diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5443 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "global_step": 451623,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.920484962997953e-07,
+      "loss": 10.3537,
+      "step": 500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.840969925995906e-07,
+      "loss": 3.8759,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4761454888993861e-06,
+      "loss": 3.3657,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9681939851991812e-06,
+      "loss": 3.2292,
+      "step": 2000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4602424814989765e-06,
+      "loss": 3.1654,
+      "step": 2500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9522909777987723e-06,
+      "loss": 3.1135,
+      "step": 3000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444339474098567e-06,
+      "loss": 3.0643,
+      "step": 3500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9363879703983625e-06,
+      "loss": 3.0263,
+      "step": 4000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428436466698158e-06,
+      "loss": 2.9939,
+      "step": 4500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.920484962997953e-06,
+      "loss": 2.9628,
+      "step": 5000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.412533459297749e-06,
+      "loss": 2.9285,
+      "step": 5500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9045819555975445e-06,
+      "loss": 2.897,
+      "step": 6000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.396630451897339e-06,
+      "loss": 2.8763,
+      "step": 6500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.888678948197134e-06,
+      "loss": 2.8517,
+      "step": 7000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.380727444496929e-06,
+      "loss": 2.8288,
+      "step": 7500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.872775940796725e-06,
+      "loss": 2.8043,
+      "step": 8000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.36482443709652e-06,
+      "loss": 2.7839,
+      "step": 8500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.856872933396316e-06,
+      "loss": 2.7614,
+      "step": 9000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.348921429696111e-06,
+      "loss": 2.7452,
+      "step": 9500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.840969925995906e-06,
+      "loss": 2.7313,
+      "step": 10000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0333018422295703e-05,
+      "loss": 2.7144,
+      "step": 10500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0825066918595498e-05,
+      "loss": 2.6997,
+      "step": 11000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1317115414895292e-05,
+      "loss": 2.6785,
+      "step": 11500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1809163911195089e-05,
+      "loss": 2.6693,
+      "step": 12000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2301212407494884e-05,
+      "loss": 2.6494,
+      "step": 12500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.2793260903794679e-05,
+      "loss": 2.6353,
+      "step": 13000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.3285309400094472e-05,
+      "loss": 2.6276,
+      "step": 13500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.3777357896394269e-05,
+      "loss": 2.6187,
+      "step": 14000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.4269406392694065e-05,
+      "loss": 2.6011,
+      "step": 14500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.4761454888993858e-05,
+      "loss": 2.5932,
+      "step": 15000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.5253503385293655e-05,
+      "loss": 2.5815,
+      "step": 15500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.574555188159345e-05,
+      "loss": 2.5736,
+      "step": 16000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.6237600377893248e-05,
+      "loss": 2.5592,
+      "step": 16500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.672964887419304e-05,
+      "loss": 2.5488,
+      "step": 17000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.7221697370492838e-05,
+      "loss": 2.5394,
+      "step": 17500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.7713745866792633e-05,
+      "loss": 2.5307,
+      "step": 18000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.8205794363092428e-05,
+      "loss": 2.5219,
+      "step": 18500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.8697842859392223e-05,
+      "loss": 2.514,
+      "step": 19000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9189891355692017e-05,
+      "loss": 2.5038,
+      "step": 19500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9681939851991812e-05,
+      "loss": 2.4932,
+      "step": 20000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.0173988348291607e-05,
+      "loss": 2.4895,
+      "step": 20500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.0666036844591405e-05,
+      "loss": 2.4754,
+      "step": 21000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.1158085340891197e-05,
+      "loss": 2.4703,
+      "step": 21500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.1650133837190995e-05,
+      "loss": 2.4602,
+      "step": 22000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.214218233349079e-05,
+      "loss": 2.4609,
+      "step": 22500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.2634230829790585e-05,
+      "loss": 2.4458,
+      "step": 23000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.312627932609038e-05,
+      "loss": 2.439,
+      "step": 23500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.3618327822390178e-05,
+      "loss": 2.4354,
+      "step": 24000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.411037631868997e-05,
+      "loss": 2.4263,
+      "step": 24500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.4602424814989768e-05,
+      "loss": 2.4238,
+      "step": 25000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.5094473311289563e-05,
+      "loss": 2.4135,
+      "step": 25500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.5586521807589358e-05,
+      "loss": 2.4098,
+      "step": 26000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.6078570303889156e-05,
+      "loss": 2.4012,
+      "step": 26500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.6570618800188944e-05,
+      "loss": 2.3966,
+      "step": 27000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.7062667296488742e-05,
+      "loss": 2.386,
+      "step": 27500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.7554715792788537e-05,
+      "loss": 2.3838,
+      "step": 28000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.8046764289088335e-05,
+      "loss": 2.3787,
+      "step": 28500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.853881278538813e-05,
+      "loss": 2.3711,
+      "step": 29000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.9030861281687925e-05,
+      "loss": 2.365,
+      "step": 29500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.9522909777987717e-05,
+      "loss": 2.3625,
+      "step": 30000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.0014958274287515e-05,
+      "loss": 2.3555,
+      "step": 30500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.050700677058731e-05,
+      "loss": 2.349,
+      "step": 31000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.099905526688711e-05,
+      "loss": 2.3439,
+      "step": 31500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.14911037631869e-05,
+      "loss": 2.3382,
+      "step": 32000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.19831522594867e-05,
+      "loss": 2.3355,
+      "step": 32500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.2475200755786496e-05,
+      "loss": 2.3265,
+      "step": 33000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.296724925208629e-05,
+      "loss": 2.3289,
+      "step": 33500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.345929774838608e-05,
+      "loss": 2.3275,
+      "step": 34000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.395134624468588e-05,
+      "loss": 2.3242,
+      "step": 34500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.4443394740985676e-05,
+      "loss": 2.311,
+      "step": 35000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.493544323728547e-05,
+      "loss": 2.3148,
+      "step": 35500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.5427491733585265e-05,
+      "loss": 2.3088,
+      "step": 36000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.591954022988506e-05,
+      "loss": 2.301,
+      "step": 36500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.6411588726184855e-05,
+      "loss": 2.2969,
+      "step": 37000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.690363722248465e-05,
+      "loss": 2.2989,
+      "step": 37500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.7395685718784445e-05,
+      "loss": 2.2882,
+      "step": 38000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.788773421508424e-05,
+      "loss": 2.2869,
+      "step": 38500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.8379782711384035e-05,
+      "loss": 2.279,
+      "step": 39000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.8871831207683826e-05,
+      "loss": 2.2817,
+      "step": 39500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.9363879703983625e-05,
+      "loss": 2.2769,
+      "step": 40000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.985592820028342e-05,
+      "loss": 2.271,
+      "step": 40500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.0347976696583214e-05,
+      "loss": 2.2703,
+      "step": 41000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.084002519288301e-05,
+      "loss": 2.2644,
+      "step": 41500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.133207368918281e-05,
+      "loss": 2.2578,
+      "step": 42000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.18241221854826e-05,
+      "loss": 2.2568,
+      "step": 42500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.2316170681782394e-05,
+      "loss": 2.2565,
+      "step": 43000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.280821917808219e-05,
+      "loss": 2.2539,
+      "step": 43500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.330026767438199e-05,
+      "loss": 2.2493,
+      "step": 44000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.379231617068179e-05,
+      "loss": 2.2488,
+      "step": 44500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.428436466698158e-05,
+      "loss": 2.2388,
+      "step": 45000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.477641316328138e-05,
+      "loss": 2.2368,
+      "step": 45500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.526846165958117e-05,
+      "loss": 2.2366,
+      "step": 46000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.576051015588096e-05,
+      "loss": 2.2323,
+      "step": 46500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.625255865218076e-05,
+      "loss": 2.2288,
+      "step": 47000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.674460714848056e-05,
+      "loss": 2.23,
+      "step": 47500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.7236655644780356e-05,
+      "loss": 2.223,
+      "step": 48000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.772870414108015e-05,
+      "loss": 2.2232,
+      "step": 48500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.822075263737994e-05,
+      "loss": 2.2188,
+      "step": 49000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.871280113367974e-05,
+      "loss": 2.217,
+      "step": 49500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9204849629979536e-05,
+      "loss": 2.2118,
+      "step": 50000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.969689812627933e-05,
+      "loss": 2.2089,
+      "step": 50500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.997900574717671e-05,
+      "loss": 2.2101,
+      "step": 51000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.992433321378273e-05,
+      "loss": 2.2048,
+      "step": 51500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.986966068038875e-05,
+      "loss": 2.2026,
+      "step": 52000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.981498814699476e-05,
+      "loss": 2.2,
+      "step": 52500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.976031561360078e-05,
+      "loss": 2.1998,
+      "step": 53000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.97056430802068e-05,
+      "loss": 2.1905,
+      "step": 53500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.965097054681281e-05,
+      "loss": 2.1896,
+      "step": 54000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.959629801341883e-05,
+      "loss": 2.1886,
+      "step": 54500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.954162548002484e-05,
+      "loss": 2.1825,
+      "step": 55000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.948695294663086e-05,
+      "loss": 2.1807,
+      "step": 55500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.943228041323688e-05,
+      "loss": 2.1814,
+      "step": 56000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9377607879842896e-05,
+      "loss": 2.1797,
+      "step": 56500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9322935346448914e-05,
+      "loss": 2.1714,
+      "step": 57000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9268262813054925e-05,
+      "loss": 2.1766,
+      "step": 57500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.921359027966094e-05,
+      "loss": 2.1725,
+      "step": 58000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.915891774626696e-05,
+      "loss": 2.1672,
+      "step": 58500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.910424521287298e-05,
+      "loss": 2.1707,
+      "step": 59000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9049572679479e-05,
+      "loss": 2.1636,
+      "step": 59500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.899490014608501e-05,
+      "loss": 2.1627,
+      "step": 60000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.894022761269103e-05,
+      "loss": 2.1598,
+      "step": 60500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8885555079297044e-05,
+      "loss": 2.1535,
+      "step": 61000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.883088254590306e-05,
+      "loss": 2.1531,
+      "step": 61500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8776210012509074e-05,
+      "loss": 2.1548,
+      "step": 62000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.872153747911509e-05,
+      "loss": 2.1451,
+      "step": 62500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8666864945721116e-05,
+      "loss": 2.1497,
+      "step": 63000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.861219241232713e-05,
+      "loss": 2.1474,
+      "step": 63500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8557519878933146e-05,
+      "loss": 2.1442,
+      "step": 64000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.850284734553916e-05,
+      "loss": 2.1417,
+      "step": 64500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8448174812145175e-05,
+      "loss": 2.1426,
+      "step": 65000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.839350227875119e-05,
+      "loss": 2.1396,
+      "step": 65500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.833882974535721e-05,
+      "loss": 2.1371,
+      "step": 66000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.828415721196323e-05,
+      "loss": 2.1352,
+      "step": 66500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.822948467856924e-05,
+      "loss": 2.1342,
+      "step": 67000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8174812145175265e-05,
+      "loss": 2.1329,
+      "step": 67500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8120139611781276e-05,
+      "loss": 2.1272,
+      "step": 68000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8065467078387294e-05,
+      "loss": 2.1278,
+      "step": 68500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.801079454499331e-05,
+      "loss": 2.1251,
+      "step": 69000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7956122011599323e-05,
+      "loss": 2.1222,
+      "step": 69500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.790144947820535e-05,
+      "loss": 2.1157,
+      "step": 70000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.784677694481136e-05,
+      "loss": 2.1234,
+      "step": 70500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.779210441141738e-05,
+      "loss": 2.1167,
+      "step": 71000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.773743187802339e-05,
+      "loss": 2.1182,
+      "step": 71500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.768275934462941e-05,
+      "loss": 2.121,
+      "step": 72000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.762808681123543e-05,
+      "loss": 2.1115,
+      "step": 72500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.757341427784144e-05,
+      "loss": 2.1171,
+      "step": 73000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.751874174444746e-05,
+      "loss": 2.11,
+      "step": 73500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.746406921105347e-05,
+      "loss": 2.1108,
+      "step": 74000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.74093966776595e-05,
+      "loss": 2.1058,
+      "step": 74500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.735472414426551e-05,
+      "loss": 2.1005,
+      "step": 75000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7300051610871526e-05,
+      "loss": 2.1043,
+      "step": 75500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7245379077477544e-05,
+      "loss": 2.1027,
+      "step": 76000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7190706544083555e-05,
+      "loss": 2.0973,
+      "step": 76500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.713603401068958e-05,
+      "loss": 2.0947,
+      "step": 77000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.708136147729559e-05,
+      "loss": 2.1008,
+      "step": 77500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.702668894390161e-05,
+      "loss": 2.0995,
+      "step": 78000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.697201641050763e-05,
+      "loss": 2.0932,
+      "step": 78500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.691734387711364e-05,
+      "loss": 2.0966,
+      "step": 79000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.686267134371966e-05,
+      "loss": 2.0932,
+      "step": 79500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.6807998810325674e-05,
+      "loss": 2.0949,
+      "step": 80000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.675332627693169e-05,
+      "loss": 2.0912,
+      "step": 80500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.669865374353771e-05,
+      "loss": 2.0885,
+      "step": 81000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.664398121014373e-05,
+      "loss": 2.0904,
+      "step": 81500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.6589308676749746e-05,
+      "loss": 2.0858,
+      "step": 82000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.653463614335576e-05,
+      "loss": 2.0828,
+      "step": 82500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.6479963609961776e-05,
+      "loss": 2.0827,
+      "step": 83000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.642529107656779e-05,
+      "loss": 2.0816,
+      "step": 83500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.637061854317381e-05,
+      "loss": 2.0789,
+      "step": 84000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.631594600977982e-05,
+      "loss": 2.0789,
+      "step": 84500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.626127347638584e-05,
+      "loss": 2.0811,
+      "step": 85000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.620660094299186e-05,
+      "loss": 2.0819,
+      "step": 85500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.615192840959787e-05,
+      "loss": 2.0768,
+      "step": 86000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.6097255876203895e-05,
+      "loss": 2.0715,
+      "step": 86500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.6042583342809906e-05,
+      "loss": 2.0742,
+      "step": 87000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.5987910809415924e-05,
+      "loss": 2.0699,
+      "step": 87500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.593323827602194e-05,
+      "loss": 2.0743,
+      "step": 88000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.587856574262796e-05,
+      "loss": 2.0694,
+      "step": 88500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.582389320923398e-05,
+      "loss": 2.0675,
+      "step": 89000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.576922067583999e-05,
+      "loss": 2.0656,
+      "step": 89500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.571454814244601e-05,
+      "loss": 2.0637,
+      "step": 90000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.5659875609052025e-05,
+      "loss": 2.0667,
+      "step": 90500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.560520307565804e-05,
+      "loss": 2.062,
+      "step": 91000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.555053054226406e-05,
+      "loss": 2.0659,
+      "step": 91500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.549585800887007e-05,
+      "loss": 2.0597,
+      "step": 92000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.544118547547609e-05,
+      "loss": 2.067,
+      "step": 92500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.53865129420821e-05,
+      "loss": 2.0593,
+      "step": 93000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.5331840408688126e-05,
+      "loss": 2.0562,
+      "step": 93500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.527716787529414e-05,
+      "loss": 2.0589,
+      "step": 94000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.5222495341900156e-05,
+      "loss": 2.0566,
+      "step": 94500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.5167822808506174e-05,
+      "loss": 2.0576,
+      "step": 95000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.511315027511219e-05,
+      "loss": 2.058,
+      "step": 95500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.505847774171821e-05,
+      "loss": 2.0528,
+      "step": 96000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.500380520832422e-05,
+      "loss": 2.0562,
+      "step": 96500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.494913267493024e-05,
+      "loss": 2.049,
+      "step": 97000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.489446014153626e-05,
+      "loss": 2.0536,
+      "step": 97500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.4839787608142275e-05,
+      "loss": 2.0538,
+      "step": 98000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.478511507474829e-05,
+      "loss": 2.0466,
+      "step": 98500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.4730442541354304e-05,
+      "loss": 2.049,
+      "step": 99000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.467577000796032e-05,
+      "loss": 2.0455,
+      "step": 99500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.462109747456634e-05,
+      "loss": 2.0441,
+      "step": 100000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.456642494117236e-05,
+      "loss": 2.0477,
+      "step": 100500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.4511752407778376e-05,
+      "loss": 2.0425,
+      "step": 101000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.445707987438439e-05,
+      "loss": 2.0475,
+      "step": 101500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.4402407340990405e-05,
+      "loss": 2.0482,
+      "step": 102000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.434773480759642e-05,
+      "loss": 2.0483,
+      "step": 102500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.429306227420244e-05,
+      "loss": 2.0375,
+      "step": 103000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.423838974080845e-05,
+      "loss": 2.0384,
+      "step": 103500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.418371720741447e-05,
+      "loss": 2.0383,
+      "step": 104000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.412904467402049e-05,
+      "loss": 2.037,
+      "step": 104500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.4074372140626507e-05,
+      "loss": 2.0315,
+      "step": 105000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.4019699607232525e-05,
+      "loss": 2.0375,
+      "step": 105500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.3965027073838536e-05,
+      "loss": 2.0414,
+      "step": 106000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.3910354540444554e-05,
+      "loss": 2.0324,
+      "step": 106500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.385568200705057e-05,
+      "loss": 2.0316,
+      "step": 107000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.380100947365659e-05,
+      "loss": 2.0343,
+      "step": 107500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.374633694026261e-05,
+      "loss": 2.0314,
+      "step": 108000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.369166440686862e-05,
+      "loss": 2.0347,
+      "step": 108500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.363699187347464e-05,
+      "loss": 2.0268,
+      "step": 109000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.3582319340080655e-05,
+      "loss": 2.0323,
+      "step": 109500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.352764680668667e-05,
+      "loss": 2.03,
+      "step": 110000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.347297427329269e-05,
+      "loss": 2.0231,
+      "step": 110500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.34183017398987e-05,
+      "loss": 2.0279,
+      "step": 111000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.336362920650472e-05,
+      "loss": 2.0298,
+      "step": 111500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.330895667311074e-05,
+      "loss": 2.0408,
+      "step": 112000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.3254284139716756e-05,
+      "loss": 2.0199,
+      "step": 112500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.319961160632277e-05,
+      "loss": 2.0297,
+      "step": 113000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.3144939072928786e-05,
+      "loss": 2.0287,
+      "step": 113500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.309026653953481e-05,
+      "loss": 2.0256,
+      "step": 114000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.303559400614082e-05,
+      "loss": 2.0197,
+      "step": 114500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.298092147274684e-05,
+      "loss": 2.0234,
+      "step": 115000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.292624893935285e-05,
+      "loss": 2.0391,
+      "step": 115500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.287157640595887e-05,
+      "loss": 2.0772,
+      "step": 116000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.281690387256489e-05,
+      "loss": 2.0586,
+      "step": 116500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.2762231339170905e-05,
+      "loss": 2.0336,
+      "step": 117000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.270755880577692e-05,
+      "loss": 2.0207,
+      "step": 117500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.2652886272382934e-05,
+      "loss": 2.0197,
+      "step": 118000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.259821373898895e-05,
+      "loss": 2.0184,
+      "step": 118500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.254354120559497e-05,
+      "loss": 2.0162,
+      "step": 119000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.248886867220099e-05,
+      "loss": 2.0139,
+      "step": 119500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.2434196138807006e-05,
+      "loss": 2.0141,
+      "step": 120000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.237952360541302e-05,
+      "loss": 2.0119,
+      "step": 120500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.232485107201904e-05,
+      "loss": 2.0109,
+      "step": 121000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.227017853862505e-05,
+      "loss": 2.0112,
+      "step": 121500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.221550600523107e-05,
+      "loss": 2.0106,
+      "step": 122000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.216083347183708e-05,
+      "loss": 2.0115,
+      "step": 122500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.21061609384431e-05,
+      "loss": 2.0164,
+      "step": 123000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.2051488405049125e-05,
+      "loss": 2.0055,
+      "step": 123500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.1996815871655136e-05,
+      "loss": 2.0117,
+      "step": 124000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.1942143338261154e-05,
+      "loss": 2.008,
+      "step": 124500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.1887470804867166e-05,
+      "loss": 2.0077,
+      "step": 125000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.1832798271473184e-05,
+      "loss": 2.009,
+      "step": 125500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.17781257380792e-05,
+      "loss": 2.0096,
+      "step": 126000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.172345320468522e-05,
+      "loss": 2.0025,
+      "step": 126500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.166878067129124e-05,
+      "loss": 2.0013,
+      "step": 127000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.161410813789725e-05,
+      "loss": 2.0031,
+      "step": 127500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.1559435604503274e-05,
+      "loss": 2.0039,
+      "step": 128000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.1504763071109285e-05,
+      "loss": 2.0002,
+      "step": 128500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.14500905377153e-05,
+      "loss": 2.0007,
+      "step": 129000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.139541800432132e-05,
+      "loss": 2.0078,
+      "step": 129500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.134074547092733e-05,
+      "loss": 1.9981,
+      "step": 130000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.128607293753336e-05,
+      "loss": 1.9989,
+      "step": 130500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.123140040413937e-05,
+      "loss": 1.997,
+      "step": 131000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.1176727870745386e-05,
+      "loss": 2.0004,
+      "step": 131500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.11220553373514e-05,
+      "loss": 1.9983,
+      "step": 132000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.1067382803957415e-05,
+      "loss": 1.9995,
+      "step": 132500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.101271027056344e-05,
+      "loss": 1.9994,
+      "step": 133000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.095803773716945e-05,
+      "loss": 1.9956,
+      "step": 133500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.090336520377547e-05,
+      "loss": 1.994,
+      "step": 134000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.084869267038148e-05,
+      "loss": 1.9955,
+      "step": 134500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.0794020136987505e-05,
+      "loss": 1.9994,
+      "step": 135000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.0739347603593517e-05,
+      "loss": 1.9897,
+      "step": 135500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.0684675070199535e-05,
+      "loss": 1.9961,
+      "step": 136000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.063000253680555e-05,
+      "loss": 1.9946,
+      "step": 136500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.0575330003411564e-05,
+      "loss": 1.991,
+      "step": 137000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.052065747001759e-05,
+      "loss": 1.9952,
+      "step": 137500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.04659849366236e-05,
+      "loss": 1.9848,
+      "step": 138000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.041131240322962e-05,
+      "loss": 1.9937,
+      "step": 138500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.0356639869835636e-05,
+      "loss": 1.9893,
+      "step": 139000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.030196733644165e-05,
+      "loss": 1.9871,
+      "step": 139500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.024729480304767e-05,
+      "loss": 1.9855,
+      "step": 140000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.019262226965368e-05,
+      "loss": 1.9887,
+      "step": 140500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.01379497362597e-05,
+      "loss": 1.9863,
+      "step": 141000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.008327720286571e-05,
+      "loss": 1.9904,
+      "step": 141500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.002860466947174e-05,
+      "loss": 1.9886,
+      "step": 142000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.9973932136077755e-05,
+      "loss": 1.9871,
+      "step": 142500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.9919259602683766e-05,
+      "loss": 1.9843,
+      "step": 143000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.9864587069289784e-05,
+      "loss": 1.9891,
+      "step": 143500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.9809914535895795e-05,
+      "loss": 1.985,
+      "step": 144000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.975524200250182e-05,
+      "loss": 1.9826,
+      "step": 144500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.970056946910783e-05,
+      "loss": 1.9839,
+      "step": 145000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.964589693571385e-05,
+      "loss": 1.9831,
+      "step": 145500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.959122440231987e-05,
+      "loss": 1.9843,
+      "step": 146000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.953655186892588e-05,
+      "loss": 1.9825,
+      "step": 146500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.9481879335531903e-05,
+      "loss": 1.9752,
+      "step": 147000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.9427206802137915e-05,
+      "loss": 1.9787,
+      "step": 147500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.937253426874393e-05,
+      "loss": 1.984,
+      "step": 148000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.931786173534995e-05,
+      "loss": 1.9783,
+      "step": 148500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.926318920195597e-05,
+      "loss": 1.9785,
+      "step": 149000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.920851666856199e-05,
+      "loss": 1.9759,
+      "step": 149500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.9153844135168e-05,
+      "loss": 1.9803,
+      "step": 150000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.9099171601774016e-05,
+      "loss": 1.9733,
+      "step": 150500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.904449906838003e-05,
+      "loss": 1.9759,
+      "step": 151000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.898982653498605e-05,
+      "loss": 1.9767,
+      "step": 151500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.893515400159207e-05,
+      "loss": 1.9773,
+      "step": 152000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.888048146819808e-05,
+      "loss": 1.9736,
+      "step": 152500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.88258089348041e-05,
+      "loss": 1.9729,
+      "step": 153000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.877113640141012e-05,
+      "loss": 1.9756,
+      "step": 153500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.8716463868016135e-05,
+      "loss": 1.9747,
+      "step": 154000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.8661791334622146e-05,
+      "loss": 1.9732,
+      "step": 154500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.8607118801228164e-05,
+      "loss": 1.9696,
+      "step": 155000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.855244626783418e-05,
+      "loss": 1.9731,
+      "step": 155500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.84977737344402e-05,
+      "loss": 1.9741,
+      "step": 156000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.844310120104622e-05,
+      "loss": 1.9747,
+      "step": 156500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.838842866765223e-05,
+      "loss": 1.9717,
+      "step": 157000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.833375613425825e-05,
+      "loss": 1.9724,
+      "step": 157500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.8279083600864266e-05,
+      "loss": 1.9663,
+      "step": 158000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.8224411067470284e-05,
+      "loss": 1.9671,
+      "step": 158500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.81697385340763e-05,
+      "loss": 1.9684,
+      "step": 159000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.811506600068231e-05,
+      "loss": 1.9683,
+      "step": 159500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.806039346728833e-05,
+      "loss": 1.9682,
+      "step": 160000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.800572093389435e-05,
+      "loss": 1.9673,
+      "step": 160500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.795104840050037e-05,
+      "loss": 1.9639,
+      "step": 161000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.7896375867106385e-05,
+      "loss": 1.9633,
+      "step": 161500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.7841703333712396e-05,
+      "loss": 1.964,
+      "step": 162000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.7787030800318414e-05,
+      "loss": 1.9648,
+      "step": 162500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.773235826692443e-05,
+      "loss": 1.9636,
+      "step": 163000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.767768573353045e-05,
+      "loss": 1.9623,
+      "step": 163500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.762301320013646e-05,
+      "loss": 1.9656,
+      "step": 164000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.756834066674248e-05,
+      "loss": 1.9636,
+      "step": 164500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.75136681333485e-05,
+      "loss": 1.9672,
+      "step": 165000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.7458995599954515e-05,
+      "loss": 1.9639,
+      "step": 165500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.740432306656053e-05,
+      "loss": 1.9628,
+      "step": 166000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.7349650533166545e-05,
+      "loss": 1.9647,
+      "step": 166500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.729497799977256e-05,
+      "loss": 1.9633,
+      "step": 167000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.724030546637858e-05,
+      "loss": 1.9585,
+      "step": 167500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.71856329329846e-05,
+      "loss": 1.9599,
+      "step": 168000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.7130960399590617e-05,
+      "loss": 1.9601,
+      "step": 168500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.707628786619663e-05,
+      "loss": 1.9617,
+      "step": 169000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.7021615332802646e-05,
+      "loss": 1.9583,
+      "step": 169500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.6966942799408664e-05,
+      "loss": 1.9606,
+      "step": 170000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.691227026601468e-05,
+      "loss": 1.955,
+      "step": 170500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.68575977326207e-05,
+      "loss": 1.956,
+      "step": 171000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.680292519922671e-05,
+      "loss": 1.9584,
+      "step": 171500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.674825266583273e-05,
+      "loss": 1.9575,
+      "step": 172000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.669358013243875e-05,
+      "loss": 1.9531,
+      "step": 172500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.6638907599044765e-05,
+      "loss": 1.9603,
+      "step": 173000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.6584235065650776e-05,
+      "loss": 1.9542,
+      "step": 173500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.6529562532256794e-05,
+      "loss": 1.9515,
+      "step": 174000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.647488999886282e-05,
+      "loss": 1.9594,
+      "step": 174500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.642021746546883e-05,
+      "loss": 1.955,
+      "step": 175000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.636554493207485e-05,
+      "loss": 1.9521,
+      "step": 175500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.631087239868086e-05,
+      "loss": 1.9564,
+      "step": 176000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.625619986528688e-05,
+      "loss": 1.9556,
+      "step": 176500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.6201527331892895e-05,
+      "loss": 1.9517,
+      "step": 177000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.6146854798498913e-05,
+      "loss": 1.9569,
+      "step": 177500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.609218226510493e-05,
+      "loss": 1.9482,
+      "step": 178000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.603750973171094e-05,
+      "loss": 1.9496,
+      "step": 178500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.598283719831696e-05,
+      "loss": 1.95,
+      "step": 179000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.592816466492298e-05,
+      "loss": 1.9519,
+      "step": 179500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.5873492131529e-05,
+      "loss": 1.9477,
+      "step": 180000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.5818819598135015e-05,
+      "loss": 1.9482,
+      "step": 180500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.5764147064741026e-05,
+      "loss": 1.9511,
+      "step": 181000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.570947453134705e-05,
+      "loss": 1.9464,
+      "step": 181500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.565480199795306e-05,
+      "loss": 1.9493,
+      "step": 182000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.560012946455908e-05,
+      "loss": 1.9462,
+      "step": 182500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.554545693116509e-05,
+      "loss": 1.9493,
+      "step": 183000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.549078439777111e-05,
+      "loss": 1.9472,
+      "step": 183500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.5436111864377134e-05,
+      "loss": 1.9475,
+      "step": 184000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.5381439330983145e-05,
+      "loss": 1.9461,
+      "step": 184500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.532676679758916e-05,
+      "loss": 1.9409,
+      "step": 185000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.5272094264195174e-05,
+      "loss": 1.9498,
+      "step": 185500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.521742173080119e-05,
+      "loss": 1.9475,
+      "step": 186000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.516274919740721e-05,
+      "loss": 1.9491,
+      "step": 186500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.510807666401323e-05,
+      "loss": 1.9433,
+      "step": 187000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.5053404130619246e-05,
+      "loss": 1.9436,
+      "step": 187500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.499873159722526e-05,
+      "loss": 1.9425,
+      "step": 188000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.494405906383128e-05,
+      "loss": 1.9384,
+      "step": 188500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.4889386530437294e-05,
+      "loss": 1.9398,
+      "step": 189000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.483471399704331e-05,
+      "loss": 1.9428,
+      "step": 189500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.478004146364933e-05,
+      "loss": 1.9416,
+      "step": 190000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.472536893025534e-05,
+      "loss": 1.9409,
+      "step": 190500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.4670696396861366e-05,
+      "loss": 1.9423,
+      "step": 191000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.461602386346738e-05,
+      "loss": 1.9409,
+      "step": 191500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.4561351330073395e-05,
+      "loss": 1.9399,
+      "step": 192000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.4506678796679406e-05,
+      "loss": 1.94,
+      "step": 192500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.4452006263285424e-05,
+      "loss": 1.9386,
+      "step": 193000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.439733372989145e-05,
+      "loss": 1.9391,
+      "step": 193500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.434266119649746e-05,
+      "loss": 1.9412,
+      "step": 194000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.428798866310348e-05,
+      "loss": 1.9384,
+      "step": 194500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.423331612970949e-05,
+      "loss": 1.9364,
+      "step": 195000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.4178643596315514e-05,
+      "loss": 1.9354,
+      "step": 195500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.4123971062921525e-05,
+      "loss": 1.9374,
+      "step": 196000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.406929852952754e-05,
+      "loss": 1.9363,
+      "step": 196500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.401462599613356e-05,
+      "loss": 1.9333,
+      "step": 197000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.395995346273957e-05,
+      "loss": 1.9375,
+      "step": 197500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.39052809293456e-05,
+      "loss": 1.9422,
+      "step": 198000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.385060839595161e-05,
+      "loss": 1.9363,
+      "step": 198500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.3795935862557626e-05,
+      "loss": 1.9335,
+      "step": 199000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.3741263329163644e-05,
+      "loss": 1.9378,
+      "step": 199500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.3686590795769656e-05,
+      "loss": 1.9394,
+      "step": 200000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.363191826237568e-05,
+      "loss": 1.927,
+      "step": 200500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.357724572898169e-05,
+      "loss": 1.9303,
+      "step": 201000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.352257319558771e-05,
+      "loss": 1.9333,
+      "step": 201500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.346790066219372e-05,
+      "loss": 1.9285,
+      "step": 202000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.3413228128799746e-05,
+      "loss": 1.9337,
+      "step": 202500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.3358555595405764e-05,
+      "loss": 1.9339,
+      "step": 203000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.3303883062011775e-05,
+      "loss": 1.9368,
+      "step": 203500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.324921052861779e-05,
+      "loss": 1.934,
+      "step": 204000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.3194537995223804e-05,
+      "loss": 1.9356,
+      "step": 204500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.313986546182983e-05,
+      "loss": 1.9305,
+      "step": 205000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.308519292843584e-05,
+      "loss": 1.9291,
+      "step": 205500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.303052039504186e-05,
+      "loss": 1.9323,
+      "step": 206000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.2975847861647876e-05,
+      "loss": 1.9343,
+      "step": 206500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.2921175328253894e-05,
+      "loss": 1.9315,
+      "step": 207000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.286650279485991e-05,
+      "loss": 1.9278,
+      "step": 207500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.2811830261465923e-05,
+      "loss": 1.9295,
+      "step": 208000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.275715772807194e-05,
+      "loss": 1.9323,
+      "step": 208500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.270248519467796e-05,
+      "loss": 1.9287,
+      "step": 209000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.264781266128398e-05,
+      "loss": 1.9327,
+      "step": 209500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.2593140127889995e-05,
+      "loss": 1.9311,
+      "step": 210000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.253846759449601e-05,
+      "loss": 1.9262,
+      "step": 210500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.2483795061102025e-05,
+      "loss": 1.9273,
+      "step": 211000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.2429122527708036e-05,
+      "loss": 1.9276,
+      "step": 211500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.237444999431406e-05,
+      "loss": 1.9307,
+      "step": 212000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.231977746092008e-05,
+      "loss": 1.9261,
+      "step": 212500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.226510492752609e-05,
+      "loss": 1.9286,
+      "step": 213000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.221043239413211e-05,
+      "loss": 1.9315,
+      "step": 213500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.2155759860738126e-05,
+      "loss": 1.9297,
+      "step": 214000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.2101087327344144e-05,
+      "loss": 1.9317,
+      "step": 214500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.2046414793950155e-05,
+      "loss": 1.927,
+      "step": 215000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.199174226055617e-05,
+      "loss": 1.9279,
+      "step": 215500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.193706972716219e-05,
+      "loss": 1.922,
+      "step": 216000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.188239719376821e-05,
+      "loss": 1.9259,
+      "step": 216500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.182772466037423e-05,
+      "loss": 1.9234,
+      "step": 217000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.177305212698024e-05,
+      "loss": 1.9232,
+      "step": 217500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.1718379593586256e-05,
+      "loss": 1.9226,
+      "step": 218000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.1663707060192274e-05,
+      "loss": 1.922,
+      "step": 218500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.160903452679829e-05,
+      "loss": 1.9215,
+      "step": 219000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.155436199340431e-05,
+      "loss": 1.9207,
+      "step": 219500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.149968946001032e-05,
+      "loss": 1.9232,
+      "step": 220000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.144501692661634e-05,
+      "loss": 1.9223,
+      "step": 220500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.139034439322236e-05,
+      "loss": 1.9224,
+      "step": 221000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.1335671859828376e-05,
+      "loss": 1.9219,
+      "step": 221500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.1280999326434394e-05,
+      "loss": 1.9235,
+      "step": 222000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.1226326793040405e-05,
+      "loss": 1.9225,
+      "step": 222500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.117165425964642e-05,
+      "loss": 1.92,
+      "step": 223000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.111698172625244e-05,
+      "loss": 1.9148,
+      "step": 223500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.106230919285846e-05,
+      "loss": 1.9245,
+      "step": 224000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.100763665946447e-05,
+      "loss": 1.9184,
+      "step": 224500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.095296412607049e-05,
+      "loss": 1.9193,
+      "step": 225000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.0898291592676506e-05,
+      "loss": 1.917,
+      "step": 225500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.0843619059282524e-05,
+      "loss": 1.9092,
+      "step": 226000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.078894652588854e-05,
+      "loss": 1.9185,
+      "step": 226500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.073427399249455e-05,
+      "loss": 1.9165,
+      "step": 227000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.067960145910057e-05,
+      "loss": 1.9171,
+      "step": 227500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.062492892570659e-05,
+      "loss": 1.9182,
+      "step": 228000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.057025639231261e-05,
+      "loss": 1.9182,
+      "step": 228500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.0515583858918622e-05,
+      "loss": 1.9155,
+      "step": 229000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.0460911325524636e-05,
+      "loss": 1.9137,
+      "step": 229500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.0406238792130654e-05,
+      "loss": 1.9162,
+      "step": 230000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.0351566258736676e-05,
+      "loss": 1.9165,
+      "step": 230500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.029689372534269e-05,
+      "loss": 1.9184,
+      "step": 231000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.0242221191948705e-05,
+      "loss": 1.9127,
+      "step": 231500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.018754865855472e-05,
+      "loss": 1.9197,
+      "step": 232000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.0132876125160738e-05,
+      "loss": 1.915,
+      "step": 232500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.0078203591766756e-05,
+      "loss": 1.9189,
+      "step": 233000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.0023531058372774e-05,
+      "loss": 1.9154,
+      "step": 233500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.9968858524978788e-05,
+      "loss": 1.9117,
+      "step": 234000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.9914185991584803e-05,
+      "loss": 1.911,
+      "step": 234500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.9859513458190824e-05,
+      "loss": 1.9191,
+      "step": 235000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.980484092479684e-05,
+      "loss": 1.9158,
+      "step": 235500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.9750168391402854e-05,
+      "loss": 1.9149,
+      "step": 236000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.969549585800887e-05,
+      "loss": 1.9146,
+      "step": 236500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.9640823324614886e-05,
+      "loss": 1.9111,
+      "step": 237000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.9586150791220908e-05,
+      "loss": 1.9105,
+      "step": 237500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.9531478257826922e-05,
+      "loss": 1.9125,
+      "step": 238000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.9476805724432937e-05,
+      "loss": 1.9051,
+      "step": 238500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.942213319103895e-05,
+      "loss": 1.913,
+      "step": 239000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.936746065764497e-05,
+      "loss": 1.9067,
+      "step": 239500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.931278812425099e-05,
+      "loss": 1.9093,
+      "step": 240000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.9258115590857005e-05,
+      "loss": 1.9095,
+      "step": 240500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.920344305746302e-05,
+      "loss": 1.9129,
+      "step": 241000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.9148770524069035e-05,
+      "loss": 1.9119,
+      "step": 241500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.9094097990675056e-05,
+      "loss": 1.911,
+      "step": 242000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.903942545728107e-05,
+      "loss": 1.9078,
+      "step": 242500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.898475292388709e-05,
+      "loss": 1.9113,
+      "step": 243000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8930080390493103e-05,
+      "loss": 1.9058,
+      "step": 243500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8875407857099118e-05,
+      "loss": 1.9114,
+      "step": 244000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.882073532370514e-05,
+      "loss": 1.9073,
+      "step": 244500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8766062790311154e-05,
+      "loss": 1.906,
+      "step": 245000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.871139025691717e-05,
+      "loss": 1.9072,
+      "step": 245500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8656717723523186e-05,
+      "loss": 1.9083,
+      "step": 246000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.86020451901292e-05,
+      "loss": 1.9047,
+      "step": 246500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8547372656735222e-05,
+      "loss": 1.9051,
+      "step": 247000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8492700123341237e-05,
+      "loss": 1.9053,
+      "step": 247500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.843802758994725e-05,
+      "loss": 1.9107,
+      "step": 248000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.838335505655327e-05,
+      "loss": 1.9094,
+      "step": 248500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8328682523159288e-05,
+      "loss": 1.9081,
+      "step": 249000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8274009989765306e-05,
+      "loss": 1.9067,
+      "step": 249500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.821933745637132e-05,
+      "loss": 1.9054,
+      "step": 250000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8164664922977335e-05,
+      "loss": 1.9007,
+      "step": 250500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.810999238958335e-05,
+      "loss": 1.9064,
+      "step": 251000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.805531985618937e-05,
+      "loss": 1.8992,
+      "step": 251500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8000647322795386e-05,
+      "loss": 1.9074,
+      "step": 252000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.7945974789401404e-05,
+      "loss": 1.9061,
+      "step": 252500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.7891302256007418e-05,
+      "loss": 1.9017,
+      "step": 253000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.7836629722613433e-05,
+      "loss": 1.9018,
+      "step": 253500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.7781957189219454e-05,
+      "loss": 1.9027,
+      "step": 254000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.772728465582547e-05,
+      "loss": 1.9062,
+      "step": 254500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.7672612122431483e-05,
+      "loss": 1.9026,
+      "step": 255000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.76179395890375e-05,
+      "loss": 1.9,
+      "step": 255500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.7563267055643523e-05,
+      "loss": 1.9043,
+      "step": 256000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.7508594522249537e-05,
+      "loss": 1.9041,
+      "step": 256500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7453921988855552e-05,
+      "loss": 1.9025,
+      "step": 257000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7399249455461567e-05,
+      "loss": 1.8992,
+      "step": 257500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7344576922067585e-05,
+      "loss": 1.9006,
+      "step": 258000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7289904388673603e-05,
+      "loss": 1.8998,
+      "step": 258500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.723523185527962e-05,
+      "loss": 1.8978,
+      "step": 259000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7180559321885635e-05,
+      "loss": 1.8991,
+      "step": 259500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.712588678849165e-05,
+      "loss": 1.898,
+      "step": 260000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.707121425509767e-05,
+      "loss": 1.9011,
+      "step": 260500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7016541721703686e-05,
+      "loss": 1.8968,
+      "step": 261000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.69618691883097e-05,
+      "loss": 1.9062,
+      "step": 261500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.690719665491572e-05,
+      "loss": 1.9018,
+      "step": 262000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.6852524121521733e-05,
+      "loss": 1.9018,
+      "step": 262500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.6797851588127754e-05,
+      "loss": 1.8995,
+      "step": 263000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.674317905473377e-05,
+      "loss": 1.8954,
+      "step": 263500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.6688506521339784e-05,
+      "loss": 1.8992,
+      "step": 264000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.6633833987945798e-05,
+      "loss": 1.8957,
+      "step": 264500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.6579161454551816e-05,
+      "loss": 1.9017,
+      "step": 265000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.6524488921157838e-05,
+      "loss": 1.8975,
+      "step": 265500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.6469816387763852e-05,
+      "loss": 1.8987,
+      "step": 266000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.6415143854369867e-05,
+      "loss": 1.8962,
+      "step": 266500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.636047132097588e-05,
+      "loss": 1.9013,
+      "step": 267000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.6305798787581903e-05,
+      "loss": 1.9004,
+      "step": 267500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.6251126254187917e-05,
+      "loss": 1.8955,
+      "step": 268000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.6196453720793935e-05,
+      "loss": 1.8956,
+      "step": 268500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.614178118739995e-05,
+      "loss": 1.8941,
+      "step": 269000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.6087108654005965e-05,
+      "loss": 1.9004,
+      "step": 269500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.6032436120611986e-05,
+      "loss": 1.8978,
+      "step": 270000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.5977763587218e-05,
+      "loss": 1.8954,
+      "step": 270500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.5923091053824015e-05,
+      "loss": 1.8948,
+      "step": 271000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.5868418520430033e-05,
+      "loss": 1.8946,
+      "step": 271500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5813745987036048e-05,
+      "loss": 1.897,
+      "step": 272000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.575907345364207e-05,
+      "loss": 1.897,
+      "step": 272500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5704400920248084e-05,
+      "loss": 1.8961,
+      "step": 273000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.56497283868541e-05,
+      "loss": 1.8925,
+      "step": 273500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5595055853460113e-05,
+      "loss": 1.8938,
+      "step": 274000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5540383320066135e-05,
+      "loss": 1.8928,
+      "step": 274500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5485710786672153e-05,
+      "loss": 1.8963,
+      "step": 275000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5431038253278167e-05,
+      "loss": 1.8923,
+      "step": 275500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5376365719884182e-05,
+      "loss": 1.8882,
+      "step": 276000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5321693186490196e-05,
+      "loss": 1.8887,
+      "step": 276500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.5267020653096218e-05,
+      "loss": 1.8903,
+      "step": 277000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.5212348119702232e-05,
+      "loss": 1.8941,
+      "step": 277500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.515767558630825e-05,
+      "loss": 1.8943,
+      "step": 278000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.5103003052914265e-05,
+      "loss": 1.8924,
+      "step": 278500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.504833051952028e-05,
+      "loss": 1.8855,
+      "step": 279000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.4993657986126298e-05,
+      "loss": 1.8918,
+      "step": 279500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.4938985452732316e-05,
+      "loss": 1.8891,
+      "step": 280000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.488431291933833e-05,
+      "loss": 1.8874,
+      "step": 280500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.4829640385944348e-05,
+      "loss": 1.8907,
+      "step": 281000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.4774967852550366e-05,
+      "loss": 1.8913,
+      "step": 281500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.472029531915638e-05,
+      "loss": 1.8888,
+      "step": 282000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.46656227857624e-05,
+      "loss": 1.8924,
+      "step": 282500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.4610950252368413e-05,
+      "loss": 1.8889,
+      "step": 283000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.455627771897443e-05,
+      "loss": 1.8866,
+      "step": 283500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.450160518558045e-05,
+      "loss": 1.8915,
+      "step": 284000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.4446932652186467e-05,
+      "loss": 1.8896,
+      "step": 284500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.4392260118792482e-05,
+      "loss": 1.8921,
+      "step": 285000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.43375875853985e-05,
+      "loss": 1.8854,
+      "step": 285500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.4282915052004515e-05,
+      "loss": 1.8887,
+      "step": 286000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.422824251861053e-05,
+      "loss": 1.8892,
+      "step": 286500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.4173569985216547e-05,
+      "loss": 1.8908,
+      "step": 287000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.4118897451822565e-05,
+      "loss": 1.8901,
+      "step": 287500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.4064224918428583e-05,
+      "loss": 1.8906,
+      "step": 288000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.4009552385034598e-05,
+      "loss": 1.8862,
+      "step": 288500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.3954879851640616e-05,
+      "loss": 1.8915,
+      "step": 289000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.390020731824663e-05,
+      "loss": 1.8891,
+      "step": 289500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.3845534784852645e-05,
+      "loss": 1.8855,
+      "step": 290000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.3790862251458667e-05,
+      "loss": 1.8828,
+      "step": 290500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.373618971806468e-05,
+      "loss": 1.8845,
+      "step": 291000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.36815171846707e-05,
+      "loss": 1.8824,
+      "step": 291500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.3626844651276714e-05,
+      "loss": 1.8832,
+      "step": 292000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.3572172117882732e-05,
+      "loss": 1.8862,
+      "step": 292500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.3517499584488746e-05,
+      "loss": 1.8862,
+      "step": 293000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.3462827051094764e-05,
+      "loss": 1.8837,
+      "step": 293500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.3408154517700782e-05,
+      "loss": 1.8864,
+      "step": 294000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.3353481984306797e-05,
+      "loss": 1.886,
+      "step": 294500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.3298809450912815e-05,
+      "loss": 1.884,
+      "step": 295000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.324413691751883e-05,
+      "loss": 1.8834,
+      "step": 295500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.3189464384124848e-05,
+      "loss": 1.8848,
+      "step": 296000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.3134791850730862e-05,
+      "loss": 1.8856,
+      "step": 296500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.308011931733688e-05,
+      "loss": 1.8874,
+      "step": 297000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.3025446783942898e-05,
+      "loss": 1.8775,
+      "step": 297500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.2970774250548913e-05,
+      "loss": 1.8859,
+      "step": 298000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.291610171715493e-05,
+      "loss": 1.8799,
+      "step": 298500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.2861429183760945e-05,
+      "loss": 1.8833,
+      "step": 299000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.2806756650366963e-05,
+      "loss": 1.882,
+      "step": 299500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.275208411697298e-05,
+      "loss": 1.8872,
+      "step": 300000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8590144171628205e-05,
+      "loss": 1.8825,
+      "step": 300500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.852863750430547e-05,
+      "loss": 1.8818,
+      "step": 301000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.846713083698273e-05,
+      "loss": 1.8859,
+      "step": 301500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8405624169659992e-05,
+      "loss": 1.8773,
+      "step": 302000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8344117502337256e-05,
+      "loss": 1.8764,
+      "step": 302500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8282610835014516e-05,
+      "loss": 1.8826,
+      "step": 303000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.822110416769178e-05,
+      "loss": 1.8805,
+      "step": 303500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.815959750036904e-05,
+      "loss": 1.8787,
+      "step": 304000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8098090833046304e-05,
+      "loss": 1.8793,
+      "step": 304500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8036584165723567e-05,
+      "loss": 1.8797,
+      "step": 305000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7975077498400827e-05,
+      "loss": 1.875,
+      "step": 305500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7913570831078088e-05,
+      "loss": 1.8747,
+      "step": 306000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.785206416375535e-05,
+      "loss": 1.8795,
+      "step": 306500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7790557496432615e-05,
+      "loss": 1.8761,
+      "step": 307000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7729050829109875e-05,
+      "loss": 1.878,
+      "step": 307500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.766754416178714e-05,
+      "loss": 1.8743,
+      "step": 308000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.76060374944644e-05,
+      "loss": 1.8762,
+      "step": 308500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7544530827141666e-05,
+      "loss": 1.8739,
+      "step": 309000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7483024159818926e-05,
+      "loss": 1.8747,
+      "step": 309500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7421517492496186e-05,
+      "loss": 1.8703,
+      "step": 310000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.736001082517345e-05,
+      "loss": 1.8709,
+      "step": 310500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.729850415785071e-05,
+      "loss": 1.8757,
+      "step": 311000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7236997490527974e-05,
+      "loss": 1.8746,
+      "step": 311500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7175490823205237e-05,
+      "loss": 1.8744,
+      "step": 312000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7113984155882498e-05,
+      "loss": 1.8742,
+      "step": 312500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7052477488559758e-05,
+      "loss": 1.8791,
+      "step": 313000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.6990970821237025e-05,
+      "loss": 1.871,
+      "step": 313500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6929464153914285e-05,
+      "loss": 1.8742,
+      "step": 314000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.686795748659155e-05,
+      "loss": 1.873,
+      "step": 314500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.680645081926881e-05,
+      "loss": 1.8775,
+      "step": 315000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6744944151946072e-05,
+      "loss": 1.8746,
+      "step": 315500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6683437484623336e-05,
+      "loss": 1.8686,
+      "step": 316000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6621930817300596e-05,
+      "loss": 1.8667,
+      "step": 316500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6560424149977856e-05,
+      "loss": 1.8716,
+      "step": 317000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.649891748265512e-05,
+      "loss": 1.8693,
+      "step": 317500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6437410815332384e-05,
+      "loss": 1.8752,
+      "step": 318000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6375904148009644e-05,
+      "loss": 1.8677,
+      "step": 318500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6314397480686908e-05,
+      "loss": 1.8724,
+      "step": 319000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6252890813364168e-05,
+      "loss": 1.8675,
+      "step": 319500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.619138414604143e-05,
+      "loss": 1.8674,
+      "step": 320000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6129877478718695e-05,
+      "loss": 1.8689,
+      "step": 320500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6068370811395955e-05,
+      "loss": 1.8717,
+      "step": 321000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.600686414407322e-05,
+      "loss": 1.8695,
+      "step": 321500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.594535747675048e-05,
+      "loss": 1.869,
+      "step": 322000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.5883850809427743e-05,
+      "loss": 1.8737,
+      "step": 322500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.5822344142105006e-05,
+      "loss": 1.868,
+      "step": 323000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.5760837474782266e-05,
+      "loss": 1.87,
+      "step": 323500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.5699330807459527e-05,
+      "loss": 1.8717,
+      "step": 324000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.5637824140136794e-05,
+      "loss": 1.8675,
+      "step": 324500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.5576317472814054e-05,
+      "loss": 1.8716,
+      "step": 325000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.5514810805491317e-05,
+      "loss": 1.8706,
+      "step": 325500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.5453304138168578e-05,
+      "loss": 1.867,
+      "step": 326000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.5391797470845838e-05,
+      "loss": 1.8638,
+      "step": 326500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.5330290803523105e-05,
+      "loss": 1.8642,
+      "step": 327000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.5268784136200365e-05,
+      "loss": 1.8672,
+      "step": 327500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.5207277468877625e-05,
+      "loss": 1.8677,
+      "step": 328000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.514577080155489e-05,
+      "loss": 1.867,
+      "step": 328500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.508426413423215e-05,
+      "loss": 1.8665,
+      "step": 329000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.5022757466909413e-05,
+      "loss": 1.8673,
+      "step": 329500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4961250799586676e-05,
+      "loss": 1.8719,
+      "step": 330000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4899744132263938e-05,
+      "loss": 1.8652,
+      "step": 330500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4838237464941202e-05,
+      "loss": 1.865,
+      "step": 331000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4776730797618462e-05,
+      "loss": 1.8645,
+      "step": 331500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4715224130295724e-05,
+      "loss": 1.8661,
+      "step": 332000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4653717462972988e-05,
+      "loss": 1.8658,
+      "step": 332500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.459221079565025e-05,
+      "loss": 1.8679,
+      "step": 333000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.453070412832751e-05,
+      "loss": 1.867,
+      "step": 333500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4469197461004775e-05,
+      "loss": 1.8648,
+      "step": 334000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4407690793682035e-05,
+      "loss": 1.8678,
+      "step": 334500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4346184126359297e-05,
+      "loss": 1.8665,
+      "step": 335000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.428467745903656e-05,
+      "loss": 1.8717,
+      "step": 335500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4223170791713821e-05,
+      "loss": 1.8683,
+      "step": 336000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.4161664124391086e-05,
+      "loss": 1.8657,
+      "step": 336500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.4100157457068347e-05,
+      "loss": 1.864,
+      "step": 337000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.4038650789745608e-05,
+      "loss": 1.8622,
+      "step": 337500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3977144122422872e-05,
+      "loss": 1.8679,
+      "step": 338000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3915637455100134e-05,
+      "loss": 1.8636,
+      "step": 338500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3854130787777394e-05,
+      "loss": 1.8688,
+      "step": 339000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3792624120454658e-05,
+      "loss": 1.8667,
+      "step": 339500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.373111745313192e-05,
+      "loss": 1.8619,
+      "step": 340000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3669610785809183e-05,
+      "loss": 1.8677,
+      "step": 340500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3608104118486445e-05,
+      "loss": 1.8635,
+      "step": 341000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3546597451163705e-05,
+      "loss": 1.8602,
+      "step": 341500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.348509078384097e-05,
+      "loss": 1.8627,
+      "step": 342000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3423584116518231e-05,
+      "loss": 1.8631,
+      "step": 342500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3362077449195493e-05,
+      "loss": 1.8664,
+      "step": 343000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3300570781872756e-05,
+      "loss": 1.8601,
+      "step": 343500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3239064114550018e-05,
+      "loss": 1.8602,
+      "step": 344000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3177557447227279e-05,
+      "loss": 1.8646,
+      "step": 344500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3116050779904542e-05,
+      "loss": 1.8606,
+      "step": 345000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3054544112581804e-05,
+      "loss": 1.8605,
+      "step": 345500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2993037445259068e-05,
+      "loss": 1.8622,
+      "step": 346000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.293153077793633e-05,
+      "loss": 1.8609,
+      "step": 346500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.287002411061359e-05,
+      "loss": 1.8543,
+      "step": 347000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2808517443290855e-05,
+      "loss": 1.8598,
+      "step": 347500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2747010775968115e-05,
+      "loss": 1.8589,
+      "step": 348000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2685504108645377e-05,
+      "loss": 1.8633,
+      "step": 348500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2623997441322641e-05,
+      "loss": 1.8633,
+      "step": 349000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2562490773999903e-05,
+      "loss": 1.8596,
+      "step": 349500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2500984106677163e-05,
+      "loss": 1.8577,
+      "step": 350000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2439477439354427e-05,
+      "loss": 1.8595,
+      "step": 350500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2377970772031689e-05,
+      "loss": 1.8702,
+      "step": 351000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.231646410470895e-05,
+      "loss": 1.8531,
+      "step": 351500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2254957437386214e-05,
+      "loss": 1.8599,
+      "step": 352000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2193450770063474e-05,
+      "loss": 1.862,
+      "step": 352500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2131944102740738e-05,
+      "loss": 1.8601,
+      "step": 353000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2070437435418e-05,
+      "loss": 1.8608,
+      "step": 353500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2008930768095263e-05,
+      "loss": 1.8589,
+      "step": 354000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1947424100772524e-05,
+      "loss": 1.8623,
+      "step": 354500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1885917433449786e-05,
+      "loss": 1.8616,
+      "step": 355000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1824410766127049e-05,
+      "loss": 1.8555,
+      "step": 355500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1762904098804311e-05,
+      "loss": 1.8579,
+      "step": 356000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1701397431481573e-05,
+      "loss": 1.8634,
+      "step": 356500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1639890764158835e-05,
+      "loss": 1.8557,
+      "step": 357000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1578384096836098e-05,
+      "loss": 1.8579,
+      "step": 357500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1516877429513359e-05,
+      "loss": 1.8614,
+      "step": 358000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1455370762190622e-05,
+      "loss": 1.8598,
+      "step": 358500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1393864094867884e-05,
+      "loss": 1.8567,
+      "step": 359000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.1332357427545146e-05,
+      "loss": 1.855,
+      "step": 359500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.1270850760222408e-05,
+      "loss": 1.8578,
+      "step": 360000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.120934409289967e-05,
+      "loss": 1.856,
+      "step": 360500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.1147837425576934e-05,
+      "loss": 1.8532,
+      "step": 361000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.1086330758254195e-05,
+      "loss": 1.8625,
+      "step": 361500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.1024824090931457e-05,
+      "loss": 1.8591,
+      "step": 362000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.096331742360872e-05,
+      "loss": 1.8595,
+      "step": 362500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0901810756285983e-05,
+      "loss": 1.8557,
+      "step": 363000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0840304088963243e-05,
+      "loss": 1.8576,
+      "step": 363500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.0778797421640507e-05,
+      "loss": 1.8548,
+      "step": 364000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.0717290754317769e-05,
+      "loss": 1.8605,
+      "step": 364500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.065578408699503e-05,
+      "loss": 1.8505,
+      "step": 365000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.0594277419672292e-05,
+      "loss": 1.8578,
+      "step": 365500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.0532770752349554e-05,
+      "loss": 1.857,
+      "step": 366000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.0471264085026818e-05,
+      "loss": 1.8545,
+      "step": 366500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.040975741770408e-05,
+      "loss": 1.8557,
+      "step": 367000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.0348250750381342e-05,
+      "loss": 1.8554,
+      "step": 367500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.0286744083058604e-05,
+      "loss": 1.8548,
+      "step": 368000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.0225237415735867e-05,
+      "loss": 1.8558,
+      "step": 368500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.0163730748413128e-05,
+      "loss": 1.8564,
+      "step": 369000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.0102224081090391e-05,
+      "loss": 1.8581,
+      "step": 369500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.0040717413767653e-05,
+      "loss": 1.8541,
+      "step": 370000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.979210746444915e-06,
+      "loss": 1.8522,
+      "step": 370500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.917704079122177e-06,
+      "loss": 1.8551,
+      "step": 371000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.856197411799439e-06,
+      "loss": 1.8572,
+      "step": 371500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.794690744476702e-06,
+      "loss": 1.8544,
+      "step": 372000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.733184077153964e-06,
+      "loss": 1.8509,
+      "step": 372500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.671677409831226e-06,
+      "loss": 1.8538,
+      "step": 373000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.610170742508488e-06,
+      "loss": 1.8561,
+      "step": 373500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.54866407518575e-06,
+      "loss": 1.8559,
+      "step": 374000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.487157407863014e-06,
+      "loss": 1.8559,
+      "step": 374500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.425650740540274e-06,
+      "loss": 1.8507,
+      "step": 375000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.364144073217537e-06,
+      "loss": 1.8526,
+      "step": 375500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.3026374058948e-06,
+      "loss": 1.8552,
+      "step": 376000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.241130738572061e-06,
+      "loss": 1.8526,
+      "step": 376500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.179624071249323e-06,
+      "loss": 1.8534,
+      "step": 377000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.118117403926587e-06,
+      "loss": 1.8539,
+      "step": 377500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.056610736603849e-06,
+      "loss": 1.8558,
+      "step": 378000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.99510406928111e-06,
+      "loss": 1.8532,
+      "step": 378500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.933597401958373e-06,
+      "loss": 1.8557,
+      "step": 379000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.872090734635634e-06,
+      "loss": 1.8528,
+      "step": 379500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.810584067312898e-06,
+      "loss": 1.8554,
+      "step": 380000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.749077399990158e-06,
+      "loss": 1.8508,
+      "step": 380500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.687570732667422e-06,
+      "loss": 1.8505,
+      "step": 381000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.626064065344684e-06,
+      "loss": 1.8489,
+      "step": 381500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.564557398021946e-06,
+      "loss": 1.8519,
+      "step": 382000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.503050730699208e-06,
+      "loss": 1.8565,
+      "step": 382500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.441544063376471e-06,
+      "loss": 1.852,
+      "step": 383000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.380037396053733e-06,
+      "loss": 1.8553,
+      "step": 383500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.318530728730995e-06,
+      "loss": 1.8512,
+      "step": 384000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.257024061408257e-06,
+      "loss": 1.8521,
+      "step": 384500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.195517394085519e-06,
+      "loss": 1.8495,
+      "step": 385000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.134010726762783e-06,
+      "loss": 1.8563,
+      "step": 385500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.072504059440043e-06,
+      "loss": 1.8524,
+      "step": 386000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.010997392117306e-06,
+      "loss": 1.8537,
+      "step": 386500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.949490724794568e-06,
+      "loss": 1.8481,
+      "step": 387000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.88798405747183e-06,
+      "loss": 1.8521,
+      "step": 387500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.826477390149092e-06,
+      "loss": 1.8488,
+      "step": 388000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.764970722826356e-06,
+      "loss": 1.856,
+      "step": 388500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.703464055503618e-06,
+      "loss": 1.8502,
+      "step": 389000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.64195738818088e-06,
+      "loss": 1.8534,
+      "step": 389500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.580450720858141e-06,
+      "loss": 1.8481,
+      "step": 390000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.518944053535404e-06,
+      "loss": 1.8516,
+      "step": 390500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.457437386212666e-06,
+      "loss": 1.8508,
+      "step": 391000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.395930718889928e-06,
+      "loss": 1.8442,
+      "step": 391500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.33442405156719e-06,
+      "loss": 1.8469,
+      "step": 392000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.272917384244453e-06,
+      "loss": 1.85,
+      "step": 392500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.211410716921714e-06,
+      "loss": 1.8454,
+      "step": 393000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.1499040495989765e-06,
+      "loss": 1.8523,
+      "step": 393500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.088397382276239e-06,
+      "loss": 1.8479,
+      "step": 394000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.026890714953502e-06,
+      "loss": 1.8438,
+      "step": 394500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.965384047630763e-06,
+      "loss": 1.8491,
+      "step": 395000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.903877380308026e-06,
+      "loss": 1.8492,
+      "step": 395500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.842370712985289e-06,
+      "loss": 1.8506,
+      "step": 396000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.7808640456625505e-06,
+      "loss": 1.8511,
+      "step": 396500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.719357378339812e-06,
+      "loss": 1.8479,
+      "step": 397000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.657850711017074e-06,
+      "loss": 1.8474,
+      "step": 397500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.596344043694337e-06,
+      "loss": 1.8472,
+      "step": 398000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.534837376371598e-06,
+      "loss": 1.8536,
+      "step": 398500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.473330709048861e-06,
+      "loss": 1.8487,
+      "step": 399000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.411824041726124e-06,
+      "loss": 1.8507,
+      "step": 399500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.3503173744033864e-06,
+      "loss": 1.8478,
+      "step": 400000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.2888107070806475e-06,
+      "loss": 1.8488,
+      "step": 400500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.22730403975791e-06,
+      "loss": 1.8462,
+      "step": 401000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.165797372435172e-06,
+      "loss": 1.8501,
+      "step": 401500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.104290705112434e-06,
+      "loss": 1.8495,
+      "step": 402000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.042784037789697e-06,
+      "loss": 1.8479,
+      "step": 402500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.981277370466959e-06,
+      "loss": 1.8474,
+      "step": 403000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.919770703144221e-06,
+      "loss": 1.851,
+      "step": 403500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.8582640358214834e-06,
+      "loss": 1.8451,
+      "step": 404000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.796757368498745e-06,
+      "loss": 1.8458,
+      "step": 404500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.735250701176008e-06,
+      "loss": 1.8485,
+      "step": 405000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.67374403385327e-06,
+      "loss": 1.8494,
+      "step": 405500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.612237366530533e-06,
+      "loss": 1.8437,
+      "step": 406000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.550730699207794e-06,
+      "loss": 1.8435,
+      "step": 406500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.489224031885057e-06,
+      "loss": 1.8472,
+      "step": 407000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.4277173645623185e-06,
+      "loss": 1.8469,
+      "step": 407500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.366210697239581e-06,
+      "loss": 1.845,
+      "step": 408000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.304704029916843e-06,
+      "loss": 1.8451,
+      "step": 408500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.243197362594105e-06,
+      "loss": 1.85,
+      "step": 409000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.181690695271368e-06,
+      "loss": 1.8436,
+      "step": 409500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.12018402794863e-06,
+      "loss": 1.8435,
+      "step": 410000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.0586773606258925e-06,
+      "loss": 1.8447,
+      "step": 410500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.9971706933031544e-06,
+      "loss": 1.847,
+      "step": 411000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.935664025980416e-06,
+      "loss": 1.8522,
+      "step": 411500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.874157358657678e-06,
+      "loss": 1.8474,
+      "step": 412000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.812650691334941e-06,
+      "loss": 1.8473,
+      "step": 412500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.751144024012203e-06,
+      "loss": 1.8446,
+      "step": 413000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.689637356689466e-06,
+      "loss": 1.8471,
+      "step": 413500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.628130689366728e-06,
+      "loss": 1.8472,
+      "step": 414000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.5666240220439895e-06,
+      "loss": 1.8456,
+      "step": 414500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.505117354721252e-06,
+      "loss": 1.8446,
+      "step": 415000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.443610687398514e-06,
+      "loss": 1.8441,
+      "step": 415500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.382104020075776e-06,
+      "loss": 1.8466,
+      "step": 416000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.320597352753038e-06,
+      "loss": 1.8424,
+      "step": 416500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.259090685430301e-06,
+      "loss": 1.8472,
+      "step": 417000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.197584018107563e-06,
+      "loss": 1.8423,
+      "step": 417500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.1360773507848255e-06,
+      "loss": 1.8463,
+      "step": 418000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.074570683462087e-06,
+      "loss": 1.8438,
+      "step": 418500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.01306401613935e-06,
+      "loss": 1.8399,
+      "step": 419000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.951557348816612e-06,
+      "loss": 1.8463,
+      "step": 419500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.890050681493874e-06,
+      "loss": 1.8404,
+      "step": 420000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.828544014171137e-06,
+      "loss": 1.8457,
+      "step": 420500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.767037346848398e-06,
+      "loss": 1.8451,
+      "step": 421000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.705530679525661e-06,
+      "loss": 1.8447,
+      "step": 421500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.644024012202923e-06,
+      "loss": 1.8427,
+      "step": 422000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.582517344880185e-06,
+      "loss": 1.8406,
+      "step": 422500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.521010677557447e-06,
+      "loss": 1.842,
+      "step": 423000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.45950401023471e-06,
+      "loss": 1.8426,
+      "step": 423500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.397997342911972e-06,
+      "loss": 1.8455,
+      "step": 424000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.336490675589234e-06,
+      "loss": 1.841,
+      "step": 424500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.274984008266496e-06,
+      "loss": 1.8418,
+      "step": 425000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.213477340943759e-06,
+      "loss": 1.8415,
+      "step": 425500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.1519706736210207e-06,
+      "loss": 1.8452,
+      "step": 426000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.090464006298283e-06,
+      "loss": 1.8472,
+      "step": 426500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.0289573389755454e-06,
+      "loss": 1.8403,
+      "step": 427000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.9674506716528073e-06,
+      "loss": 1.8406,
+      "step": 427500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.9059440043300696e-06,
+      "loss": 1.8425,
+      "step": 428000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.8444373370073315e-06,
+      "loss": 1.8417,
+      "step": 428500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.782930669684594e-06,
+      "loss": 1.8402,
+      "step": 429000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7214240023618562e-06,
+      "loss": 1.8465,
+      "step": 429500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.659917335039118e-06,
+      "loss": 1.8454,
+      "step": 430000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.5984106677163805e-06,
+      "loss": 1.8389,
+      "step": 430500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.536904000393643e-06,
+      "loss": 1.8439,
+      "step": 431000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.475397333070905e-06,
+      "loss": 1.8451,
+      "step": 431500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.413890665748167e-06,
+      "loss": 1.8393,
+      "step": 432000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.3523839984254294e-06,
+      "loss": 1.8428,
+      "step": 432500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.2908773311026917e-06,
+      "loss": 1.8397,
+      "step": 433000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.229370663779954e-06,
+      "loss": 1.8435,
+      "step": 433500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.1678639964572164e-06,
+      "loss": 1.8411,
+      "step": 434000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.1063573291344783e-06,
+      "loss": 1.8435,
+      "step": 434500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.0448506618117402e-06,
+      "loss": 1.8427,
+      "step": 435000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9833439944890026e-06,
+      "loss": 1.8371,
+      "step": 435500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.921837327166265e-06,
+      "loss": 1.8378,
+      "step": 436000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.860330659843527e-06,
+      "loss": 1.8403,
+      "step": 436500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.7988239925207894e-06,
+      "loss": 1.8427,
+      "step": 437000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.7373173251980517e-06,
+      "loss": 1.8414,
+      "step": 437500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.6758106578753138e-06,
+      "loss": 1.8371,
+      "step": 438000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.6143039905525761e-06,
+      "loss": 1.8388,
+      "step": 438500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.552797323229838e-06,
+      "loss": 1.8447,
+      "step": 439000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.4912906559071004e-06,
+      "loss": 1.8438,
+      "step": 439500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.4297839885843625e-06,
+      "loss": 1.84,
+      "step": 440000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.3682773212616249e-06,
+      "loss": 1.844,
+      "step": 440500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.3067706539388872e-06,
+      "loss": 1.8395,
+      "step": 441000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.245263986616149e-06,
+      "loss": 1.8408,
+      "step": 441500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.1837573192934114e-06,
+      "loss": 1.8426,
+      "step": 442000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.1222506519706736e-06,
+      "loss": 1.8412,
+      "step": 442500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.060743984647936e-06,
+      "loss": 1.8389,
+      "step": 443000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 9.992373173251982e-07,
+      "loss": 1.839,
+      "step": 443500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 9.377306500024604e-07,
+      "loss": 1.8411,
+      "step": 444000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.762239826797225e-07,
+      "loss": 1.843,
+      "step": 444500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.147173153569847e-07,
+      "loss": 1.84,
+      "step": 445000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 7.532106480342469e-07,
+      "loss": 1.847,
+      "step": 445500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.917039807115092e-07,
+      "loss": 1.8363,
+      "step": 446000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.301973133887713e-07,
+      "loss": 1.8402,
+      "step": 446500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.686906460660336e-07,
+      "loss": 1.8397,
+      "step": 447000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.071839787432959e-07,
+      "loss": 1.8424,
+      "step": 447500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.45677311420558e-07,
+      "loss": 1.8357,
+      "step": 448000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.841706440978202e-07,
+      "loss": 1.843,
+      "step": 448500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.2266397677508245e-07,
+      "loss": 1.8371,
+      "step": 449000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6115730945234463e-07,
+      "loss": 1.8429,
+      "step": 449500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9965064212960688e-07,
+      "loss": 1.8424,
+      "step": 450000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.381439748068691e-07,
+      "loss": 1.8414,
+      "step": 450500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.663730748413129e-08,
+      "loss": 1.8446,
+      "step": 451000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5130640161393495e-08,
+      "loss": 1.8381,
+      "step": 451500
+    },
+    {
+      "epoch": 1.0,
+      "step": 451623,
+      "total_flos": 1.5894400168611545e+19,
+      "train_loss": 0.6229404193166468,
+      "train_runtime": 298967.217,
+      "train_samples_per_second": 870.111,
+      "train_steps_per_second": 1.511
+    }
+  ],
+  "max_steps": 451623,
+  "num_train_epochs": 1,
+  "total_flos": 1.5894400168611545e+19,
+  "trial_name": null,
+  "trial_params": null
+}