diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7221 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9261155815654716,
+  "eval_steps": 500,
+  "global_step": 12000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 5.6785,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 4.5104,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-06,
+      "loss": 4.054,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.3259,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-06,
+      "loss": 2.6502,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6e-06,
+      "loss": 3.112,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 3.557,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 4.2576,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9e-06,
+      "loss": 2.8817,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 3.4098,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 2.8188,
+      "step": 110
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.2e-05,
+      "loss": 1.8579,
+      "step": 120
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 4.1002,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 2.9713,
+      "step": 140
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.5e-05,
+      "loss": 2.7547,
+      "step": 150
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 3.3592,
+      "step": 160
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 3.3475,
+      "step": 170
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.8e-05,
+      "loss": 2.6595,
+      "step": 180
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9e-05,
+      "loss": 2.5173,
+      "step": 190
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 1.8036,
+      "step": 200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.1e-05,
+      "loss": 3.8063,
+      "step": 210
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 3.1958,
+      "step": 220
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 2.5966,
+      "step": 230
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.4e-05,
+      "loss": 1.2437,
+      "step": 240
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.5e-05,
+      "loss": 3.9446,
+      "step": 250
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 2.9969,
+      "step": 260
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 2.8512,
+      "step": 270
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 2.083,
+      "step": 280
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9e-05,
+      "loss": 2.5073,
+      "step": 290
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3e-05,
+      "loss": 3.2805,
+      "step": 300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.1e-05,
+      "loss": 2.9954,
+      "step": 310
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 3.0859,
+      "step": 320
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.3e-05,
+      "loss": 2.8681,
+      "step": 330
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 1.8659,
+      "step": 340
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 3.5e-05,
+      "loss": 3.2076,
+      "step": 350
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 3.6e-05,
+      "loss": 4.5292,
+      "step": 360
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 3.7e-05,
+      "loss": 3.1894,
+      "step": 370
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 3.8e-05,
+      "loss": 3.1197,
+      "step": 380
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 3.6814,
+      "step": 390
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4e-05,
+      "loss": 2.9482,
+      "step": 400
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.1e-05,
+      "loss": 3.0854,
+      "step": 410
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.2e-05,
+      "loss": 2.2873,
+      "step": 420
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.3e-05,
+      "loss": 3.2374,
+      "step": 430
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 3.1769,
+      "step": 440
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.5e-05,
+      "loss": 4.7101,
+      "step": 450
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 2.7169,
+      "step": 460
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.7e-05,
+      "loss": 1.9965,
+      "step": 470
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8e-05,
+      "loss": 3.5535,
+      "step": 480
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9e-05,
+      "loss": 3.3874,
+      "step": 490
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 4.1868,
+      "step": 500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9957637888672374e-05,
+      "loss": 2.5748,
+      "step": 510
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9915275777344746e-05,
+      "loss": 2.8385,
+      "step": 520
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.987291366601712e-05,
+      "loss": 3.2211,
+      "step": 530
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.983055155468949e-05,
+      "loss": 3.3657,
+      "step": 540
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.978818944336186e-05,
+      "loss": 2.5346,
+      "step": 550
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9745827332034226e-05,
+      "loss": 4.6905,
+      "step": 560
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9703465220706605e-05,
+      "loss": 2.579,
+      "step": 570
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.966110310937898e-05,
+      "loss": 3.1246,
+      "step": 580
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.961874099805134e-05,
+      "loss": 4.3711,
+      "step": 590
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.957637888672372e-05,
+      "loss": 3.7318,
+      "step": 600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9534016775396085e-05,
+      "loss": 3.3223,
+      "step": 610
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.949165466406846e-05,
+      "loss": 2.5632,
+      "step": 620
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9449292552740836e-05,
+      "loss": 1.8259,
+      "step": 630
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.94069304414132e-05,
+      "loss": 2.3178,
+      "step": 640
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.936456833008557e-05,
+      "loss": 3.5661,
+      "step": 650
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9322206218757944e-05,
+      "loss": 2.3049,
+      "step": 660
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9279844107430316e-05,
+      "loss": 2.941,
+      "step": 670
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.923748199610269e-05,
+      "loss": 1.7203,
+      "step": 680
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.919511988477506e-05,
+      "loss": 3.4478,
+      "step": 690
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.915275777344743e-05,
+      "loss": 2.8344,
+      "step": 700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.91103956621198e-05,
+      "loss": 2.909,
+      "step": 710
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9068033550792175e-05,
+      "loss": 1.5403,
+      "step": 720
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.902567143946454e-05,
+      "loss": 3.4486,
+      "step": 730
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.898330932813692e-05,
+      "loss": 2.3368,
+      "step": 740
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.894094721680929e-05,
+      "loss": 4.352,
+      "step": 750
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.889858510548166e-05,
+      "loss": 3.3411,
+      "step": 760
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.8856222994154034e-05,
+      "loss": 3.125,
+      "step": 770
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.88138608828264e-05,
+      "loss": 3.0494,
+      "step": 780
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.877149877149878e-05,
+      "loss": 2.3695,
+      "step": 790
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.872913666017114e-05,
+      "loss": 2.8533,
+      "step": 800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.8686774548843514e-05,
+      "loss": 4.6261,
+      "step": 810
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.864441243751589e-05,
+      "loss": 3.8034,
+      "step": 820
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.860205032618826e-05,
+      "loss": 3.7497,
+      "step": 830
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.855968821486063e-05,
+      "loss": 3.9485,
+      "step": 840
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.8517326103533e-05,
+      "loss": 4.0835,
+      "step": 850
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.847496399220537e-05,
+      "loss": 1.7266,
+      "step": 860
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.8432601880877745e-05,
+      "loss": 3.3918,
+      "step": 870
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.839023976955012e-05,
+      "loss": 1.7983,
+      "step": 880
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.834787765822249e-05,
+      "loss": 3.6065,
+      "step": 890
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.830551554689486e-05,
+      "loss": 1.6316,
+      "step": 900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.826315343556723e-05,
+      "loss": 3.5684,
+      "step": 910
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.82207913242396e-05,
+      "loss": 3.4992,
+      "step": 920
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.8178429212911976e-05,
+      "loss": 2.5107,
+      "step": 930
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.813606710158435e-05,
+      "loss": 3.8318,
+      "step": 940
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.809370499025672e-05,
+      "loss": 3.6797,
+      "step": 950
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.805134287892909e-05,
+      "loss": 2.384,
+      "step": 960
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.8008980767601456e-05,
+      "loss": 3.0802,
+      "step": 970
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.7966618656273835e-05,
+      "loss": 3.3293,
+      "step": 980
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.79242565449462e-05,
+      "loss": 3.0357,
+      "step": 990
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.788189443361857e-05,
+      "loss": 3.1906,
+      "step": 1000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.783953232229095e-05,
+      "loss": 3.8859,
+      "step": 1010
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.7797170210963315e-05,
+      "loss": 2.3375,
+      "step": 1020
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.775480809963569e-05,
+      "loss": 3.1464,
+      "step": 1030
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.771244598830806e-05,
+      "loss": 3.2851,
+      "step": 1040
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.767008387698043e-05,
+      "loss": 3.2028,
+      "step": 1050
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.76277217656528e-05,
+      "loss": 3.0606,
+      "step": 1060
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.7585359654325174e-05,
+      "loss": 3.7212,
+      "step": 1070
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.7542997542997546e-05,
+      "loss": 3.4396,
+      "step": 1080
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.750063543166992e-05,
+      "loss": 4.0281,
+      "step": 1090
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.745827332034229e-05,
+      "loss": 2.8205,
+      "step": 1100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.7415911209014654e-05,
+      "loss": 3.118,
+      "step": 1110
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.737354909768703e-05,
+      "loss": 3.895,
+      "step": 1120
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.7331186986359405e-05,
+      "loss": 4.0991,
+      "step": 1130
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.728882487503177e-05,
+      "loss": 2.5872,
+      "step": 1140
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.724646276370415e-05,
+      "loss": 2.4609,
+      "step": 1150
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.720410065237651e-05,
+      "loss": 3.2963,
+      "step": 1160
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.716173854104889e-05,
+      "loss": 2.6754,
+      "step": 1170
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.711937642972126e-05,
+      "loss": 2.713,
+      "step": 1180
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.707701431839363e-05,
+      "loss": 2.8602,
+      "step": 1190
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.703465220706601e-05,
+      "loss": 3.3341,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.699229009573837e-05,
+      "loss": 3.1514,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.6949927984410744e-05,
+      "loss": 3.8893,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.6907565873083116e-05,
+      "loss": 2.2375,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.686520376175549e-05,
+      "loss": 2.2121,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.682284165042786e-05,
+      "loss": 2.2958,
+      "step": 1250
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.678047953910023e-05,
+      "loss": 4.2912,
+      "step": 1260
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.67381174277726e-05,
+      "loss": 3.4638,
+      "step": 1270
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.6695755316444975e-05,
+      "loss": 2.4403,
+      "step": 1280
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.6653393205117346e-05,
+      "loss": 4.1816,
+      "step": 1290
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.661103109378971e-05,
+      "loss": 3.2218,
+      "step": 1300
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.656866898246209e-05,
+      "loss": 3.0637,
+      "step": 1310
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.652630687113446e-05,
+      "loss": 1.8839,
+      "step": 1320
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.648394475980683e-05,
+      "loss": 4.1917,
+      "step": 1330
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.6441582648479205e-05,
+      "loss": 3.7383,
+      "step": 1340
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.639922053715157e-05,
+      "loss": 3.1253,
+      "step": 1350
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.635685842582395e-05,
+      "loss": 2.4917,
+      "step": 1360
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.6314496314496314e-05,
+      "loss": 3.9342,
+      "step": 1370
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.6272134203168686e-05,
+      "loss": 3.93,
+      "step": 1380
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.6229772091841064e-05,
+      "loss": 3.9129,
+      "step": 1390
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.618740998051343e-05,
+      "loss": 3.419,
+      "step": 1400
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.61450478691858e-05,
+      "loss": 4.4223,
+      "step": 1410
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.610268575785817e-05,
+      "loss": 2.353,
+      "step": 1420
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.6060323646530545e-05,
+      "loss": 3.5424,
+      "step": 1430
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.6017961535202917e-05,
+      "loss": 2.0227,
+      "step": 1440
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.597559942387529e-05,
+      "loss": 3.3002,
+      "step": 1450
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.593323731254766e-05,
+      "loss": 2.7871,
+      "step": 1460
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.589087520122003e-05,
+      "loss": 2.3579,
+      "step": 1470
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.5848513089892404e-05,
+      "loss": 2.7544,
+      "step": 1480
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.580615097856477e-05,
+      "loss": 3.8925,
+      "step": 1490
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.576378886723715e-05,
+      "loss": 4.5524,
+      "step": 1500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.572142675590952e-05,
+      "loss": 2.9189,
+      "step": 1510
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.5679064644581884e-05,
+      "loss": 3.3411,
+      "step": 1520
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.563670253325426e-05,
+      "loss": 2.4151,
+      "step": 1530
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.559434042192663e-05,
+      "loss": 2.5622,
+      "step": 1540
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.5551978310599e-05,
+      "loss": 2.7203,
+      "step": 1550
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.550961619927138e-05,
+      "loss": 2.8707,
+      "step": 1560
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.546725408794374e-05,
+      "loss": 3.0078,
+      "step": 1570
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.542489197661612e-05,
+      "loss": 3.2162,
+      "step": 1580
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.5382529865288487e-05,
+      "loss": 4.1395,
+      "step": 1590
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.534016775396086e-05,
+      "loss": 1.4732,
+      "step": 1600
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.529780564263323e-05,
+      "loss": 3.9653,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.52554435313056e-05,
+      "loss": 2.3421,
+      "step": 1620
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5213081419977974e-05,
+      "loss": 2.1206,
+      "step": 1630
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5170719308650345e-05,
+      "loss": 4.0762,
+      "step": 1640
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.512835719732272e-05,
+      "loss": 4.574,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.508599508599509e-05,
+      "loss": 3.4856,
+      "step": 1660
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.504363297466746e-05,
+      "loss": 3.3381,
+      "step": 1670
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.5001270863339826e-05,
+      "loss": 0.9077,
+      "step": 1680
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.4958908752012204e-05,
+      "loss": 2.7744,
+      "step": 1690
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.4916546640684576e-05,
+      "loss": 3.734,
+      "step": 1700
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.487418452935694e-05,
+      "loss": 3.8027,
+      "step": 1710
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.483182241802932e-05,
+      "loss": 2.5385,
+      "step": 1720
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.4789460306701685e-05,
+      "loss": 4.9046,
+      "step": 1730
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.4747098195374057e-05,
+      "loss": 2.9602,
+      "step": 1740
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.4704736084046435e-05,
+      "loss": 2.9869,
+      "step": 1750
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.46623739727188e-05,
+      "loss": 3.1115,
+      "step": 1760
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.462001186139118e-05,
+      "loss": 3.5951,
+      "step": 1770
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.4577649750063544e-05,
+      "loss": 3.4571,
+      "step": 1780
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4535287638735916e-05,
+      "loss": 2.5486,
+      "step": 1790
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.449292552740829e-05,
+      "loss": 3.6386,
+      "step": 1800
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.445056341608066e-05,
+      "loss": 4.2667,
+      "step": 1810
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.440820130475303e-05,
+      "loss": 2.4345,
+      "step": 1820
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.43658391934254e-05,
+      "loss": 3.87,
+      "step": 1830
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4323477082097774e-05,
+      "loss": 3.5761,
+      "step": 1840
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4281114970770146e-05,
+      "loss": 3.8179,
+      "step": 1850
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.423875285944252e-05,
+      "loss": 2.7925,
+      "step": 1860
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.419639074811489e-05,
+      "loss": 5.4179,
+      "step": 1870
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.415402863678726e-05,
+      "loss": 2.41,
+      "step": 1880
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.411166652545963e-05,
+      "loss": 2.5791,
+      "step": 1890
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.4069304414132e-05,
+      "loss": 2.0424,
+      "step": 1900
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.402694230280438e-05,
+      "loss": 2.3137,
+      "step": 1910
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.398458019147674e-05,
+      "loss": 2.0938,
+      "step": 1920
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3942218080149114e-05,
+      "loss": 3.4396,
+      "step": 1930
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.389985596882149e-05,
+      "loss": 2.0333,
+      "step": 1940
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.385749385749386e-05,
+      "loss": 3.4806,
+      "step": 1950
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3815131746166236e-05,
+      "loss": 3.8209,
+      "step": 1960
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.37727696348386e-05,
+      "loss": 3.1992,
+      "step": 1970
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.373040752351097e-05,
+      "loss": 2.7737,
+      "step": 1980
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3688045412183344e-05,
+      "loss": 2.8133,
+      "step": 1990
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3645683300855716e-05,
+      "loss": 2.9959,
+      "step": 2000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.360332118952809e-05,
+      "loss": 3.0946,
+      "step": 2010
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.356095907820046e-05,
+      "loss": 2.2239,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.351859696687283e-05,
+      "loss": 2.9569,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.3476234855545203e-05,
+      "loss": 3.9401,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.3433872744217575e-05,
+      "loss": 2.447,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.339151063288995e-05,
+      "loss": 4.3506,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.334914852156232e-05,
+      "loss": 3.7485,
+      "step": 2070
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.330678641023469e-05,
+      "loss": 3.132,
+      "step": 2080
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.3264424298907056e-05,
+      "loss": 2.5788,
+      "step": 2090
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.3222062187579434e-05,
+      "loss": 2.9854,
+      "step": 2100
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.31797000762518e-05,
+      "loss": 3.765,
+      "step": 2110
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.313733796492417e-05,
+      "loss": 1.9939,
+      "step": 2120
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.309497585359655e-05,
+      "loss": 2.3252,
+      "step": 2130
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.3052613742268915e-05,
+      "loss": 2.6576,
+      "step": 2140
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.3010251630941286e-05,
+      "loss": 3.2442,
+      "step": 2150
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.296788951961366e-05,
+      "loss": 3.2068,
+      "step": 2160
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.292552740828603e-05,
+      "loss": 2.89,
+      "step": 2170
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.288316529695841e-05,
+      "loss": 3.0358,
+      "step": 2180
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.2840803185630773e-05,
+      "loss": 2.8249,
+      "step": 2190
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.2798441074303145e-05,
+      "loss": 2.6096,
+      "step": 2200
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.275607896297552e-05,
+      "loss": 3.9853,
+      "step": 2210
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.271371685164789e-05,
+      "loss": 3.0829,
+      "step": 2220
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.267135474032026e-05,
+      "loss": 3.3061,
+      "step": 2230
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.262899262899263e-05,
+      "loss": 3.7588,
+      "step": 2240
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.2586630517665004e-05,
+      "loss": 2.9968,
+      "step": 2250
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.2544268406337376e-05,
+      "loss": 3.6007,
+      "step": 2260
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.250190629500975e-05,
+      "loss": 2.1882,
+      "step": 2270
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.245954418368211e-05,
+      "loss": 3.2399,
+      "step": 2280
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.241718207235449e-05,
+      "loss": 3.012,
+      "step": 2290
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.2374819961026856e-05,
+      "loss": 3.4545,
+      "step": 2300
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.233245784969923e-05,
+      "loss": 2.9524,
+      "step": 2310
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.229009573837161e-05,
+      "loss": 3.6387,
+      "step": 2320
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.224773362704397e-05,
+      "loss": 2.0361,
+      "step": 2330
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.2205371515716343e-05,
+      "loss": 2.8897,
+      "step": 2340
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.2163009404388715e-05,
+      "loss": 2.303,
+      "step": 2350
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.212064729306109e-05,
+      "loss": 2.8729,
+      "step": 2360
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.2078285181733466e-05,
+      "loss": 3.1418,
+      "step": 2370
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.203592307040583e-05,
+      "loss": 3.6455,
+      "step": 2380
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.19935609590782e-05,
+      "loss": 4.3795,
+      "step": 2390
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.1951198847750574e-05,
+      "loss": 2.5988,
+      "step": 2400
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.1908836736422946e-05,
+      "loss": 3.6995,
+      "step": 2410
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.186647462509531e-05,
+      "loss": 3.7109,
+      "step": 2420
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.182411251376769e-05,
+      "loss": 3.3371,
+      "step": 2430
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.178175040244006e-05,
+      "loss": 2.6656,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.173938829111243e-05,
+      "loss": 3.7572,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.1697026179784805e-05,
+      "loss": 3.3319,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.165466406845717e-05,
+      "loss": 2.3181,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.161230195712955e-05,
+      "loss": 3.4949,
+      "step": 2480
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.1569939845801914e-05,
+      "loss": 2.8937,
+      "step": 2490
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.1527577734474285e-05,
+      "loss": 3.3363,
+      "step": 2500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.1485215623146664e-05,
+      "loss": 3.0678,
+      "step": 2510
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.144285351181903e-05,
+      "loss": 2.1284,
+      "step": 2520
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.14004914004914e-05,
+      "loss": 3.0226,
+      "step": 2530
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.135812928916377e-05,
+      "loss": 2.0553,
+      "step": 2540
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.1315767177836144e-05,
+      "loss": 2.1608,
+      "step": 2550
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.1273405066508516e-05,
+      "loss": 3.9251,
+      "step": 2560
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.123104295518089e-05,
+      "loss": 3.5469,
+      "step": 2570
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.118868084385326e-05,
+      "loss": 2.627,
+      "step": 2580
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.114631873252563e-05,
+      "loss": 2.791,
+      "step": 2590
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.1103956621198e-05,
+      "loss": 3.2762,
+      "step": 2600
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.106159450987037e-05,
+      "loss": 2.4096,
+      "step": 2610
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.101923239854275e-05,
+      "loss": 3.8677,
+      "step": 2620
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.097687028721512e-05,
+      "loss": 4.0316,
+      "step": 2630
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.093450817588749e-05,
+      "loss": 3.8613,
+      "step": 2640
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.089214606455986e-05,
+      "loss": 2.6343,
+      "step": 2650
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.084978395323223e-05,
+      "loss": 3.0759,
+      "step": 2660
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.0807421841904606e-05,
+      "loss": 2.6594,
+      "step": 2670
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.076505973057698e-05,
+      "loss": 4.5179,
+      "step": 2680
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.072269761924934e-05,
+      "loss": 1.9108,
+      "step": 2690
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.068033550792172e-05,
+      "loss": 3.1447,
+      "step": 2700
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.0637973396594086e-05,
+      "loss": 3.3925,
+      "step": 2710
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.059561128526646e-05,
+      "loss": 2.6961,
+      "step": 2720
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.055324917393883e-05,
+      "loss": 2.4436,
+      "step": 2730
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.05108870626112e-05,
+      "loss": 4.1392,
+      "step": 2740
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.046852495128357e-05,
+      "loss": 3.049,
+      "step": 2750
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.0426162839955945e-05,
+      "loss": 2.8152,
+      "step": 2760
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.038380072862832e-05,
+      "loss": 3.9789,
+      "step": 2770
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.034143861730069e-05,
+      "loss": 2.3605,
+      "step": 2780
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.029907650597306e-05,
+      "loss": 2.4767,
+      "step": 2790
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.0256714394645425e-05,
+      "loss": 3.7781,
+      "step": 2800
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.0214352283317804e-05,
+      "loss": 3.3173,
+      "step": 2810
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.0171990171990176e-05,
+      "loss": 4.4773,
+      "step": 2820
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.012962806066254e-05,
+      "loss": 2.5059,
+      "step": 2830
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.008726594933492e-05,
+      "loss": 2.9141,
+      "step": 2840
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.0044903838007284e-05,
+      "loss": 2.5059,
+      "step": 2850
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.000254172667966e-05,
+      "loss": 2.6355,
+      "step": 2860
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.9960179615352035e-05,
+      "loss": 2.057,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.99178175040244e-05,
+      "loss": 0.5472,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.987545539269678e-05,
+      "loss": 3.1453,
+      "step": 2890
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.983309328136914e-05,
+      "loss": 2.9714,
+      "step": 2900
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.9790731170041515e-05,
+      "loss": 2.1662,
+      "step": 2910
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.974836905871389e-05,
+      "loss": 3.4159,
+      "step": 2920
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.970600694738626e-05,
+      "loss": 2.4046,
+      "step": 2930
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.966364483605863e-05,
+      "loss": 2.9214,
+      "step": 2940
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.9621282724731e-05,
+      "loss": 3.9307,
+      "step": 2950
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.9578920613403374e-05,
+      "loss": 2.5037,
+      "step": 2960
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.9536558502075746e-05,
+      "loss": 3.6345,
+      "step": 2970
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.949419639074812e-05,
+      "loss": 2.9814,
+      "step": 2980
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.945183427942049e-05,
+      "loss": 2.3943,
+      "step": 2990
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.940947216809286e-05,
+      "loss": 2.4179,
+      "step": 3000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.936711005676523e-05,
+      "loss": 1.6241,
+      "step": 3010
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.93247479454376e-05,
+      "loss": 4.0399,
+      "step": 3020
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.9282385834109976e-05,
+      "loss": 3.1503,
+      "step": 3030
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.924002372278234e-05,
+      "loss": 3.9505,
+      "step": 3040
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.919766161145472e-05,
+      "loss": 2.486,
+      "step": 3050
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.915529950012709e-05,
+      "loss": 1.4641,
+      "step": 3060
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.911293738879946e-05,
+      "loss": 3.2979,
+      "step": 3070
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.9070575277471835e-05,
+      "loss": 2.6891,
+      "step": 3080
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.90282131661442e-05,
+      "loss": 3.5378,
+      "step": 3090
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.898585105481657e-05,
+      "loss": 2.5907,
+      "step": 3100
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.8943488943488944e-05,
+      "loss": 2.5752,
+      "step": 3110
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.8901126832161316e-05,
+      "loss": 2.4972,
+      "step": 3120
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.885876472083369e-05,
+      "loss": 1.5583,
+      "step": 3130
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.881640260950606e-05,
+      "loss": 3.3501,
+      "step": 3140
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.877404049817843e-05,
+      "loss": 3.8679,
+      "step": 3150
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.87316783868508e-05,
+      "loss": 1.9064,
+      "step": 3160
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.8689316275523175e-05,
+      "loss": 3.5752,
+      "step": 3170
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.8646954164195547e-05,
+      "loss": 2.0187,
+      "step": 3180
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.860459205286792e-05,
+      "loss": 3.9003,
+      "step": 3190
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.856222994154029e-05,
+      "loss": 3.5827,
+      "step": 3200
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.8519867830212655e-05,
+      "loss": 2.6393,
+      "step": 3210
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.8477505718885034e-05,
+      "loss": 3.8133,
+      "step": 3220
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.84351436075574e-05,
+      "loss": 3.5699,
+      "step": 3230
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.839278149622977e-05,
+      "loss": 4.5841,
+      "step": 3240
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.835041938490215e-05,
+      "loss": 3.1499,
+      "step": 3250
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.8308057273574514e-05,
+      "loss": 4.2794,
+      "step": 3260
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.826569516224689e-05,
+      "loss": 2.5589,
+      "step": 3270
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.822333305091926e-05,
+      "loss": 4.8515,
+      "step": 3280
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.818097093959163e-05,
+      "loss": 2.6834,
+      "step": 3290
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.813860882826401e-05,
+      "loss": 4.4515,
+      "step": 3300
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.809624671693637e-05,
+      "loss": 2.4279,
+      "step": 3310
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.8053884605608745e-05,
+      "loss": 2.6435,
+      "step": 3320
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.8011522494281117e-05,
+      "loss": 3.7042,
+      "step": 3330
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.796916038295349e-05,
+      "loss": 2.6835,
+      "step": 3340
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.792679827162586e-05,
+      "loss": 3.2614,
+      "step": 3350
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.788443616029823e-05,
+      "loss": 2.062,
+      "step": 3360
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.7842074048970604e-05,
+      "loss": 3.26,
+      "step": 3370
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.7799711937642975e-05,
+      "loss": 3.1542,
+      "step": 3380
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.775734982631535e-05,
+      "loss": 2.2394,
+      "step": 3390
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.771498771498771e-05,
+      "loss": 3.841,
+      "step": 3400
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.767262560366009e-05,
+      "loss": 3.338,
+      "step": 3410
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.7630263492332456e-05,
+      "loss": 2.9295,
+      "step": 3420
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.758790138100483e-05,
+      "loss": 3.2942,
+      "step": 3430
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.7545539269677206e-05,
+      "loss": 2.1179,
+      "step": 3440
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.750317715834957e-05,
+      "loss": 2.3783,
+      "step": 3450
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.746081504702195e-05,
+      "loss": 3.2622,
+      "step": 3460
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.7418452935694315e-05,
+      "loss": 3.7284,
+      "step": 3470
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.7376090824366687e-05,
+      "loss": 2.1249,
+      "step": 3480
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.7333728713039065e-05,
+      "loss": 2.644,
+      "step": 3490
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.729136660171143e-05,
+      "loss": 2.9117,
+      "step": 3500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.72490044903838e-05,
+      "loss": 2.982,
+      "step": 3510
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.7206642379056174e-05,
+      "loss": 2.6437,
+      "step": 3520
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.7164280267728546e-05,
+      "loss": 2.6276,
+      "step": 3530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.712191815640092e-05,
+      "loss": 3.1733,
+      "step": 3540
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.707955604507329e-05,
+      "loss": 3.3072,
+      "step": 3550
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.703719393374566e-05,
+      "loss": 4.058,
+      "step": 3560
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.699483182241803e-05,
+      "loss": 2.3998,
+      "step": 3570
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6952469711090404e-05,
+      "loss": 3.4927,
+      "step": 3580
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.691010759976277e-05,
+      "loss": 2.4766,
+      "step": 3590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.686774548843515e-05,
+      "loss": 4.301,
+      "step": 3600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.682538337710751e-05,
+      "loss": 3.0968,
+      "step": 3610
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6783021265779885e-05,
+      "loss": 3.337,
+      "step": 3620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.674065915445226e-05,
+      "loss": 2.2142,
+      "step": 3630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.669829704312463e-05,
+      "loss": 2.8435,
+      "step": 3640
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.6655934931797e-05,
+      "loss": 3.8649,
+      "step": 3650
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.661357282046937e-05,
+      "loss": 3.7533,
+      "step": 3660
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.6571210709141744e-05,
+      "loss": 3.4148,
+      "step": 3670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.652884859781412e-05,
+      "loss": 3.059,
+      "step": 3680
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.648648648648649e-05,
+      "loss": 2.9798,
+      "step": 3690
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.644412437515886e-05,
+      "loss": 3.2666,
+      "step": 3700
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.640176226383123e-05,
+      "loss": 3.118,
+      "step": 3710
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.63594001525036e-05,
+      "loss": 3.0364,
+      "step": 3720
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.6317038041175974e-05,
+      "loss": 2.304,
+      "step": 3730
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.6274675929848346e-05,
+      "loss": 3.1448,
+      "step": 3740
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.623231381852072e-05,
+      "loss": 3.4461,
+      "step": 3750
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.618995170719309e-05,
+      "loss": 3.7228,
+      "step": 3760
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.614758959586546e-05,
+      "loss": 2.9702,
+      "step": 3770
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.610522748453783e-05,
+      "loss": 3.3596,
+      "step": 3780
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.6062865373210205e-05,
+      "loss": 3.9703,
+      "step": 3790
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.602050326188258e-05,
+      "loss": 3.0907,
+      "step": 3800
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.597814115055494e-05,
+      "loss": 3.6026,
+      "step": 3810
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.593577903922732e-05,
+      "loss": 3.3953,
+      "step": 3820
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5893416927899686e-05,
+      "loss": 3.9044,
+      "step": 3830
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.585105481657206e-05,
+      "loss": 2.7496,
+      "step": 3840
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.580869270524443e-05,
+      "loss": 3.2217,
+      "step": 3850
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.57663305939168e-05,
+      "loss": 2.1367,
+      "step": 3860
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.572396848258918e-05,
+      "loss": 3.8704,
+      "step": 3870
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.5681606371261544e-05,
+      "loss": 2.8283,
+      "step": 3880
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.5639244259933916e-05,
+      "loss": 2.2891,
+      "step": 3890
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.559688214860629e-05,
+      "loss": 2.4959,
+      "step": 3900
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.555452003727866e-05,
+      "loss": 2.7823,
+      "step": 3910
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.551215792595103e-05,
+      "loss": 3.633,
+      "step": 3920
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.5469795814623403e-05,
+      "loss": 1.952,
+      "step": 3930
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.5427433703295775e-05,
+      "loss": 3.1351,
+      "step": 3940
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.538507159196815e-05,
+      "loss": 2.3038,
+      "step": 3950
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.534270948064052e-05,
+      "loss": 3.1765,
+      "step": 3960
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.5300347369312884e-05,
+      "loss": 3.4554,
+      "step": 3970
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.525798525798526e-05,
+      "loss": 2.1874,
+      "step": 3980
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.5215623146657634e-05,
+      "loss": 3.5274,
+      "step": 3990
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.517326103533e-05,
+      "loss": 3.0357,
+      "step": 4000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.513089892400238e-05,
+      "loss": 2.3487,
+      "step": 4010
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.508853681267474e-05,
+      "loss": 2.6817,
+      "step": 4020
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.5046174701347115e-05,
+      "loss": 3.598,
+      "step": 4030
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.5003812590019486e-05,
+      "loss": 2.0949,
+      "step": 4040
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.496145047869186e-05,
+      "loss": 3.5464,
+      "step": 4050
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.491908836736423e-05,
+      "loss": 2.5463,
+      "step": 4060
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.48767262560366e-05,
+      "loss": 3.8892,
+      "step": 4070
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.4834364144708973e-05,
+      "loss": 3.7547,
+      "step": 4080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4792002033381345e-05,
+      "loss": 2.6305,
+      "step": 4090
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.474963992205372e-05,
+      "loss": 2.4977,
+      "step": 4100
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.470727781072609e-05,
+      "loss": 0.9359,
+      "step": 4110
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.466491569939846e-05,
+      "loss": 2.0838,
+      "step": 4120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.462255358807083e-05,
+      "loss": 2.9107,
+      "step": 4130
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4580191476743204e-05,
+      "loss": 2.2076,
+      "step": 4140
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4537829365415576e-05,
+      "loss": 3.1737,
+      "step": 4150
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.449546725408794e-05,
+      "loss": 1.1921,
+      "step": 4160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.445310514276032e-05,
+      "loss": 2.1966,
+      "step": 4170
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.441074303143269e-05,
+      "loss": 2.0372,
+      "step": 4180
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4368380920105056e-05,
+      "loss": 2.5099,
+      "step": 4190
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4326018808777435e-05,
+      "loss": 2.0307,
+      "step": 4200
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.42836566974498e-05,
+      "loss": 2.4915,
+      "step": 4210
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.424129458612217e-05,
+      "loss": 2.2316,
+      "step": 4220
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.4198932474794543e-05,
+      "loss": 1.9918,
+      "step": 4230
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.4156570363466915e-05,
+      "loss": 1.8922,
+      "step": 4240
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.411420825213929e-05,
+      "loss": 2.1347,
+      "step": 4250
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.407184614081166e-05,
+      "loss": 2.0886,
+      "step": 4260
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.402948402948403e-05,
+      "loss": 1.5643,
+      "step": 4270
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.39871219181564e-05,
+      "loss": 1.8818,
+      "step": 4280
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.3944759806828774e-05,
+      "loss": 2.9763,
+      "step": 4290
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.3902397695501146e-05,
+      "loss": 1.2643,
+      "step": 4300
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.386003558417352e-05,
+      "loss": 1.3101,
+      "step": 4310
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.381767347284589e-05,
+      "loss": 1.8863,
+      "step": 4320
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.377531136151826e-05,
+      "loss": 1.551,
+      "step": 4330
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.373294925019063e-05,
+      "loss": 1.7621,
+      "step": 4340
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.3690587138863e-05,
+      "loss": 2.3386,
+      "step": 4350
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.364822502753538e-05,
+      "loss": 2.0052,
+      "step": 4360
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.360586291620775e-05,
+      "loss": 2.0421,
+      "step": 4370
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.3563500804880114e-05,
+      "loss": 1.9337,
+      "step": 4380
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.352113869355249e-05,
+      "loss": 2.0312,
+      "step": 4390
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.347877658222486e-05,
+      "loss": 1.9996,
+      "step": 4400
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.343641447089723e-05,
+      "loss": 1.6679,
+      "step": 4410
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.339405235956961e-05,
+      "loss": 2.9902,
+      "step": 4420
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.335169024824197e-05,
+      "loss": 1.8793,
+      "step": 4430
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.3309328136914344e-05,
+      "loss": 2.0604,
+      "step": 4440
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.3266966025586716e-05,
+      "loss": 1.666,
+      "step": 4450
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.322460391425909e-05,
+      "loss": 1.8598,
+      "step": 4460
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.318224180293146e-05,
+      "loss": 1.7999,
+      "step": 4470
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.313987969160383e-05,
+      "loss": 2.5453,
+      "step": 4480
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.30975175802762e-05,
+      "loss": 2.2554,
+      "step": 4490
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.3055155468948575e-05,
+      "loss": 2.1116,
+      "step": 4500
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.301279335762095e-05,
+      "loss": 2.1907,
+      "step": 4510
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.297043124629331e-05,
+      "loss": 1.411,
+      "step": 4520
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.292806913496569e-05,
+      "loss": 0.5621,
+      "step": 4530
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.2885707023638055e-05,
+      "loss": 2.702,
+      "step": 4540
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.2843344912310434e-05,
+      "loss": 2.6155,
+      "step": 4550
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.2800982800982806e-05,
+      "loss": 1.7449,
+      "step": 4560
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.275862068965517e-05,
+      "loss": 1.8469,
+      "step": 4570
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.271625857832755e-05,
+      "loss": 2.1547,
+      "step": 4580
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.2673896466999914e-05,
+      "loss": 1.7365,
+      "step": 4590
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.2631534355672286e-05,
+      "loss": 0.9322,
+      "step": 4600
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.2589172244344665e-05,
+      "loss": 1.9555,
+      "step": 4610
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.254681013301703e-05,
+      "loss": 2.4951,
+      "step": 4620
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.25044480216894e-05,
+      "loss": 2.8419,
+      "step": 4630
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.246208591036177e-05,
+      "loss": 2.5815,
+      "step": 4640
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.2419723799034145e-05,
+      "loss": 1.892,
+      "step": 4650
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.237736168770652e-05,
+      "loss": 1.4939,
+      "step": 4660
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.233499957637889e-05,
+      "loss": 1.4957,
+      "step": 4670
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.229263746505126e-05,
+      "loss": 1.8644,
+      "step": 4680
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.225027535372363e-05,
+      "loss": 1.4489,
+      "step": 4690
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.2207913242396004e-05,
+      "loss": 2.3615,
+      "step": 4700
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.216555113106837e-05,
+      "loss": 2.6989,
+      "step": 4710
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.212318901974075e-05,
+      "loss": 1.2804,
+      "step": 4720
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.208082690841311e-05,
+      "loss": 1.5119,
+      "step": 4730
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.203846479708549e-05,
+      "loss": 1.8845,
+      "step": 4740
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.199610268575786e-05,
+      "loss": 2.6107,
+      "step": 4750
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.195374057443023e-05,
+      "loss": 1.7699,
+      "step": 4760
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.1911378463102606e-05,
+      "loss": 1.7155,
+      "step": 4770
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.186901635177497e-05,
+      "loss": 1.9242,
+      "step": 4780
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.182665424044734e-05,
+      "loss": 1.7136,
+      "step": 4790
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.178429212911972e-05,
+      "loss": 2.3578,
+      "step": 4800
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.174193001779209e-05,
+      "loss": 2.2565,
+      "step": 4810
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.169956790646446e-05,
+      "loss": 1.1242,
+      "step": 4820
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.165720579513683e-05,
+      "loss": 2.6882,
+      "step": 4830
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.16148436838092e-05,
+      "loss": 2.0054,
+      "step": 4840
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.1572481572481574e-05,
+      "loss": 1.2765,
+      "step": 4850
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.1530119461153946e-05,
+      "loss": 1.9453,
+      "step": 4860
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.148775734982632e-05,
+      "loss": 2.2615,
+      "step": 4870
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.144539523849869e-05,
+      "loss": 1.0669,
+      "step": 4880
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.140303312717106e-05,
+      "loss": 1.6441,
+      "step": 4890
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.1360671015843426e-05,
+      "loss": 2.5318,
+      "step": 4900
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.1318308904515805e-05,
+      "loss": 1.0544,
+      "step": 4910
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.1275946793188176e-05,
+      "loss": 1.8715,
+      "step": 4920
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.123358468186054e-05,
+      "loss": 1.4182,
+      "step": 4930
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.119122257053292e-05,
+      "loss": 1.7556,
+      "step": 4940
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.1148860459205285e-05,
+      "loss": 1.8498,
+      "step": 4950
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.1106498347877664e-05,
+      "loss": 2.068,
+      "step": 4960
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.106413623655003e-05,
+      "loss": 1.5508,
+      "step": 4970
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.10217741252224e-05,
+      "loss": 1.8953,
+      "step": 4980
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.097941201389478e-05,
+      "loss": 2.3344,
+      "step": 4990
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.0937049902567144e-05,
+      "loss": 1.3442,
+      "step": 5000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.0894687791239516e-05,
+      "loss": 1.9394,
+      "step": 5010
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.085232567991189e-05,
+      "loss": 1.1371,
+      "step": 5020
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.080996356858426e-05,
+      "loss": 1.8254,
+      "step": 5030
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.076760145725663e-05,
+      "loss": 2.1171,
+      "step": 5040
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.0725239345929e-05,
+      "loss": 1.8488,
+      "step": 5050
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.0682877234601375e-05,
+      "loss": 1.2931,
+      "step": 5060
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.0640515123273747e-05,
+      "loss": 1.9931,
+      "step": 5070
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.059815301194612e-05,
+      "loss": 2.8016,
+      "step": 5080
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.055579090061848e-05,
+      "loss": 1.5954,
+      "step": 5090
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.051342878929086e-05,
+      "loss": 1.1403,
+      "step": 5100
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.0471066677963234e-05,
+      "loss": 2.2751,
+      "step": 5110
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.0428704566635602e-05,
+      "loss": 2.1418,
+      "step": 5120
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.0386342455307977e-05,
+      "loss": 2.1881,
+      "step": 5130
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.0343980343980342e-05,
+      "loss": 2.0472,
+      "step": 5140
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.0301618232652717e-05,
+      "loss": 2.7806,
+      "step": 5150
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.0259256121325086e-05,
+      "loss": 1.6729,
+      "step": 5160
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.021689400999746e-05,
+      "loss": 1.2469,
+      "step": 5170
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.0174531898669833e-05,
+      "loss": 3.1278,
+      "step": 5180
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.01321697873422e-05,
+      "loss": 1.8955,
+      "step": 5190
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.0089807676014576e-05,
+      "loss": 2.0903,
+      "step": 5200
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.0047445564686945e-05,
+      "loss": 1.8173,
+      "step": 5210
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.0005083453359317e-05,
+      "loss": 2.2454,
+      "step": 5220
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.9962721342031692e-05,
+      "loss": 2.2454,
+      "step": 5230
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.992035923070406e-05,
+      "loss": 3.5885,
+      "step": 5240
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.9877997119376432e-05,
+      "loss": 1.803,
+      "step": 5250
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.98356350080488e-05,
+      "loss": 2.0046,
+      "step": 5260
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.9793272896721175e-05,
+      "loss": 2.3274,
+      "step": 5270
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.9750910785393544e-05,
+      "loss": 1.1254,
+      "step": 5280
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.9708548674065916e-05,
+      "loss": 2.5921,
+      "step": 5290
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.966618656273829e-05,
+      "loss": 1.8062,
+      "step": 5300
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.962382445141066e-05,
+      "loss": 2.5566,
+      "step": 5310
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.958146234008303e-05,
+      "loss": 3.2912,
+      "step": 5320
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.95391002287554e-05,
+      "loss": 1.7776,
+      "step": 5330
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.9496738117427775e-05,
+      "loss": 1.8023,
+      "step": 5340
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.9454376006100143e-05,
+      "loss": 2.234,
+      "step": 5350
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.9412013894772518e-05,
+      "loss": 2.0186,
+      "step": 5360
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.936965178344489e-05,
+      "loss": 1.4665,
+      "step": 5370
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.932728967211726e-05,
+      "loss": 1.738,
+      "step": 5380
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.9284927560789634e-05,
+      "loss": 1.7414,
+      "step": 5390
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.9242565449462002e-05,
+      "loss": 2.3558,
+      "step": 5400
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.9200203338134374e-05,
+      "loss": 2.0124,
+      "step": 5410
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.915784122680675e-05,
+      "loss": 1.7629,
+      "step": 5420
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.9115479115479117e-05,
+      "loss": 2.7381,
+      "step": 5430
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.907311700415149e-05,
+      "loss": 2.6545,
+      "step": 5440
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.9030754892823857e-05,
+      "loss": 1.9099,
+      "step": 5450
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.8988392781496233e-05,
+      "loss": 2.024,
+      "step": 5460
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.89460306701686e-05,
+      "loss": 1.9093,
+      "step": 5470
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.8903668558840973e-05,
+      "loss": 2.265,
+      "step": 5480
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.8861306447513348e-05,
+      "loss": 1.6859,
+      "step": 5490
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.8818944336185716e-05,
+      "loss": 2.4686,
+      "step": 5500
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.8776582224858088e-05,
+      "loss": 2.0404,
+      "step": 5510
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.8734220113530457e-05,
+      "loss": 2.0324,
+      "step": 5520
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.8691858002202832e-05,
+      "loss": 2.8294,
+      "step": 5530
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.8649495890875207e-05,
+      "loss": 2.0527,
+      "step": 5540
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.8607133779547572e-05,
+      "loss": 0.9728,
+      "step": 5550
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.8564771668219947e-05,
+      "loss": 1.856,
+      "step": 5560
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.8522409556892316e-05,
+      "loss": 2.2293,
+      "step": 5570
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.848004744556469e-05,
+      "loss": 2.2294,
+      "step": 5580
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.843768533423706e-05,
+      "loss": 1.9026,
+      "step": 5590
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.839532322290943e-05,
+      "loss": 0.8172,
+      "step": 5600
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.8352961111581806e-05,
+      "loss": 1.7732,
+      "step": 5610
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.8310599000254174e-05,
+      "loss": 1.5588,
+      "step": 5620
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.8268236888926546e-05,
+      "loss": 3.0688,
+      "step": 5630
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8225874777598915e-05,
+      "loss": 2.2315,
+      "step": 5640
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.818351266627129e-05,
+      "loss": 1.4582,
+      "step": 5650
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8141150554943658e-05,
+      "loss": 2.1769,
+      "step": 5660
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.809878844361603e-05,
+      "loss": 2.0149,
+      "step": 5670
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.8056426332288405e-05,
+      "loss": 2.3662,
+      "step": 5680
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.8014064220960774e-05,
+      "loss": 2.29,
+      "step": 5690
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.7971702109633145e-05,
+      "loss": 1.7657,
+      "step": 5700
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.7929339998305514e-05,
+      "loss": 2.3735,
+      "step": 5710
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.788697788697789e-05,
+      "loss": 2.6198,
+      "step": 5720
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.784461577565026e-05,
+      "loss": 1.4926,
+      "step": 5730
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.780225366432263e-05,
+      "loss": 1.4625,
+      "step": 5740
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.7759891552995004e-05,
+      "loss": 1.7667,
+      "step": 5750
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.7717529441667373e-05,
+      "loss": 1.4901,
+      "step": 5760
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.7675167330339748e-05,
+      "loss": 1.5037,
+      "step": 5770
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.7632805219012113e-05,
+      "loss": 2.4814,
+      "step": 5780
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.7590443107684488e-05,
+      "loss": 2.2755,
+      "step": 5790
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.7548080996356863e-05,
+      "loss": 2.6432,
+      "step": 5800
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.750571888502923e-05,
+      "loss": 2.5837,
+      "step": 5810
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7463356773701603e-05,
+      "loss": 2.214,
+      "step": 5820
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7420994662373972e-05,
+      "loss": 1.9176,
+      "step": 5830
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7378632551046347e-05,
+      "loss": 2.1772,
+      "step": 5840
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.7336270439718715e-05,
+      "loss": 1.3768,
+      "step": 5850
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.7293908328391087e-05,
+      "loss": 2.3045,
+      "step": 5860
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.7251546217063462e-05,
+      "loss": 1.8353,
+      "step": 5870
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.720918410573583e-05,
+      "loss": 2.2037,
+      "step": 5880
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.7166821994408203e-05,
+      "loss": 1.4456,
+      "step": 5890
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.712445988308057e-05,
+      "loss": 1.8822,
+      "step": 5900
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.7082097771752946e-05,
+      "loss": 2.1746,
+      "step": 5910
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.7039735660425318e-05,
+      "loss": 2.4628,
+      "step": 5920
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.6997373549097686e-05,
+      "loss": 2.813,
+      "step": 5930
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.695501143777006e-05,
+      "loss": 1.7021,
+      "step": 5940
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.691264932644243e-05,
+      "loss": 1.6366,
+      "step": 5950
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.6870287215114802e-05,
+      "loss": 1.9514,
+      "step": 5960
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.682792510378717e-05,
+      "loss": 3.4098,
+      "step": 5970
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.6785562992459545e-05,
+      "loss": 2.4604,
+      "step": 5980
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.674320088113192e-05,
+      "loss": 2.2009,
+      "step": 5990
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.670083876980429e-05,
+      "loss": 3.0638,
+      "step": 6000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.665847665847666e-05,
+      "loss": 1.3695,
+      "step": 6010
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.661611454714903e-05,
+      "loss": 1.8349,
+      "step": 6020
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.6573752435821404e-05,
+      "loss": 2.045,
+      "step": 6030
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.6531390324493776e-05,
+      "loss": 2.5156,
+      "step": 6040
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.6489028213166144e-05,
+      "loss": 2.0776,
+      "step": 6050
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.644666610183852e-05,
+      "loss": 2.0897,
+      "step": 6060
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.6404303990510888e-05,
+      "loss": 1.8716,
+      "step": 6070
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.636194187918326e-05,
+      "loss": 2.086,
+      "step": 6080
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.6319579767855628e-05,
+      "loss": 2.2685,
+      "step": 6090
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.6277217656528003e-05,
+      "loss": 2.7301,
+      "step": 6100
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.6234855545200375e-05,
+      "loss": 2.5793,
+      "step": 6110
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.6192493433872744e-05,
+      "loss": 1.4742,
+      "step": 6120
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.615013132254512e-05,
+      "loss": 2.116,
+      "step": 6130
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.6107769211217487e-05,
+      "loss": 1.151,
+      "step": 6140
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.606540709988986e-05,
+      "loss": 2.3006,
+      "step": 6150
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.6023044988562227e-05,
+      "loss": 1.462,
+      "step": 6160
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.5980682877234602e-05,
+      "loss": 0.769,
+      "step": 6170
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5938320765906978e-05,
+      "loss": 1.0338,
+      "step": 6180
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5895958654579343e-05,
+      "loss": 2.3097,
+      "step": 6190
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5853596543251718e-05,
+      "loss": 2.0747,
+      "step": 6200
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5811234431924086e-05,
+      "loss": 2.0228,
+      "step": 6210
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.576887232059646e-05,
+      "loss": 2.0715,
+      "step": 6220
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.5726510209268833e-05,
+      "loss": 2.3442,
+      "step": 6230
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.56841480979412e-05,
+      "loss": 1.7032,
+      "step": 6240
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.5641785986613577e-05,
+      "loss": 2.2626,
+      "step": 6250
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5599423875285945e-05,
+      "loss": 1.9353,
+      "step": 6260
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5557061763958317e-05,
+      "loss": 2.025,
+      "step": 6270
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5514699652630685e-05,
+      "loss": 2.027,
+      "step": 6280
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.547233754130306e-05,
+      "loss": 2.2029,
+      "step": 6290
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5429975429975432e-05,
+      "loss": 1.8088,
+      "step": 6300
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.53876133186478e-05,
+      "loss": 2.4682,
+      "step": 6310
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5345251207320176e-05,
+      "loss": 1.7378,
+      "step": 6320
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5302889095992544e-05,
+      "loss": 2.603,
+      "step": 6330
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5260526984664916e-05,
+      "loss": 3.0907,
+      "step": 6340
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.521816487333729e-05,
+      "loss": 1.8687,
+      "step": 6350
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.517580276200966e-05,
+      "loss": 3.1415,
+      "step": 6360
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.513344065068203e-05,
+      "loss": 2.0988,
+      "step": 6370
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.50910785393544e-05,
+      "loss": 1.9755,
+      "step": 6380
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.5048716428026775e-05,
+      "loss": 3.0538,
+      "step": 6390
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.5006354316699143e-05,
+      "loss": 1.8245,
+      "step": 6400
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.496399220537152e-05,
+      "loss": 2.2791,
+      "step": 6410
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4921630094043887e-05,
+      "loss": 2.2188,
+      "step": 6420
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.487926798271626e-05,
+      "loss": 2.1376,
+      "step": 6430
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.483690587138863e-05,
+      "loss": 1.9605,
+      "step": 6440
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4794543760061002e-05,
+      "loss": 3.2437,
+      "step": 6450
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4752181648733374e-05,
+      "loss": 2.3829,
+      "step": 6460
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4709819537405746e-05,
+      "loss": 1.8132,
+      "step": 6470
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4667457426078118e-05,
+      "loss": 2.1358,
+      "step": 6480
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4625095314750486e-05,
+      "loss": 2.1904,
+      "step": 6490
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4582733203422858e-05,
+      "loss": 1.8758,
+      "step": 6500
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.4540371092095233e-05,
+      "loss": 2.6824,
+      "step": 6510
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.4498008980767605e-05,
+      "loss": 2.533,
+      "step": 6520
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.4455646869439973e-05,
+      "loss": 0.8647,
+      "step": 6530
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.4413284758112345e-05,
+      "loss": 0.7595,
+      "step": 6540
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.4370922646784717e-05,
+      "loss": 1.8398,
+      "step": 6550
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.432856053545709e-05,
+      "loss": 1.8411,
+      "step": 6560
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.428619842412946e-05,
+      "loss": 1.9194,
+      "step": 6570
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.4243836312801832e-05,
+      "loss": 1.7175,
+      "step": 6580
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.4201474201474204e-05,
+      "loss": 2.0469,
+      "step": 6590
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.4159112090146572e-05,
+      "loss": 1.5491,
+      "step": 6600
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.4116749978818944e-05,
+      "loss": 2.6109,
+      "step": 6610
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.4074387867491316e-05,
+      "loss": 2.8283,
+      "step": 6620
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.403202575616369e-05,
+      "loss": 1.565,
+      "step": 6630
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.398966364483606e-05,
+      "loss": 1.4641,
+      "step": 6640
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.394730153350843e-05,
+      "loss": 1.4535,
+      "step": 6650
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.3904939422180803e-05,
+      "loss": 2.6998,
+      "step": 6660
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.3862577310853175e-05,
+      "loss": 2.4404,
+      "step": 6670
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.3820215199525543e-05,
+      "loss": 2.6695,
+      "step": 6680
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.3777853088197915e-05,
+      "loss": 2.2211,
+      "step": 6690
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.373549097687029e-05,
+      "loss": 1.2419,
+      "step": 6700
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.369312886554266e-05,
+      "loss": 2.2412,
+      "step": 6710
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.365076675421503e-05,
+      "loss": 1.7556,
+      "step": 6720
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.3608404642887402e-05,
+      "loss": 2.45,
+      "step": 6730
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.3566042531559774e-05,
+      "loss": 1.8402,
+      "step": 6740
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.3523680420232146e-05,
+      "loss": 2.3175,
+      "step": 6750
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.3481318308904518e-05,
+      "loss": 1.9969,
+      "step": 6760
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.343895619757689e-05,
+      "loss": 2.8838,
+      "step": 6770
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.339659408624926e-05,
+      "loss": 1.2146,
+      "step": 6780
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.335423197492163e-05,
+      "loss": 2.8859,
+      "step": 6790
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.3311869863594e-05,
+      "loss": 2.4621,
+      "step": 6800
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.3269507752266373e-05,
+      "loss": 2.5522,
+      "step": 6810
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.3227145640938748e-05,
+      "loss": 1.5923,
+      "step": 6820
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.3184783529611117e-05,
+      "loss": 2.0304,
+      "step": 6830
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.314242141828349e-05,
+      "loss": 1.5219,
+      "step": 6840
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.310005930695586e-05,
+      "loss": 2.1149,
+      "step": 6850
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.3057697195628232e-05,
+      "loss": 2.4139,
+      "step": 6860
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.30153350843006e-05,
+      "loss": 2.8719,
+      "step": 6870
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2972972972972976e-05,
+      "loss": 1.6354,
+      "step": 6880
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2930610861645347e-05,
+      "loss": 1.9854,
+      "step": 6890
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2888248750317716e-05,
+      "loss": 2.0286,
+      "step": 6900
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2845886638990088e-05,
+      "loss": 2.4089,
+      "step": 6910
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.280352452766246e-05,
+      "loss": 2.0499,
+      "step": 6920
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.276116241633483e-05,
+      "loss": 1.5199,
+      "step": 6930
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.2718800305007203e-05,
+      "loss": 1.7327,
+      "step": 6940
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.2676438193679575e-05,
+      "loss": 1.8151,
+      "step": 6950
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.2634076082351947e-05,
+      "loss": 2.7421,
+      "step": 6960
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.259171397102432e-05,
+      "loss": 2.3864,
+      "step": 6970
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.2549351859696687e-05,
+      "loss": 1.5677,
+      "step": 6980
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.250698974836906e-05,
+      "loss": 2.0365,
+      "step": 6990
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.246462763704143e-05,
+      "loss": 2.4206,
+      "step": 7000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.2422265525713802e-05,
+      "loss": 2.1518,
+      "step": 7010
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.2379903414386174e-05,
+      "loss": 2.3625,
+      "step": 7020
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.2337541303058546e-05,
+      "loss": 2.6026,
+      "step": 7030
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.2295179191730917e-05,
+      "loss": 1.8608,
+      "step": 7040
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.225281708040329e-05,
+      "loss": 2.3889,
+      "step": 7050
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.2210454969075658e-05,
+      "loss": 2.3806,
+      "step": 7060
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.2168092857748033e-05,
+      "loss": 0.5251,
+      "step": 7070
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.2125730746420405e-05,
+      "loss": 1.5179,
+      "step": 7080
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.2083368635092773e-05,
+      "loss": 2.7728,
+      "step": 7090
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.2041006523765145e-05,
+      "loss": 1.5799,
+      "step": 7100
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1998644412437517e-05,
+      "loss": 2.7693,
+      "step": 7110
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.195628230110989e-05,
+      "loss": 3.0033,
+      "step": 7120
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.191392018978226e-05,
+      "loss": 1.2471,
+      "step": 7130
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1871558078454632e-05,
+      "loss": 2.0973,
+      "step": 7140
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1829195967127004e-05,
+      "loss": 1.7758,
+      "step": 7150
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1786833855799376e-05,
+      "loss": 2.1995,
+      "step": 7160
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1744471744471744e-05,
+      "loss": 2.8555,
+      "step": 7170
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1702109633144116e-05,
+      "loss": 2.4562,
+      "step": 7180
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.165974752181649e-05,
+      "loss": 1.9584,
+      "step": 7190
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.161738541048886e-05,
+      "loss": 1.5362,
+      "step": 7200
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.157502329916123e-05,
+      "loss": 2.0313,
+      "step": 7210
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.1532661187833603e-05,
+      "loss": 1.6663,
+      "step": 7220
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.1490299076505975e-05,
+      "loss": 2.133,
+      "step": 7230
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.1447936965178343e-05,
+      "loss": 2.1854,
+      "step": 7240
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.1405574853850715e-05,
+      "loss": 1.3735,
+      "step": 7250
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.136321274252309e-05,
+      "loss": 1.503,
+      "step": 7260
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.1320850631195462e-05,
+      "loss": 2.5417,
+      "step": 7270
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.127848851986783e-05,
+      "loss": 1.3044,
+      "step": 7280
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.1236126408540202e-05,
+      "loss": 2.8509,
+      "step": 7290
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.1193764297212574e-05,
+      "loss": 0.8797,
+      "step": 7300
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.1151402185884946e-05,
+      "loss": 2.4059,
+      "step": 7310
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.1109040074557317e-05,
+      "loss": 1.7712,
+      "step": 7320
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.106667796322969e-05,
+      "loss": 2.2861,
+      "step": 7330
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.102431585190206e-05,
+      "loss": 2.2852,
+      "step": 7340
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.098195374057443e-05,
+      "loss": 2.1867,
+      "step": 7350
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.09395916292468e-05,
+      "loss": 1.9676,
+      "step": 7360
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0897229517919173e-05,
+      "loss": 1.8575,
+      "step": 7370
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0854867406591548e-05,
+      "loss": 2.0654,
+      "step": 7380
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0812505295263916e-05,
+      "loss": 1.2704,
+      "step": 7390
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0770143183936288e-05,
+      "loss": 2.0859,
+      "step": 7400
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.072778107260866e-05,
+      "loss": 1.9481,
+      "step": 7410
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.0685418961281032e-05,
+      "loss": 1.4198,
+      "step": 7420
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.06430568499534e-05,
+      "loss": 1.1006,
+      "step": 7430
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.0600694738625775e-05,
+      "loss": 1.6572,
+      "step": 7440
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.0558332627298147e-05,
+      "loss": 2.4729,
+      "step": 7450
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.051597051597052e-05,
+      "loss": 1.7319,
+      "step": 7460
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.0473608404642887e-05,
+      "loss": 1.0956,
+      "step": 7470
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.043124629331526e-05,
+      "loss": 2.0044,
+      "step": 7480
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.038888418198763e-05,
+      "loss": 2.6305,
+      "step": 7490
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.0346522070660003e-05,
+      "loss": 1.9486,
+      "step": 7500
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.0304159959332375e-05,
+      "loss": 1.6576,
+      "step": 7510
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.0261797848004746e-05,
+      "loss": 1.2474,
+      "step": 7520
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0219435736677118e-05,
+      "loss": 2.1858,
+      "step": 7530
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0177073625349486e-05,
+      "loss": 2.4697,
+      "step": 7540
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0134711514021858e-05,
+      "loss": 1.484,
+      "step": 7550
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.009234940269423e-05,
+      "loss": 0.8322,
+      "step": 7560
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.0049987291366605e-05,
+      "loss": 2.0463,
+      "step": 7570
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.0007625180038974e-05,
+      "loss": 1.7798,
+      "step": 7580
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.9965263068711345e-05,
+      "loss": 2.1085,
+      "step": 7590
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.9922900957383717e-05,
+      "loss": 1.3203,
+      "step": 7600
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.988053884605609e-05,
+      "loss": 1.8849,
+      "step": 7610
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9838176734728457e-05,
+      "loss": 1.7881,
+      "step": 7620
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9795814623400833e-05,
+      "loss": 1.6501,
+      "step": 7630
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9753452512073204e-05,
+      "loss": 1.9652,
+      "step": 7640
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.9711090400745573e-05,
+      "loss": 2.5812,
+      "step": 7650
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.9668728289417945e-05,
+      "loss": 3.0978,
+      "step": 7660
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.9626366178090316e-05,
+      "loss": 2.7041,
+      "step": 7670
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.9584004066762688e-05,
+      "loss": 2.3167,
+      "step": 7680
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.954164195543506e-05,
+      "loss": 2.0777,
+      "step": 7690
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.949927984410743e-05,
+      "loss": 2.0212,
+      "step": 7700
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.9456917732779803e-05,
+      "loss": 1.4816,
+      "step": 7710
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.9414555621452175e-05,
+      "loss": 2.3649,
+      "step": 7720
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.9372193510124544e-05,
+      "loss": 1.6498,
+      "step": 7730
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.9329831398796915e-05,
+      "loss": 1.2647,
+      "step": 7740
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.928746928746929e-05,
+      "loss": 2.8664,
+      "step": 7750
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.924510717614166e-05,
+      "loss": 2.6277,
+      "step": 7760
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.920274506481403e-05,
+      "loss": 1.3932,
+      "step": 7770
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.9160382953486403e-05,
+      "loss": 0.8097,
+      "step": 7780
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.9118020842158774e-05,
+      "loss": 2.7805,
+      "step": 7790
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.9075658730831146e-05,
+      "loss": 1.7193,
+      "step": 7800
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.9033296619503515e-05,
+      "loss": 1.9944,
+      "step": 7810
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.899093450817589e-05,
+      "loss": 1.8126,
+      "step": 7820
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.894857239684826e-05,
+      "loss": 1.883,
+      "step": 7830
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.890621028552063e-05,
+      "loss": 1.8189,
+      "step": 7840
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.8863848174193002e-05,
+      "loss": 2.6088,
+      "step": 7850
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8821486062865374e-05,
+      "loss": 3.0362,
+      "step": 7860
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8779123951537745e-05,
+      "loss": 1.724,
+      "step": 7870
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8736761840210117e-05,
+      "loss": 2.0149,
+      "step": 7880
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.869439972888249e-05,
+      "loss": 1.7307,
+      "step": 7890
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.865203761755486e-05,
+      "loss": 2.1296,
+      "step": 7900
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8609675506227232e-05,
+      "loss": 2.8909,
+      "step": 7910
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.85673133948996e-05,
+      "loss": 2.4823,
+      "step": 7920
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8524951283571973e-05,
+      "loss": 1.9871,
+      "step": 7930
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8482589172244348e-05,
+      "loss": 2.1547,
+      "step": 7940
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8440227060916716e-05,
+      "loss": 2.1964,
+      "step": 7950
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8397864949589088e-05,
+      "loss": 2.6151,
+      "step": 7960
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.835550283826146e-05,
+      "loss": 1.2912,
+      "step": 7970
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.831314072693383e-05,
+      "loss": 1.1387,
+      "step": 7980
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.82707786156062e-05,
+      "loss": 3.1345,
+      "step": 7990
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.8228416504278575e-05,
+      "loss": 1.571,
+      "step": 8000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.8186054392950947e-05,
+      "loss": 1.6557,
+      "step": 8010
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.814369228162332e-05,
+      "loss": 2.221,
+      "step": 8020
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.8101330170295687e-05,
+      "loss": 1.6428,
+      "step": 8030
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.805896805896806e-05,
+      "loss": 1.3434,
+      "step": 8040
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.801660594764043e-05,
+      "loss": 2.0555,
+      "step": 8050
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7974243836312802e-05,
+      "loss": 1.3324,
+      "step": 8060
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7931881724985174e-05,
+      "loss": 1.9456,
+      "step": 8070
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7889519613657546e-05,
+      "loss": 1.8268,
+      "step": 8080
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7847157502329918e-05,
+      "loss": 1.2563,
+      "step": 8090
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.780479539100229e-05,
+      "loss": 2.0412,
+      "step": 8100
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7762433279674658e-05,
+      "loss": 1.6659,
+      "step": 8110
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.772007116834703e-05,
+      "loss": 2.7998,
+      "step": 8120
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7677709057019405e-05,
+      "loss": 1.5926,
+      "step": 8130
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7635346945691773e-05,
+      "loss": 2.2987,
+      "step": 8140
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7592984834364145e-05,
+      "loss": 2.2893,
+      "step": 8150
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7550622723036517e-05,
+      "loss": 2.4805,
+      "step": 8160
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.750826061170889e-05,
+      "loss": 1.856,
+      "step": 8170
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7465898500381257e-05,
+      "loss": 2.046,
+      "step": 8180
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.7423536389053632e-05,
+      "loss": 2.7798,
+      "step": 8190
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.7381174277726004e-05,
+      "loss": 2.484,
+      "step": 8200
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.7338812166398376e-05,
+      "loss": 1.4067,
+      "step": 8210
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.7296450055070744e-05,
+      "loss": 0.7482,
+      "step": 8220
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.7254087943743116e-05,
+      "loss": 0.4303,
+      "step": 8230
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.7211725832415488e-05,
+      "loss": 0.8174,
+      "step": 8240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.716936372108786e-05,
+      "loss": 0.2746,
+      "step": 8250
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.712700160976023e-05,
+      "loss": 0.566,
+      "step": 8260
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7084639498432603e-05,
+      "loss": 0.8012,
+      "step": 8270
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7042277387104975e-05,
+      "loss": 0.7246,
+      "step": 8280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.6999915275777343e-05,
+      "loss": 0.6704,
+      "step": 8290
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.6957553164449715e-05,
+      "loss": 0.3289,
+      "step": 8300
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.691519105312209e-05,
+      "loss": 0.5398,
+      "step": 8310
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.6872828941794462e-05,
+      "loss": 0.594,
+      "step": 8320
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.683046683046683e-05,
+      "loss": 0.389,
+      "step": 8330
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.6788104719139202e-05,
+      "loss": 0.5755,
+      "step": 8340
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.6745742607811574e-05,
+      "loss": 0.23,
+      "step": 8350
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.6703380496483946e-05,
+      "loss": 0.4805,
+      "step": 8360
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.6661018385156314e-05,
+      "loss": 0.5814,
+      "step": 8370
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.661865627382869e-05,
+      "loss": 0.5806,
+      "step": 8380
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.657629416250106e-05,
+      "loss": 0.5084,
+      "step": 8390
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.653393205117343e-05,
+      "loss": 0.599,
+      "step": 8400
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.64915699398458e-05,
+      "loss": 0.7653,
+      "step": 8410
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.6449207828518173e-05,
+      "loss": 0.4046,
+      "step": 8420
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.6406845717190545e-05,
+      "loss": 0.9935,
+      "step": 8430
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.6364483605862917e-05,
+      "loss": 0.4776,
+      "step": 8440
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.632212149453529e-05,
+      "loss": 0.6331,
+      "step": 8450
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.627975938320766e-05,
+      "loss": 0.7177,
+      "step": 8460
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.6237397271880032e-05,
+      "loss": 0.4219,
+      "step": 8470
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.61950351605524e-05,
+      "loss": 0.7412,
+      "step": 8480
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.6152673049224772e-05,
+      "loss": 0.8571,
+      "step": 8490
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.6110310937897148e-05,
+      "loss": 0.4699,
+      "step": 8500
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.606794882656952e-05,
+      "loss": 0.6861,
+      "step": 8510
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.6025586715241888e-05,
+      "loss": 0.6181,
+      "step": 8520
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.598322460391426e-05,
+      "loss": 0.715,
+      "step": 8530
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.594086249258663e-05,
+      "loss": 0.5667,
+      "step": 8540
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.5898500381259003e-05,
+      "loss": 0.9669,
+      "step": 8550
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.5856138269931375e-05,
+      "loss": 0.5971,
+      "step": 8560
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.5813776158603747e-05,
+      "loss": 0.3442,
+      "step": 8570
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.577141404727612e-05,
+      "loss": 0.347,
+      "step": 8580
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.5729051935948487e-05,
+      "loss": 0.493,
+      "step": 8590
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.568668982462086e-05,
+      "loss": 0.6026,
+      "step": 8600
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.564432771329323e-05,
+      "loss": 0.6687,
+      "step": 8610
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.5601965601965606e-05,
+      "loss": 0.4297,
+      "step": 8620
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.5559603490637974e-05,
+      "loss": 0.3964,
+      "step": 8630
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.5517241379310346e-05,
+      "loss": 0.2003,
+      "step": 8640
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.5474879267982718e-05,
+      "loss": 0.6641,
+      "step": 8650
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.543251715665509e-05,
+      "loss": 0.4963,
+      "step": 8660
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.5390155045327458e-05,
+      "loss": 0.5852,
+      "step": 8670
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.534779293399983e-05,
+      "loss": 0.3761,
+      "step": 8680
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.5305430822672205e-05,
+      "loss": 0.6719,
+      "step": 8690
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.5263068711344573e-05,
+      "loss": 0.6055,
+      "step": 8700
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.5220706600016945e-05,
+      "loss": 0.7134,
+      "step": 8710
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.5178344488689317e-05,
+      "loss": 0.3932,
+      "step": 8720
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.5135982377361689e-05,
+      "loss": 0.6178,
+      "step": 8730
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.5093620266034059e-05,
+      "loss": 0.3793,
+      "step": 8740
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.5051258154706432e-05,
+      "loss": 0.4163,
+      "step": 8750
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.5008896043378804e-05,
+      "loss": 0.618,
+      "step": 8760
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.4966533932051174e-05,
+      "loss": 0.5041,
+      "step": 8770
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.4924171820723546e-05,
+      "loss": 0.847,
+      "step": 8780
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.4881809709395916e-05,
+      "loss": 0.3593,
+      "step": 8790
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.4839447598068288e-05,
+      "loss": 0.4644,
+      "step": 8800
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.4797085486740661e-05,
+      "loss": 0.9418,
+      "step": 8810
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.4754723375413033e-05,
+      "loss": 0.3844,
+      "step": 8820
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.4712361264085403e-05,
+      "loss": 0.6866,
+      "step": 8830
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4669999152757775e-05,
+      "loss": 0.8426,
+      "step": 8840
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4627637041430145e-05,
+      "loss": 0.4384,
+      "step": 8850
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4585274930102517e-05,
+      "loss": 0.555,
+      "step": 8860
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.454291281877489e-05,
+      "loss": 0.777,
+      "step": 8870
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.450055070744726e-05,
+      "loss": 0.812,
+      "step": 8880
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4458188596119632e-05,
+      "loss": 0.5145,
+      "step": 8890
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4415826484792002e-05,
+      "loss": 0.42,
+      "step": 8900
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4373464373464374e-05,
+      "loss": 0.5528,
+      "step": 8910
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4331102262136744e-05,
+      "loss": 0.6014,
+      "step": 8920
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4288740150809116e-05,
+      "loss": 0.3353,
+      "step": 8930
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.424637803948149e-05,
+      "loss": 0.4894,
+      "step": 8940
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4204015928153861e-05,
+      "loss": 0.6702,
+      "step": 8950
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4161653816826231e-05,
+      "loss": 0.4726,
+      "step": 8960
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.4119291705498603e-05,
+      "loss": 0.7141,
+      "step": 8970
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.4076929594170973e-05,
+      "loss": 0.6713,
+      "step": 8980
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.4034567482843345e-05,
+      "loss": 0.7595,
+      "step": 8990
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3992205371515718e-05,
+      "loss": 0.695,
+      "step": 9000
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3949843260188088e-05,
+      "loss": 0.4454,
+      "step": 9010
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.390748114886046e-05,
+      "loss": 0.4785,
+      "step": 9020
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.386511903753283e-05,
+      "loss": 0.6736,
+      "step": 9030
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3822756926205202e-05,
+      "loss": 0.3806,
+      "step": 9040
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.3780394814877574e-05,
+      "loss": 0.9536,
+      "step": 9050
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.3738032703549947e-05,
+      "loss": 0.5646,
+      "step": 9060
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.3695670592222317e-05,
+      "loss": 0.3187,
+      "step": 9070
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.365330848089469e-05,
+      "loss": 0.382,
+      "step": 9080
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.361094636956706e-05,
+      "loss": 0.8692,
+      "step": 9090
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.3568584258239431e-05,
+      "loss": 0.5565,
+      "step": 9100
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.3526222146911801e-05,
+      "loss": 0.6482,
+      "step": 9110
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.3483860035584175e-05,
+      "loss": 0.2117,
+      "step": 9120
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.3441497924256546e-05,
+      "loss": 0.3319,
+      "step": 9130
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.3399135812928918e-05,
+      "loss": 0.7229,
+      "step": 9140
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.3356773701601288e-05,
+      "loss": 0.3715,
+      "step": 9150
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.331441159027366e-05,
+      "loss": 0.5898,
+      "step": 9160
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.327204947894603e-05,
+      "loss": 0.5303,
+      "step": 9170
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.3229687367618404e-05,
+      "loss": 0.6634,
+      "step": 9180
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.3187325256290775e-05,
+      "loss": 0.5,
+      "step": 9190
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.3144963144963146e-05,
+      "loss": 0.5164,
+      "step": 9200
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.3102601033635517e-05,
+      "loss": 0.4212,
+      "step": 9210
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.3060238922307887e-05,
+      "loss": 0.6097,
+      "step": 9220
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.301787681098026e-05,
+      "loss": 0.5178,
+      "step": 9230
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.297551469965263e-05,
+      "loss": 0.8441,
+      "step": 9240
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2933152588325004e-05,
+      "loss": 0.4995,
+      "step": 9250
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2890790476997375e-05,
+      "loss": 0.6676,
+      "step": 9260
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2848428365669746e-05,
+      "loss": 0.6302,
+      "step": 9270
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2806066254342116e-05,
+      "loss": 0.4195,
+      "step": 9280
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2763704143014488e-05,
+      "loss": 1.0051,
+      "step": 9290
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2721342031686858e-05,
+      "loss": 0.7695,
+      "step": 9300
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2678979920359232e-05,
+      "loss": 0.9529,
+      "step": 9310
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2636617809031604e-05,
+      "loss": 0.479,
+      "step": 9320
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2594255697703974e-05,
+      "loss": 0.4358,
+      "step": 9330
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2551893586376345e-05,
+      "loss": 0.469,
+      "step": 9340
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2509531475048716e-05,
+      "loss": 0.7724,
+      "step": 9350
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2467169363721089e-05,
+      "loss": 0.193,
+      "step": 9360
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.242480725239346e-05,
+      "loss": 0.7209,
+      "step": 9370
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.2382445141065831e-05,
+      "loss": 0.4119,
+      "step": 9380
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.2340083029738203e-05,
+      "loss": 0.5697,
+      "step": 9390
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.2297720918410575e-05,
+      "loss": 0.6327,
+      "step": 9400
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.2255358807082945e-05,
+      "loss": 0.7893,
+      "step": 9410
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.2212996695755318e-05,
+      "loss": 0.5615,
+      "step": 9420
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.2170634584427688e-05,
+      "loss": 0.4562,
+      "step": 9430
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.212827247310006e-05,
+      "loss": 0.5147,
+      "step": 9440
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.2085910361772432e-05,
+      "loss": 0.6618,
+      "step": 9450
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.2043548250444804e-05,
+      "loss": 0.4668,
+      "step": 9460
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.2001186139117174e-05,
+      "loss": 0.788,
+      "step": 9470
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.1958824027789545e-05,
+      "loss": 0.6433,
+      "step": 9480
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.1916461916461917e-05,
+      "loss": 0.5399,
+      "step": 9490
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1874099805134287e-05,
+      "loss": 0.4714,
+      "step": 9500
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.183173769380666e-05,
+      "loss": 0.4613,
+      "step": 9510
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1789375582479031e-05,
+      "loss": 0.3284,
+      "step": 9520
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1747013471151403e-05,
+      "loss": 0.4049,
+      "step": 9530
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.1704651359823774e-05,
+      "loss": 0.5543,
+      "step": 9540
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.1662289248496146e-05,
+      "loss": 0.4532,
+      "step": 9550
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.1619927137168516e-05,
+      "loss": 0.6403,
+      "step": 9560
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.157756502584089e-05,
+      "loss": 0.4485,
+      "step": 9570
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.153520291451326e-05,
+      "loss": 0.8185,
+      "step": 9580
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1492840803185632e-05,
+      "loss": 0.5941,
+      "step": 9590
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1450478691858003e-05,
+      "loss": 0.6509,
+      "step": 9600
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1408116580530374e-05,
+      "loss": 0.6686,
+      "step": 9610
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1365754469202745e-05,
+      "loss": 0.7639,
+      "step": 9620
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1323392357875117e-05,
+      "loss": 0.44,
+      "step": 9630
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1281030246547489e-05,
+      "loss": 0.7185,
+      "step": 9640
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1238668135219859e-05,
+      "loss": 0.6479,
+      "step": 9650
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.119630602389223e-05,
+      "loss": 0.2792,
+      "step": 9660
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1153943912564603e-05,
+      "loss": 0.5244,
+      "step": 9670
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1111581801236974e-05,
+      "loss": 0.3198,
+      "step": 9680
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1069219689909344e-05,
+      "loss": 0.7125,
+      "step": 9690
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.1026857578581718e-05,
+      "loss": 0.4181,
+      "step": 9700
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0984495467254088e-05,
+      "loss": 0.2214,
+      "step": 9710
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.094213335592646e-05,
+      "loss": 0.3217,
+      "step": 9720
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0899771244598832e-05,
+      "loss": 0.6054,
+      "step": 9730
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0857409133271203e-05,
+      "loss": 0.5259,
+      "step": 9740
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0815047021943574e-05,
+      "loss": 0.4686,
+      "step": 9750
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0772684910615945e-05,
+      "loss": 0.5126,
+      "step": 9760
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0730322799288317e-05,
+      "loss": 0.6182,
+      "step": 9770
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0687960687960689e-05,
+      "loss": 0.7905,
+      "step": 9780
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.064559857663306e-05,
+      "loss": 0.4709,
+      "step": 9790
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.060323646530543e-05,
+      "loss": 0.5607,
+      "step": 9800
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.0560874353977803e-05,
+      "loss": 0.366,
+      "step": 9810
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.0518512242650174e-05,
+      "loss": 0.6703,
+      "step": 9820
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.0476150131322546e-05,
+      "loss": 0.505,
+      "step": 9830
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.0433788019994916e-05,
+      "loss": 0.6329,
+      "step": 9840
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.039142590866729e-05,
+      "loss": 0.7426,
+      "step": 9850
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.034906379733966e-05,
+      "loss": 0.6777,
+      "step": 9860
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.0306701686012032e-05,
+      "loss": 0.4268,
+      "step": 9870
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.0264339574684403e-05,
+      "loss": 0.5794,
+      "step": 9880
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.0221977463356775e-05,
+      "loss": 0.6381,
+      "step": 9890
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.0179615352029145e-05,
+      "loss": 0.5742,
+      "step": 9900
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.0137253240701517e-05,
+      "loss": 0.7118,
+      "step": 9910
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.0094891129373889e-05,
+      "loss": 0.6782,
+      "step": 9920
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.0052529018046259e-05,
+      "loss": 0.503,
+      "step": 9930
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.001016690671863e-05,
+      "loss": 0.5714,
+      "step": 9940
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 9.967804795391002e-06,
+      "loss": 0.938,
+      "step": 9950
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 9.925442684063374e-06,
+      "loss": 0.403,
+      "step": 9960
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 9.883080572735744e-06,
+      "loss": 0.4648,
+      "step": 9970
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 9.840718461408118e-06,
+      "loss": 0.4648,
+      "step": 9980
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.798356350080488e-06,
+      "loss": 0.5522,
+      "step": 9990
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.75599423875286e-06,
+      "loss": 0.3334,
+      "step": 10000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.713632127425232e-06,
+      "loss": 0.3977,
+      "step": 10010
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.671270016097603e-06,
+      "loss": 0.9446,
+      "step": 10020
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 9.628907904769973e-06,
+      "loss": 0.8682,
+      "step": 10030
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 9.586545793442347e-06,
+      "loss": 0.9647,
+      "step": 10040
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 9.544183682114717e-06,
+      "loss": 0.1914,
+      "step": 10050
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 9.501821570787089e-06,
+      "loss": 0.4032,
+      "step": 10060
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 0.5889,
+      "step": 10070
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.41709734813183e-06,
+      "loss": 0.4782,
+      "step": 10080
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.374735236804202e-06,
+      "loss": 0.9566,
+      "step": 10090
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.332373125476574e-06,
+      "loss": 1.6344,
+      "step": 10100
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 9.290011014148946e-06,
+      "loss": 0.4026,
+      "step": 10110
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 9.247648902821316e-06,
+      "loss": 1.4795,
+      "step": 10120
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 9.20528679149369e-06,
+      "loss": 0.811,
+      "step": 10130
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 9.16292468016606e-06,
+      "loss": 0.9765,
+      "step": 10140
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 9.120562568838431e-06,
+      "loss": 1.0907,
+      "step": 10150
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 9.078200457510803e-06,
+      "loss": 1.6697,
+      "step": 10160
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 9.035838346183175e-06,
+      "loss": 1.2014,
+      "step": 10170
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 8.993476234855545e-06,
+      "loss": 0.5545,
+      "step": 10180
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 8.951114123527917e-06,
+      "loss": 0.4729,
+      "step": 10190
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.908752012200289e-06,
+      "loss": 0.9808,
+      "step": 10200
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.86638990087266e-06,
+      "loss": 0.2746,
+      "step": 10210
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.82402778954503e-06,
+      "loss": 0.4114,
+      "step": 10220
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.781665678217402e-06,
+      "loss": 0.2869,
+      "step": 10230
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.739303566889774e-06,
+      "loss": 0.2721,
+      "step": 10240
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.696941455562144e-06,
+      "loss": 0.2916,
+      "step": 10250
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.654579344234518e-06,
+      "loss": 1.1184,
+      "step": 10260
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.612217232906888e-06,
+      "loss": 1.1453,
+      "step": 10270
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.56985512157926e-06,
+      "loss": 2.0171,
+      "step": 10280
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.527493010251631e-06,
+      "loss": 0.64,
+      "step": 10290
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.485130898924003e-06,
+      "loss": 0.4109,
+      "step": 10300
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.442768787596373e-06,
+      "loss": 0.1617,
+      "step": 10310
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.400406676268747e-06,
+      "loss": 1.0423,
+      "step": 10320
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.358044564941117e-06,
+      "loss": 1.6148,
+      "step": 10330
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.315682453613489e-06,
+      "loss": 0.3669,
+      "step": 10340
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.27332034228586e-06,
+      "loss": 0.825,
+      "step": 10350
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 8.230958230958232e-06,
+      "loss": 0.3359,
+      "step": 10360
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 8.188596119630602e-06,
+      "loss": 0.1691,
+      "step": 10370
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 8.146234008302974e-06,
+      "loss": 1.1247,
+      "step": 10380
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 8.103871896975346e-06,
+      "loss": 0.4599,
+      "step": 10390
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 8.061509785647716e-06,
+      "loss": 0.3186,
+      "step": 10400
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 8.01914767432009e-06,
+      "loss": 0.3601,
+      "step": 10410
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 7.97678556299246e-06,
+      "loss": 1.071,
+      "step": 10420
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 7.934423451664831e-06,
+      "loss": 0.9437,
+      "step": 10430
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.892061340337203e-06,
+      "loss": 0.7318,
+      "step": 10440
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.849699229009575e-06,
+      "loss": 1.8935,
+      "step": 10450
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.807337117681945e-06,
+      "loss": 0.2494,
+      "step": 10460
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.764975006354317e-06,
+      "loss": 0.4978,
+      "step": 10470
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.722612895026689e-06,
+      "loss": 2.9353,
+      "step": 10480
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.68025078369906e-06,
+      "loss": 0.2079,
+      "step": 10490
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.63788867237143e-06,
+      "loss": 0.8816,
+      "step": 10500
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.595526561043803e-06,
+      "loss": 0.335,
+      "step": 10510
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.553164449716174e-06,
+      "loss": 1.3264,
+      "step": 10520
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.510802338388545e-06,
+      "loss": 0.3081,
+      "step": 10530
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.468440227060918e-06,
+      "loss": 1.8423,
+      "step": 10540
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.4260781157332886e-06,
+      "loss": 1.4617,
+      "step": 10550
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.3837160044056595e-06,
+      "loss": 0.2261,
+      "step": 10560
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.341353893078032e-06,
+      "loss": 1.7769,
+      "step": 10570
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.298991781750403e-06,
+      "loss": 0.5638,
+      "step": 10580
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.256629670422774e-06,
+      "loss": 1.8723,
+      "step": 10590
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.214267559095146e-06,
+      "loss": 0.1413,
+      "step": 10600
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 7.171905447767517e-06,
+      "loss": 1.026,
+      "step": 10610
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 7.129543336439888e-06,
+      "loss": 0.9169,
+      "step": 10620
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 7.08718122511226e-06,
+      "loss": 0.3298,
+      "step": 10630
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 7.044819113784631e-06,
+      "loss": 0.5623,
+      "step": 10640
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 7.002457002457002e-06,
+      "loss": 0.1858,
+      "step": 10650
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 6.960094891129375e-06,
+      "loss": 0.8963,
+      "step": 10660
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 6.917732779801746e-06,
+      "loss": 0.3163,
+      "step": 10670
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 6.875370668474117e-06,
+      "loss": 0.2722,
+      "step": 10680
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 6.833008557146489e-06,
+      "loss": 1.2305,
+      "step": 10690
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 6.79064644581886e-06,
+      "loss": 0.4875,
+      "step": 10700
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 6.748284334491231e-06,
+      "loss": 0.1613,
+      "step": 10710
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 6.705922223163604e-06,
+      "loss": 0.5625,
+      "step": 10720
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 6.663560111835975e-06,
+      "loss": 0.1873,
+      "step": 10730
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 6.621198000508346e-06,
+      "loss": 0.4334,
+      "step": 10740
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 6.578835889180717e-06,
+      "loss": 0.9462,
+      "step": 10750
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 6.5364737778530885e-06,
+      "loss": 1.4734,
+      "step": 10760
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 6.494111666525459e-06,
+      "loss": 1.0356,
+      "step": 10770
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 6.45174955519783e-06,
+      "loss": 0.153,
+      "step": 10780
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 6.409387443870203e-06,
+      "loss": 0.8046,
+      "step": 10790
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 6.367025332542574e-06,
+      "loss": 0.9482,
+      "step": 10800
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.324663221214945e-06,
+      "loss": 1.0965,
+      "step": 10810
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.2823011098873175e-06,
+      "loss": 1.57,
+      "step": 10820
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.2399389985596884e-06,
+      "loss": 0.2467,
+      "step": 10830
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.19757688723206e-06,
+      "loss": 0.4729,
+      "step": 10840
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.155214775904431e-06,
+      "loss": 0.2895,
+      "step": 10850
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.112852664576803e-06,
+      "loss": 1.0783,
+      "step": 10860
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.070490553249175e-06,
+      "loss": 0.501,
+      "step": 10870
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.028128441921546e-06,
+      "loss": 0.5623,
+      "step": 10880
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.9857663305939175e-06,
+      "loss": 1.205,
+      "step": 10890
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.943404219266288e-06,
+      "loss": 0.8954,
+      "step": 10900
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.90104210793866e-06,
+      "loss": 0.2346,
+      "step": 10910
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.858679996611031e-06,
+      "loss": 0.9149,
+      "step": 10920
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.816317885283403e-06,
+      "loss": 0.3965,
+      "step": 10930
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.773955773955774e-06,
+      "loss": 0.3429,
+      "step": 10940
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.731593662628146e-06,
+      "loss": 0.6026,
+      "step": 10950
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.689231551300517e-06,
+      "loss": 0.9188,
+      "step": 10960
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.646869439972888e-06,
+      "loss": 0.8711,
+      "step": 10970
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.60450732864526e-06,
+      "loss": 1.1473,
+      "step": 10980
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.562145217317631e-06,
+      "loss": 1.3842,
+      "step": 10990
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.519783105990003e-06,
+      "loss": 0.8322,
+      "step": 11000
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.477420994662375e-06,
+      "loss": 0.7533,
+      "step": 11010
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.435058883334746e-06,
+      "loss": 0.5009,
+      "step": 11020
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.392696772007117e-06,
+      "loss": 0.3193,
+      "step": 11030
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.350334660679489e-06,
+      "loss": 1.1132,
+      "step": 11040
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.30797254935186e-06,
+      "loss": 0.3861,
+      "step": 11050
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.265610438024231e-06,
+      "loss": 2.5128,
+      "step": 11060
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.223248326696603e-06,
+      "loss": 1.2422,
+      "step": 11070
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.180886215368974e-06,
+      "loss": 0.3001,
+      "step": 11080
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.138524104041346e-06,
+      "loss": 0.6416,
+      "step": 11090
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.0961619927137165e-06,
+      "loss": 0.2177,
+      "step": 11100
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.053799881386088e-06,
+      "loss": 1.2584,
+      "step": 11110
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.01143777005846e-06,
+      "loss": 0.4628,
+      "step": 11120
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.969075658730831e-06,
+      "loss": 0.5163,
+      "step": 11130
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.926713547403203e-06,
+      "loss": 0.45,
+      "step": 11140
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.884351436075575e-06,
+      "loss": 0.3469,
+      "step": 11150
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.8419893247479455e-06,
+      "loss": 0.3414,
+      "step": 11160
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.799627213420317e-06,
+      "loss": 0.5382,
+      "step": 11170
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.757265102092689e-06,
+      "loss": 0.348,
+      "step": 11180
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.71490299076506e-06,
+      "loss": 0.4124,
+      "step": 11190
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.672540879437432e-06,
+      "loss": 1.0529,
+      "step": 11200
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.630178768109803e-06,
+      "loss": 0.3426,
+      "step": 11210
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.587816656782174e-06,
+      "loss": 0.8135,
+      "step": 11220
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.9044,
+      "step": 11230
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.5030924341269165e-06,
+      "loss": 0.3513,
+      "step": 11240
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.460730322799288e-06,
+      "loss": 0.4668,
+      "step": 11250
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.41836821147166e-06,
+      "loss": 1.15,
+      "step": 11260
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.376006100144031e-06,
+      "loss": 0.3603,
+      "step": 11270
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.333643988816403e-06,
+      "loss": 1.0308,
+      "step": 11280
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.2912818774887745e-06,
+      "loss": 0.3113,
+      "step": 11290
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.2489197661611455e-06,
+      "loss": 0.4548,
+      "step": 11300
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.206557654833517e-06,
+      "loss": 1.1289,
+      "step": 11310
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.164195543505889e-06,
+      "loss": 0.5599,
+      "step": 11320
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.12183343217826e-06,
+      "loss": 1.1617,
+      "step": 11330
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.079471320850632e-06,
+      "loss": 1.0273,
+      "step": 11340
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.037109209523003e-06,
+      "loss": 0.4952,
+      "step": 11350
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.9947470981953745e-06,
+      "loss": 0.9921,
+      "step": 11360
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.9523849868677455e-06,
+      "loss": 0.4659,
+      "step": 11370
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.910022875540116e-06,
+      "loss": 0.5762,
+      "step": 11380
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.867660764212488e-06,
+      "loss": 0.4774,
+      "step": 11390
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.82529865288486e-06,
+      "loss": 0.5106,
+      "step": 11400
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.782936541557231e-06,
+      "loss": 0.329,
+      "step": 11410
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.7405744302296027e-06,
+      "loss": 0.1065,
+      "step": 11420
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.6982123189019745e-06,
+      "loss": 1.1783,
+      "step": 11430
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.6558502075743454e-06,
+      "loss": 0.8846,
+      "step": 11440
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.6134880962467172e-06,
+      "loss": 0.2988,
+      "step": 11450
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.571125984919089e-06,
+      "loss": 0.3167,
+      "step": 11460
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.52876387359146e-06,
+      "loss": 0.3847,
+      "step": 11470
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.4864017622638313e-06,
+      "loss": 0.3489,
+      "step": 11480
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.444039650936203e-06,
+      "loss": 0.4234,
+      "step": 11490
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.401677539608574e-06,
+      "loss": 0.8811,
+      "step": 11500
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.359315428280946e-06,
+      "loss": 1.2944,
+      "step": 11510
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.3169533169533168e-06,
+      "loss": 0.2166,
+      "step": 11520
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.2745912056256885e-06,
+      "loss": 1.534,
+      "step": 11530
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.2322290942980603e-06,
+      "loss": 1.0961,
+      "step": 11540
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.1898669829704313e-06,
+      "loss": 0.4982,
+      "step": 11550
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.1475048716428026e-06,
+      "loss": 1.2357,
+      "step": 11560
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.105142760315174e-06,
+      "loss": 0.2432,
+      "step": 11570
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.062780648987546e-06,
+      "loss": 1.4383,
+      "step": 11580
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 3.020418537659917e-06,
+      "loss": 0.3697,
+      "step": 11590
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.9780564263322885e-06,
+      "loss": 0.5363,
+      "step": 11600
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.93569431500466e-06,
+      "loss": 1.0902,
+      "step": 11610
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.8933322036770317e-06,
+      "loss": 0.0907,
+      "step": 11620
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.8509700923494026e-06,
+      "loss": 0.3343,
+      "step": 11630
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.808607981021774e-06,
+      "loss": 2.0042,
+      "step": 11640
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.7662458696941458e-06,
+      "loss": 0.9941,
+      "step": 11650
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.723883758366517e-06,
+      "loss": 0.4094,
+      "step": 11660
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.6815216470388885e-06,
+      "loss": 1.0563,
+      "step": 11670
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.63915953571126e-06,
+      "loss": 0.3991,
+      "step": 11680
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.5967974243836316e-06,
+      "loss": 0.3047,
+      "step": 11690
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.554435313056003e-06,
+      "loss": 0.1611,
+      "step": 11700
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.512073201728374e-06,
+      "loss": 1.9046,
+      "step": 11710
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.4697110904007457e-06,
+      "loss": 0.1787,
+      "step": 11720
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.427348979073117e-06,
+      "loss": 1.0945,
+      "step": 11730
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.3849868677454885e-06,
+      "loss": 0.3679,
+      "step": 11740
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.34262475641786e-06,
+      "loss": 0.2251,
+      "step": 11750
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.3002626450902316e-06,
+      "loss": 1.8121,
+      "step": 11760
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.257900533762603e-06,
+      "loss": 0.1915,
+      "step": 11770
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.2155384224349743e-06,
+      "loss": 0.4249,
+      "step": 11780
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.1731763111073457e-06,
+      "loss": 1.4603,
+      "step": 11790
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.130814199779717e-06,
+      "loss": 1.1997,
+      "step": 11800
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.0884520884520884e-06,
+      "loss": 0.1811,
+      "step": 11810
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.0460899771244598e-06,
+      "loss": 1.2228,
+      "step": 11820
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.0037278657968316e-06,
+      "loss": 0.2999,
+      "step": 11830
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.961365754469203e-06,
+      "loss": 0.5039,
+      "step": 11840
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.9190036431415743e-06,
+      "loss": 0.3262,
+      "step": 11850
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.8766415318139459e-06,
+      "loss": 0.8737,
+      "step": 11860
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.8342794204863172e-06,
+      "loss": 1.5924,
+      "step": 11870
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.7919173091586884e-06,
+      "loss": 0.9632,
+      "step": 11880
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.7495551978310598e-06,
+      "loss": 0.498,
+      "step": 11890
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.7071930865034315e-06,
+      "loss": 0.3852,
+      "step": 11900
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.664830975175803e-06,
+      "loss": 0.1534,
+      "step": 11910
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.622468863848174e-06,
+      "loss": 0.3076,
+      "step": 11920
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.5801067525205458e-06,
+      "loss": 0.3339,
+      "step": 11930
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.5377446411929172e-06,
+      "loss": 0.9119,
+      "step": 11940
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.4953825298652886e-06,
+      "loss": 0.3734,
+      "step": 11950
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.45302041853766e-06,
+      "loss": 1.0711,
+      "step": 11960
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.4106583072100315e-06,
+      "loss": 0.5645,
+      "step": 11970
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.3682961958824029e-06,
+      "loss": 1.631,
+      "step": 11980
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.3259340845547742e-06,
+      "loss": 1.018,
+      "step": 11990
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.2835719732271456e-06,
+      "loss": 0.2324,
+      "step": 12000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 12303,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 9534066130944000.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}