173516,
-      "grad_norm": 76.80335235595703,
-      "learning_rate": 2.523972602739726e-05,
-      "loss": 0.9386,
+      "epoch": 0.634703196347032,
+      "grad_norm": 19.299884796142578,
+      "learning_rate": 2.0479452054794522e-05,
+      "loss": 0.9861,
       "step": 6950
     },
     {
-      "epoch": 0.319634703196347,
-      "grad_norm": 50.3594970703125,
-      "learning_rate": 2.5205479452054793e-05,
-      "loss": 1.0721,
+      "epoch": 0.639269406392694,
+      "grad_norm": 15.215560913085938,
+      "learning_rate": 2.041095890410959e-05,
+      "loss": 1.0685,
       "step": 7000
     },
     {
-      "epoch": 0.3219178082191781,
-      "grad_norm": 26.805213928222656,
-      "learning_rate": 2.517123287671233e-05,
-      "loss": 1.2563,
+      "epoch": 0.6438356164383562,
+      "grad_norm": 12.113781929016113,
+      "learning_rate": 2.034246575342466e-05,
+      "loss": 1.0452,
       "step": 7050
     },
     {
-      "epoch": 0.3242009132420091,
-      "grad_norm": 39.83240509033203,
-      "learning_rate": 2.5136986301369862e-05,
-      "loss": 1.0924,
+      "epoch": 0.6484018264840182,
+      "grad_norm": 53.0004768371582,
+      "learning_rate": 2.0273972602739724e-05,
+      "loss": 1.0177,
       "step": 7100
     },
     {
-      "epoch": 0.3264840182648402,
-      "grad_norm": 41.219818115234375,
-      "learning_rate": 2.51027397260274e-05,
-      "loss": 1.0771,
+      "epoch": 0.6529680365296804,
+      "grad_norm": 10.609742164611816,
+      "learning_rate": 2.0205479452054793e-05,
+      "loss": 0.9896,
       "step": 7150
     },
     {
-      "epoch": 0.3287671232876712,
-      "grad_norm": 63.68681335449219,
-      "learning_rate": 2.5068493150684932e-05,
-      "loss": 1.1793,
+      "epoch": 0.6575342465753424,
+      "grad_norm": 13.45383071899414,
+      "learning_rate": 2.0136986301369863e-05,
+      "loss": 0.7549,
       "step": 7200
     },
     {
-      "epoch": 0.3310502283105023,
-      "grad_norm": 96.54421997070312,
-      "learning_rate": 2.5034246575342465e-05,
-      "loss": 1.185,
+      "epoch": 0.6621004566210046,
+      "grad_norm": 21.812686920166016,
+      "learning_rate": 2.0068493150684932e-05,
+      "loss": 1.0402,
       "step": 7250
     },
     {
-      "epoch": 0.3333333333333333,
-      "grad_norm": 35.43101119995117,
-      "learning_rate": 2.5e-05,
-      "loss": 1.3505,
+      "epoch": 0.6666666666666666,
+      "grad_norm": 7.622537612915039,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 0.9468,
       "step": 7300
     },
     {
-      "epoch": 0.3356164383561644,
-      "grad_norm": 26.726581573486328,
-      "learning_rate": 2.4965753424657534e-05,
-      "loss": 1.0832,
+      "epoch": 0.6712328767123288,
+      "grad_norm": 14.563289642333984,
+      "learning_rate": 1.9931506849315068e-05,
+      "loss": 1.0154,
       "step": 7350
     },
     {
-      "epoch": 0.3378995433789954,
-      "grad_norm": 22.066526412963867,
-      "learning_rate": 2.4931506849315067e-05,
-      "loss": 1.2593,
+      "epoch": 0.6757990867579908,
+      "grad_norm": 29.9089298248291,
+      "learning_rate": 1.9863013698630137e-05,
+      "loss": 0.9872,
       "step": 7400
     },
     {
-      "epoch": 0.3401826484018265,
-      "grad_norm": 9.076361656188965,
-      "learning_rate": 2.4897260273972604e-05,
-      "loss": 0.9935,
+      "epoch": 0.680365296803653,
+      "grad_norm": 7.758576393127441,
+      "learning_rate": 1.9794520547945207e-05,
+      "loss": 0.8618,
       "step": 7450
     },
     {
-      "epoch": 0.3424657534246575,
-      "grad_norm": 32.710853576660156,
-      "learning_rate": 2.4863013698630137e-05,
-      "loss": 1.2397,
+      "epoch": 0.684931506849315,
+      "grad_norm": 15.849843978881836,
+      "learning_rate": 1.9726027397260273e-05,
+      "loss": 0.9981,
       "step": 7500
     },
     {
-      "epoch": 0.3447488584474886,
-      "grad_norm": 24.153223037719727,
-      "learning_rate": 2.4828767123287673e-05,
-      "loss": 1.1828,
+      "epoch": 0.6894977168949772,
+      "grad_norm": 13.798691749572754,
+      "learning_rate": 1.9657534246575342e-05,
+      "loss": 1.0578,
       "step": 7550
     },
     {
-      "epoch": 0.3470319634703196,
-      "grad_norm": 21.1640682220459,
-      "learning_rate": 2.4794520547945206e-05,
-      "loss": 1.2872,
+      "epoch": 0.6940639269406392,
+      "grad_norm": 12.977076530456543,
+      "learning_rate": 1.9589041095890412e-05,
+      "loss": 1.0081,
       "step": 7600
     },
     {
-      "epoch": 0.3493150684931507,
-      "grad_norm": 8.183709144592285,
-      "learning_rate": 2.476027397260274e-05,
-      "loss": 1.0255,
+      "epoch": 0.6986301369863014,
+      "grad_norm": 25.823368072509766,
+      "learning_rate": 1.952054794520548e-05,
+      "loss": 0.9483,
       "step": 7650
     },
     {
-      "epoch": 0.3515981735159817,
-      "grad_norm": 32.05094528198242,
-      "learning_rate": 2.4726027397260276e-05,
-      "loss": 1.1408,
+      "epoch": 0.7031963470319634,
+      "grad_norm": 1.7260483503341675,
+      "learning_rate": 1.9452054794520547e-05,
+      "loss": 0.9396,
       "step": 7700
     },
     {
-      "epoch": 0.3538812785388128,
-      "grad_norm": 42.780330657958984,
-      "learning_rate": 2.469178082191781e-05,
-      "loss": 1.0889,
+      "epoch": 0.7077625570776256,
+      "grad_norm": 28.217559814453125,
+      "learning_rate": 1.9383561643835617e-05,
+      "loss": 1.0159,
       "step": 7750
     },
     {
-      "epoch": 0.3561643835616438,
-      "grad_norm": 12.330619812011719,
-      "learning_rate": 2.4657534246575342e-05,
-      "loss": 1.1555,
+      "epoch": 0.7123287671232876,
+      "grad_norm": 15.566514015197754,
+      "learning_rate": 1.9315068493150686e-05,
+      "loss": 1.0673,
       "step": 7800
     },
     {
-      "epoch": 0.3584474885844749,
-      "grad_norm": 52.48133850097656,
-      "learning_rate": 2.4623287671232878e-05,
-      "loss": 0.9871,
+      "epoch": 0.7168949771689498,
+      "grad_norm": 49.06792449951172,
+      "learning_rate": 1.9246575342465756e-05,
+      "loss": 0.9817,
       "step": 7850
     },
     {
-      "epoch": 0.3607305936073059,
-      "grad_norm": 17.115123748779297,
-      "learning_rate": 2.458904109589041e-05,
-      "loss": 1.2171,
+      "epoch": 0.7214611872146118,
+      "grad_norm": 35.709224700927734,
+      "learning_rate": 1.9178082191780822e-05,
+      "loss": 0.9818,
       "step": 7900
     },
     {
-      "epoch": 0.363013698630137,
-      "grad_norm": 24.72197914123535,
-      "learning_rate": 2.4554794520547948e-05,
-      "loss": 0.9396,
+      "epoch": 0.726027397260274,
+      "grad_norm": 20.94626808166504,
+      "learning_rate": 1.910958904109589e-05,
+      "loss": 0.9113,
       "step": 7950
     },
     {
-      "epoch": 0.365296803652968,
-      "grad_norm": 59.79161834716797,
-      "learning_rate": 2.452054794520548e-05,
-      "loss": 1.161,
+      "epoch": 0.730593607305936,
+      "grad_norm": 25.5698184967041,
+      "learning_rate": 1.904109589041096e-05,
+      "loss": 1.0525,
       "step": 8000
     },
     {
-      "epoch": 0.3675799086757991,
-      "grad_norm": 27.012365341186523,
-      "learning_rate": 2.4486301369863014e-05,
-      "loss": 1.2269,
+      "epoch": 0.7351598173515982,
+      "grad_norm": 36.41669845581055,
+      "learning_rate": 1.897260273972603e-05,
+      "loss": 1.007,
       "step": 8050
     },
     {
-      "epoch": 0.3698630136986301,
-      "grad_norm": 10.282275199890137,
-      "learning_rate": 2.445205479452055e-05,
-      "loss": 1.0582,
+      "epoch": 0.7397260273972602,
+      "grad_norm": 17.15513038635254,
+      "learning_rate": 1.8904109589041096e-05,
+      "loss": 0.9568,
       "step": 8100
     },
     {
-      "epoch": 0.3721461187214612,
-      "grad_norm": 16.697723388671875,
-      "learning_rate": 2.4417808219178083e-05,
-      "loss": 1.1504,
+      "epoch": 0.7442922374429224,
+      "grad_norm": 24.435543060302734,
+      "learning_rate": 1.8835616438356166e-05,
+      "loss": 1.0073,
       "step": 8150
     },
     {
-      "epoch": 0.3744292237442922,
-      "grad_norm": 19.30514144897461,
-      "learning_rate": 2.4383561643835616e-05,
-      "loss": 1.0279,
+      "epoch": 0.7488584474885844,
+      "grad_norm": 26.5279483795166,
+      "learning_rate": 1.8767123287671235e-05,
+      "loss": 0.9768,
       "step": 8200
     },
-    {
-      "epoch": 0.3767123287671233,
-      "grad_norm": 49.45153045654297,
-      "learning_rate": 2.4349315068493153e-05,
-      "loss": 1.0799,
-      "step": 8250
-    },
-    {
-      "epoch": 0.3789954337899543,
-      "grad_norm": 33.181060791015625,
-      "learning_rate": 2.4315068493150686e-05,
-      "loss": 1.1191,
-      "step": 8300
-    },
-    {
-      "epoch": 0.3812785388127854,
-      "grad_norm": 79.8065185546875,
-      "learning_rate": 2.4280821917808222e-05,
-      "loss": 1.1442,
-      "step": 8350
-    },
-    {
-      "epoch": 0.3835616438356164,
-      "grad_norm": 48.79990005493164,
-      "learning_rate": 2.4246575342465755e-05,
-      "loss": 1.349,
-      "step": 8400
-    },
-    {
-      "epoch": 0.3858447488584475,
-      "grad_norm": 56.24677658081055,
-      "learning_rate": 2.4212328767123288e-05,
-      "loss": 1.0238,
-      "step": 8450
-    },
-    {
-      "epoch": 0.3881278538812785,
-      "grad_norm": 12.791101455688477,
-      "learning_rate": 2.4178082191780825e-05,
-      "loss": 1.0045,
-      "step": 8500
-    },
-    {
-      "epoch": 0.3904109589041096,
-      "grad_norm": 13.716742515563965,
-      "learning_rate": 2.4143835616438358e-05,
-      "loss": 1.0964,
-      "step": 8550
-    },
-    {
-      "epoch": 0.3926940639269406,
-      "grad_norm": 6.694267272949219,
-      "learning_rate": 2.410958904109589e-05,
-      "loss": 0.9396,
-      "step": 8600
-    },
-    {
-      "epoch": 0.3949771689497717,
-      "grad_norm": 19.237689971923828,
-      "learning_rate": 2.4075342465753427e-05,
-      "loss": 1.0931,
-      "step": 8650
-    },
-    {
-      "epoch": 0.3972602739726027,
-      "grad_norm": 23.344566345214844,
-      "learning_rate": 2.404109589041096e-05,
-      "loss": 1.3482,
-      "step": 8700
-    },
-    {
-      "epoch": 0.3995433789954338,
-      "grad_norm": 9.369404792785645,
-      "learning_rate": 2.4006849315068497e-05,
-      "loss": 1.0205,
-      "step": 8750
-    },
-    {
-      "epoch": 0.4018264840182648,
-      "grad_norm": 20.9238224029541,
-      "learning_rate": 2.397260273972603e-05,
-      "loss": 1.0606,
-      "step": 8800
-    },
-    {
-      "epoch": 0.4041095890410959,
-      "grad_norm": 66.00227355957031,
-      "learning_rate": 2.3938356164383563e-05,
-      "loss": 1.103,
-      "step": 8850
-    },
-    {
-      "epoch": 0.4063926940639269,
-      "grad_norm": 5.158641815185547,
-      "learning_rate": 2.39041095890411e-05,
-      "loss": 1.0207,
-      "step": 8900
-    },
-    {
-      "epoch": 0.408675799086758,
-      "grad_norm": 19.67837142944336,
-      "learning_rate": 2.3869863013698632e-05,
-      "loss": 1.108,
-      "step": 8950
-    },
-    {
-      "epoch": 0.410958904109589,
-      "grad_norm": 48.31447219848633,
-      "learning_rate": 2.3835616438356165e-05,
-      "loss": 1.0708,
-      "step": 9000
-    },
-    {
-      "epoch": 0.4132420091324201,
-      "grad_norm": 18.173908233642578,
-      "learning_rate": 2.38013698630137e-05,
-      "loss": 1.0449,
-      "step": 9050
-    },
-    {
-      "epoch": 0.4155251141552511,
-      "grad_norm": 57.505226135253906,
-      "learning_rate": 2.376712328767123e-05,
-      "loss": 1.1911,
-      "step": 9100
-    },
-    {
-      "epoch": 0.4178082191780822,
-      "grad_norm": 39.649169921875,
-      "learning_rate": 2.3732876712328768e-05,
-      "loss": 1.1708,
-      "step": 9150
-    },
-    {
-      "epoch": 0.4200913242009132,
-      "grad_norm": 7.622274398803711,
-      "learning_rate": 2.36986301369863e-05,
-      "loss": 0.9928,
-      "step": 9200
-    },
-    {
-      "epoch": 0.4223744292237443,
-      "grad_norm": 17.452634811401367,
-      "learning_rate": 2.3664383561643834e-05,
-      "loss": 0.9558,
-      "step": 9250
-    },
-    {
-      "epoch": 0.4246575342465753,
-      "grad_norm": 59.85329055786133,
-      "learning_rate": 2.363013698630137e-05,
-      "loss": 1.0805,
-      "step": 9300
-    },
-    {
-      "epoch": 0.4269406392694064,
-      "grad_norm": 9.634559631347656,
-      "learning_rate": 2.3595890410958903e-05,
-      "loss": 0.9829,
-      "step": 9350
-    },
-    {
-      "epoch": 0.4292237442922374,
-      "grad_norm": 1.924180507659912,
-      "learning_rate": 2.3561643835616436e-05,
-      "loss": 1.0737,
-      "step": 9400
-    },
-    {
-      "epoch": 0.4315068493150685,
-      "grad_norm": 31.276365280151367,
-      "learning_rate": 2.3527397260273973e-05,
-      "loss": 1.0905,
-      "step": 9450
-    },
-    {
-      "epoch": 0.4337899543378995,
-      "grad_norm": 29.218658447265625,
-      "learning_rate": 2.3493150684931506e-05,
-      "loss": 1.2167,
-      "step": 9500
-    },
-    {
-      "epoch": 0.4360730593607306,
-      "grad_norm": 42.0178108215332,
-      "learning_rate": 2.3458904109589042e-05,
-      "loss": 1.0009,
-      "step": 9550
-    },
-    {
-      "epoch": 0.4383561643835616,
-      "grad_norm": 36.67307662963867,
-      "learning_rate": 2.3424657534246575e-05,
-      "loss": 1.1016,
-      "step": 9600
-    },
-    {
-      "epoch": 0.4406392694063927,
-      "grad_norm": 21.634462356567383,
-      "learning_rate": 2.339041095890411e-05,
-      "loss": 1.0895,
-      "step": 9650
-    },
-    {
-      "epoch": 0.4429223744292237,
-      "grad_norm": 20.32236671447754,
-      "learning_rate": 2.3356164383561645e-05,
-      "loss": 1.1135,
-      "step": 9700
-    },
-    {
-      "epoch": 0.4452054794520548,
-      "grad_norm": 5.574302673339844,
-      "learning_rate": 2.3321917808219178e-05,
-      "loss": 1.1313,
-      "step": 9750
-    },
-    {
-      "epoch": 0.4474885844748858,
-      "grad_norm": 5.582201957702637,
-      "learning_rate": 2.328767123287671e-05,
-      "loss": 1.0072,
-      "step": 9800
-    },
-    {
-      "epoch": 0.4497716894977169,
-      "grad_norm": 58.499603271484375,
-      "learning_rate": 2.3253424657534247e-05,
-      "loss": 1.2044,
-      "step": 9850
-    },
-    {
-      "epoch": 0.4520547945205479,
-      "grad_norm": 18.887069702148438,
-      "learning_rate": 2.321917808219178e-05,
-      "loss": 1.0175,
-      "step": 9900
-    },
-    {
-      "epoch": 0.454337899543379,
-      "grad_norm": 33.383880615234375,
-      "learning_rate": 2.3184931506849317e-05,
-      "loss": 1.1733,
-      "step": 9950
-    },
-    {
-      "epoch": 0.45662100456621,
-      "grad_norm": 85.34075927734375,
-      "learning_rate": 2.315068493150685e-05,
-      "loss": 1.1876,
-      "step": 10000
-    },
-    {
-      "epoch": 0.4589041095890411,
-      "grad_norm": 22.957134246826172,
-      "learning_rate": 2.3116438356164383e-05,
-      "loss": 1.074,
-      "step": 10050
-    },
-    {
-      "epoch": 0.4611872146118721,
-      "grad_norm": 57.47038269042969,
-      "learning_rate": 2.308219178082192e-05,
-      "loss": 0.9007,
-      "step": 10100
-    },
-    {
-      "epoch": 0.4634703196347032,
-      "grad_norm": 11.614636421203613,
-      "learning_rate": 2.3047945205479452e-05,
-      "loss": 1.0877,
-      "step": 10150
-    },
-    {
-      "epoch": 0.4657534246575342,
-      "grad_norm": 8.970778465270996,
-      "learning_rate": 2.3013698630136985e-05,
-      "loss": 1.3286,
-      "step": 10200
-    },
-    {
-      "epoch": 0.4680365296803653,
-      "grad_norm": 43.425514221191406,
-      "learning_rate": 2.2979452054794522e-05,
-      "loss": 1.1883,
-      "step": 10250
-    },
-    {
-      "epoch": 0.4703196347031963,
-      "grad_norm": 12.85531234741211,
-      "learning_rate": 2.2945205479452055e-05,
-      "loss": 1.0921,
-      "step": 10300
-    },
-    {
-      "epoch": 0.4726027397260274,
-      "grad_norm": 6.603804111480713,
-      "learning_rate": 2.291095890410959e-05,
-      "loss": 1.1883,
-      "step": 10350
-    },
-    {
-      "epoch": 0.4748858447488584,
-      "grad_norm": 47.8962516784668,
-      "learning_rate": 2.2876712328767124e-05,
-      "loss": 1.1814,
-      "step": 10400
-    },
-    {
-      "epoch": 0.4771689497716895,
-      "grad_norm": 59.406280517578125,
-      "learning_rate": 2.2842465753424657e-05,
-      "loss": 1.115,
-      "step": 10450
-    },
-    {
-      "epoch": 0.4794520547945205,
-      "grad_norm": 43.54423522949219,
-      "learning_rate": 2.2808219178082194e-05,
-      "loss": 0.998,
-      "step": 10500
-    },
-    {
-      "epoch": 0.4817351598173516,
-      "grad_norm": 17.37261962890625,
-      "learning_rate": 2.2773972602739727e-05,
-      "loss": 0.9542,
-      "step": 10550
-    },
-    {
-      "epoch": 0.4840182648401826,
-      "grad_norm": 38.04684829711914,
-      "learning_rate": 2.273972602739726e-05,
-      "loss": 1.0195,
-      "step": 10600
-    },
-    {
-      "epoch": 0.4863013698630137,
-      "grad_norm": 17.202608108520508,
-      "learning_rate": 2.2705479452054796e-05,
-      "loss": 1.109,
-      "step": 10650
-    },
-    {
-      "epoch": 0.4885844748858447,
-      "grad_norm": 4.1588897705078125,
-      "learning_rate": 2.267123287671233e-05,
-      "loss": 1.0125,
-      "step": 10700
-    },
-    {
-      "epoch": 0.4908675799086758,
-      "grad_norm": 10.435503005981445,
-      "learning_rate": 2.2636986301369866e-05,
-      "loss": 1.0044,
-      "step": 10750
-    },
-    {
-      "epoch": 0.4931506849315068,
-      "grad_norm": 62.7014045715332,
-      "learning_rate": 2.26027397260274e-05,
-      "loss": 1.0457,
-      "step": 10800
-    },
-    {
-      "epoch": 0.4954337899543379,
-      "grad_norm": 19.47702407836914,
-      "learning_rate": 2.2568493150684932e-05,
-      "loss": 1.0468,
-      "step": 10850
-    },
-    {
-      "epoch": 0.4977168949771689,
-      "grad_norm": 42.62397766113281,
-      "learning_rate": 2.253424657534247e-05,
-      "loss": 1.0374,
-      "step": 10900
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 52.27951431274414,
-      "learning_rate": 2.25e-05,
-      "loss": 1.0967,
-      "step": 10950
-    },
-    {
-      "epoch": 0.502283105022831,
-      "grad_norm": 8.714099884033203,
-      "learning_rate": 2.2465753424657534e-05,
-      "loss": 1.2492,
-      "step": 11000
-    },
-    {
-      "epoch": 0.5045662100456622,
-      "grad_norm": 10.670646667480469,
-      "learning_rate": 2.243150684931507e-05,
-      "loss": 1.23,
-      "step": 11050
-    },
-    {
-      "epoch": 0.5068493150684932,
-      "grad_norm": 30.391632080078125,
-      "learning_rate": 2.2397260273972604e-05,
-      "loss": 1.3576,
-      "step": 11100
-    },
-    {
-      "epoch": 0.5091324200913242,
-      "grad_norm": 36.597408294677734,
-      "learning_rate": 2.2363013698630137e-05,
-      "loss": 0.9834,
-      "step": 11150
-    },
-    {
-      "epoch": 0.5114155251141552,
-      "grad_norm": 58.925941467285156,
-      "learning_rate": 2.2328767123287673e-05,
-      "loss": 1.139,
-      "step": 11200
-    },
-    {
-      "epoch": 0.5136986301369864,
-      "grad_norm": 52.20880126953125,
-      "learning_rate": 2.2294520547945206e-05,
-      "loss": 1.1037,
-      "step": 11250
-    },
-    {
-      "epoch": 0.5159817351598174,
-      "grad_norm": 104.13248443603516,
-      "learning_rate": 2.2260273972602743e-05,
-      "loss": 1.0124,
-      "step": 11300
-    },
-    {
-      "epoch": 0.5182648401826484,
-      "grad_norm": 37.718406677246094,
-      "learning_rate": 2.2226027397260276e-05,
-      "loss": 1.1873,
-      "step": 11350
-    },
-    {
-      "epoch": 0.5205479452054794,
-      "grad_norm": 31.402559280395508,
-      "learning_rate": 2.219178082191781e-05,
-      "loss": 1.1506,
-      "step": 11400
-    },
-    {
-      "epoch": 0.5228310502283106,
-      "grad_norm": 7.707183361053467,
-      "learning_rate": 2.2157534246575345e-05,
-      "loss": 1.2423,
-      "step": 11450
-    },
-    {
-      "epoch": 0.5251141552511416,
-      "grad_norm": 45.80585861206055,
-      "learning_rate": 2.212328767123288e-05,
-      "loss": 1.1457,
-      "step": 11500
-    },
-    {
-      "epoch": 0.5273972602739726,
-      "grad_norm": 58.7805061340332,
-      "learning_rate": 2.208904109589041e-05,
-      "loss": 1.0302,
-      "step": 11550
-    },
-    {
-      "epoch": 0.5296803652968036,
-      "grad_norm": 29.0611515045166,
-      "learning_rate": 2.2054794520547945e-05,
-      "loss": 1.1852,
-      "step": 11600
-    },
-    {
-      "epoch": 0.5319634703196348,
-      "grad_norm": 17.17108917236328,
-      "learning_rate": 2.2020547945205478e-05,
-      "loss": 1.0164,
-      "step": 11650
-    },
-    {
-      "epoch": 0.5342465753424658,
-      "grad_norm": 28.332836151123047,
-      "learning_rate": 2.1986301369863014e-05,
-      "loss": 0.968,
-      "step": 11700
-    },
-    {
-      "epoch": 0.5365296803652968,
-      "grad_norm": 7.278358459472656,
-      "learning_rate": 2.1952054794520547e-05,
-      "loss": 1.2058,
-      "step": 11750
-    },
-    {
-      "epoch": 0.5388127853881278,
-      "grad_norm": 15.584250450134277,
-      "learning_rate": 2.191780821917808e-05,
-      "loss": 1.0892,
-      "step": 11800
-    },
-    {
-      "epoch": 0.541095890410959,
-      "grad_norm": 22.4052791595459,
-      "learning_rate": 2.1883561643835617e-05,
-      "loss": 0.9984,
-      "step": 11850
-    },
-    {
-      "epoch": 0.54337899543379,
-      "grad_norm": 31.585800170898438,
-      "learning_rate": 2.184931506849315e-05,
-      "loss": 1.2338,
-      "step": 11900
-    },
-    {
-      "epoch": 0.545662100456621,
-      "grad_norm": 19.224178314208984,
-      "learning_rate": 2.1815068493150683e-05,
-      "loss": 1.0994,
-      "step": 11950
-    },
-    {
-      "epoch": 0.547945205479452,
-      "grad_norm": 15.44568157196045,
-      "learning_rate": 2.178082191780822e-05,
-      "loss": 1.0938,
-      "step": 12000
-    },
-    {
-      "epoch": 0.5502283105022832,
-      "grad_norm": 9.826760292053223,
-      "learning_rate": 2.1746575342465752e-05,
-      "loss": 1.1607,
-      "step": 12050
-    },
-    {
-      "epoch": 0.5525114155251142,
-      "grad_norm": 8.327254295349121,
-      "learning_rate": 2.171232876712329e-05,
-      "loss": 1.0973,
-      "step": 12100
-    },
-    {
-      "epoch": 0.5547945205479452,
-      "grad_norm": 39.339962005615234,
-      "learning_rate": 2.167808219178082e-05,
-      "loss": 1.1066,
-      "step": 12150
-    },
-    {
-      "epoch": 0.5570776255707762,
-      "grad_norm": 27.773502349853516,
-      "learning_rate": 2.1643835616438355e-05,
-      "loss": 0.9664,
-      "step": 12200
-    },
-    {
-      "epoch": 0.5593607305936074,
-      "grad_norm": 34.17534637451172,
-      "learning_rate": 2.160958904109589e-05,
-      "loss": 1.2806,
-      "step": 12250
-    },
-    {
-      "epoch": 0.5616438356164384,
-      "grad_norm": 14.751806259155273,
-      "learning_rate": 2.1575342465753424e-05,
-      "loss": 1.0973,
-      "step": 12300
-    },
-    {
-      "epoch": 0.5639269406392694,
-      "grad_norm": 39.10814666748047,
-      "learning_rate": 2.1541095890410957e-05,
-      "loss": 1.0821,
-      "step": 12350
-    },
-    {
-      "epoch": 0.5662100456621004,
-      "grad_norm": 14.468631744384766,
-      "learning_rate": 2.1506849315068494e-05,
-      "loss": 0.8218,
-      "step": 12400
-    },
-    {
-      "epoch": 0.5684931506849316,
-      "grad_norm": 30.599267959594727,
-      "learning_rate": 2.1472602739726027e-05,
-      "loss": 1.0412,
-      "step": 12450
-    },
-    {
-      "epoch": 0.5707762557077626,
-      "grad_norm": 15.382964134216309,
-      "learning_rate": 2.1438356164383563e-05,
-      "loss": 0.9351,
-      "step": 12500
-    },
-    {
-      "epoch": 0.5730593607305936,
-      "grad_norm": 44.646480560302734,
-      "learning_rate": 2.1404109589041096e-05,
-      "loss": 1.1423,
-      "step": 12550
-    },
-    {
-      "epoch": 0.5753424657534246,
-      "grad_norm": 40.618309020996094,
-      "learning_rate": 2.136986301369863e-05,
-      "loss": 1.0657,
-      "step": 12600
-    },
-    {
-      "epoch": 0.5776255707762558,
-      "grad_norm": 50.3376579284668,
-      "learning_rate": 2.1335616438356166e-05,
-      "loss": 0.9775,
-      "step": 12650
-    },
-    {
-      "epoch": 0.5799086757990868,
-      "grad_norm": 60.17688751220703,
-      "learning_rate": 2.13013698630137e-05,
-      "loss": 1.1291,
-      "step": 12700
-    },
-    {
-      "epoch": 0.5821917808219178,
-      "grad_norm": 11.148223876953125,
-      "learning_rate": 2.126712328767123e-05,
-      "loss": 1.1556,
-      "step": 12750
-    },
-    {
-      "epoch": 0.5844748858447488,
-      "grad_norm": 20.931018829345703,
-      "learning_rate": 2.1232876712328768e-05,
-      "loss": 1.0614,
-      "step": 12800
-    },
-    {
-      "epoch": 0.58675799086758,
-      "grad_norm": 29.79435157775879,
-      "learning_rate": 2.11986301369863e-05,
-      "loss": 1.2341,
-      "step": 12850
-    },
-    {
-      "epoch": 0.589041095890411,
-      "grad_norm": 27.72561264038086,
-      "learning_rate": 2.1164383561643838e-05,
-      "loss": 0.9783,
-      "step": 12900
-    },
-    {
-      "epoch": 0.591324200913242,
-      "grad_norm": 6.564560890197754,
-      "learning_rate": 2.113013698630137e-05,
-      "loss": 1.0293,
-      "step": 12950
-    },
-    {
-      "epoch": 0.593607305936073,
-      "grad_norm": 68.2934799194336,
-      "learning_rate": 2.1095890410958904e-05,
-      "loss": 1.1324,
-      "step": 13000
-    },
-    {
-      "epoch": 0.5958904109589042,
-      "grad_norm": 8.194302558898926,
-      "learning_rate": 2.106164383561644e-05,
-      "loss": 0.9099,
-      "step": 13050
-    },
-    {
-      "epoch": 0.5981735159817352,
-      "grad_norm": 25.813316345214844,
-      "learning_rate": 2.1027397260273973e-05,
-      "loss": 1.0368,
-      "step": 13100
-    },
-    {
-      "epoch": 0.6004566210045662,
-      "grad_norm": 19.83176612854004,
-      "learning_rate": 2.0993150684931506e-05,
-      "loss": 0.9496,
-      "step": 13150
-    },
-    {
-      "epoch": 0.6027397260273972,
-      "grad_norm": 17.05171012878418,
-      "learning_rate": 2.0958904109589043e-05,
-      "loss": 1.1386,
-      "step": 13200
-    },
-    {
-      "epoch": 0.6050228310502284,
-      "grad_norm": 12.28995418548584,
-      "learning_rate": 2.0924657534246576e-05,
-      "loss": 1.1897,
-      "step": 13250
-    },
-    {
-      "epoch": 0.6073059360730594,
-      "grad_norm": 22.48783302307129,
-      "learning_rate": 2.0890410958904112e-05,
-      "loss": 1.1413,
-      "step": 13300
-    },
-    {
-      "epoch": 0.6095890410958904,
-      "grad_norm": 37.44598388671875,
-      "learning_rate": 2.0856164383561645e-05,
-      "loss": 1.1138,
-      "step": 13350
-    },
-    {
-      "epoch": 0.6118721461187214,
-      "grad_norm": 19.184656143188477,
-      "learning_rate": 2.0821917808219178e-05,
-      "loss": 1.0924,
-      "step": 13400
-    },
-    {
-      "epoch": 0.6141552511415526,
-      "grad_norm": 30.34543228149414,
-      "learning_rate": 2.0787671232876715e-05,
-      "loss": 1.0382,
-      "step": 13450
-    },
-    {
-      "epoch": 0.6164383561643836,
-      "grad_norm": 18.820228576660156,
-      "learning_rate": 2.0753424657534248e-05,
-      "loss": 1.0618,
-      "step": 13500
-    },
-    {
-      "epoch": 0.6187214611872146,
-      "grad_norm": 3.0790977478027344,
-      "learning_rate": 2.071917808219178e-05,
-      "loss": 1.0412,
-      "step": 13550
-    },
-    {
-      "epoch": 0.6210045662100456,
-      "grad_norm": 42.972923278808594,
-      "learning_rate": 2.0684931506849317e-05,
-      "loss": 0.9701,
-      "step": 13600
-    },
-    {
-      "epoch": 0.6232876712328768,
-      "grad_norm": 13.229798316955566,
-      "learning_rate": 2.065068493150685e-05,
-      "loss": 0.9746,
-      "step": 13650
-    },
-    {
-      "epoch": 0.6255707762557078,
-      "grad_norm": 87.67366790771484,
-      "learning_rate": 2.0616438356164387e-05,
-      "loss": 1.1411,
-      "step": 13700
-    },
-    {
-      "epoch": 0.6278538812785388,
-      "grad_norm": 12.436323165893555,
-      "learning_rate": 2.058219178082192e-05,
-      "loss": 1.208,
-      "step": 13750
-    },
-    {
-      "epoch": 0.6301369863013698,
-      "grad_norm": 19.009395599365234,
-      "learning_rate": 2.0547945205479453e-05,
-      "loss": 1.0635,
-      "step": 13800
-    },
-    {
-      "epoch": 0.632420091324201,
-      "grad_norm": 6.680107593536377,
-      "learning_rate": 2.051369863013699e-05,
-      "loss": 1.0463,
-      "step": 13850
-    },
-    {
-      "epoch": 0.634703196347032,
-      "grad_norm": 11.398260116577148,
-      "learning_rate": 2.0479452054794522e-05,
-      "loss": 0.9179,
-      "step": 13900
-    },
-    {
-      "epoch": 0.636986301369863,
-      "grad_norm": 4.983315467834473,
-      "learning_rate": 2.0445205479452055e-05,
-      "loss": 1.0699,
-      "step": 13950
-    },
-    {
-      "epoch": 0.639269406392694,
-      "grad_norm": 29.882274627685547,
-      "learning_rate": 2.041095890410959e-05,
-      "loss": 1.1485,
-      "step": 14000
-    },
-    {
-      "epoch": 0.6415525114155252,
-      "grad_norm": 9.104654312133789,
-      "learning_rate": 2.0376712328767125e-05,
-      "loss": 0.9783,
-      "step": 14050
-    },
-    {
-      "epoch": 0.6438356164383562,
-      "grad_norm": 8.866716384887695,
-      "learning_rate": 2.034246575342466e-05,
-      "loss": 0.9843,
-      "step": 14100
-    },
-    {
-      "epoch": 0.6461187214611872,
-      "grad_norm": 20.7504825592041,
-      "learning_rate": 2.0308219178082194e-05,
-      "loss": 0.9999,
-      "step": 14150
-    },
-    {
-      "epoch": 0.6484018264840182,
-      "grad_norm": 16.239980697631836,
-      "learning_rate": 2.0273972602739724e-05,
-      "loss": 0.9305,
-      "step": 14200
-    },
-    {
-      "epoch": 0.6506849315068494,
-      "grad_norm": 52.316673278808594,
-      "learning_rate": 2.023972602739726e-05,
-      "loss": 0.9185,
-      "step": 14250
-    },
-    {
-      "epoch": 0.6529680365296804,
-      "grad_norm": 13.330124855041504,
-      "learning_rate": 2.0205479452054793e-05,
-      "loss": 0.9943,
-      "step": 14300
-    },
-    {
-      "epoch": 0.6552511415525114,
-      "grad_norm": 13.177003860473633,
-      "learning_rate": 2.0171232876712326e-05,
-      "loss": 0.7542,
-      "step": 14350
-    },
-    {
-      "epoch": 0.6575342465753424,
-      "grad_norm": 36.442108154296875,
-      "learning_rate": 2.0136986301369863e-05,
-      "loss": 0.8722,
-      "step": 14400
-    },
-    {
-      "epoch": 0.6598173515981736,
-      "grad_norm": 26.251983642578125,
-      "learning_rate": 2.0102739726027396e-05,
-      "loss": 0.9657,
-      "step": 14450
-    },
-    {
-      "epoch": 0.6621004566210046,
-      "grad_norm": 35.95085144042969,
-      "learning_rate": 2.0068493150684932e-05,
-      "loss": 1.0656,
-      "step": 14500
-    },
-    {
-      "epoch": 0.6643835616438356,
-      "grad_norm": 29.169719696044922,
-      "learning_rate": 2.0034246575342465e-05,
-      "loss": 1.076,
-      "step": 14550
-    },
-    {
-      "epoch": 0.6666666666666666,
-      "grad_norm": 49.64678955078125,
-      "learning_rate": 1.9999999999999998e-05,
-      "loss": 0.9267,
-      "step": 14600
-    },
-    {
-      "epoch": 0.6689497716894978,
-      "grad_norm": 25.249004364013672,
-      "learning_rate": 1.9965753424657535e-05,
-      "loss": 1.1205,
-      "step": 14650
-    },
-    {
-      "epoch": 0.6712328767123288,
-      "grad_norm": 21.96654510498047,
-      "learning_rate": 1.9931506849315068e-05,
-      "loss": 0.8928,
-      "step": 14700
-    },
-    {
-      "epoch": 0.6735159817351598,
-      "grad_norm": 78.21800994873047,
-      "learning_rate": 1.98972602739726e-05,
-      "loss": 0.9646,
-      "step": 14750
-    },
-    {
-      "epoch": 0.6757990867579908,
-      "grad_norm": 49.2692756652832,
-      "learning_rate": 1.9863013698630137e-05,
-      "loss": 1.0115,
-      "step": 14800
-    },
-    {
-      "epoch": 0.678082191780822,
-      "grad_norm": 35.43125915527344,
-      "learning_rate": 1.982876712328767e-05,
-      "loss": 1.0473,
-      "step": 14850
-    },
-    {
-      "epoch": 0.680365296803653,
-      "grad_norm": 18.292205810546875,
-      "learning_rate": 1.9794520547945207e-05,
-      "loss": 0.7609,
-      "step": 14900
-    },
-    {
-      "epoch": 0.682648401826484,
-      "grad_norm": 18.42896842956543,
-      "learning_rate": 1.976027397260274e-05,
-      "loss": 0.989,
-      "step": 14950
-    },
-    {
-      "epoch": 0.684931506849315,
-      "grad_norm": 25.584640502929688,
-      "learning_rate": 1.9726027397260273e-05,
-      "loss": 1.1391,
-      "step": 15000
-    },
-    {
-      "epoch": 0.6872146118721462,
-      "grad_norm": 12.116101264953613,
-      "learning_rate": 1.969178082191781e-05,
-      "loss": 1.09,
-      "step": 15050
-    },
-    {
-      "epoch": 0.6894977168949772,
-      "grad_norm": 4.520615100860596,
-      "learning_rate": 1.9657534246575342e-05,
-      "loss": 0.9834,
-      "step": 15100
-    },
-    {
-      "epoch": 0.6917808219178082,
-      "grad_norm": 16.461069107055664,
-      "learning_rate": 1.9623287671232875e-05,
-      "loss": 0.8777,
-      "step": 15150
-    },
-    {
-      "epoch": 0.6940639269406392,
-      "grad_norm": 41.71174240112305,
-      "learning_rate": 1.9589041095890412e-05,
-      "loss": 1.2619,
-      "step": 15200
-    },
-    {
-      "epoch": 0.6963470319634704,
-      "grad_norm": 10.733747482299805,
-      "learning_rate": 1.9554794520547945e-05,
-      "loss": 1.0843,
-      "step": 15250
-    },
-    {
-      "epoch": 0.6986301369863014,
-      "grad_norm": 34.77432632446289,
-      "learning_rate": 1.952054794520548e-05,
-      "loss": 0.9292,
-      "step": 15300
-    },
-    {
-      "epoch": 0.7009132420091324,
-      "grad_norm": 46.95319747924805,
-      "learning_rate": 1.9486301369863014e-05,
-      "loss": 1.0058,
-      "step": 15350
-    },
-    {
-      "epoch": 0.7031963470319634,
-      "grad_norm": 1.2955800294876099,
-      "learning_rate": 1.9452054794520547e-05,
-      "loss": 0.9672,
-      "step": 15400
-    },
-    {
-      "epoch": 0.7054794520547946,
-      "grad_norm": 16.37827491760254,
-      "learning_rate": 1.9417808219178084e-05,
-      "loss": 0.9386,
-      "step": 15450
-    },
-    {
-      "epoch": 0.7077625570776256,
-      "grad_norm": 41.87287139892578,
-      "learning_rate": 1.9383561643835617e-05,
-      "loss": 1.0772,
-      "step": 15500
-    },
-    {
-      "epoch": 0.7100456621004566,
-      "grad_norm": 8.469085693359375,
-      "learning_rate": 1.934931506849315e-05,
-      "loss": 0.9201,
-      "step": 15550
-    },
-    {
-      "epoch": 0.7123287671232876,
-      "grad_norm": 5.279160976409912,
-      "learning_rate": 1.9315068493150686e-05,
-      "loss": 1.2076,
-      "step": 15600
-    },
-    {
-      "epoch": 0.7146118721461188,
-      "grad_norm": 24.052440643310547,
-      "learning_rate": 1.928082191780822e-05,
-      "loss": 1.079,
-      "step": 15650
-    },
-    {
-      "epoch": 0.7168949771689498,
-      "grad_norm": 43.76991271972656,
-      "learning_rate": 1.9246575342465756e-05,
-      "loss": 0.9909,
-      "step": 15700
-    },
-    {
-      "epoch": 0.7191780821917808,
-      "grad_norm": 0.4643230438232422,
-      "learning_rate": 1.921232876712329e-05,
-      "loss": 1.0314,
-      "step": 15750
-    },
-    {
-      "epoch": 0.7214611872146118,
-      "grad_norm": 31.16683006286621,
-      "learning_rate": 1.9178082191780822e-05,
-      "loss": 1.0446,
-      "step": 15800
-    },
-    {
-      "epoch": 0.723744292237443,
-      "grad_norm": 2.468339443206787,
-      "learning_rate": 1.9143835616438358e-05,
-      "loss": 0.9801,
-      "step": 15850
-    },
-    {
-      "epoch": 0.726027397260274,
-      "grad_norm": 18.055511474609375,
-      "learning_rate": 1.910958904109589e-05,
-      "loss": 0.9679,
-      "step": 15900
-    },
-    {
-      "epoch": 0.728310502283105,
-      "grad_norm": 11.843624114990234,
-      "learning_rate": 1.9075342465753424e-05,
-      "loss": 1.049,
-      "step": 15950
-    },
-    {
-      "epoch": 0.730593607305936,
-      "grad_norm": 28.83637237548828,
-      "learning_rate": 1.904109589041096e-05,
-      "loss": 1.0862,
-      "step": 16000
-    },
-    {
-      "epoch": 0.7328767123287672,
-      "grad_norm": 41.827579498291016,
-      "learning_rate": 1.9006849315068494e-05,
-      "loss": 0.9812,
-      "step": 16050
-    },
-    {
-      "epoch": 0.7351598173515982,
-      "grad_norm": 19.110126495361328,
-      "learning_rate": 1.897260273972603e-05,
-      "loss": 1.2085,
-      "step": 16100
-    },
-    {
-      "epoch": 0.7374429223744292,
-      "grad_norm": 7.440555572509766,
-      "learning_rate": 1.8938356164383563e-05,
-      "loss": 1.0045,
-      "step": 16150
-    },
-    {
-      "epoch": 0.7397260273972602,
-      "grad_norm": 14.894591331481934,
-      "learning_rate": 1.8904109589041096e-05,
-      "loss": 1.0214,
-      "step": 16200
-    },
-    {
-      "epoch": 0.7420091324200914,
-      "grad_norm": 50.265140533447266,
-      "learning_rate": 1.8869863013698633e-05,
-      "loss": 1.0115,
-      "step": 16250
-    },
-    {
-      "epoch": 0.7442922374429224,
-      "grad_norm": 48.24840545654297,
-      "learning_rate": 1.8835616438356166e-05,
-      "loss": 1.022,
-      "step": 16300
-    },
-    {
-      "epoch": 0.7465753424657534,
-      "grad_norm": 23.087678909301758,
-      "learning_rate": 1.88013698630137e-05,
-      "loss": 0.9043,
-      "step": 16350
-    },
-    {
-      "epoch": 0.7488584474885844,
-      "grad_norm": 41.646461486816406,
-      "learning_rate": 1.8767123287671235e-05,
-      "loss": 1.0964,
-      "step": 16400
-    },
-    {
-      "epoch": 0.7511415525114156,
-      "grad_norm": 18.607566833496094,
-      "learning_rate": 1.8732876712328768e-05,
-      "loss": 1.1307,
-      "step": 16450
-    },
     {
       "epoch": 0.7534246575342466,
-      "grad_norm": 19.668004989624023,
+      "grad_norm": 18.786693572998047,
       "learning_rate": 1.8698630136986305e-05,
-      "loss": 0.9917,
-      "step": 16500
-    },
-    {
-      "epoch": 0.7557077625570776,
-      "grad_norm": 81.84849548339844,
-      "learning_rate": 1.8664383561643838e-05,
-      "loss": 1.1245,
-      "step": 16550
+      "loss": 1.0579,
+      "step": 8250
     },
     {
       "epoch": 0.7579908675799086,
-      "grad_norm": 17.170801162719727,
+      "grad_norm": 22.201034545898438,
       "learning_rate": 1.863013698630137e-05,
-      "loss": 0.8943,
-      "step": 16600
-    },
-    {
-      "epoch": 0.7602739726027398,
-      "grad_norm": 20.587200164794922,
-      "learning_rate": 1.8595890410958907e-05,
-      "loss": 0.9715,
-      "step": 16650
+      "loss": 1.003,
+      "step": 8300
     },
     {
       "epoch": 0.7625570776255708,
-      "grad_norm": 65.86396026611328,
+      "grad_norm": 28.934587478637695,
       "learning_rate": 1.856164383561644e-05,
-      "loss": 1.0448,
-      "step": 16700
-    },
-    {
-      "epoch": 0.7648401826484018,
-      "grad_norm": 16.04901695251465,
-      "learning_rate": 1.8527397260273973e-05,
-      "loss": 1.0682,
-      "step": 16750
+      "loss": 0.9753,
+      "step": 8350
     },
     {
       "epoch": 0.7671232876712328,
-      "grad_norm": 19.407167434692383,
+      "grad_norm": 17.935937881469727,
       "learning_rate": 1.8493150684931506e-05,
-      "loss": 0.9607,
-      "step": 16800
-    },
-    {
-      "epoch": 0.769406392694064,
-      "grad_norm": 13.71429443359375,
-      "learning_rate": 1.845890410958904e-05,
-      "loss": 1.0604,
-      "step": 16850
+      "loss": 0.9038,
+      "step": 8400
     },
     {
       "epoch": 0.771689497716895,
-      "grad_norm": 51.34487533569336,
+      "grad_norm": 29.648881912231445,
       "learning_rate": 1.8424657534246576e-05,
-      "loss": 0.8779,
-      "step": 16900
-    },
-    {
-      "epoch": 0.773972602739726,
-      "grad_norm": 13.411955833435059,
-      "learning_rate": 1.839041095890411e-05,
-      "loss": 1.0547,
-      "step": 16950
+      "loss": 1.0109,
+      "step": 8450
     },
     {
       "epoch": 0.776255707762557,
-      "grad_norm": 60.41900634765625,
+      "grad_norm": 32.65835189819336,
       "learning_rate": 1.8356164383561642e-05,
-      "loss": 1.0141,
-      "step": 17000
-    },
-    {
-      "epoch": 0.7785388127853882,
-      "grad_norm": 65.25275421142578,
-      "learning_rate": 1.832191780821918e-05,
-      "loss": 1.0541,
-      "step": 17050
+      "loss": 1.0207,
+      "step": 8500
     },
     {
       "epoch": 0.7808219178082192,
-      "grad_norm": 64.50421905517578,
+      "grad_norm": 21.674053192138672,
       "learning_rate": 1.828767123287671e-05,
-      "loss": 1.0467,
-      "step": 17100
-    },
-    {
-      "epoch": 0.7831050228310502,
-      "grad_norm": 35.63545608520508,
-      "learning_rate": 1.8253424657534244e-05,
-      "loss": 1.0431,
-      "step": 17150
+      "loss": 1.0573,
+      "step": 8550
     },
     {
       "epoch": 0.7853881278538812,
-      "grad_norm": 8.089126586914062,
+      "grad_norm": 19.038619995117188,
       "learning_rate": 1.821917808219178e-05,
-      "loss": 1.0258,
-      "step": 17200
-    },
-    {
-      "epoch": 0.7876712328767124,
-      "grad_norm": 33.44828414916992,
-      "learning_rate": 1.8184931506849314e-05,
-      "loss": 0.9516,
-      "step": 17250
+      "loss": 1.042,
+      "step": 8600
     },
     {
       "epoch": 0.7899543378995434,
-      "grad_norm": 52.08647537231445,
+      "grad_norm": 23.994884490966797,
       "learning_rate": 1.815068493150685e-05,
-      "loss": 1.0529,
-      "step": 17300
-    },
-    {
-      "epoch": 0.7922374429223744,
-      "grad_norm": 33.359886169433594,
-      "learning_rate": 1.8116438356164383e-05,
-      "loss": 1.0985,
-      "step": 17350
+      "loss": 1.0316,
+      "step": 8650
     },
     {
       "epoch": 0.7945205479452054,
-      "grad_norm": 43.1573486328125,
+      "grad_norm": 26.702028274536133,
       "learning_rate": 1.8082191780821916e-05,
-      "loss": 0.987,
-      "step": 17400
-    },
-    {
-      "epoch": 0.7968036529680366,
-      "grad_norm": 41.887718200683594,
-      "learning_rate": 1.8047945205479453e-05,
-      "loss": 0.9562,
-      "step": 17450
+      "loss": 0.9592,
+      "step": 8700
     },
     {
       "epoch": 0.7990867579908676,
-      "grad_norm": 4.935914993286133,
+      "grad_norm": 7.485787391662598,
       "learning_rate": 1.8013698630136986e-05,
-      "loss": 1.011,
-      "step": 17500
-    },
-    {
-      "epoch": 0.8013698630136986,
-      "grad_norm": 17.437244415283203,
-      "learning_rate": 1.797945205479452e-05,
-      "loss": 1.1939,
-      "step": 17550
+      "loss": 1.0175,
+      "step": 8750
     },
     {
       "epoch": 0.8036529680365296,
-      "grad_norm": 38.83684158325195,
+      "grad_norm": 24.249893188476562,
       "learning_rate": 1.7945205479452055e-05,
-      "loss": 1.0831,
-      "step": 17600
-    },
-    {
-      "epoch": 0.8059360730593608,
-      "grad_norm": 389.5946960449219,
-      "learning_rate": 1.791095890410959e-05,
-      "loss": 0.9442,
-      "step": 17650
+      "loss": 1.0643,
+      "step": 8800
     },
     {
       "epoch": 0.8082191780821918,
-      "grad_norm": 55.63192367553711,
+      "grad_norm": 31.136962890625,
       "learning_rate": 1.7876712328767125e-05,
-      "loss": 0.8581,
-      "step": 17700
-    },
-    {
-      "epoch": 0.8105022831050228,
-      "grad_norm": 12.910749435424805,
-      "learning_rate": 1.7842465753424658e-05,
-      "loss": 0.9023,
-      "step": 17750
+      "loss": 0.8719,
+      "step": 8850
     },
     {
       "epoch": 0.8127853881278538,
-      "grad_norm": 25.35328483581543,
+      "grad_norm": 18.02696418762207,
       "learning_rate": 1.780821917808219e-05,
-      "loss": 1.0642,
-      "step": 17800
-    },
-    {
-      "epoch": 0.815068493150685,
-      "grad_norm": 14.35108470916748,
-      "learning_rate": 1.7773972602739727e-05,
-      "loss": 0.9517,
-      "step": 17850
+      "loss": 0.9169,
+      "step": 8900
     },
     {
       "epoch": 0.817351598173516,
-      "grad_norm": 16.878704071044922,
+      "grad_norm": 7.9297261238098145,
       "learning_rate": 1.773972602739726e-05,
-      "loss": 1.0333,
-      "step": 17900
-    },
-    {
-      "epoch": 0.819634703196347,
-      "grad_norm": 4.217007160186768,
-      "learning_rate": 1.7705479452054793e-05,
-      "loss": 1.048,
-      "step": 17950
+      "loss": 0.9241,
+      "step": 8950
     },
     {
       "epoch": 0.821917808219178,
-      "grad_norm": 7.6629767417907715,
+      "grad_norm": 15.147187232971191,
       "learning_rate": 1.767123287671233e-05,
-      "loss": 0.9847,
-      "step": 18000
-    },
-    {
-      "epoch": 0.8242009132420092,
-      "grad_norm": 8.383101463317871,
-      "learning_rate": 1.7636986301369863e-05,
-      "loss": 0.8269,
-      "step": 18050
+      "loss": 0.9451,
+      "step": 9000
     },
     {
       "epoch": 0.8264840182648402,
-      "grad_norm": 21.882862091064453,
+      "grad_norm": 26.369773864746094,
       "learning_rate": 1.76027397260274e-05,
-      "loss": 1.0138,
-      "step": 18100
-    },
-    {
-      "epoch": 0.8287671232876712,
-      "grad_norm": 15.08014965057373,
-      "learning_rate": 1.7568493150684932e-05,
-      "loss": 1.0072,
-      "step": 18150
+      "loss": 0.9198,
+      "step": 9050
     },
     {
       "epoch": 0.8310502283105022,
-      "grad_norm": 9.188797950744629,
+      "grad_norm": 10.24289321899414,
       "learning_rate": 1.7534246575342465e-05,
-      "loss": 0.998,
-      "step": 18200
-    },
-    {
-      "epoch": 0.8333333333333334,
-      "grad_norm": 8.108163833618164,
-      "learning_rate": 1.7500000000000002e-05,
-      "loss": 0.8591,
-      "step": 18250
+      "loss": 0.9423,
+      "step": 9100
     },
     {
       "epoch": 0.8356164383561644,
-      "grad_norm": 8.082106590270996,
+      "grad_norm": 17.285388946533203,
       "learning_rate": 1.7465753424657535e-05,
-      "loss": 0.9486,
-      "step": 18300
-    },
-    {
-      "epoch": 0.8378995433789954,
-      "grad_norm": 9.28232479095459,
-      "learning_rate": 1.7431506849315068e-05,
-      "loss": 1.1733,
-      "step": 18350
+      "loss": 0.8636,
+      "step": 9150
     },
     {
       "epoch": 0.8401826484018264,
-      "grad_norm": 42.009498596191406,
+      "grad_norm": 26.444520950317383,
       "learning_rate": 1.7397260273972604e-05,
-      "loss": 1.0863,
-      "step": 18400
-    },
-    {
-      "epoch": 0.8424657534246576,
-      "grad_norm": 16.338623046875,
-      "learning_rate": 1.7363013698630137e-05,
-      "loss": 0.8428,
-      "step": 18450
+      "loss": 1.0647,
+      "step": 9200
     },
     {
       "epoch": 0.8447488584474886,
-      "grad_norm": 22.031343460083008,
+      "grad_norm": 37.353084564208984,
       "learning_rate": 1.7328767123287674e-05,
-      "loss": 0.9678,
-      "step": 18500
-    },
-    {
-      "epoch": 0.8470319634703196,
-      "grad_norm": 60.101600646972656,
-      "learning_rate": 1.7294520547945207e-05,
-      "loss": 1.0961,
-      "step": 18550
+      "loss": 0.9484,
+      "step": 9250
     },
     {
       "epoch": 0.8493150684931506,
-      "grad_norm": 17.335065841674805,
+      "grad_norm": 9.881476402282715,
       "learning_rate": 1.726027397260274e-05,
-      "loss": 0.9917,
-      "step": 18600
-    },
-    {
-      "epoch": 0.8515981735159818,
-      "grad_norm": 1.8705586194992065,
-      "learning_rate": 1.7226027397260276e-05,
-      "loss": 0.9031,
-      "step": 18650
+      "loss": 1.014,
+      "step": 9300
     },
     {
       "epoch": 0.8538812785388128,
-      "grad_norm": 8.452926635742188,
+      "grad_norm": 27.257709503173828,
       "learning_rate": 1.719178082191781e-05,
-      "loss": 0.9161,
-      "step": 18700
-    },
-    {
-      "epoch": 0.8561643835616438,
-      "grad_norm": 22.091020584106445,
-      "learning_rate": 1.7157534246575342e-05,
-      "loss": 0.9697,
-      "step": 18750
+      "loss": 0.9169,
+      "step": 9350
     },
     {
       "epoch": 0.8584474885844748,
-      "grad_norm": 14.892813682556152,
+      "grad_norm": 20.704347610473633,
       "learning_rate": 1.712328767123288e-05,
-      "loss": 1.1043,
-      "step": 18800
-    },
-    {
-      "epoch": 0.860730593607306,
-      "grad_norm": 5.6726274490356445,
-      "learning_rate": 1.7089041095890412e-05,
-      "loss": 0.9272,
-      "step": 18850
+      "loss": 0.9597,
+      "step": 9400
     },
     {
       "epoch": 0.863013698630137,
-      "grad_norm": 4.598392963409424,
+      "grad_norm": 20.994293212890625,
       "learning_rate": 1.705479452054795e-05,
-      "loss": 1.0438,
-      "step": 18900
-    },
-    {
-      "epoch": 0.865296803652968,
-      "grad_norm": 10.54848575592041,
-      "learning_rate": 1.702054794520548e-05,
-      "loss": 1.068,
-      "step": 18950
+      "loss": 0.9328,
+      "step": 9450
     },
     {
       "epoch": 0.867579908675799,
-      "grad_norm": 43.7364616394043,
+      "grad_norm": 31.617778778076172,
       "learning_rate": 1.6986301369863014e-05,
-      "loss": 0.9575,
-      "step": 19000
-    },
-    {
-      "epoch": 0.8698630136986302,
-      "grad_norm": 17.5739803314209,
-      "learning_rate": 1.695205479452055e-05,
-      "loss": 0.933,
-      "step": 19050
+      "loss": 0.9988,
+      "step": 9500
     },
     {
       "epoch": 0.8721461187214612,
-      "grad_norm": 2.4345619678497314,
+      "grad_norm": 24.76189422607422,
       "learning_rate": 1.6917808219178084e-05,
-      "loss": 0.8702,
-      "step": 19100
-    },
-    {
-      "epoch": 0.8744292237442922,
-      "grad_norm": 27.58513069152832,
-      "learning_rate": 1.6883561643835617e-05,
-      "loss": 0.9842,
-      "step": 19150
+      "loss": 0.9338,
+      "step": 9550
     },
     {
       "epoch": 0.8767123287671232,
-      "grad_norm": 44.58627700805664,
+      "grad_norm": 27.70781135559082,
       "learning_rate": 1.6849315068493153e-05,
-      "loss": 0.9117,
-      "step": 19200
-    },
-    {
-      "epoch": 0.8789954337899544,
-      "grad_norm": 23.16618537902832,
-      "learning_rate": 1.6815068493150686e-05,
-      "loss": 0.9699,
-      "step": 19250
+      "loss": 0.8994,
+      "step": 9600
     },
     {
       "epoch": 0.8812785388127854,
-      "grad_norm": 10.979578971862793,
+      "grad_norm": 6.332718372344971,
       "learning_rate": 1.6780821917808223e-05,
-      "loss": 0.9397,
-      "step": 19300
-    },
-    {
-      "epoch": 0.8835616438356164,
-      "grad_norm": 12.799473762512207,
-      "learning_rate": 1.6746575342465753e-05,
-      "loss": 1.0797,
-      "step": 19350
+      "loss": 0.8815,
+      "step": 9650
     },
     {
       "epoch": 0.8858447488584474,
-      "grad_norm": 6.065539836883545,
+      "grad_norm": 25.22089195251465,
       "learning_rate": 1.6712328767123286e-05,
-      "loss": 1.1177,
-      "step": 19400
-    },
-    {
-      "epoch": 0.8881278538812786,
-      "grad_norm": 14.423439979553223,
-      "learning_rate": 1.6678082191780822e-05,
-      "loss": 0.8818,
-      "step": 19450
+      "loss": 1.0243,
+      "step": 9700
     },
     {
       "epoch": 0.8904109589041096,
-      "grad_norm": 1.1675946712493896,
+      "grad_norm": 6.610196590423584,
       "learning_rate": 1.6643835616438355e-05,
-      "loss": 0.8819,
-      "step": 19500
-    },
-    {
-      "epoch": 0.8926940639269406,
-      "grad_norm": 18.09140396118164,
-      "learning_rate": 1.6609589041095888e-05,
-      "loss": 0.9183,
-      "step": 19550
+      "loss": 0.856,
+      "step": 9750
     },
     {
       "epoch": 0.8949771689497716,
-      "grad_norm": 11.98416805267334,
+      "grad_norm": 9.7604341506958,
       "learning_rate": 1.6575342465753425e-05,
-      "loss": 1.0715,
-      "step": 19600
-    },
-    {
-      "epoch": 0.8972602739726028,
-      "grad_norm": 16.66376495361328,
-      "learning_rate": 1.6541095890410958e-05,
-      "loss": 0.7896,
-      "step": 19650
+      "loss": 0.9275,
+      "step": 9800
     },
     {
       "epoch": 0.8995433789954338,
-      "grad_norm": 17.945493698120117,
+      "grad_norm": 13.222633361816406,
       "learning_rate": 1.6506849315068494e-05,
-      "loss": 0.8961,
-      "step": 19700
-    },
-    {
-      "epoch": 0.9018264840182648,
-      "grad_norm": 34.556827545166016,
-      "learning_rate": 1.6472602739726027e-05,
-      "loss": 0.9543,
-      "step": 19750
+      "loss": 0.7912,
+      "step": 9850
     },
     {
       "epoch": 0.9041095890410958,
-      "grad_norm": 16.06093406677246,
+      "grad_norm": 8.07390022277832,
       "learning_rate": 1.643835616438356e-05,
-      "loss": 0.913,
-      "step": 19800
-    },
-    {
-      "epoch": 0.906392694063927,
-      "grad_norm": 28.835208892822266,
-      "learning_rate": 1.6404109589041096e-05,
-      "loss": 1.068,
-      "step": 19850
+      "loss": 0.9296,
+      "step": 9900
     },
     {
       "epoch": 0.908675799086758,
-      "grad_norm": 3.6414895057678223,
+      "grad_norm": 8.25124740600586,
       "learning_rate": 1.636986301369863e-05,
-      "loss": 0.9695,
-      "step": 19900
-    },
-    {
-      "epoch": 0.910958904109589,
-      "grad_norm": 16.142841339111328,
-      "learning_rate": 1.6335616438356163e-05,
-      "loss": 1.0057,
-      "step": 19950
+      "loss": 0.9269,
+      "step": 9950
     },
     {
       "epoch": 0.91324200913242,
-      "grad_norm": 15.585680961608887,
+      "grad_norm": 22.882007598876953,
       "learning_rate": 1.63013698630137e-05,
-      "loss": 0.964,
-      "step": 20000
-    },
-    {
-      "epoch": 0.9155251141552512,
-      "grad_norm": 23.79058265686035,
-      "learning_rate": 1.6267123287671232e-05,
-      "loss": 0.8243,
-      "step": 20050
+      "loss": 0.9694,
+      "step": 10000
     },
     {
       "epoch": 0.9178082191780822,
-      "grad_norm": 14.258800506591797,
+      "grad_norm": 8.665299415588379,
       "learning_rate": 1.623287671232877e-05,
-      "loss": 0.9247,
-      "step": 20100
-    },
-    {
-      "epoch": 0.9200913242009132,
-      "grad_norm": 16.076623916625977,
-      "learning_rate": 1.61986301369863e-05,
-      "loss": 0.8446,
-      "step": 20150
+      "loss": 0.8643,
+      "step": 10050
     },
     {
       "epoch": 0.9223744292237442,
-      "grad_norm": 19.330589294433594,
+      "grad_norm": 13.816045761108398,
       "learning_rate": 1.6164383561643835e-05,
-      "loss": 1.0295,
-      "step": 20200
-    },
-    {
-      "epoch": 0.9246575342465754,
-      "grad_norm": 29.074434280395508,
-      "learning_rate": 1.613013698630137e-05,
-      "loss": 1.0043,
-      "step": 20250
+      "loss": 0.9476,
+      "step": 10100
     },
     {
       "epoch": 0.9269406392694064,
-      "grad_norm": 19.37755012512207,
+      "grad_norm": 11.227286338806152,
       "learning_rate": 1.6095890410958904e-05,
-      "loss": 0.8284,
-      "step": 20300
-    },
-    {
-      "epoch": 0.9292237442922374,
-      "grad_norm": 17.540924072265625,
-      "learning_rate": 1.6061643835616437e-05,
-      "loss": 0.9585,
-      "step": 20350
+      "loss": 0.8805,
+      "step": 10150
     },
     {
       "epoch": 0.9315068493150684,
-      "grad_norm": 15.943336486816406,
+      "grad_norm": 7.756823539733887,
       "learning_rate": 1.6027397260273974e-05,
-      "loss": 0.9534,
-      "step": 20400
-    },
-    {
-      "epoch": 0.9337899543378996,
-      "grad_norm": 14.471035957336426,
-      "learning_rate": 1.5993150684931507e-05,
-      "loss": 0.9346,
-      "step": 20450
+      "loss": 0.9185,
+      "step": 10200
     },
     {
       "epoch": 0.9360730593607306,
-      "grad_norm": 2.5866785049438477,
+      "grad_norm": 3.170640230178833,
       "learning_rate": 1.5958904109589043e-05,
-      "loss": 0.874,
-      "step": 20500
-    },
-    {
-      "epoch": 0.9383561643835616,
-      "grad_norm": 23.115089416503906,
-      "learning_rate": 1.5924657534246576e-05,
-      "loss": 1.0657,
-      "step": 20550
+      "loss": 0.8761,
+      "step": 10250
     },
     {
       "epoch": 0.9406392694063926,
-      "grad_norm": 37.99090576171875,
+      "grad_norm": 11.567336082458496,
       "learning_rate": 1.589041095890411e-05,
-      "loss": 0.8484,
-      "step": 20600
-    },
-    {
-      "epoch": 0.9429223744292238,
-      "grad_norm": 8.818378448486328,
-      "learning_rate": 1.5856164383561646e-05,
-      "loss": 0.8682,
-      "step": 20650
+      "loss": 0.9298,
+      "step": 10300
     },
     {
       "epoch": 0.9452054794520548,
-      "grad_norm": 64.82261657714844,
+      "grad_norm": 23.299612045288086,
       "learning_rate": 1.582191780821918e-05,
-      "loss": 0.9238,
-      "step": 20700
-    },
-    {
-      "epoch": 0.9474885844748858,
-      "grad_norm": 4.82705545425415,
-      "learning_rate": 1.578767123287671e-05,
-      "loss": 1.0183,
-      "step": 20750
+      "loss": 0.8361,
+      "step": 10350
     },
     {
       "epoch": 0.9497716894977168,
-      "grad_norm": 35.37221908569336,
+      "grad_norm": 14.57529067993164,
       "learning_rate": 1.5753424657534248e-05,
-      "loss": 0.9661,
-      "step": 20800
-    },
-    {
-      "epoch": 0.952054794520548,
-      "grad_norm": 59.23238754272461,
-      "learning_rate": 1.571917808219178e-05,
-      "loss": 1.1756,
-      "step": 20850
+      "loss": 0.9669,
+      "step": 10400
     },
     {
       "epoch": 0.954337899543379,
-      "grad_norm": 5.600244045257568,
+      "grad_norm": 10.777979850769043,
       "learning_rate": 1.5684931506849318e-05,
-      "loss": 0.9528,
-      "step": 20900
-    },
-    {
-      "epoch": 0.95662100456621,
-      "grad_norm": 36.54751968383789,
-      "learning_rate": 1.565068493150685e-05,
-      "loss": 0.8113,
-      "step": 20950
+      "loss": 0.9632,
+      "step": 10450
     },
     {
       "epoch": 0.958904109589041,
-      "grad_norm": 21.206743240356445,
+      "grad_norm": 25.46247100830078,
       "learning_rate": 1.5616438356164384e-05,
-      "loss": 1.0184,
-      "step": 21000
-    },
-    {
-      "epoch": 0.9611872146118722,
-      "grad_norm": 5.483078479766846,
-      "learning_rate": 1.558219178082192e-05,
-      "loss": 1.0164,
-      "step": 21050
+      "loss": 0.8625,
+      "step": 10500
     },
     {
       "epoch": 0.9634703196347032,
-      "grad_norm": 10.294635772705078,
-      "learning_rate": 1.5547945205479453e-05,
-      "loss": 0.8765,
-      "step": 21100
-    },
-    {
-      "epoch": 0.9657534246575342,
-      "grad_norm": 15.811431884765625,
-      "learning_rate": 1.5513698630136986e-05,
-      "loss": 1.0893,
-      "step": 21150
+      "grad_norm": 8.925729751586914,
+      "learning_rate": 1.5547945205479453e-05,
+      "loss": 0.903,
+      "step": 10550
     },
     {
       "epoch": 0.9680365296803652,
-      "grad_norm": 12.22139835357666,
+      "grad_norm": 11.709217071533203,
       "learning_rate": 1.5479452054794523e-05,
-      "loss": 0.8607,
-      "step": 21200
-    },
-    {
-      "epoch": 0.9703196347031964,
-      "grad_norm": 32.24748229980469,
-      "learning_rate": 1.5445205479452056e-05,
-      "loss": 0.9461,
-      "step": 21250
+      "loss": 0.9446,
+      "step": 10600
     },
     {
       "epoch": 0.9726027397260274,
-      "grad_norm": 105.09614562988281,
+      "grad_norm": 36.75910568237305,
       "learning_rate": 1.5410958904109592e-05,
-      "loss": 1.0007,
-      "step": 21300
-    },
-    {
-      "epoch": 0.9748858447488584,
-      "grad_norm": 39.1617546081543,
-      "learning_rate": 1.5376712328767125e-05,
-      "loss": 0.8074,
-      "step": 21350
+      "loss": 0.9568,
+      "step": 10650
     },
     {
       "epoch": 0.9771689497716894,
-      "grad_norm": 54.6515007019043,
+      "grad_norm": 30.059825897216797,
       "learning_rate": 1.5342465753424658e-05,
-      "loss": 1.0728,
-      "step": 21400
-    },
-    {
-      "epoch": 0.9794520547945206,
-      "grad_norm": 11.429344177246094,
-      "learning_rate": 1.5308219178082195e-05,
-      "loss": 1.0846,
-      "step": 21450
+      "loss": 0.9603,
+      "step": 10700
     },
     {
       "epoch": 0.9817351598173516,
-      "grad_norm": 35.74282455444336,
+      "grad_norm": 23.478477478027344,
       "learning_rate": 1.5273972602739728e-05,
-      "loss": 0.9805,
-      "step": 21500
-    },
-    {
-      "epoch": 0.9840182648401826,
-      "grad_norm": 28.15215301513672,
-      "learning_rate": 1.5239726027397259e-05,
-      "loss": 1.057,
-      "step": 21550
+      "loss": 0.9828,
+      "step": 10750
     },
     {
       "epoch": 0.9863013698630136,
-      "grad_norm": 24.915042877197266,
+      "grad_norm": 34.233699798583984,
       "learning_rate": 1.5205479452054795e-05,
-      "loss": 0.9953,
-      "step": 21600
-    },
-    {
-      "epoch": 0.9885844748858448,
-      "grad_norm": 54.70051574707031,
-      "learning_rate": 1.5171232876712328e-05,
-      "loss": 0.8781,
-      "step": 21650
+      "loss": 0.9867,
+      "step": 10800
     },
     {
       "epoch": 0.9908675799086758,
-      "grad_norm": 57.645572662353516,
+      "grad_norm": 19.399288177490234,
       "learning_rate": 1.5136986301369865e-05,
-      "loss": 0.9031,
-      "step": 21700
-    },
-    {
-      "epoch": 0.9931506849315068,
-      "grad_norm": 49.95241165161133,
-      "learning_rate": 1.5102739726027398e-05,
-      "loss": 1.1313,
-      "step": 21750
+      "loss": 0.8618,
+      "step": 10850
     },
     {
       "epoch": 0.9954337899543378,
-      "grad_norm": 11.169427871704102,
+      "grad_norm": 9.319437026977539,
       "learning_rate": 1.5068493150684931e-05,
-      "loss": 0.7831,
-      "step": 21800
-    },
-    {
-      "epoch": 0.997716894977169,
-      "grad_norm": 1.9000864028930664,
-      "learning_rate": 1.5034246575342467e-05,
-      "loss": 0.9135,
-      "step": 21850
+      "loss": 0.9352,
+      "step": 10900
     },
     {
       "epoch": 1.0,
-      "grad_norm": 8.12338638305664,
+      "grad_norm": 8.767410278320312,
       "learning_rate": 1.5e-05,
-      "loss": 0.8517,
-      "step": 21900
-    },
-    {
-      "epoch": 1.0022831050228311,
-      "grad_norm": 9.123151779174805,
-      "learning_rate": 1.4965753424657535e-05,
-      "loss": 0.8109,
-      "step": 21950
+      "loss": 0.8511,
+      "step": 10950
     },
     {
       "epoch": 1.004566210045662,
-      "grad_norm": 1.5693093538284302,
+      "grad_norm": 15.091998100280762,
       "learning_rate": 1.4931506849315068e-05,
-      "loss": 0.637,
-      "step": 22000
-    },
-    {
-      "epoch": 1.0068493150684932,
-      "grad_norm": 57.09550094604492,
-      "learning_rate": 1.4897260273972603e-05,
-      "loss": 0.7138,
-      "step": 22050
+      "loss": 0.6653,
+      "step": 11000
     },
     {
       "epoch": 1.009132420091324,
-      "grad_norm": 22.930618286132812,
+      "grad_norm": 23.510337829589844,
       "learning_rate": 1.4863013698630138e-05,
-      "loss": 0.5267,
-      "step": 22100
-    },
-    {
-      "epoch": 1.0114155251141552,
-      "grad_norm": 6.480524063110352,
-      "learning_rate": 1.4828767123287672e-05,
-      "loss": 0.583,
-      "step": 22150
+      "loss": 0.6163,
+      "step": 11050
     },
     {
       "epoch": 1.0136986301369864,
-      "grad_norm": 8.956890106201172,
+      "grad_norm": 8.980317115783691,
       "learning_rate": 1.4794520547945205e-05,
-      "loss": 0.838,
-      "step": 22200
-    },
-    {
-      "epoch": 1.0159817351598173,
-      "grad_norm": 28.80997657775879,
-      "learning_rate": 1.476027397260274e-05,
-      "loss": 0.7791,
-      "step": 22250
+      "loss": 0.7783,
+      "step": 11100
     },
     {
       "epoch": 1.0182648401826484,
-      "grad_norm": 51.440643310546875,
+      "grad_norm": 36.67283248901367,
       "learning_rate": 1.4726027397260275e-05,
-      "loss": 0.6699,
-      "step": 22300
-    },
-    {
-      "epoch": 1.0205479452054795,
-      "grad_norm": 16.62401580810547,
-      "learning_rate": 1.469178082191781e-05,
-      "loss": 0.6056,
-      "step": 22350
+      "loss": 0.7681,
+      "step": 11150
     },
     {
       "epoch": 1.0228310502283104,
-      "grad_norm": 6.103922367095947,
+      "grad_norm": 4.734091758728027,
       "learning_rate": 1.4657534246575343e-05,
-      "loss": 0.7163,
-      "step": 22400
-    },
-    {
-      "epoch": 1.0251141552511416,
-      "grad_norm": 13.347216606140137,
-      "learning_rate": 1.4623287671232877e-05,
-      "loss": 0.8158,
-      "step": 22450
+      "loss": 0.6181,
+      "step": 11200
     },
     {
       "epoch": 1.0273972602739727,
-      "grad_norm": 10.218559265136719,
+      "grad_norm": 7.513982772827148,
       "learning_rate": 1.4589041095890412e-05,
-      "loss": 0.7067,
-      "step": 22500
-    },
-    {
-      "epoch": 1.0296803652968036,
-      "grad_norm": 20.360658645629883,
-      "learning_rate": 1.4554794520547945e-05,
-      "loss": 0.5413,
-      "step": 22550
+      "loss": 0.7297,
+      "step": 11250
     },
     {
       "epoch": 1.0319634703196348,
-      "grad_norm": 14.813282012939453,
+      "grad_norm": 16.1832275390625,
       "learning_rate": 1.4520547945205478e-05,
-      "loss": 0.7121,
-      "step": 22600
-    },
-    {
-      "epoch": 1.0342465753424657,
-      "grad_norm": 4.041282653808594,
-      "learning_rate": 1.4486301369863013e-05,
-      "loss": 0.6259,
-      "step": 22650
+      "loss": 0.6487,
+      "step": 11300
     },
     {
       "epoch": 1.0365296803652968,
-      "grad_norm": 8.718100547790527,
+      "grad_norm": 25.557220458984375,
       "learning_rate": 1.4452054794520548e-05,
-      "loss": 0.679,
-      "step": 22700
-    },
-    {
-      "epoch": 1.038812785388128,
-      "grad_norm": 24.674726486206055,
-      "learning_rate": 1.4417808219178082e-05,
-      "loss": 0.5784,
-      "step": 22750
+      "loss": 0.7089,
+      "step": 11350
     },
     {
       "epoch": 1.0410958904109588,
-      "grad_norm": 18.543848037719727,
+      "grad_norm": 11.07717227935791,
       "learning_rate": 1.4383561643835615e-05,
-      "loss": 0.6259,
-      "step": 22800
-    },
-    {
-      "epoch": 1.04337899543379,
-      "grad_norm": 11.60615348815918,
-      "learning_rate": 1.434931506849315e-05,
-      "loss": 0.6859,
-      "step": 22850
+      "loss": 0.6266,
+      "step": 11400
     },
     {
       "epoch": 1.045662100456621,
-      "grad_norm": 19.409120559692383,
+      "grad_norm": 11.381722450256348,
       "learning_rate": 1.4315068493150685e-05,
-      "loss": 0.724,
-      "step": 22900
-    },
-    {
-      "epoch": 1.047945205479452,
-      "grad_norm": 18.087451934814453,
-      "learning_rate": 1.428082191780822e-05,
-      "loss": 0.6719,
-      "step": 22950
+      "loss": 0.6758,
+      "step": 11450
     },
     {
       "epoch": 1.0502283105022832,
-      "grad_norm": 37.11878967285156,
+      "grad_norm": 9.080512046813965,
       "learning_rate": 1.4246575342465753e-05,
-      "loss": 0.5994,
-      "step": 23000
-    },
-    {
-      "epoch": 1.052511415525114,
-      "grad_norm": 35.999149322509766,
-      "learning_rate": 1.4212328767123287e-05,
-      "loss": 0.8059,
-      "step": 23050
+      "loss": 0.6865,
+      "step": 11500
     },
     {
       "epoch": 1.0547945205479452,
-      "grad_norm": 23.739049911499023,
+      "grad_norm": 10.727092742919922,
       "learning_rate": 1.4178082191780822e-05,
-      "loss": 0.4798,
-      "step": 23100
-    },
-    {
-      "epoch": 1.0570776255707763,
-      "grad_norm": 4.680600166320801,
-      "learning_rate": 1.4143835616438357e-05,
-      "loss": 0.6557,
-      "step": 23150
+      "loss": 0.6494,
+      "step": 11550
     },
     {
       "epoch": 1.0593607305936072,
-      "grad_norm": 31.171585083007812,
+      "grad_norm": 4.969435691833496,
       "learning_rate": 1.410958904109589e-05,
-      "loss": 0.6122,
-      "step": 23200
-    },
-    {
-      "epoch": 1.0616438356164384,
-      "grad_norm": 10.616646766662598,
-      "learning_rate": 1.4075342465753425e-05,
-      "loss": 0.7256,
-      "step": 23250
+      "loss": 0.5874,
+      "step": 11600
     },
     {
       "epoch": 1.0639269406392695,
-      "grad_norm": 5.910991191864014,
+      "grad_norm": 12.03378963470459,
       "learning_rate": 1.404109589041096e-05,
-      "loss": 0.5245,
-      "step": 23300
-    },
-    {
-      "epoch": 1.0662100456621004,
-      "grad_norm": 6.659880638122559,
-      "learning_rate": 1.4006849315068494e-05,
-      "loss": 0.6245,
-      "step": 23350
+      "loss": 0.6301,
+      "step": 11650
     },
     {
       "epoch": 1.0684931506849316,
-      "grad_norm": 28.779626846313477,
+      "grad_norm": 8.797894477844238,
       "learning_rate": 1.3972602739726027e-05,
-      "loss": 0.7664,
-      "step": 23400
-    },
-    {
-      "epoch": 1.0707762557077625,
-      "grad_norm": 42.86344909667969,
-      "learning_rate": 1.3938356164383562e-05,
-      "loss": 0.7317,
-      "step": 23450
+      "loss": 0.6839,
+      "step": 11700
     },
     {
       "epoch": 1.0730593607305936,
-      "grad_norm": 17.784263610839844,
+      "grad_norm": 9.395676612854004,
       "learning_rate": 1.3904109589041097e-05,
-      "loss": 0.7032,
-      "step": 23500
-    },
-    {
-      "epoch": 1.0753424657534247,
-      "grad_norm": 33.24246597290039,
-      "learning_rate": 1.3869863013698631e-05,
-      "loss": 0.6477,
-      "step": 23550
+      "loss": 0.6672,
+      "step": 11750
     },
     {
       "epoch": 1.0776255707762556,
-      "grad_norm": 23.703388214111328,
+      "grad_norm": 26.268247604370117,
       "learning_rate": 1.3835616438356164e-05,
-      "loss": 0.5103,
-      "step": 23600
-    },
-    {
-      "epoch": 1.0799086757990868,
-      "grad_norm": 27.350101470947266,
-      "learning_rate": 1.38013698630137e-05,
-      "loss": 0.7417,
-      "step": 23650
+      "loss": 0.6398,
+      "step": 11800
     },
     {
       "epoch": 1.0821917808219177,
-      "grad_norm": 32.2100715637207,
+      "grad_norm": 19.63582992553711,
       "learning_rate": 1.3767123287671234e-05,
-      "loss": 0.9106,
-      "step": 23700
-    },
-    {
-      "epoch": 1.0844748858447488,
-      "grad_norm": 8.949542999267578,
-      "learning_rate": 1.3732876712328769e-05,
-      "loss": 0.6949,
-      "step": 23750
+      "loss": 0.7825,
+      "step": 11850
     },
     {
       "epoch": 1.08675799086758,
-      "grad_norm": 31.792240142822266,
+      "grad_norm": 23.42495346069336,
       "learning_rate": 1.3698630136986302e-05,
-      "loss": 0.725,
-      "step": 23800
-    },
-    {
-      "epoch": 1.0890410958904109,
-      "grad_norm": 51.95265579223633,
-      "learning_rate": 1.3664383561643835e-05,
-      "loss": 0.73,
-      "step": 23850
+      "loss": 0.7078,
+      "step": 11900
     },
     {
       "epoch": 1.091324200913242,
-      "grad_norm": 19.113056182861328,
+      "grad_norm": 29.21826934814453,
       "learning_rate": 1.363013698630137e-05,
-      "loss": 0.6004,
-      "step": 23900
-    },
-    {
-      "epoch": 1.0936073059360731,
-      "grad_norm": 16.85398292541504,
-      "learning_rate": 1.3595890410958904e-05,
-      "loss": 0.5843,
-      "step": 23950
+      "loss": 0.6973,
+      "step": 11950
     },
     {
       "epoch": 1.095890410958904,
-      "grad_norm": 20.591157913208008,
+      "grad_norm": 13.69019603729248,
       "learning_rate": 1.3561643835616437e-05,
-      "loss": 0.6938,
-      "step": 24000
-    },
-    {
-      "epoch": 1.0981735159817352,
-      "grad_norm": 22.592805862426758,
-      "learning_rate": 1.3527397260273972e-05,
-      "loss": 0.6366,
-      "step": 24050
+      "loss": 0.6325,
+      "step": 12000
     },
     {
       "epoch": 1.1004566210045663,
-      "grad_norm": 77.25360870361328,
+      "grad_norm": 22.293821334838867,
       "learning_rate": 1.3493150684931507e-05,
-      "loss": 0.7245,
-      "step": 24100
-    },
-    {
-      "epoch": 1.1027397260273972,
-      "grad_norm": 35.447452545166016,
-      "learning_rate": 1.3458904109589042e-05,
-      "loss": 0.7001,
-      "step": 24150
+      "loss": 0.6833,
+      "step": 12050
     },
     {
       "epoch": 1.1050228310502284,
-      "grad_norm": 50.135650634765625,
+      "grad_norm": 9.701814651489258,
       "learning_rate": 1.3424657534246575e-05,
-      "loss": 0.6473,
-      "step": 24200
-    },
-    {
-      "epoch": 1.1073059360730593,
-      "grad_norm": 9.847333908081055,
-      "learning_rate": 1.339041095890411e-05,
-      "loss": 0.6206,
-      "step": 24250
+      "loss": 0.6518,
+      "step": 12100
     },
     {
       "epoch": 1.1095890410958904,
-      "grad_norm": 16.841964721679688,
+      "grad_norm": 25.30548667907715,
       "learning_rate": 1.3356164383561644e-05,
-      "loss": 0.754,
-      "step": 24300
-    },
-    {
-      "epoch": 1.1118721461187215,
-      "grad_norm": 25.09712028503418,
-      "learning_rate": 1.3321917808219179e-05,
-      "loss": 0.7678,
-      "step": 24350
+      "loss": 0.7458,
+      "step": 12150
     },
     {
       "epoch": 1.1141552511415524,
-      "grad_norm": 20.83018684387207,
+      "grad_norm": 11.800670623779297,
       "learning_rate": 1.3287671232876712e-05,
-      "loss": 0.7375,
-      "step": 24400
-    },
-    {
-      "epoch": 1.1164383561643836,
-      "grad_norm": 1.3051903247833252,
-      "learning_rate": 1.3253424657534247e-05,
-      "loss": 0.73,
-      "step": 24450
+      "loss": 0.755,
+      "step": 12200
     },
     {
       "epoch": 1.1187214611872145,
-      "grad_norm": 21.558069229125977,
+      "grad_norm": 15.019549369812012,
       "learning_rate": 1.3219178082191781e-05,
-      "loss": 0.6535,
-      "step": 24500
-    },
-    {
-      "epoch": 1.1210045662100456,
-      "grad_norm": 53.48995590209961,
-      "learning_rate": 1.3184931506849316e-05,
-      "loss": 0.7592,
-      "step": 24550
+      "loss": 0.6871,
+      "step": 12250
     },
     {
       "epoch": 1.1232876712328768,
-      "grad_norm": 6.8114166259765625,
+      "grad_norm": 8.90066909790039,
       "learning_rate": 1.3150684931506849e-05,
-      "loss": 0.6735,
-      "step": 24600
-    },
-    {
-      "epoch": 1.1255707762557077,
-      "grad_norm": 16.559179306030273,
-      "learning_rate": 1.3116438356164384e-05,
-      "loss": 0.4556,
-      "step": 24650
+      "loss": 0.738,
+      "step": 12300
     },
     {
       "epoch": 1.1278538812785388,
-      "grad_norm": 2.124957323074341,
+      "grad_norm": 6.738426685333252,
       "learning_rate": 1.3082191780821919e-05,
-      "loss": 0.92,
-      "step": 24700
-    },
-    {
-      "epoch": 1.13013698630137,
-      "grad_norm": 34.67999267578125,
-      "learning_rate": 1.3047945205479453e-05,
-      "loss": 0.6309,
-      "step": 24750
+      "loss": 0.6828,
+      "step": 12350
     },
     {
       "epoch": 1.1324200913242009,
-      "grad_norm": 9.184309005737305,
+      "grad_norm": 7.866949558258057,
       "learning_rate": 1.3013698630136986e-05,
-      "loss": 0.7584,
-      "step": 24800
-    },
-    {
-      "epoch": 1.134703196347032,
-      "grad_norm": 17.552547454833984,
-      "learning_rate": 1.2979452054794521e-05,
-      "loss": 0.6098,
-      "step": 24850
+      "loss": 0.708,
+      "step": 12400
     },
     {
       "epoch": 1.1369863013698631,
-      "grad_norm": 37.24542999267578,
+      "grad_norm": 26.066892623901367,
       "learning_rate": 1.2945205479452056e-05,
-      "loss": 0.8336,
-      "step": 24900
-    },
-    {
-      "epoch": 1.139269406392694,
-      "grad_norm": 52.120880126953125,
-      "learning_rate": 1.291095890410959e-05,
-      "loss": 0.6372,
-      "step": 24950
+      "loss": 0.7078,
+      "step": 12450
     },
     {
       "epoch": 1.1415525114155252,
-      "grad_norm": 13.773005485534668,
+      "grad_norm": 7.540081024169922,
       "learning_rate": 1.2876712328767124e-05,
-      "loss": 0.6734,
-      "step": 25000
-    },
-    {
-      "epoch": 1.143835616438356,
-      "grad_norm": 12.726426124572754,
-      "learning_rate": 1.2842465753424658e-05,
-      "loss": 0.5249,
-      "step": 25050
+      "loss": 0.6958,
+      "step": 12500
     },
     {
       "epoch": 1.1461187214611872,
-      "grad_norm": 8.800257682800293,
+      "grad_norm": 13.667712211608887,
       "learning_rate": 1.2808219178082193e-05,
-      "loss": 0.5094,
-      "step": 25100
-    },
-    {
-      "epoch": 1.1484018264840183,
-      "grad_norm": 77.74162292480469,
-      "learning_rate": 1.2773972602739726e-05,
-      "loss": 0.7281,
-      "step": 25150
+      "loss": 0.5233,
+      "step": 12550
     },
     {
       "epoch": 1.1506849315068493,
-      "grad_norm": 8.394207954406738,
+      "grad_norm": 17.8378849029541,
       "learning_rate": 1.2739726027397259e-05,
-      "loss": 0.615,
-      "step": 25200
-    },
-    {
-      "epoch": 1.1529680365296804,
-      "grad_norm": 62.93704605102539,
-      "learning_rate": 1.2705479452054794e-05,
-      "loss": 0.5909,
-      "step": 25250
+      "loss": 0.669,
+      "step": 12600
     },
     {
       "epoch": 1.1552511415525113,
-      "grad_norm": 18.461288452148438,
+      "grad_norm": 25.038570404052734,
       "learning_rate": 1.2671232876712329e-05,
-      "loss": 0.6578,
-      "step": 25300
-    },
-    {
-      "epoch": 1.1575342465753424,
-      "grad_norm": 17.798723220825195,
-      "learning_rate": 1.2636986301369863e-05,
-      "loss": 0.7566,
-      "step": 25350
+      "loss": 0.5989,
+      "step": 12650
     },
     {
       "epoch": 1.1598173515981736,
-      "grad_norm": 0.32706037163734436,
+      "grad_norm": 2.4400246143341064,
       "learning_rate": 1.2602739726027396e-05,
-      "loss": 0.6193,
-      "step": 25400
-    },
-    {
-      "epoch": 1.1621004566210045,
-      "grad_norm": 19.030363082885742,
-      "learning_rate": 1.2568493150684931e-05,
-      "loss": 0.6249,
-      "step": 25450
+      "loss": 0.745,
+      "step": 12700
     },
     {
       "epoch": 1.1643835616438356,
-      "grad_norm": 0.1700681447982788,
+      "grad_norm": 3.209836006164551,
       "learning_rate": 1.2534246575342466e-05,
-      "loss": 0.7782,
-      "step": 25500
-    },
-    {
-      "epoch": 1.1666666666666667,
-      "grad_norm": 147.6623077392578,
-      "learning_rate": 1.25e-05,
-      "loss": 0.6307,
-      "step": 25550
+      "loss": 0.6943,
+      "step": 12750
     },
     {
       "epoch": 1.1689497716894977,
-      "grad_norm": 13.00705623626709,
+      "grad_norm": 12.491949081420898,
       "learning_rate": 1.2465753424657534e-05,
-      "loss": 0.6638,
-      "step": 25600
-    },
-    {
-      "epoch": 1.1712328767123288,
-      "grad_norm": 10.626334190368652,
-      "learning_rate": 1.2431506849315068e-05,
-      "loss": 0.7118,
-      "step": 25650
+      "loss": 0.6115,
+      "step": 12800
     },
     {
       "epoch": 1.17351598173516,
-      "grad_norm": 5.709634304046631,
+      "grad_norm": 11.220638275146484,
       "learning_rate": 1.2397260273972603e-05,
-      "loss": 0.6299,
-      "step": 25700
-    },
-    {
-      "epoch": 1.1757990867579908,
-      "grad_norm": 30.29009246826172,
-      "learning_rate": 1.2363013698630138e-05,
-      "loss": 0.7486,
-      "step": 25750
+      "loss": 0.5898,
+      "step": 12850
     },
     {
       "epoch": 1.178082191780822,
-      "grad_norm": 18.62900733947754,
+      "grad_norm": 16.73185920715332,
       "learning_rate": 1.2328767123287671e-05,
-      "loss": 0.5936,
-      "step": 25800
-    },
-    {
-      "epoch": 1.1803652968036529,
-      "grad_norm": 14.033234596252441,
-      "learning_rate": 1.2294520547945206e-05,
-      "loss": 0.7004,
-      "step": 25850
+      "loss": 0.6734,
+      "step": 12900
     },
     {
       "epoch": 1.182648401826484,
-      "grad_norm": 38.063873291015625,
+      "grad_norm": 14.307918548583984,
       "learning_rate": 1.226027397260274e-05,
-      "loss": 0.7239,
-      "step": 25900
-    },
-    {
-      "epoch": 1.1849315068493151,
-      "grad_norm": 40.94581985473633,
-      "learning_rate": 1.2226027397260275e-05,
-      "loss": 0.6232,
-      "step": 25950
+      "loss": 0.7266,
+      "step": 12950
     },
     {
       "epoch": 1.187214611872146,
-      "grad_norm": 10.695091247558594,
+      "grad_norm": 14.836384773254395,
       "learning_rate": 1.2191780821917808e-05,
-      "loss": 0.8526,
-      "step": 26000
-    },
-    {
-      "epoch": 1.1894977168949772,
-      "grad_norm": 19.597412109375,
-      "learning_rate": 1.2157534246575343e-05,
-      "loss": 0.6878,
-      "step": 26050
+      "loss": 0.7586,
+      "step": 13000
     },
     {
       "epoch": 1.191780821917808,
-      "grad_norm": 57.820003509521484,
+      "grad_norm": 29.282228469848633,
       "learning_rate": 1.2123287671232878e-05,
-      "loss": 0.7689,
-      "step": 26100
-    },
-    {
-      "epoch": 1.1940639269406392,
-      "grad_norm": 19.449691772460938,
-      "learning_rate": 1.2089041095890412e-05,
-      "loss": 0.6267,
-      "step": 26150
+      "loss": 0.7917,
+      "step": 13050
     },
     {
       "epoch": 1.1963470319634704,
-      "grad_norm": 30.782194137573242,
+      "grad_norm": 26.005468368530273,
       "learning_rate": 1.2054794520547945e-05,
-      "loss": 0.6165,
-      "step": 26200
-    },
-    {
-      "epoch": 1.1986301369863013,
-      "grad_norm": 27.298837661743164,
-      "learning_rate": 1.202054794520548e-05,
-      "loss": 0.8057,
-      "step": 26250
+      "loss": 0.6413,
+      "step": 13100
     },
     {
       "epoch": 1.2009132420091324,
-      "grad_norm": 185.9980010986328,
+      "grad_norm": 8.961703300476074,
       "learning_rate": 1.1986301369863015e-05,
-      "loss": 0.7694,
-      "step": 26300
-    },
-    {
-      "epoch": 1.2031963470319635,
-      "grad_norm": 3.1694886684417725,
-      "learning_rate": 1.195205479452055e-05,
-      "loss": 0.6598,
-      "step": 26350
+      "loss": 0.792,
+      "step": 13150
     },
     {
       "epoch": 1.2054794520547945,
-      "grad_norm": 6.1167097091674805,
+      "grad_norm": 10.702567100524902,
       "learning_rate": 1.1917808219178083e-05,
-      "loss": 0.6464,
-      "step": 26400
-    },
-    {
-      "epoch": 1.2077625570776256,
-      "grad_norm": 15.646038055419922,
-      "learning_rate": 1.1883561643835616e-05,
-      "loss": 0.9328,
-      "step": 26450
+      "loss": 0.6594,
+      "step": 13200
     },
     {
       "epoch": 1.2100456621004567,
-      "grad_norm": 7.978808403015137,
+      "grad_norm": 13.418671607971191,
       "learning_rate": 1.184931506849315e-05,
-      "loss": 0.6384,
-      "step": 26500
-    },
-    {
-      "epoch": 1.2123287671232876,
-      "grad_norm": 21.067642211914062,
-      "learning_rate": 1.1815068493150685e-05,
-      "loss": 0.8629,
-      "step": 26550
+      "loss": 0.7547,
+      "step": 13250
     },
     {
       "epoch": 1.2146118721461188,
-      "grad_norm": 27.418428421020508,
+      "grad_norm": 17.413429260253906,
       "learning_rate": 1.1780821917808218e-05,
-      "loss": 0.6208,
-      "step": 26600
-    },
-    {
-      "epoch": 1.2168949771689497,
-      "grad_norm": 10.672882080078125,
-      "learning_rate": 1.1746575342465753e-05,
-      "loss": 0.5777,
-      "step": 26650
+      "loss": 0.7908,
+      "step": 13300
     },
     {
       "epoch": 1.2191780821917808,
-      "grad_norm": 82.25353240966797,
+      "grad_norm": 18.375572204589844,
       "learning_rate": 1.1712328767123288e-05,
-      "loss": 0.9331,
-      "step": 26700
-    },
-    {
-      "epoch": 1.221461187214612,
-      "grad_norm": 12.071101188659668,
-      "learning_rate": 1.1678082191780822e-05,
-      "loss": 0.7766,
-      "step": 26750
+      "loss": 0.7298,
+      "step": 13350
     },
     {
       "epoch": 1.2237442922374429,
-      "grad_norm": 23.268993377685547,
+      "grad_norm": 12.126824378967285,
       "learning_rate": 1.1643835616438355e-05,
-      "loss": 0.6875,
-      "step": 26800
-    },
-    {
-      "epoch": 1.226027397260274,
-      "grad_norm": 39.18050003051758,
-      "learning_rate": 1.160958904109589e-05,
-      "loss": 0.7285,
-      "step": 26850
+      "loss": 0.7482,
+      "step": 13400
     },
     {
       "epoch": 1.228310502283105,
-      "grad_norm": 3.805318832397461,
+      "grad_norm": 6.292917728424072,
       "learning_rate": 1.1575342465753425e-05,
-      "loss": 0.6275,
-      "step": 26900
-    },
-    {
-      "epoch": 1.230593607305936,
-      "grad_norm": 39.39872741699219,
-      "learning_rate": 1.154109589041096e-05,
-      "loss": 0.6889,
-      "step": 26950
+      "loss": 0.7049,
+      "step": 13450
     },
     {
       "epoch": 1.2328767123287672,
-      "grad_norm": 16.79233169555664,
+      "grad_norm": 48.449954986572266,
       "learning_rate": 1.1506849315068493e-05,
-      "loss": 0.5247,
-      "step": 27000
-    },
-    {
-      "epoch": 1.235159817351598,
-      "grad_norm": 30.6214656829834,
-      "learning_rate": 1.1472602739726027e-05,
-      "loss": 0.5567,
-      "step": 27050
+      "loss": 0.5808,
+      "step": 13500
     },
     {
       "epoch": 1.2374429223744292,
-      "grad_norm": 5.248475551605225,
+      "grad_norm": 11.724007606506348,
       "learning_rate": 1.1438356164383562e-05,
-      "loss": 0.7133,
-      "step": 27100
-    },
-    {
-      "epoch": 1.2397260273972603,
-      "grad_norm": 16.61432456970215,
-      "learning_rate": 1.1404109589041097e-05,
-      "loss": 0.6672,
-      "step": 27150
+      "loss": 0.6694,
+      "step": 13550
     },
     {
       "epoch": 1.2420091324200913,
-      "grad_norm": 16.011980056762695,
+      "grad_norm": 11.694669723510742,
       "learning_rate": 1.136986301369863e-05,
-      "loss": 0.7677,
-      "step": 27200
-    },
-    {
-      "epoch": 1.2442922374429224,
-      "grad_norm": 34.770938873291016,
-      "learning_rate": 1.1335616438356165e-05,
-      "loss": 0.6866,
-      "step": 27250
+      "loss": 0.7223,
+      "step": 13600
     },
     {
       "epoch": 1.2465753424657535,
-      "grad_norm": 39.278892517089844,
+      "grad_norm": 18.402746200561523,
       "learning_rate": 1.13013698630137e-05,
-      "loss": 0.7042,
-      "step": 27300
-    },
-    {
-      "epoch": 1.2488584474885844,
-      "grad_norm": 21.77368927001953,
-      "learning_rate": 1.1267123287671234e-05,
-      "loss": 0.6923,
-      "step": 27350
+      "loss": 0.6794,
+      "step": 13650
     },
     {
       "epoch": 1.2511415525114156,
-      "grad_norm": 8.305987358093262,
+      "grad_norm": 10.988481521606445,
       "learning_rate": 1.1232876712328767e-05,
-      "loss": 0.6222,
-      "step": 27400
-    },
-    {
-      "epoch": 1.2534246575342465,
-      "grad_norm": 27.238414764404297,
-      "learning_rate": 1.1198630136986302e-05,
-      "loss": 0.6661,
-      "step": 27450
+      "loss": 0.618,
+      "step": 13700
     },
     {
       "epoch": 1.2557077625570776,
-      "grad_norm": 25.92155647277832,
+      "grad_norm": 8.175172805786133,
       "learning_rate": 1.1164383561643837e-05,
-      "loss": 0.7408,
-      "step": 27500
-    },
-    {
-      "epoch": 1.2579908675799087,
-      "grad_norm": 16.504167556762695,
-      "learning_rate": 1.1130136986301371e-05,
-      "loss": 0.5134,
-      "step": 27550
+      "loss": 0.6552,
+      "step": 13750
     },
     {
       "epoch": 1.2602739726027397,
-      "grad_norm": 33.571617126464844,
+      "grad_norm": 17.85883903503418,
       "learning_rate": 1.1095890410958904e-05,
-      "loss": 0.7309,
-      "step": 27600
-    },
-    {
-      "epoch": 1.2625570776255708,
-      "grad_norm": 10.731171607971191,
-      "learning_rate": 1.106164383561644e-05,
-      "loss": 0.7008,
-      "step": 27650
+      "loss": 0.6645,
+      "step": 13800
     },
     {
       "epoch": 1.2648401826484017,
-      "grad_norm": 9.138223648071289,
+      "grad_norm": 6.0199198722839355,
       "learning_rate": 1.1027397260273972e-05,
-      "loss": 0.7206,
-      "step": 27700
-    },
-    {
-      "epoch": 1.2671232876712328,
-      "grad_norm": 19.61602210998535,
-      "learning_rate": 1.0993150684931507e-05,
-      "loss": 0.7169,
-      "step": 27750
+      "loss": 0.7725,
+      "step": 13850
     },
     {
       "epoch": 1.269406392694064,
-      "grad_norm": 25.721527099609375,
+      "grad_norm": 12.200461387634277,
       "learning_rate": 1.095890410958904e-05,
-      "loss": 0.6631,
-      "step": 27800
-    },
-    {
-      "epoch": 1.271689497716895,
-      "grad_norm": 15.044477462768555,
-      "learning_rate": 1.0924657534246575e-05,
-      "loss": 0.6167,
-      "step": 27850
+      "loss": 0.6685,
+      "step": 13900
     },
     {
       "epoch": 1.273972602739726,
-      "grad_norm": 13.365096092224121,
+      "grad_norm": 11.29808235168457,
       "learning_rate": 1.089041095890411e-05,
-      "loss": 0.5629,
-      "step": 27900
-    },
-    {
-      "epoch": 1.2762557077625571,
-      "grad_norm": 26.571229934692383,
-      "learning_rate": 1.0856164383561644e-05,
-      "loss": 0.6986,
-      "step": 27950
+      "loss": 0.5974,
+      "step": 13950
     },
     {
       "epoch": 1.278538812785388,
-      "grad_norm": 23.07392692565918,
+      "grad_norm": 20.522750854492188,
       "learning_rate": 1.0821917808219177e-05,
-      "loss": 0.6209,
-      "step": 28000
-    },
-    {
-      "epoch": 1.2808219178082192,
-      "grad_norm": 1.8312214612960815,
-      "learning_rate": 1.0787671232876712e-05,
-      "loss": 0.7247,
-      "step": 28050
+      "loss": 0.6497,
+      "step": 14000
     },
     {
       "epoch": 1.2831050228310503,
-      "grad_norm": 3.6338565349578857,
+      "grad_norm": 4.903714656829834,
       "learning_rate": 1.0753424657534247e-05,
-      "loss": 0.896,
-      "step": 28100
-    },
-    {
-      "epoch": 1.2853881278538812,
-      "grad_norm": 22.483678817749023,
-      "learning_rate": 1.0719178082191782e-05,
-      "loss": 0.7087,
-      "step": 28150
+      "loss": 0.805,
+      "step": 14050
     },
     {
       "epoch": 1.2876712328767124,
-      "grad_norm": 29.26844024658203,
+      "grad_norm": 14.09408950805664,
       "learning_rate": 1.0684931506849315e-05,
-      "loss": 0.7095,
-      "step": 28200
-    },
-    {
-      "epoch": 1.2899543378995433,
-      "grad_norm": 0.29415565729141235,
-      "learning_rate": 1.065068493150685e-05,
-      "loss": 0.5849,
-      "step": 28250
+      "loss": 0.6975,
+      "step": 14100
     },
     {
       "epoch": 1.2922374429223744,
-      "grad_norm": 0.7176687121391296,
+      "grad_norm": 5.306293964385986,
       "learning_rate": 1.0616438356164384e-05,
-      "loss": 0.657,
-      "step": 28300
-    },
-    {
-      "epoch": 1.2945205479452055,
-      "grad_norm": 11.345915794372559,
-      "learning_rate": 1.0582191780821919e-05,
-      "loss": 0.5598,
-      "step": 28350
+      "loss": 0.6181,
+      "step": 14150
     },
     {
       "epoch": 1.2968036529680365,
-      "grad_norm": 35.81932067871094,
+      "grad_norm": 15.217981338500977,
       "learning_rate": 1.0547945205479452e-05,
-      "loss": 0.8267,
-      "step": 28400
-    },
-    {
-      "epoch": 1.2990867579908676,
-      "grad_norm": 30.678194046020508,
-      "learning_rate": 1.0513698630136987e-05,
-      "loss": 0.6508,
-      "step": 28450
+      "loss": 0.6363,
+      "step": 14200
     },
     {
       "epoch": 1.3013698630136985,
-      "grad_norm": 3.55430006980896,
+      "grad_norm": 2.86833119392395,
       "learning_rate": 1.0479452054794521e-05,
-      "loss": 0.6641,
-      "step": 28500
-    },
-    {
-      "epoch": 1.3036529680365296,
-      "grad_norm": 18.566883087158203,
-      "learning_rate": 1.0445205479452056e-05,
-      "loss": 0.7889,
-      "step": 28550
+      "loss": 0.632,
+      "step": 14250
     },
     {
       "epoch": 1.3059360730593608,
-      "grad_norm": 19.019390106201172,
+      "grad_norm": 12.824779510498047,
       "learning_rate": 1.0410958904109589e-05,
-      "loss": 0.6177,
-      "step": 28600
-    },
-    {
-      "epoch": 1.308219178082192,
-      "grad_norm": 17.26028823852539,
-      "learning_rate": 1.0376712328767124e-05,
-      "loss": 0.7045,
-      "step": 28650
+      "loss": 0.7499,
+      "step": 14300
     },
     {
       "epoch": 1.3105022831050228,
-      "grad_norm": 33.43575668334961,
+      "grad_norm": 17.47176170349121,
       "learning_rate": 1.0342465753424659e-05,
-      "loss": 0.6046,
-      "step": 28700
-    },
-    {
-      "epoch": 1.312785388127854,
-      "grad_norm": 51.09929275512695,
-      "learning_rate": 1.0308219178082193e-05,
-      "loss": 0.5056,
-      "step": 28750
+      "loss": 0.581,
+      "step": 14350
     },
     {
       "epoch": 1.3150684931506849,
-      "grad_norm": 17.530790328979492,
+      "grad_norm": 7.620934009552002,
       "learning_rate": 1.0273972602739726e-05,
-      "loss": 0.6507,
-      "step": 28800
-    },
-    {
-      "epoch": 1.317351598173516,
-      "grad_norm": 26.142900466918945,
-      "learning_rate": 1.0239726027397261e-05,
-      "loss": 0.5859,
-      "step": 28850
+      "loss": 0.5913,
+      "step": 14400
     },
     {
       "epoch": 1.3196347031963471,
-      "grad_norm": 16.933874130249023,
+      "grad_norm": 11.1875,
       "learning_rate": 1.0205479452054796e-05,
-      "loss": 0.7274,
-      "step": 28900
-    },
-    {
-      "epoch": 1.321917808219178,
-      "grad_norm": 5.472013473510742,
-      "learning_rate": 1.017123287671233e-05,
-      "loss": 0.6445,
-      "step": 28950
+      "loss": 0.6465,
+      "step": 14450
     },
     {
       "epoch": 1.3242009132420092,
-      "grad_norm": 20.826509475708008,
+      "grad_norm": 24.72510528564453,
       "learning_rate": 1.0136986301369862e-05,
-      "loss": 0.7194,
-      "step": 29000
-    },
-    {
-      "epoch": 1.32648401826484,
-      "grad_norm": 36.42399597167969,
-      "learning_rate": 1.0102739726027397e-05,
-      "loss": 0.6405,
-      "step": 29050
+      "loss": 0.6796,
+      "step": 14500
     },
     {
       "epoch": 1.3287671232876712,
-      "grad_norm": 17.444469451904297,
+      "grad_norm": 6.102226734161377,
       "learning_rate": 1.0068493150684931e-05,
-      "loss": 0.7348,
-      "step": 29100
-    },
-    {
-      "epoch": 1.3310502283105023,
-      "grad_norm": 12.610512733459473,
-      "learning_rate": 1.0034246575342466e-05,
-      "loss": 0.6521,
-      "step": 29150
+      "loss": 0.7074,
+      "step": 14550
     },
     {
       "epoch": 1.3333333333333333,
-      "grad_norm": 14.737506866455078,
+      "grad_norm": 6.332953929901123,
       "learning_rate": 9.999999999999999e-06,
-      "loss": 0.6979,
-      "step": 29200
-    },
-    {
-      "epoch": 1.3356164383561644,
-      "grad_norm": 27.808490753173828,
-      "learning_rate": 9.965753424657534e-06,
-      "loss": 0.7839,
-      "step": 29250
+      "loss": 0.6338,
+      "step": 14600
     },
     {
       "epoch": 1.3378995433789953,
-      "grad_norm": 7.368427753448486,
+      "grad_norm": 28.30670738220215,
       "learning_rate": 9.931506849315069e-06,
-      "loss": 0.6117,
-      "step": 29300
-    },
-    {
-      "epoch": 1.3401826484018264,
-      "grad_norm": 27.834564208984375,
-      "learning_rate": 9.897260273972603e-06,
-      "loss": 0.6565,
-      "step": 29350
+      "loss": 0.7109,
+      "step": 14650
     },
     {
       "epoch": 1.3424657534246576,
-      "grad_norm": 33.823081970214844,
+      "grad_norm": 40.727230072021484,
       "learning_rate": 9.863013698630136e-06,
-      "loss": 0.6681,
-      "step": 29400
-    },
-    {
-      "epoch": 1.3447488584474887,
-      "grad_norm": 26.18691635131836,
-      "learning_rate": 9.828767123287671e-06,
-      "loss": 0.9009,
-      "step": 29450
+      "loss": 0.7342,
+      "step": 14700
     },
     {
       "epoch": 1.3470319634703196,
-      "grad_norm": 20.92985725402832,
+      "grad_norm": 11.026389122009277,
       "learning_rate": 9.794520547945206e-06,
-      "loss": 0.5297,
-      "step": 29500
-    },
-    {
-      "epoch": 1.3493150684931507,
-      "grad_norm": 4.244528770446777,
-      "learning_rate": 9.76027397260274e-06,
-      "loss": 0.6733,
-      "step": 29550
+      "loss": 0.6972,
+      "step": 14750
     },
     {
       "epoch": 1.3515981735159817,
-      "grad_norm": 67.45909118652344,
+      "grad_norm": 16.95206642150879,
       "learning_rate": 9.726027397260274e-06,
-      "loss": 0.575,
-      "step": 29600
-    },
-    {
-      "epoch": 1.3538812785388128,
-      "grad_norm": 26.713943481445312,
-      "learning_rate": 9.691780821917808e-06,
-      "loss": 0.7428,
-      "step": 29650
+      "loss": 0.6509,
+      "step": 14800
     },
     {
       "epoch": 1.356164383561644,
-      "grad_norm": 8.588297843933105,
+      "grad_norm": 24.887845993041992,
       "learning_rate": 9.657534246575343e-06,
-      "loss": 0.5567,
-      "step": 29700
-    },
-    {
-      "epoch": 1.3584474885844748,
-      "grad_norm": 22.027597427368164,
-      "learning_rate": 9.623287671232878e-06,
-      "loss": 0.7798,
-      "step": 29750
+      "loss": 0.6608,
+      "step": 14850
     },
     {
       "epoch": 1.360730593607306,
-      "grad_norm": 10.512967109680176,
+      "grad_norm": 5.1824421882629395,
       "learning_rate": 9.589041095890411e-06,
-      "loss": 0.5646,
-      "step": 29800
-    },
-    {
-      "epoch": 1.3630136986301369,
-      "grad_norm": 9.836832046508789,
-      "learning_rate": 9.554794520547946e-06,
-      "loss": 0.6904,
-      "step": 29850
+      "loss": 0.664,
+      "step": 14900
     },
     {
       "epoch": 1.365296803652968,
-      "grad_norm": 74.73491668701172,
+      "grad_norm": 35.08380889892578,
       "learning_rate": 9.52054794520548e-06,
-      "loss": 0.7029,
-      "step": 29900
-    },
-    {
-      "epoch": 1.3675799086757991,
-      "grad_norm": 36.467491149902344,
-      "learning_rate": 9.486301369863015e-06,
-      "loss": 0.7266,
-      "step": 29950
+      "loss": 0.7375,
+      "step": 14950
     },
     {
       "epoch": 1.36986301369863,
-      "grad_norm": 21.993133544921875,
+      "grad_norm": 13.273919105529785,
       "learning_rate": 9.452054794520548e-06,
-      "loss": 0.6316,
-      "step": 30000
-    },
-    {
-      "epoch": 1.3721461187214612,
-      "grad_norm": 1.2455904483795166,
-      "learning_rate": 9.417808219178083e-06,
-      "loss": 0.5555,
-      "step": 30050
+      "loss": 0.6581,
+      "step": 15000
     },
     {
       "epoch": 1.374429223744292,
-      "grad_norm": 13.85091495513916,
+      "grad_norm": 20.243751525878906,
       "learning_rate": 9.383561643835618e-06,
-      "loss": 0.7729,
-      "step": 30100
-    },
-    {
-      "epoch": 1.3767123287671232,
-      "grad_norm": 23.61090850830078,
-      "learning_rate": 9.349315068493152e-06,
-      "loss": 0.6255,
-      "step": 30150
+      "loss": 0.7028,
+      "step": 15050
     },
     {
       "epoch": 1.3789954337899544,
-      "grad_norm": 40.24787139892578,
+      "grad_norm": 6.9884934425354,
       "learning_rate": 9.315068493150685e-06,
-      "loss": 0.7646,
-      "step": 30200
-    },
-    {
-      "epoch": 1.3812785388127855,
-      "grad_norm": 23.249425888061523,
-      "learning_rate": 9.28082191780822e-06,
-      "loss": 0.5965,
-      "step": 30250
+      "loss": 0.6699,
+      "step": 15100
     },
     {
       "epoch": 1.3835616438356164,
-      "grad_norm": 8.090579986572266,
+      "grad_norm": 10.110499382019043,
       "learning_rate": 9.246575342465753e-06,
-      "loss": 0.4478,
-      "step": 30300
-    },
-    {
-      "epoch": 1.3858447488584476,
-      "grad_norm": 20.011905670166016,
-      "learning_rate": 9.212328767123288e-06,
-      "loss": 0.7777,
-      "step": 30350
+      "loss": 0.5804,
+      "step": 15150
     },
     {
       "epoch": 1.3881278538812785,
-      "grad_norm": 8.697684288024902,
+      "grad_norm": 5.272585868835449,
       "learning_rate": 9.178082191780821e-06,
-      "loss": 0.6304,
-      "step": 30400
-    },
-    {
-      "epoch": 1.3904109589041096,
-      "grad_norm": 1.545689582824707,
-      "learning_rate": 9.143835616438356e-06,
-      "loss": 0.7816,
-      "step": 30450
+      "loss": 0.7046,
+      "step": 15200
     },
     {
       "epoch": 1.3926940639269407,
-      "grad_norm": 0.6150842308998108,
+      "grad_norm": 3.3239293098449707,
       "learning_rate": 9.10958904109589e-06,
-      "loss": 0.5802,
-      "step": 30500
-    },
-    {
-      "epoch": 1.3949771689497716,
-      "grad_norm": 21.5743350982666,
-      "learning_rate": 9.075342465753425e-06,
-      "loss": 0.6369,
-      "step": 30550
+      "loss": 0.6621,
+      "step": 15250
     },
     {
       "epoch": 1.3972602739726028,
-      "grad_norm": 10.7164306640625,
+      "grad_norm": 9.12402057647705,
       "learning_rate": 9.041095890410958e-06,
-      "loss": 0.4902,
-      "step": 30600
-    },
-    {
-      "epoch": 1.3995433789954337,
-      "grad_norm": 28.598312377929688,
-      "learning_rate": 9.006849315068493e-06,
-      "loss": 0.613,
-      "step": 30650
+      "loss": 0.5613,
+      "step": 15300
     },
     {
       "epoch": 1.4018264840182648,
-      "grad_norm": 24.138431549072266,
+      "grad_norm": 9.260445594787598,
       "learning_rate": 8.972602739726028e-06,
-      "loss": 0.6153,
-      "step": 30700
-    },
-    {
-      "epoch": 1.404109589041096,
-      "grad_norm": 22.55198860168457,
-      "learning_rate": 8.938356164383562e-06,
-      "loss": 0.5419,
-      "step": 30750
+      "loss": 0.6179,
+      "step": 15350
     },
     {
       "epoch": 1.4063926940639269,
-      "grad_norm": 26.992374420166016,
+      "grad_norm": 40.01394271850586,
       "learning_rate": 8.904109589041095e-06,
-      "loss": 0.5726,
-      "step": 30800
-    },
-    {
-      "epoch": 1.408675799086758,
-      "grad_norm": 1.8640153408050537,
-      "learning_rate": 8.86986301369863e-06,
-      "loss": 0.6794,
-      "step": 30850
+      "loss": 0.5897,
+      "step": 15400
     },
     {
       "epoch": 1.410958904109589,
-      "grad_norm": 14.333259582519531,
+      "grad_norm": 43.27082443237305,
       "learning_rate": 8.835616438356165e-06,
-      "loss": 0.4972,
-      "step": 30900
-    },
-    {
-      "epoch": 1.41324200913242,
-      "grad_norm": 64.34575653076172,
-      "learning_rate": 8.8013698630137e-06,
-      "loss": 0.9213,
-      "step": 30950
+      "loss": 0.5804,
+      "step": 15450
     },
     {
       "epoch": 1.4155251141552512,
-      "grad_norm": 9.60162353515625,
+      "grad_norm": 8.149184226989746,
       "learning_rate": 8.767123287671233e-06,
-      "loss": 0.6868,
-      "step": 31000
-    },
-    {
-      "epoch": 1.4178082191780823,
-      "grad_norm": 8.155502319335938,
-      "learning_rate": 8.732876712328767e-06,
-      "loss": 0.6873,
-      "step": 31050
+      "loss": 0.7865,
+      "step": 15500
     },
     {
       "epoch": 1.4200913242009132,
-      "grad_norm": 5.085892677307129,
+      "grad_norm": 16.708332061767578,
       "learning_rate": 8.698630136986302e-06,
-      "loss": 0.7768,
-      "step": 31100
-    },
-    {
-      "epoch": 1.4223744292237444,
-      "grad_norm": 52.48747634887695,
-      "learning_rate": 8.664383561643837e-06,
-      "loss": 0.687,
-      "step": 31150
+      "loss": 0.6983,
+      "step": 15550
     },
     {
       "epoch": 1.4246575342465753,
-      "grad_norm": 1.8491209745407104,
+      "grad_norm": 2.826059103012085,
       "learning_rate": 8.63013698630137e-06,
-      "loss": 0.5669,
-      "step": 31200
-    },
-    {
-      "epoch": 1.4269406392694064,
-      "grad_norm": 12.143204689025879,
-      "learning_rate": 8.595890410958905e-06,
-      "loss": 0.5568,
-      "step": 31250
+      "loss": 0.6238,
+      "step": 15600
     },
     {
       "epoch": 1.4292237442922375,
-      "grad_norm": 2.939903974533081,
+      "grad_norm": 13.411745071411133,
       "learning_rate": 8.56164383561644e-06,
-      "loss": 0.7732,
-      "step": 31300
-    },
-    {
-      "epoch": 1.4315068493150684,
-      "grad_norm": 17.773130416870117,
-      "learning_rate": 8.527397260273974e-06,
-      "loss": 0.7444,
-      "step": 31350
+      "loss": 0.6448,
+      "step": 15650
     },
     {
       "epoch": 1.4337899543378996,
-      "grad_norm": 47.27958679199219,
+      "grad_norm": 12.094820976257324,
       "learning_rate": 8.493150684931507e-06,
-      "loss": 0.6621,
-      "step": 31400
-    },
-    {
-      "epoch": 1.4360730593607305,
-      "grad_norm": 50.40327453613281,
-      "learning_rate": 8.458904109589042e-06,
-      "loss": 0.84,
-      "step": 31450
+      "loss": 0.6868,
+      "step": 15700
     },
     {
       "epoch": 1.4383561643835616,
-      "grad_norm": 8.335402488708496,
+      "grad_norm": 13.272956848144531,
       "learning_rate": 8.424657534246577e-06,
-      "loss": 0.6762,
-      "step": 31500
-    },
-    {
-      "epoch": 1.4406392694063928,
-      "grad_norm": 12.027316093444824,
-      "learning_rate": 8.390410958904111e-06,
-      "loss": 0.6736,
-      "step": 31550
+      "loss": 0.6758,
+      "step": 15750
     },
     {
       "epoch": 1.4429223744292237,
-      "grad_norm": 17.410192489624023,
+      "grad_norm": 7.710869312286377,
       "learning_rate": 8.356164383561643e-06,
-      "loss": 0.5072,
-      "step": 31600
-    },
-    {
-      "epoch": 1.4452054794520548,
-      "grad_norm": 48.263450622558594,
-      "learning_rate": 8.321917808219178e-06,
-      "loss": 0.6268,
-      "step": 31650
+      "loss": 0.6154,
+      "step": 15800
     },
     {
       "epoch": 1.4474885844748857,
-      "grad_norm": 3.8568694591522217,
+      "grad_norm": 24.845901489257812,
       "learning_rate": 8.287671232876712e-06,
-      "loss": 0.5454,
-      "step": 31700
-    },
-    {
-      "epoch": 1.4497716894977168,
-      "grad_norm": 13.764704704284668,
-      "learning_rate": 8.253424657534247e-06,
-      "loss": 0.6823,
-      "step": 31750
+      "loss": 0.6228,
+      "step": 15850
     },
     {
       "epoch": 1.452054794520548,
-      "grad_norm": 13.48620319366455,
+      "grad_norm": 10.101279258728027,
       "learning_rate": 8.21917808219178e-06,
-      "loss": 0.7103,
-      "step": 31800
-    },
-    {
-      "epoch": 1.454337899543379,
-      "grad_norm": 17.291501998901367,
-      "learning_rate": 8.184931506849315e-06,
-      "loss": 0.7011,
-      "step": 31850
+      "loss": 0.7418,
+      "step": 15900
     },
     {
       "epoch": 1.45662100456621,
-      "grad_norm": 1.461418867111206,
+      "grad_norm": 71.19440460205078,
       "learning_rate": 8.15068493150685e-06,
-      "loss": 0.6667,
-      "step": 31900
-    },
-    {
-      "epoch": 1.4589041095890412,
-      "grad_norm": 16.34942626953125,
-      "learning_rate": 8.116438356164384e-06,
-      "loss": 0.7885,
-      "step": 31950
+      "loss": 0.7128,
+      "step": 15950
     },
     {
       "epoch": 1.461187214611872,
-      "grad_norm": 14.74634075164795,
+      "grad_norm": 6.2137041091918945,
       "learning_rate": 8.082191780821917e-06,
-      "loss": 0.644,
-      "step": 32000
-    },
-    {
-      "epoch": 1.4634703196347032,
-      "grad_norm": 6.794888973236084,
-      "learning_rate": 8.047945205479452e-06,
-      "loss": 0.6738,
-      "step": 32050
+      "loss": 0.7121,
+      "step": 16000
     },
     {
       "epoch": 1.4657534246575343,
-      "grad_norm": 31.303226470947266,
+      "grad_norm": 20.703536987304688,
       "learning_rate": 8.013698630136987e-06,
-      "loss": 0.8235,
-      "step": 32100
-    },
-    {
-      "epoch": 1.4680365296803652,
-      "grad_norm": 42.993648529052734,
-      "learning_rate": 7.979452054794521e-06,
-      "loss": 0.4712,
-      "step": 32150
+      "loss": 0.7115,
+      "step": 16050
     },
     {
       "epoch": 1.4703196347031964,
-      "grad_norm": 7.875132083892822,
+      "grad_norm": 26.53441619873047,
       "learning_rate": 7.945205479452055e-06,
-      "loss": 0.5951,
-      "step": 32200
-    },
-    {
-      "epoch": 1.4726027397260273,
-      "grad_norm": 9.124963760375977,
-      "learning_rate": 7.91095890410959e-06,
-      "loss": 0.5968,
-      "step": 32250
+      "loss": 0.5534,
+      "step": 16100
     },
     {
       "epoch": 1.4748858447488584,
-      "grad_norm": 13.793811798095703,
+      "grad_norm": 20.233125686645508,
       "learning_rate": 7.876712328767124e-06,
-      "loss": 0.6484,
-      "step": 32300
-    },
-    {
-      "epoch": 1.4771689497716896,
-      "grad_norm": 2.1718921661376953,
-      "learning_rate": 7.842465753424659e-06,
-      "loss": 0.6279,
-      "step": 32350
+      "loss": 0.6204,
+      "step": 16150
     },
     {
       "epoch": 1.4794520547945205,
-      "grad_norm": 60.621543884277344,
+      "grad_norm": 23.36627769470215,
       "learning_rate": 7.808219178082192e-06,
-      "loss": 0.6718,
-      "step": 32400
-    },
-    {
-      "epoch": 1.4817351598173516,
-      "grad_norm": 6.748918533325195,
-      "learning_rate": 7.773972602739727e-06,
-      "loss": 0.6333,
-      "step": 32450
+      "loss": 0.6914,
+      "step": 16200
     },
     {
       "epoch": 1.4840182648401825,
-      "grad_norm": 10.061300277709961,
+      "grad_norm": 11.94163703918457,
       "learning_rate": 7.739726027397261e-06,
-      "loss": 0.5952,
-      "step": 32500
-    },
-    {
-      "epoch": 1.4863013698630136,
-      "grad_norm": 55.56220245361328,
-      "learning_rate": 7.705479452054796e-06,
-      "loss": 0.6297,
-      "step": 32550
+      "loss": 0.6062,
+      "step": 16250
     },
     {
       "epoch": 1.4885844748858448,
-      "grad_norm": 31.07186508178711,
+      "grad_norm": 15.757901191711426,
       "learning_rate": 7.671232876712329e-06,
-      "loss": 0.5302,
-      "step": 32600
-    },
-    {
-      "epoch": 1.490867579908676,
-      "grad_norm": 30.925626754760742,
-      "learning_rate": 7.636986301369864e-06,
-      "loss": 0.6608,
-      "step": 32650
+      "loss": 0.6142,
+      "step": 16300
     },
     {
       "epoch": 1.4931506849315068,
-      "grad_norm": 21.15188217163086,
+      "grad_norm": 12.007556915283203,
       "learning_rate": 7.602739726027398e-06,
-      "loss": 0.6838,
-      "step": 32700
-    },
-    {
-      "epoch": 1.495433789954338,
-      "grad_norm": 15.808161735534668,
-      "learning_rate": 7.568493150684932e-06,
-      "loss": 0.5405,
-      "step": 32750
+      "loss": 0.6708,
+      "step": 16350
     },
     {
       "epoch": 1.4977168949771689,
-      "grad_norm": 11.866249084472656,
+      "grad_norm": 9.127739906311035,
       "learning_rate": 7.5342465753424655e-06,
-      "loss": 0.6924,
-      "step": 32800
-    },
-    {
-      "epoch": 1.5,
-      "grad_norm": 29.02684783935547,
-      "learning_rate": 7.5e-06,
-      "loss": 0.6354,
-      "step": 32850
+      "loss": 0.6319,
+      "step": 16400
     },
     {
       "epoch": 1.5022831050228311,
-      "grad_norm": 20.26506996154785,
+      "grad_norm": 14.81264877319336,
       "learning_rate": 7.465753424657534e-06,
-      "loss": 0.6467,
-      "step": 32900
-    },
-    {
-      "epoch": 1.5045662100456623,
-      "grad_norm": 23.63490867614746,
-      "learning_rate": 7.431506849315069e-06,
-      "loss": 0.6682,
-      "step": 32950
+      "loss": 0.6143,
+      "step": 16450
     },
     {
       "epoch": 1.5068493150684932,
-      "grad_norm": 16.075380325317383,
+      "grad_norm": 8.986160278320312,
       "learning_rate": 7.397260273972603e-06,
-      "loss": 0.6922,
-      "step": 33000
-    },
-    {
-      "epoch": 1.509132420091324,
-      "grad_norm": 14.159255027770996,
-      "learning_rate": 7.3630136986301374e-06,
-      "loss": 0.6063,
-      "step": 33050
+      "loss": 0.7508,
+      "step": 16500
     },
     {
       "epoch": 1.5114155251141552,
-      "grad_norm": 22.143796920776367,
+      "grad_norm": 23.52107810974121,
       "learning_rate": 7.328767123287671e-06,
-      "loss": 0.7155,
-      "step": 33100
-    },
-    {
-      "epoch": 1.5136986301369864,
-      "grad_norm": 98.22097778320312,
-      "learning_rate": 7.294520547945206e-06,
-      "loss": 0.7567,
-      "step": 33150
+      "loss": 0.6377,
+      "step": 16550
     },
     {
       "epoch": 1.5159817351598175,
-      "grad_norm": 0.7336256504058838,
+      "grad_norm": 4.14312219619751,
       "learning_rate": 7.260273972602739e-06,
-      "loss": 0.5414,
-      "step": 33200
-    },
-    {
-      "epoch": 1.5182648401826484,
-      "grad_norm": 0.3773713707923889,
-      "learning_rate": 7.226027397260274e-06,
-      "loss": 0.5702,
-      "step": 33250
+      "loss": 0.595,
+      "step": 16600
     },
     {
       "epoch": 1.5205479452054793,
-      "grad_norm": 8.909625053405762,
+      "grad_norm": 32.40318298339844,
       "learning_rate": 7.191780821917808e-06,
-      "loss": 0.7228,
-      "step": 33300
-    },
-    {
-      "epoch": 1.5228310502283104,
-      "grad_norm": 21.098960876464844,
-      "learning_rate": 7.1575342465753425e-06,
-      "loss": 0.6001,
-      "step": 33350
+      "loss": 0.6177,
+      "step": 16650
     },
     {
       "epoch": 1.5251141552511416,
-      "grad_norm": 15.906450271606445,
+      "grad_norm": 13.583507537841797,
       "learning_rate": 7.123287671232876e-06,
-      "loss": 0.6303,
-      "step": 33400
-    },
-    {
-      "epoch": 1.5273972602739727,
-      "grad_norm": 24.9348201751709,
-      "learning_rate": 7.089041095890411e-06,
-      "loss": 0.673,
-      "step": 33450
+      "loss": 0.6513,
+      "step": 16700
     },
     {
       "epoch": 1.5296803652968036,
-      "grad_norm": 8.255683898925781,
+      "grad_norm": 27.02487564086914,
       "learning_rate": 7.054794520547945e-06,
-      "loss": 0.5831,
-      "step": 33500
-    },
-    {
-      "epoch": 1.5319634703196348,
-      "grad_norm": 73.46847534179688,
-      "learning_rate": 7.02054794520548e-06,
-      "loss": 0.7362,
-      "step": 33550
+      "loss": 0.6136,
+      "step": 16750
     },
     {
       "epoch": 1.5342465753424657,
-      "grad_norm": 88.85016632080078,
+      "grad_norm": 46.82355880737305,
       "learning_rate": 6.986301369863014e-06,
-      "loss": 0.6682,
-      "step": 33600
-    },
-    {
-      "epoch": 1.5365296803652968,
-      "grad_norm": 52.53008270263672,
-      "learning_rate": 6.952054794520548e-06,
-      "loss": 0.5667,
-      "step": 33650
+      "loss": 0.6545,
+      "step": 16800
     },
     {
       "epoch": 1.538812785388128,
-      "grad_norm": 18.00398826599121,
+      "grad_norm": 11.266030311584473,
       "learning_rate": 6.917808219178082e-06,
-      "loss": 0.6214,
-      "step": 33700
-    },
-    {
-      "epoch": 1.541095890410959,
-      "grad_norm": 27.124656677246094,
-      "learning_rate": 6.883561643835617e-06,
-      "loss": 0.6737,
-      "step": 33750
+      "loss": 0.6438,
+      "step": 16850
     },
     {
       "epoch": 1.54337899543379,
-      "grad_norm": 39.45083999633789,
+      "grad_norm": 21.652450561523438,
       "learning_rate": 6.849315068493151e-06,
-      "loss": 0.703,
-      "step": 33800
-    },
-    {
-      "epoch": 1.545662100456621,
-      "grad_norm": 0.20495979487895966,
-      "learning_rate": 6.815068493150685e-06,
-      "loss": 0.6271,
-      "step": 33850
+      "loss": 0.6617,
+      "step": 16900
     },
     {
       "epoch": 1.547945205479452,
-      "grad_norm": 0.8208453059196472,
+      "grad_norm": 18.254072189331055,
       "learning_rate": 6.780821917808219e-06,
-      "loss": 0.6804,
-      "step": 33900
-    },
-    {
-      "epoch": 1.5502283105022832,
-      "grad_norm": 12.416110038757324,
-      "learning_rate": 6.746575342465753e-06,
-      "loss": 0.7786,
-      "step": 33950
+      "loss": 0.645,
+      "step": 16950
     },
     {
       "epoch": 1.5525114155251143,
-      "grad_norm": 21.69839096069336,
+      "grad_norm": 7.959854602813721,
       "learning_rate": 6.712328767123287e-06,
-      "loss": 0.7048,
-      "step": 34000
-    },
-    {
-      "epoch": 1.5547945205479452,
-      "grad_norm": 57.30881881713867,
-      "learning_rate": 6.678082191780822e-06,
-      "loss": 0.7398,
-      "step": 34050
+      "loss": 0.7444,
+      "step": 17000
     },
     {
       "epoch": 1.5570776255707761,
-      "grad_norm": 22.58492088317871,
+      "grad_norm": 15.169598579406738,
       "learning_rate": 6.643835616438356e-06,
-      "loss": 0.7027,
-      "step": 34100
-    },
-    {
-      "epoch": 1.5593607305936072,
-      "grad_norm": 8.803092002868652,
-      "learning_rate": 6.609589041095891e-06,
-      "loss": 0.7296,
-      "step": 34150
+      "loss": 0.6788,
+      "step": 17050
     },
     {
       "epoch": 1.5616438356164384,
-      "grad_norm": 18.931156158447266,
+      "grad_norm": 14.872618675231934,
       "learning_rate": 6.5753424657534245e-06,
-      "loss": 0.7622,
-      "step": 34200
-    },
-    {
-      "epoch": 1.5639269406392695,
-      "grad_norm": 8.858073234558105,
-      "learning_rate": 6.541095890410959e-06,
-      "loss": 0.615,
-      "step": 34250
+      "loss": 0.7433,
+      "step": 17100
     },
     {
       "epoch": 1.5662100456621004,
-      "grad_norm": 6.284381866455078,
+      "grad_norm": 12.479876518249512,
       "learning_rate": 6.506849315068493e-06,
-      "loss": 0.7147,
-      "step": 34300
-    },
-    {
-      "epoch": 1.5684931506849316,
-      "grad_norm": 21.08570098876953,
-      "learning_rate": 6.472602739726028e-06,
-      "loss": 0.6153,
-      "step": 34350
+      "loss": 0.676,
+      "step": 17150
     },
     {
       "epoch": 1.5707762557077625,
-      "grad_norm": 2.0850419998168945,
+      "grad_norm": 4.30610990524292,
       "learning_rate": 6.438356164383562e-06,
-      "loss": 0.588,
-      "step": 34400
-    },
-    {
-      "epoch": 1.5730593607305936,
-      "grad_norm": 0.25530076026916504,
-      "learning_rate": 6.4041095890410965e-06,
-      "loss": 0.7249,
-      "step": 34450
+      "loss": 0.597,
+      "step": 17200
     },
     {
       "epoch": 1.5753424657534247,
-      "grad_norm": 31.284807205200195,
+      "grad_norm": 9.52687931060791,
       "learning_rate": 6.3698630136986296e-06,
-      "loss": 0.6106,
-      "step": 34500
-    },
-    {
-      "epoch": 1.5776255707762559,
-      "grad_norm": 19.524412155151367,
-      "learning_rate": 6.335616438356164e-06,
-      "loss": 0.7259,
-      "step": 34550
+      "loss": 0.6084,
+      "step": 17250
     },
     {
       "epoch": 1.5799086757990868,
-      "grad_norm": 6.005446910858154,
+      "grad_norm": 6.067666053771973,
       "learning_rate": 6.301369863013698e-06,
-      "loss": 0.5231,
-      "step": 34600
-    },
-    {
-      "epoch": 1.5821917808219177,
-      "grad_norm": 17.577402114868164,
-      "learning_rate": 6.267123287671233e-06,
-      "loss": 0.847,
-      "step": 34650
+      "loss": 0.61,
+      "step": 17300
     },
     {
       "epoch": 1.5844748858447488,
-      "grad_norm": 31.817855834960938,
+      "grad_norm": 15.737329483032227,
       "learning_rate": 6.232876712328767e-06,
-      "loss": 0.5509,
-      "step": 34700
-    },
-    {
-      "epoch": 1.58675799086758,
-      "grad_norm": 4.6908063888549805,
-      "learning_rate": 6.1986301369863016e-06,
-      "loss": 0.752,
-      "step": 34750
+      "loss": 0.7011,
+      "step": 17350
     },
     {
       "epoch": 1.589041095890411,
-      "grad_norm": 2.8228561878204346,
+      "grad_norm": 4.7880730628967285,
       "learning_rate": 6.1643835616438354e-06,
-      "loss": 0.7156,
-      "step": 34800
-    },
-    {
-      "epoch": 1.591324200913242,
-      "grad_norm": 7.878891468048096,
-      "learning_rate": 6.13013698630137e-06,
-      "loss": 0.6926,
-      "step": 34850
+      "loss": 0.694,
+      "step": 17400
     },
     {
       "epoch": 1.593607305936073,
-      "grad_norm": 30.530006408691406,
+      "grad_norm": 19.6992130279541,
       "learning_rate": 6.095890410958904e-06,
-      "loss": 0.7552,
-      "step": 34900
-    },
-    {
-      "epoch": 1.595890410958904,
-      "grad_norm": 29.396806716918945,
-      "learning_rate": 6.061643835616439e-06,
-      "loss": 0.7008,
-      "step": 34950
+      "loss": 0.706,
+      "step": 17450
     },
     {
       "epoch": 1.5981735159817352,
-      "grad_norm": 10.500929832458496,
+      "grad_norm": 10.835814476013184,
       "learning_rate": 6.027397260273973e-06,
-      "loss": 0.4952,
-      "step": 35000
-    },
-    {
-      "epoch": 1.6004566210045663,
-      "grad_norm": 2.337519407272339,
-      "learning_rate": 5.9931506849315074e-06,
-      "loss": 0.5864,
-      "step": 35050
+      "loss": 0.6298,
+      "step": 17500
     },
     {
       "epoch": 1.6027397260273972,
-      "grad_norm": 8.646376609802246,
+      "grad_norm": 13.219555854797363,
       "learning_rate": 5.958904109589041e-06,
-      "loss": 0.7454,
-      "step": 35100
-    },
-    {
-      "epoch": 1.6050228310502284,
-      "grad_norm": 17.153099060058594,
-      "learning_rate": 5.924657534246575e-06,
-      "loss": 0.6943,
-      "step": 35150
+      "loss": 0.6496,
+      "step": 17550
     },
     {
       "epoch": 1.6073059360730593,
-      "grad_norm": 25.350088119506836,
+      "grad_norm": 22.237091064453125,
       "learning_rate": 5.890410958904109e-06,
-      "loss": 0.6803,
-      "step": 35200
-    },
-    {
-      "epoch": 1.6095890410958904,
-      "grad_norm": 4.12929105758667,
-      "learning_rate": 5.856164383561644e-06,
-      "loss": 0.6976,
-      "step": 35250
+      "loss": 0.6585,
+      "step": 17600
     },
     {
       "epoch": 1.6118721461187215,
-      "grad_norm": 14.61955451965332,
+      "grad_norm": 12.173138618469238,
       "learning_rate": 5.821917808219178e-06,
-      "loss": 0.7487,
-      "step": 35300
-    },
-    {
-      "epoch": 1.6141552511415527,
-      "grad_norm": 6.208589553833008,
-      "learning_rate": 5.7876712328767125e-06,
-      "loss": 0.6569,
-      "step": 35350
+      "loss": 0.712,
+      "step": 17650
     },
     {
       "epoch": 1.6164383561643836,
-      "grad_norm": 9.5521240234375,
+      "grad_norm": 6.5948896408081055,
       "learning_rate": 5.753424657534246e-06,
-      "loss": 0.6505,
-      "step": 35400
-    },
-    {
-      "epoch": 1.6187214611872145,
-      "grad_norm": 14.391396522521973,
-      "learning_rate": 5.719178082191781e-06,
-      "loss": 0.5558,
-      "step": 35450
+      "loss": 0.6253,
+      "step": 17700
     },
     {
       "epoch": 1.6210045662100456,
-      "grad_norm": 1.2627131938934326,
+      "grad_norm": 5.447400093078613,
       "learning_rate": 5.684931506849315e-06,
-      "loss": 0.6405,
-      "step": 35500
-    },
-    {
-      "epoch": 1.6232876712328768,
-      "grad_norm": 83.7956314086914,
-      "learning_rate": 5.65068493150685e-06,
-      "loss": 0.5483,
-      "step": 35550
+      "loss": 0.5762,
+      "step": 17750
     },
     {
       "epoch": 1.625570776255708,
-      "grad_norm": 15.18497085571289,
+      "grad_norm": 12.413744926452637,
       "learning_rate": 5.616438356164384e-06,
-      "loss": 0.6692,
-      "step": 35600
-    },
-    {
-      "epoch": 1.6278538812785388,
-      "grad_norm": 33.16044998168945,
-      "learning_rate": 5.582191780821918e-06,
-      "loss": 0.6939,
-      "step": 35650
+      "loss": 0.6272,
+      "step": 17800
     },
     {
       "epoch": 1.6301369863013697,
-      "grad_norm": 12.063103675842285,
+      "grad_norm": 48.877052307128906,
       "learning_rate": 5.547945205479452e-06,
-      "loss": 0.5414,
-      "step": 35700
-    },
-    {
-      "epoch": 1.6324200913242009,
-      "grad_norm": 26.803749084472656,
-      "learning_rate": 5.513698630136986e-06,
-      "loss": 0.5353,
-      "step": 35750
+      "loss": 0.589,
+      "step": 17850
     },
     {
       "epoch": 1.634703196347032,
-      "grad_norm": 6.5856523513793945,
+      "grad_norm": 21.571834564208984,
       "learning_rate": 5.47945205479452e-06,
-      "loss": 0.7678,
-      "step": 35800
-    },
-    {
-      "epoch": 1.6369863013698631,
-      "grad_norm": 14.661989212036133,
-      "learning_rate": 5.445205479452055e-06,
-      "loss": 0.67,
-      "step": 35850
+      "loss": 0.6592,
+      "step": 17900
     },
     {
       "epoch": 1.639269406392694,
-      "grad_norm": 2.4577255249023438,
+      "grad_norm": 7.417336940765381,
       "learning_rate": 5.410958904109589e-06,
-      "loss": 0.4579,
-      "step": 35900
-    },
-    {
-      "epoch": 1.6415525114155252,
-      "grad_norm": 17.849546432495117,
-      "learning_rate": 5.376712328767123e-06,
-      "loss": 0.7221,
-      "step": 35950
+      "loss": 0.5248,
+      "step": 17950
     },
     {
       "epoch": 1.643835616438356,
-      "grad_norm": 6.453017234802246,
+      "grad_norm": 48.4299430847168,
       "learning_rate": 5.342465753424657e-06,
-      "loss": 0.6889,
-      "step": 36000
-    },
-    {
-      "epoch": 1.6461187214611872,
-      "grad_norm": 15.138044357299805,
-      "learning_rate": 5.308219178082192e-06,
-      "loss": 0.6234,
-      "step": 36050
+      "loss": 0.6622,
+      "step": 18000
     },
     {
       "epoch": 1.6484018264840183,
-      "grad_norm": 11.393730163574219,
+      "grad_norm": 17.760732650756836,
       "learning_rate": 5.273972602739726e-06,
-      "loss": 0.6353,
-      "step": 36100
-    },
-    {
-      "epoch": 1.6506849315068495,
-      "grad_norm": 2.726991891860962,
-      "learning_rate": 5.239726027397261e-06,
-      "loss": 0.5305,
-      "step": 36150
+      "loss": 0.626,
+      "step": 18050
     },
     {
       "epoch": 1.6529680365296804,
-      "grad_norm": 14.54566478729248,
+      "grad_norm": 14.000800132751465,
       "learning_rate": 5.2054794520547945e-06,
-      "loss": 0.8252,
-      "step": 36200
-    },
-    {
-      "epoch": 1.6552511415525113,
-      "grad_norm": 0.952422022819519,
-      "learning_rate": 5.171232876712329e-06,
-      "loss": 0.6773,
-      "step": 36250
+      "loss": 0.6084,
+      "step": 18100
     },
     {
       "epoch": 1.6575342465753424,
-      "grad_norm": 21.44168472290039,
+      "grad_norm": 8.742088317871094,
       "learning_rate": 5.136986301369863e-06,
-      "loss": 0.6106,
-      "step": 36300
-    },
-    {
-      "epoch": 1.6598173515981736,
-      "grad_norm": 11.663095474243164,
-      "learning_rate": 5.102739726027398e-06,
-      "loss": 0.6514,
-      "step": 36350
+      "loss": 0.6029,
+      "step": 18150
     },
     {
       "epoch": 1.6621004566210047,
-      "grad_norm": 23.589557647705078,
+      "grad_norm": 26.16733169555664,
       "learning_rate": 5.068493150684931e-06,
-      "loss": 0.708,
-      "step": 36400
-    },
-    {
-      "epoch": 1.6643835616438356,
-      "grad_norm": 10.283199310302734,
-      "learning_rate": 5.034246575342466e-06,
-      "loss": 0.5563,
-      "step": 36450
+      "loss": 0.6828,
+      "step": 18200
     },
     {
       "epoch": 1.6666666666666665,
-      "grad_norm": 44.115047454833984,
+      "grad_norm": 29.90041160583496,
       "learning_rate": 4.9999999999999996e-06,
-      "loss": 0.5912,
-      "step": 36500
-    },
-    {
-      "epoch": 1.6689497716894977,
-      "grad_norm": 71.8247299194336,
-      "learning_rate": 4.965753424657534e-06,
-      "loss": 0.7993,
-      "step": 36550
+      "loss": 0.5702,
+      "step": 18250
     },
     {
       "epoch": 1.6712328767123288,
-      "grad_norm": 11.808229446411133,
+      "grad_norm": 15.568696022033691,
       "learning_rate": 4.931506849315068e-06,
-      "loss": 0.5851,
-      "step": 36600
-    },
-    {
-      "epoch": 1.67351598173516,
-      "grad_norm": 74.73955535888672,
-      "learning_rate": 4.897260273972603e-06,
-      "loss": 0.7004,
-      "step": 36650
+      "loss": 0.6376,
+      "step": 18300
     },
     {
       "epoch": 1.6757990867579908,
-      "grad_norm": 36.06229019165039,
+      "grad_norm": 15.59715747833252,
       "learning_rate": 4.863013698630137e-06,
-      "loss": 0.7382,
-      "step": 36700
-    },
-    {
-      "epoch": 1.678082191780822,
-      "grad_norm": 53.29566955566406,
-      "learning_rate": 4.8287671232876716e-06,
-      "loss": 0.5035,
-      "step": 36750
+      "loss": 0.6776,
+      "step": 18350
     },
     {
       "epoch": 1.6803652968036529,
-      "grad_norm": 21.9272403717041,
+      "grad_norm": 11.07044506072998,
       "learning_rate": 4.7945205479452054e-06,
-      "loss": 0.5309,
-      "step": 36800
-    },
-    {
-      "epoch": 1.682648401826484,
-      "grad_norm": 39.56712341308594,
-      "learning_rate": 4.76027397260274e-06,
-      "loss": 0.4754,
-      "step": 36850
+      "loss": 0.529,
+      "step": 18400
     },
     {
       "epoch": 1.6849315068493151,
-      "grad_norm": 2.814680576324463,
+      "grad_norm": 5.349613666534424,
       "learning_rate": 4.726027397260274e-06,
-      "loss": 0.7033,
-      "step": 36900
-    },
-    {
-      "epoch": 1.6872146118721463,
-      "grad_norm": 95.83110809326172,
-      "learning_rate": 4.691780821917809e-06,
-      "loss": 0.6999,
-      "step": 36950
+      "loss": 0.692,
+      "step": 18450
     },
     {
       "epoch": 1.6894977168949772,
-      "grad_norm": 27.638185501098633,
+      "grad_norm": 18.147794723510742,
       "learning_rate": 4.657534246575343e-06,
-      "loss": 0.6522,
-      "step": 37000
-    },
-    {
-      "epoch": 1.691780821917808,
-      "grad_norm": 10.899153709411621,
-      "learning_rate": 4.623287671232877e-06,
-      "loss": 0.6386,
-      "step": 37050
+      "loss": 0.7057,
+      "step": 18500
     },
     {
       "epoch": 1.6940639269406392,
-      "grad_norm": 21.410276412963867,
+      "grad_norm": 13.486499786376953,
       "learning_rate": 4.5890410958904105e-06,
-      "loss": 0.695,
-      "step": 37100
-    },
-    {
-      "epoch": 1.6963470319634704,
-      "grad_norm": 15.208582878112793,
-      "learning_rate": 4.554794520547945e-06,
-      "loss": 0.7636,
-      "step": 37150
+      "loss": 0.6415,
+      "step": 18550
     },
     {
       "epoch": 1.6986301369863015,
-      "grad_norm": 19.083850860595703,
+      "grad_norm": 10.16304874420166,
       "learning_rate": 4.520547945205479e-06,
-      "loss": 0.6331,
-      "step": 37200
-    },
-    {
-      "epoch": 1.7009132420091324,
-      "grad_norm": 4.408557415008545,
-      "learning_rate": 4.486301369863014e-06,
-      "loss": 0.6886,
-      "step": 37250
+      "loss": 0.6604,
+      "step": 18600
     },
     {
       "epoch": 1.7031963470319633,
-      "grad_norm": 10.206310272216797,
+      "grad_norm": 29.04235076904297,
       "learning_rate": 4.452054794520548e-06,
-      "loss": 0.639,
-      "step": 37300
-    },
-    {
-      "epoch": 1.7054794520547945,
-      "grad_norm": 18.985891342163086,
-      "learning_rate": 4.4178082191780825e-06,
-      "loss": 0.6884,
-      "step": 37350
+      "loss": 0.6773,
+      "step": 18650
     },
     {
       "epoch": 1.7077625570776256,
-      "grad_norm": 16.533288955688477,
+      "grad_norm": 9.932119369506836,
       "learning_rate": 4.383561643835616e-06,
-      "loss": 0.6446,
-      "step": 37400
-    },
-    {
-      "epoch": 1.7100456621004567,
-      "grad_norm": 25.728469848632812,
-      "learning_rate": 4.349315068493151e-06,
-      "loss": 0.4863,
-      "step": 37450
+      "loss": 0.617,
+      "step": 18700
     },
     {
       "epoch": 1.7123287671232876,
-      "grad_norm": 88.2020492553711,
+      "grad_norm": 70.96830749511719,
       "learning_rate": 4.315068493150685e-06,
-      "loss": 0.6831,
-      "step": 37500
-    },
-    {
-      "epoch": 1.7146118721461188,
-      "grad_norm": 44.737815856933594,
-      "learning_rate": 4.28082191780822e-06,
-      "loss": 0.4687,
-      "step": 37550
+      "loss": 0.6157,
+      "step": 18750
     },
     {
       "epoch": 1.7168949771689497,
-      "grad_norm": 1.9043503999710083,
+      "grad_norm": 15.635278701782227,
       "learning_rate": 4.246575342465754e-06,
-      "loss": 0.6608,
-      "step": 37600
-    },
-    {
-      "epoch": 1.7191780821917808,
-      "grad_norm": 11.180625915527344,
-      "learning_rate": 4.212328767123288e-06,
-      "loss": 0.7672,
-      "step": 37650
+      "loss": 0.6127,
+      "step": 18800
     },
     {
       "epoch": 1.721461187214612,
-      "grad_norm": 3.448392391204834,
+      "grad_norm": 2.3553667068481445,
       "learning_rate": 4.178082191780821e-06,
-      "loss": 0.6176,
-      "step": 37700
-    },
-    {
-      "epoch": 1.723744292237443,
-      "grad_norm": 10.672887802124023,
-      "learning_rate": 4.143835616438356e-06,
-      "loss": 0.5263,
-      "step": 37750
+      "loss": 0.6847,
+      "step": 18850
     },
     {
       "epoch": 1.726027397260274,
-      "grad_norm": 15.69261360168457,
+      "grad_norm": 8.880425453186035,
       "learning_rate": 4.10958904109589e-06,
-      "loss": 0.4637,
-      "step": 37800
-    },
-    {
-      "epoch": 1.728310502283105,
-      "grad_norm": 19.786346435546875,
-      "learning_rate": 4.075342465753425e-06,
-      "loss": 0.4554,
-      "step": 37850
+      "loss": 0.5043,
+      "step": 18900
     },
     {
       "epoch": 1.730593607305936,
-      "grad_norm": 31.991483688354492,
+      "grad_norm": 22.30252456665039,
       "learning_rate": 4.041095890410959e-06,
-      "loss": 0.561,
-      "step": 37900
-    },
-    {
-      "epoch": 1.7328767123287672,
-      "grad_norm": 21.96062469482422,
-      "learning_rate": 4.006849315068493e-06,
-      "loss": 0.5969,
-      "step": 37950
+      "loss": 0.5546,
+      "step": 18950
     },
     {
       "epoch": 1.7351598173515983,
-      "grad_norm": 10.800865173339844,
+      "grad_norm": 7.0828142166137695,
       "learning_rate": 3.972602739726027e-06,
-      "loss": 0.6058,
-      "step": 38000
-    },
-    {
-      "epoch": 1.7374429223744292,
-      "grad_norm": 10.979826927185059,
-      "learning_rate": 3.938356164383562e-06,
-      "loss": 0.6034,
-      "step": 38050
+      "loss": 0.5997,
+      "step": 19000
     },
     {
       "epoch": 1.7397260273972601,
-      "grad_norm": 41.3328742980957,
+      "grad_norm": 25.635570526123047,
       "learning_rate": 3.904109589041096e-06,
-      "loss": 0.5687,
-      "step": 38100
-    },
-    {
-      "epoch": 1.7420091324200913,
-      "grad_norm": 38.379608154296875,
-      "learning_rate": 3.869863013698631e-06,
-      "loss": 0.6931,
-      "step": 38150
+      "loss": 0.6091,
+      "step": 19050
     },
     {
       "epoch": 1.7442922374429224,
-      "grad_norm": 3.292733907699585,
+      "grad_norm": 9.004444122314453,
       "learning_rate": 3.8356164383561645e-06,
-      "loss": 0.4162,
-      "step": 38200
-    },
-    {
-      "epoch": 1.7465753424657535,
-      "grad_norm": 18.47883415222168,
-      "learning_rate": 3.801369863013699e-06,
-      "loss": 0.6574,
-      "step": 38250
+      "loss": 0.6257,
+      "step": 19100
     },
     {
       "epoch": 1.7488584474885844,
-      "grad_norm": 10.917158126831055,
+      "grad_norm": 5.485713958740234,
       "learning_rate": 3.7671232876712327e-06,
-      "loss": 0.6617,
-      "step": 38300
-    },
-    {
-      "epoch": 1.7511415525114156,
-      "grad_norm": 13.783547401428223,
-      "learning_rate": 3.732876712328767e-06,
-      "loss": 0.7701,
-      "step": 38350
+      "loss": 0.6728,
+      "step": 19150
     },
     {
       "epoch": 1.7534246575342465,
-      "grad_norm": 21.937267303466797,
+      "grad_norm": 35.63444519042969,
       "learning_rate": 3.6986301369863014e-06,
-      "loss": 0.7627,
-      "step": 38400
-    },
-    {
-      "epoch": 1.7557077625570776,
-      "grad_norm": 15.421838760375977,
-      "learning_rate": 3.6643835616438357e-06,
-      "loss": 0.6636,
-      "step": 38450
+      "loss": 0.7696,
+      "step": 19200
     },
     {
       "epoch": 1.7579908675799087,
-      "grad_norm": 14.788371086120605,
+      "grad_norm": 29.199115753173828,
       "learning_rate": 3.6301369863013696e-06,
-      "loss": 0.5593,
-      "step": 38500
-    },
-    {
-      "epoch": 1.7602739726027399,
-      "grad_norm": 5.76630163192749,
-      "learning_rate": 3.595890410958904e-06,
-      "loss": 0.5112,
-      "step": 38550
+      "loss": 0.59,
+      "step": 19250
     },
     {
       "epoch": 1.7625570776255708,
-      "grad_norm": 23.72429656982422,
+      "grad_norm": 18.041336059570312,
       "learning_rate": 3.561643835616438e-06,
-      "loss": 0.5729,
-      "step": 38600
-    },
-    {
-      "epoch": 1.7648401826484017,
-      "grad_norm": 18.512802124023438,
-      "learning_rate": 3.5273972602739725e-06,
-      "loss": 0.6827,
-      "step": 38650
+      "loss": 0.5744,
+      "step": 19300
     },
     {
       "epoch": 1.7671232876712328,
-      "grad_norm": 4.7244720458984375,
+      "grad_norm": 16.391035079956055,
       "learning_rate": 3.493150684931507e-06,
-      "loss": 0.6428,
-      "step": 38700
-    },
-    {
-      "epoch": 1.769406392694064,
-      "grad_norm": 6.208735466003418,
-      "learning_rate": 3.458904109589041e-06,
-      "loss": 0.5986,
-      "step": 38750
+      "loss": 0.6831,
+      "step": 19350
     },
     {
       "epoch": 1.771689497716895,
-      "grad_norm": 2.6915433406829834,
+      "grad_norm": 9.0728759765625,
       "learning_rate": 3.4246575342465754e-06,
-      "loss": 0.5905,
-      "step": 38800
-    },
-    {
-      "epoch": 1.773972602739726,
-      "grad_norm": 4.2726969718933105,
-      "learning_rate": 3.3904109589041093e-06,
-      "loss": 0.564,
-      "step": 38850
+      "loss": 0.5779,
+      "step": 19400
     },
     {
       "epoch": 1.776255707762557,
-      "grad_norm": 26.4520263671875,
+      "grad_norm": 18.102890014648438,
       "learning_rate": 3.3561643835616436e-06,
-      "loss": 0.5738,
-      "step": 38900
-    },
-    {
-      "epoch": 1.778538812785388,
-      "grad_norm": 43.63593673706055,
-      "learning_rate": 3.321917808219178e-06,
-      "loss": 0.621,
-      "step": 38950
+      "loss": 0.5507,
+      "step": 19450
     },
     {
       "epoch": 1.7808219178082192,
-      "grad_norm": 4.847127437591553,
+      "grad_norm": 17.248735427856445,
       "learning_rate": 3.2876712328767123e-06,
-      "loss": 0.8252,
-      "step": 39000
-    },
-    {
-      "epoch": 1.7831050228310503,
-      "grad_norm": 7.686138153076172,
-      "learning_rate": 3.2534246575342466e-06,
-      "loss": 0.7211,
-      "step": 39050
+      "loss": 0.7226,
+      "step": 19500
     },
     {
       "epoch": 1.7853881278538812,
-      "grad_norm": 42.92139434814453,
+      "grad_norm": 27.942777633666992,
       "learning_rate": 3.219178082191781e-06,
-      "loss": 0.7652,
-      "step": 39100
-    },
-    {
-      "epoch": 1.7876712328767124,
-      "grad_norm": 57.05276107788086,
-      "learning_rate": 3.1849315068493148e-06,
-      "loss": 0.5234,
-      "step": 39150
+      "loss": 0.7178,
+      "step": 19550
     },
     {
       "epoch": 1.7899543378995433,
-      "grad_norm": 18.587209701538086,
+      "grad_norm": 5.3349809646606445,
       "learning_rate": 3.150684931506849e-06,
-      "loss": 0.457,
-      "step": 39200
-    },
-    {
-      "epoch": 1.7922374429223744,
-      "grad_norm": 7.19858455657959,
-      "learning_rate": 3.1164383561643834e-06,
-      "loss": 0.8795,
-      "step": 39250
+      "loss": 0.5566,
+      "step": 19600
     },
     {
       "epoch": 1.7945205479452055,
-      "grad_norm": 10.892264366149902,
+      "grad_norm": 30.73387908935547,
       "learning_rate": 3.0821917808219177e-06,
-      "loss": 0.7042,
-      "step": 39300
-    },
-    {
-      "epoch": 1.7968036529680367,
-      "grad_norm": 28.82424545288086,
-      "learning_rate": 3.047945205479452e-06,
-      "loss": 0.6396,
-      "step": 39350
+      "loss": 0.7677,
+      "step": 19650
     },
     {
       "epoch": 1.7990867579908676,
-      "grad_norm": 7.087406158447266,
+      "grad_norm": 21.230749130249023,
       "learning_rate": 3.0136986301369864e-06,
-      "loss": 0.6665,
-      "step": 39400
-    },
-    {
-      "epoch": 1.8013698630136985,
-      "grad_norm": 22.56847381591797,
-      "learning_rate": 2.9794520547945207e-06,
-      "loss": 0.7265,
-      "step": 39450
+      "loss": 0.61,
+      "step": 19700
     },
     {
       "epoch": 1.8036529680365296,
-      "grad_norm": 18.845949172973633,
+      "grad_norm": 10.363127708435059,
       "learning_rate": 2.9452054794520546e-06,
-      "loss": 0.6475,
-      "step": 39500
-    },
-    {
-      "epoch": 1.8059360730593608,
-      "grad_norm": 26.794076919555664,
-      "learning_rate": 2.910958904109589e-06,
-      "loss": 0.7632,
-      "step": 39550
+      "loss": 0.6385,
+      "step": 19750
     },
     {
       "epoch": 1.808219178082192,
-      "grad_norm": 0.44524723291397095,
+      "grad_norm": 1.5289000272750854,
       "learning_rate": 2.876712328767123e-06,
-      "loss": 0.4843,
-      "step": 39600
-    },
-    {
-      "epoch": 1.8105022831050228,
-      "grad_norm": 45.64598083496094,
-      "learning_rate": 2.8424657534246575e-06,
-      "loss": 0.5918,
-      "step": 39650
+      "loss": 0.6747,
+      "step": 19800
     },
     {
       "epoch": 1.8127853881278537,
-      "grad_norm": 75.41986846923828,
+      "grad_norm": 18.69597625732422,
       "learning_rate": 2.808219178082192e-06,
-      "loss": 0.6352,
-      "step": 39700
-    },
-    {
-      "epoch": 1.8150684931506849,
-      "grad_norm": 10.345170974731445,
-      "learning_rate": 2.773972602739726e-06,
-      "loss": 0.6204,
-      "step": 39750
+      "loss": 0.5934,
+      "step": 19850
     },
     {
       "epoch": 1.817351598173516,
-      "grad_norm": 11.58834171295166,
+      "grad_norm": 10.11032772064209,
       "learning_rate": 2.73972602739726e-06,
-      "loss": 0.5325,
-      "step": 39800
-    },
-    {
-      "epoch": 1.8196347031963471,
-      "grad_norm": 14.26885986328125,
-      "learning_rate": 2.7054794520547943e-06,
-      "loss": 0.5008,
-      "step": 39850
+      "loss": 0.5918,
+      "step": 19900
     },
     {
       "epoch": 1.821917808219178,
-      "grad_norm": 4.937170505523682,
+      "grad_norm": 30.333818435668945,
       "learning_rate": 2.6712328767123286e-06,
-      "loss": 0.7317,
-      "step": 39900
-    },
-    {
-      "epoch": 1.8242009132420092,
-      "grad_norm": 27.849742889404297,
-      "learning_rate": 2.636986301369863e-06,
-      "loss": 0.6825,
-      "step": 39950
+      "loss": 0.6381,
+      "step": 19950
     },
     {
       "epoch": 1.82648401826484,
-      "grad_norm": 38.649810791015625,
+      "grad_norm": 15.229024887084961,
       "learning_rate": 2.6027397260273973e-06,
-      "loss": 0.4422,
-      "step": 40000
-    },
-    {
-      "epoch": 1.8287671232876712,
-      "grad_norm": 52.95954895019531,
-      "learning_rate": 2.5684931506849316e-06,
-      "loss": 0.747,
-      "step": 40050
+      "loss": 0.6151,
+      "step": 20000
     },
     {
       "epoch": 1.8310502283105023,
-      "grad_norm": 10.486088752746582,
+      "grad_norm": 28.18828773498535,
       "learning_rate": 2.5342465753424655e-06,
-      "loss": 0.6757,
-      "step": 40100
-    },
-    {
-      "epoch": 1.8333333333333335,
-      "grad_norm": 0.9142507314682007,
-      "learning_rate": 2.4999999999999998e-06,
-      "loss": 0.712,
-      "step": 40150
+      "loss": 0.6968,
+      "step": 20050
     },
     {
       "epoch": 1.8356164383561644,
-      "grad_norm": 51.62909698486328,
+      "grad_norm": 48.18195724487305,
       "learning_rate": 2.465753424657534e-06,
-      "loss": 0.6107,
-      "step": 40200
-    },
-    {
-      "epoch": 1.8378995433789953,
-      "grad_norm": 31.60240936279297,
-      "learning_rate": 2.4315068493150684e-06,
-      "loss": 0.601,
-      "step": 40250
+      "loss": 0.6256,
+      "step": 20100
     },
     {
       "epoch": 1.8401826484018264,
-      "grad_norm": 55.16636657714844,
+      "grad_norm": 16.33424949645996,
       "learning_rate": 2.3972602739726027e-06,
-      "loss": 0.6121,
-      "step": 40300
-    },
-    {
-      "epoch": 1.8424657534246576,
-      "grad_norm": 34.450416564941406,
-      "learning_rate": 2.363013698630137e-06,
-      "loss": 0.6331,
-      "step": 40350
+      "loss": 0.6477,
+      "step": 20150
     },
     {
       "epoch": 1.8447488584474887,
-      "grad_norm": 27.42693328857422,
+      "grad_norm": 25.994909286499023,
       "learning_rate": 2.3287671232876713e-06,
-      "loss": 0.733,
-      "step": 40400
-    },
-    {
-      "epoch": 1.8470319634703196,
-      "grad_norm": 44.26624298095703,
-      "learning_rate": 2.2945205479452052e-06,
-      "loss": 0.7453,
-      "step": 40450
+      "loss": 0.692,
+      "step": 20200
     },
     {
       "epoch": 1.8493150684931505,
-      "grad_norm": 16.07997703552246,
+      "grad_norm": 8.117030143737793,
       "learning_rate": 2.2602739726027396e-06,
-      "loss": 0.6628,
-      "step": 40500
-    },
-    {
-      "epoch": 1.8515981735159817,
-      "grad_norm": 2.816776752471924,
-      "learning_rate": 2.226027397260274e-06,
-      "loss": 0.4991,
-      "step": 40550
+      "loss": 0.6933,
+      "step": 20250
     },
     {
       "epoch": 1.8538812785388128,
-      "grad_norm": 4.0401434898376465,
+      "grad_norm": 8.02834415435791,
       "learning_rate": 2.191780821917808e-06,
-      "loss": 0.6943,
-      "step": 40600
-    },
-    {
-      "epoch": 1.856164383561644,
-      "grad_norm": 5.525669097900391,
-      "learning_rate": 2.1575342465753425e-06,
-      "loss": 0.6887,
-      "step": 40650
+      "loss": 0.6413,
+      "step": 20300
     },
     {
       "epoch": 1.8584474885844748,
-      "grad_norm": 85.989990234375,
+      "grad_norm": 22.62827491760254,
       "learning_rate": 2.123287671232877e-06,
-      "loss": 0.7425,
-      "step": 40700
-    },
-    {
-      "epoch": 1.860730593607306,
-      "grad_norm": 4.5465779304504395,
-      "learning_rate": 2.0890410958904107e-06,
-      "loss": 0.7462,
-      "step": 40750
+      "loss": 0.679,
+      "step": 20350
     },
     {
       "epoch": 1.8630136986301369,
-      "grad_norm": 5.920977592468262,
+      "grad_norm": 9.562274932861328,
       "learning_rate": 2.054794520547945e-06,
-      "loss": 0.5546,
-      "step": 40800
-    },
-    {
-      "epoch": 1.865296803652968,
-      "grad_norm": 12.154388427734375,
-      "learning_rate": 2.0205479452054793e-06,
-      "loss": 0.6693,
-      "step": 40850
+      "loss": 0.6615,
+      "step": 20400
     },
     {
       "epoch": 1.8675799086757991,
-      "grad_norm": 16.800073623657227,
+      "grad_norm": 12.407808303833008,
       "learning_rate": 1.9863013698630136e-06,
-      "loss": 0.7908,
-      "step": 40900
-    },
-    {
-      "epoch": 1.8698630136986303,
-      "grad_norm": 43.42325973510742,
-      "learning_rate": 1.952054794520548e-06,
-      "loss": 0.5155,
-      "step": 40950
+      "loss": 0.7476,
+      "step": 20450
     },
     {
       "epoch": 1.8721461187214612,
-      "grad_norm": 29.5067138671875,
+      "grad_norm": 41.344093322753906,
       "learning_rate": 1.9178082191780823e-06,
-      "loss": 0.7588,
-      "step": 41000
-    },
-    {
-      "epoch": 1.874429223744292,
-      "grad_norm": 28.01750946044922,
-      "learning_rate": 1.8835616438356164e-06,
-      "loss": 0.6622,
-      "step": 41050
+      "loss": 0.5827,
+      "step": 20500
     },
     {
       "epoch": 1.8767123287671232,
-      "grad_norm": 16.869781494140625,
+      "grad_norm": 10.044130325317383,
       "learning_rate": 1.8493150684931507e-06,
-      "loss": 0.6737,
-      "step": 41100
-    },
-    {
-      "epoch": 1.8789954337899544,
-      "grad_norm": 8.377634048461914,
-      "learning_rate": 1.8150684931506848e-06,
-      "loss": 0.6281,
-      "step": 41150
+      "loss": 0.6566,
+      "step": 20550
     },
     {
       "epoch": 1.8812785388127855,
-      "grad_norm": 16.61414337158203,
+      "grad_norm": 9.382560729980469,
       "learning_rate": 1.780821917808219e-06,
-      "loss": 0.6251,
-      "step": 41200
-    },
-    {
-      "epoch": 1.8835616438356164,
-      "grad_norm": 16.144508361816406,
-      "learning_rate": 1.7465753424657534e-06,
-      "loss": 0.6607,
-      "step": 41250
+      "loss": 0.6259,
+      "step": 20600
     },
     {
       "epoch": 1.8858447488584473,
-      "grad_norm": 20.15201759338379,
+      "grad_norm": 9.731813430786133,
       "learning_rate": 1.7123287671232877e-06,
-      "loss": 0.5207,
-      "step": 41300
-    },
-    {
-      "epoch": 1.8881278538812785,
-      "grad_norm": 7.15456485748291,
-      "learning_rate": 1.6780821917808218e-06,
-      "loss": 0.5882,
-      "step": 41350
+      "loss": 0.5924,
+      "step": 20650
     },
     {
       "epoch": 1.8904109589041096,
-      "grad_norm": 17.336624145507812,
+      "grad_norm": 13.417922973632812,
       "learning_rate": 1.6438356164383561e-06,
-      "loss": 0.556,
-      "step": 41400
-    },
-    {
-      "epoch": 1.8926940639269407,
-      "grad_norm": 20.451026916503906,
-      "learning_rate": 1.6095890410958904e-06,
-      "loss": 0.6011,
-      "step": 41450
+      "loss": 0.6229,
+      "step": 20700
     },
     {
       "epoch": 1.8949771689497716,
-      "grad_norm": 33.44941329956055,
+      "grad_norm": 32.03701400756836,
       "learning_rate": 1.5753424657534245e-06,
-      "loss": 0.5166,
-      "step": 41500
-    },
-    {
-      "epoch": 1.8972602739726028,
-      "grad_norm": 24.176786422729492,
-      "learning_rate": 1.5410958904109589e-06,
-      "loss": 0.6837,
-      "step": 41550
+      "loss": 0.5761,
+      "step": 20750
     },
     {
       "epoch": 1.8995433789954337,
-      "grad_norm": 17.142606735229492,
+      "grad_norm": 6.067290782928467,
       "learning_rate": 1.5068493150684932e-06,
-      "loss": 0.5813,
-      "step": 41600
-    },
-    {
-      "epoch": 1.9018264840182648,
-      "grad_norm": 34.20349884033203,
-      "learning_rate": 1.4726027397260273e-06,
-      "loss": 0.786,
-      "step": 41650
+      "loss": 0.6463,
+      "step": 20800
     },
     {
       "epoch": 1.904109589041096,
-      "grad_norm": 1.5305472612380981,
+      "grad_norm": 19.67026710510254,
       "learning_rate": 1.4383561643835616e-06,
-      "loss": 0.5928,
-      "step": 41700
-    },
-    {
-      "epoch": 1.906392694063927,
-      "grad_norm": 1.1295257806777954,
-      "learning_rate": 1.404109589041096e-06,
-      "loss": 0.7055,
-      "step": 41750
+      "loss": 0.6671,
+      "step": 20850
     },
     {
       "epoch": 1.908675799086758,
-      "grad_norm": 23.80326271057129,
+      "grad_norm": 50.498802185058594,
       "learning_rate": 1.36986301369863e-06,
-      "loss": 0.5861,
-      "step": 41800
-    },
-    {
-      "epoch": 1.910958904109589,
-      "grad_norm": 3.346529960632324,
-      "learning_rate": 1.3356164383561643e-06,
-      "loss": 0.6864,
-      "step": 41850
+      "loss": 0.5811,
+      "step": 20900
     },
     {
       "epoch": 1.91324200913242,
-      "grad_norm": 34.07392883300781,
+      "grad_norm": 15.981374740600586,
       "learning_rate": 1.3013698630136986e-06,
-      "loss": 0.6754,
-      "step": 41900
-    },
-    {
-      "epoch": 1.9155251141552512,
-      "grad_norm": 42.87485122680664,
-      "learning_rate": 1.2671232876712327e-06,
-      "loss": 0.613,
-      "step": 41950
+      "loss": 0.6823,
+      "step": 20950
     },
     {
       "epoch": 1.9178082191780823,
-      "grad_norm": 9.337113380432129,
+      "grad_norm": 19.175485610961914,
       "learning_rate": 1.232876712328767e-06,
-      "loss": 0.5302,
-      "step": 42000
-    },
-    {
-      "epoch": 1.9200913242009132,
-      "grad_norm": 19.920682907104492,
-      "learning_rate": 1.1986301369863014e-06,
-      "loss": 0.4405,
-      "step": 42050
+      "loss": 0.6112,
+      "step": 21000
     },
     {
       "epoch": 1.9223744292237441,
-      "grad_norm": 24.49388313293457,
+      "grad_norm": 8.795181274414062,
       "learning_rate": 1.1643835616438357e-06,
-      "loss": 0.696,
-      "step": 42100
-    },
-    {
-      "epoch": 1.9246575342465753,
-      "grad_norm": 7.732158184051514,
-      "learning_rate": 1.1301369863013698e-06,
-      "loss": 0.758,
-      "step": 42150
+      "loss": 0.5474,
+      "step": 21050
     },
     {
       "epoch": 1.9269406392694064,
-      "grad_norm": 6.940062046051025,
+      "grad_norm": 3.590404748916626,
       "learning_rate": 1.095890410958904e-06,
-      "loss": 0.6804,
-      "step": 42200
-    },
-    {
-      "epoch": 1.9292237442922375,
-      "grad_norm": 1.062066674232483,
-      "learning_rate": 1.0616438356164384e-06,
-      "loss": 0.6059,
-      "step": 42250
+      "loss": 0.7026,
+      "step": 21100
     },
     {
       "epoch": 1.9315068493150684,
-      "grad_norm": 20.240144729614258,
+      "grad_norm": 21.619140625,
       "learning_rate": 1.0273972602739725e-06,
-      "loss": 0.7992,
-      "step": 42300
-    },
-    {
-      "epoch": 1.9337899543378996,
-      "grad_norm": 19.4890193939209,
-      "learning_rate": 9.931506849315068e-07,
-      "loss": 0.6943,
-      "step": 42350
+      "loss": 0.701,
+      "step": 21150
     },
     {
       "epoch": 1.9360730593607305,
-      "grad_norm": 25.273487091064453,
+      "grad_norm": 22.84990692138672,
       "learning_rate": 9.589041095890411e-07,
-      "loss": 0.5925,
-      "step": 42400
-    },
-    {
-      "epoch": 1.9383561643835616,
-      "grad_norm": 21.916284561157227,
-      "learning_rate": 9.246575342465753e-07,
-      "loss": 0.6607,
-      "step": 42450
+      "loss": 0.5942,
+      "step": 21200
     },
     {
       "epoch": 1.9406392694063928,
-      "grad_norm": 38.107566833496094,
+      "grad_norm": 28.548683166503906,
       "learning_rate": 8.904109589041095e-07,
-      "loss": 0.5807,
-      "step": 42500
-    },
-    {
-      "epoch": 1.9429223744292239,
-      "grad_norm": 8.979408264160156,
-      "learning_rate": 8.561643835616439e-07,
-      "loss": 0.6391,
-      "step": 42550
+      "loss": 0.6343,
+      "step": 21250
     },
     {
       "epoch": 1.9452054794520548,
-      "grad_norm": 20.871389389038086,
+      "grad_norm": 20.216583251953125,
       "learning_rate": 8.219178082191781e-07,
-      "loss": 0.764,
-      "step": 42600
-    },
-    {
-      "epoch": 1.9474885844748857,
-      "grad_norm": 19.994056701660156,
-      "learning_rate": 7.876712328767123e-07,
-      "loss": 0.7959,
-      "step": 42650
+      "loss": 0.6836,
+      "step": 21300
     },
     {
       "epoch": 1.9497716894977168,
-      "grad_norm": 25.47404670715332,
+      "grad_norm": 22.950048446655273,
       "learning_rate": 7.534246575342466e-07,
-      "loss": 0.5579,
-      "step": 42700
-    },
-    {
-      "epoch": 1.952054794520548,
-      "grad_norm": 10.598165512084961,
-      "learning_rate": 7.191780821917808e-07,
-      "loss": 0.8362,
-      "step": 42750
+      "loss": 0.6815,
+      "step": 21350
     },
     {
       "epoch": 1.954337899543379,
-      "grad_norm": 1.3319069147109985,
+      "grad_norm": 20.81789207458496,
       "learning_rate": 6.84931506849315e-07,
-      "loss": 0.8132,
-      "step": 42800
-    },
-    {
-      "epoch": 1.95662100456621,
-      "grad_norm": 4.061497211456299,
-      "learning_rate": 6.506849315068493e-07,
-      "loss": 0.5364,
-      "step": 42850
+      "loss": 0.7679,
+      "step": 21400
     },
     {
       "epoch": 1.958904109589041,
-      "grad_norm": 3.6196768283843994,
+      "grad_norm": 14.494473457336426,
       "learning_rate": 6.164383561643835e-07,
-      "loss": 0.57,
-      "step": 42900
-    },
-    {
-      "epoch": 1.961187214611872,
-      "grad_norm": 26.95933723449707,
-      "learning_rate": 5.821917808219178e-07,
-      "loss": 0.7364,
-      "step": 42950
+      "loss": 0.5561,
+      "step": 21450
     },
     {
       "epoch": 1.9634703196347032,
-      "grad_norm": 25.18138313293457,
+      "grad_norm": 17.119592666625977,
       "learning_rate": 5.47945205479452e-07,
-      "loss": 0.5801,
-      "step": 43000
-    },
-    {
-      "epoch": 1.9657534246575343,
-      "grad_norm": 77.7193374633789,
-      "learning_rate": 5.136986301369863e-07,
-      "loss": 0.6965,
-      "step": 43050
+      "loss": 0.6881,
+      "step": 21500
     },
     {
       "epoch": 1.9680365296803652,
-      "grad_norm": 10.553460121154785,
+      "grad_norm": 8.369721412658691,
       "learning_rate": 4.794520547945206e-07,
-      "loss": 0.6019,
-      "step": 43100
-    },
-    {
-      "epoch": 1.9703196347031964,
-      "grad_norm": 14.633034706115723,
-      "learning_rate": 4.4520547945205477e-07,
-      "loss": 0.5948,
-      "step": 43150
+      "loss": 0.6692,
+      "step": 21550
     },
     {
       "epoch": 1.9726027397260273,
-      "grad_norm": 10.625903129577637,
+      "grad_norm": 11.264369010925293,
       "learning_rate": 4.1095890410958903e-07,
-      "loss": 0.5868,
-      "step": 43200
-    },
-    {
-      "epoch": 1.9748858447488584,
-      "grad_norm": 11.625406265258789,
-      "learning_rate": 3.767123287671233e-07,
-      "loss": 0.7335,
-      "step": 43250
+      "loss": 0.6405,
+      "step": 21600
     },
     {
       "epoch": 1.9771689497716896,
-      "grad_norm": 11.875858306884766,
+      "grad_norm": 3.320608615875244,
       "learning_rate": 3.424657534246575e-07,
-      "loss": 0.6305,
-      "step": 43300
-    },
-    {
-      "epoch": 1.9794520547945207,
-      "grad_norm": 65.91475677490234,
-      "learning_rate": 3.0821917808219176e-07,
-      "loss": 0.5038,
-      "step": 43350
+      "loss": 0.671,
+      "step": 21650
     },
     {
       "epoch": 1.9817351598173516,
-      "grad_norm": 7.164722442626953,
+      "grad_norm": 19.23822784423828,
       "learning_rate": 2.73972602739726e-07,
-      "loss": 0.6426,
-      "step": 43400
-    },
-    {
-      "epoch": 1.9840182648401825,
-      "grad_norm": 19.229677200317383,
-      "learning_rate": 2.397260273972603e-07,
-      "loss": 0.7671,
-      "step": 43450
+      "loss": 0.5475,
+      "step": 21700
     },
     {
       "epoch": 1.9863013698630136,
-      "grad_norm": 9.592227935791016,
+      "grad_norm": 13.919721603393555,
       "learning_rate": 2.0547945205479452e-07,
-      "loss": 0.7229,
-      "step": 43500
-    },
-    {
-      "epoch": 1.9885844748858448,
-      "grad_norm": 23.81863021850586,
-      "learning_rate": 1.7123287671232875e-07,
-      "loss": 0.5238,
-      "step": 43550
+      "loss": 0.7422,
+      "step": 21750
     },
     {
       "epoch": 1.990867579908676,
-      "grad_norm": 30.481460571289062,
+      "grad_norm": 14.141480445861816,
       "learning_rate": 1.36986301369863e-07,
-      "loss": 0.5882,
-      "step": 43600
-    },
-    {
-      "epoch": 1.9931506849315068,
-      "grad_norm": 3.194218397140503,
-      "learning_rate": 1.0273972602739726e-07,
-      "loss": 0.5594,
-      "step": 43650
+      "loss": 0.5855,
+      "step": 21800
     },
     {
       "epoch": 1.9954337899543377,
-      "grad_norm": 32.75349807739258,
+      "grad_norm": 19.506189346313477,
       "learning_rate": 6.84931506849315e-08,
-      "loss": 0.7056,
-      "step": 43700
-    },
-    {
-      "epoch": 1.9977168949771689,
-      "grad_norm": 37.13774490356445,
-      "learning_rate": 3.424657534246575e-08,
-      "loss": 0.7017,
-      "step": 43750
+      "loss": 0.7103,
+      "step": 21850
     },
     {
       "epoch": 2.0,
-      "grad_norm": 38.73429489135742,
+      "grad_norm": 16.25570297241211,
       "learning_rate": 0.0,
-      "loss": 0.6852,
-      "step": 43800
+      "loss": 0.7072,
+      "step": 21900
     },
     {
       "epoch": 2.0,
-      "step": 43800,
-      "total_flos": 3.864501390676132e+17,
-      "train_loss": 0.05644623499482734,
-      "train_runtime": 4615.1991,
-      "train_samples_per_second": 37.961,
-      "train_steps_per_second": 9.49
+      "step": 21900,
+      "total_flos": 3.8619551920019866e+17,
+      "train_loss": 0.9270246051109,
+      "train_runtime": 51747.0424,
+      "train_samples_per_second": 3.386,
+      "train_steps_per_second": 0.423
     }
   ],
   "logging_steps": 50,
-  "max_steps": 43800,
+  "max_steps": 21900,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 2,
   "save_steps": 5000,
@@ -6167,8 +3101,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.864501390676132e+17,
-  "train_batch_size": 4,
+  "total_flos": 3.8619551920019866e+17,
+  "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null
 }