diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5901 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9954282041941533,
+  "eval_steps": 500,
+  "global_step": 4900,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5e-06,
+      "loss": 0.8423,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 0.7396,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.5e-06,
+      "loss": 0.6387,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.5493,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.25e-05,
+      "loss": 0.5063,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5e-05,
+      "loss": 0.4662,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.75e-05,
+      "loss": 0.4701,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 0.4564,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.25e-05,
+      "loss": 0.4671,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4337,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 0.449,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3e-05,
+      "loss": 0.4253,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 0.4148,
+      "step": 65
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.5e-05,
+      "loss": 0.4262,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.4137,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4e-05,
+      "loss": 0.4021,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.25e-05,
+      "loss": 0.3965,
+      "step": 85
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.5e-05,
+      "loss": 0.4039,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.75e-05,
+      "loss": 0.4039,
+      "step": 95
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 0.4024,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.999994156428523e-05,
+      "loss": 0.3966,
+      "step": 105
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.999976625741408e-05,
+      "loss": 0.392,
+      "step": 110
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999947408020611e-05,
+      "loss": 0.3804,
+      "step": 115
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999906503402718e-05,
+      "loss": 0.3792,
+      "step": 120
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999853912078954e-05,
+      "loss": 0.382,
+      "step": 125
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9997896342951755e-05,
+      "loss": 0.3823,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.999713670351871e-05,
+      "loss": 0.3775,
+      "step": 135
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.999626020604163e-05,
+      "loss": 0.3676,
+      "step": 140
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9995266854618e-05,
+      "loss": 0.3755,
+      "step": 145
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9994156653891596e-05,
+      "loss": 0.3669,
+      "step": 150
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.999292960905246e-05,
+      "loss": 0.3689,
+      "step": 155
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.999158572583684e-05,
+      "loss": 0.3668,
+      "step": 160
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9990125010527206e-05,
+      "loss": 0.3739,
+      "step": 165
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.998854746995219e-05,
+      "loss": 0.3581,
+      "step": 170
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9986853111486563e-05,
+      "loss": 0.3645,
+      "step": 175
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9985041943051215e-05,
+      "loss": 0.365,
+      "step": 180
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9983113973113105e-05,
+      "loss": 0.3472,
+      "step": 185
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9981069210685205e-05,
+      "loss": 0.3484,
+      "step": 190
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9978907665326493e-05,
+      "loss": 0.345,
+      "step": 195
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.997662934714189e-05,
+      "loss": 0.3623,
+      "step": 200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.997423426678221e-05,
+      "loss": 0.3575,
+      "step": 205
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.997172243544409e-05,
+      "loss": 0.3554,
+      "step": 210
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.996909386487001e-05,
+      "loss": 0.3391,
+      "step": 215
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9966348567348144e-05,
+      "loss": 0.3564,
+      "step": 220
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.996348655571237e-05,
+      "loss": 0.3454,
+      "step": 225
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9960507843342184e-05,
+      "loss": 0.3547,
+      "step": 230
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.995741244416265e-05,
+      "loss": 0.3267,
+      "step": 235
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.995420037264431e-05,
+      "loss": 0.3487,
+      "step": 240
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.995087164380313e-05,
+      "loss": 0.3563,
+      "step": 245
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9947426273200455e-05,
+      "loss": 0.3467,
+      "step": 250
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.99438642769429e-05,
+      "loss": 0.3565,
+      "step": 255
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.994018567168228e-05,
+      "loss": 0.3467,
+      "step": 260
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.993639047461556e-05,
+      "loss": 0.3405,
+      "step": 265
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.993247870348473e-05,
+      "loss": 0.3491,
+      "step": 270
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.992845037657677e-05,
+      "loss": 0.3378,
+      "step": 275
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9924305512723544e-05,
+      "loss": 0.3544,
+      "step": 280
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.992004413130168e-05,
+      "loss": 0.3504,
+      "step": 285
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.991566625223254e-05,
+      "loss": 0.3308,
+      "step": 290
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.991117189598208e-05,
+      "loss": 0.3428,
+      "step": 295
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.990656108356076e-05,
+      "loss": 0.3485,
+      "step": 300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.990183383652349e-05,
+      "loss": 0.3466,
+      "step": 305
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.989699017696947e-05,
+      "loss": 0.333,
+      "step": 310
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.989203012754211e-05,
+      "loss": 0.3316,
+      "step": 315
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.988695371142893e-05,
+      "loss": 0.3561,
+      "step": 320
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9881760952361465e-05,
+      "loss": 0.3365,
+      "step": 325
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.987645187461511e-05,
+      "loss": 0.349,
+      "step": 330
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.987102650300905e-05,
+      "loss": 0.3337,
+      "step": 335
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.986548486290612e-05,
+      "loss": 0.3315,
+      "step": 340
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9859826980212694e-05,
+      "loss": 0.3371,
+      "step": 345
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9854052881378564e-05,
+      "loss": 0.3111,
+      "step": 350
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.984816259339683e-05,
+      "loss": 0.3243,
+      "step": 355
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.984215614380374e-05,
+      "loss": 0.3456,
+      "step": 360
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9836033560678584e-05,
+      "loss": 0.3315,
+      "step": 365
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.982979487264356e-05,
+      "loss": 0.3224,
+      "step": 370
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.982344010886366e-05,
+      "loss": 0.3338,
+      "step": 375
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.981696929904648e-05,
+      "loss": 0.3405,
+      "step": 380
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9810382473442145e-05,
+      "loss": 0.3113,
+      "step": 385
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9803679662843114e-05,
+      "loss": 0.3325,
+      "step": 390
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.979686089858407e-05,
+      "loss": 0.3256,
+      "step": 395
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.978992621254178e-05,
+      "loss": 0.3121,
+      "step": 400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.978287563713488e-05,
+      "loss": 0.3242,
+      "step": 405
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9775709205323825e-05,
+      "loss": 0.333,
+      "step": 410
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.976842695061066e-05,
+      "loss": 0.3154,
+      "step": 415
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.976102890703888e-05,
+      "loss": 0.3456,
+      "step": 420
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.975351510919327e-05,
+      "loss": 0.3076,
+      "step": 425
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.974588559219978e-05,
+      "loss": 0.3131,
+      "step": 430
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9738140391725315e-05,
+      "loss": 0.3186,
+      "step": 435
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9730279543977566e-05,
+      "loss": 0.3203,
+      "step": 440
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.972230308570488e-05,
+      "loss": 0.3355,
+      "step": 445
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.971421105419606e-05,
+      "loss": 0.3231,
+      "step": 450
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9706003487280195e-05,
+      "loss": 0.3187,
+      "step": 455
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9697680423326497e-05,
+      "loss": 0.3148,
+      "step": 460
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.968924190124409e-05,
+      "loss": 0.3224,
+      "step": 465
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.968068796048186e-05,
+      "loss": 0.3185,
+      "step": 470
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.967201864102828e-05,
+      "loss": 0.314,
+      "step": 475
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.966323398341115e-05,
+      "loss": 0.3176,
+      "step": 480
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.965433402869751e-05,
+      "loss": 0.314,
+      "step": 485
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.964531881849337e-05,
+      "loss": 0.3104,
+      "step": 490
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.963618839494355e-05,
+      "loss": 0.3124,
+      "step": 495
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9626942800731476e-05,
+      "loss": 0.327,
+      "step": 500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9617582079078984e-05,
+      "loss": 0.3232,
+      "step": 505
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.96081062737461e-05,
+      "loss": 0.3223,
+      "step": 510
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.959851542903088e-05,
+      "loss": 0.3194,
+      "step": 515
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.958880958976914e-05,
+      "loss": 0.3015,
+      "step": 520
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.957898880133429e-05,
+      "loss": 0.3104,
+      "step": 525
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9569053109637117e-05,
+      "loss": 0.3066,
+      "step": 530
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9559002561125565e-05,
+      "loss": 0.3038,
+      "step": 535
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.95488372027845e-05,
+      "loss": 0.3149,
+      "step": 540
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.953855708213554e-05,
+      "loss": 0.3141,
+      "step": 545
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9528162247236775e-05,
+      "loss": 0.3024,
+      "step": 550
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.9517652746682566e-05,
+      "loss": 0.3123,
+      "step": 555
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.9507028629603326e-05,
+      "loss": 0.3036,
+      "step": 560
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.9496289945665284e-05,
+      "loss": 0.3153,
+      "step": 565
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.948543674507027e-05,
+      "loss": 0.2925,
+      "step": 570
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.947446907855543e-05,
+      "loss": 0.3018,
+      "step": 575
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9463386997393045e-05,
+      "loss": 0.3169,
+      "step": 580
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.945219055339027e-05,
+      "loss": 0.3105,
+      "step": 585
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.944087979888886e-05,
+      "loss": 0.3035,
+      "step": 590
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9429454786765e-05,
+      "loss": 0.3082,
+      "step": 595
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.941791557042897e-05,
+      "loss": 0.3051,
+      "step": 600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.9406262203824985e-05,
+      "loss": 0.3019,
+      "step": 605
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.939449474143084e-05,
+      "loss": 0.3017,
+      "step": 610
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.938261323825776e-05,
+      "loss": 0.3042,
+      "step": 615
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.9370617749850056e-05,
+      "loss": 0.3086,
+      "step": 620
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.935850833228495e-05,
+      "loss": 0.2993,
+      "step": 625
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.934628504217222e-05,
+      "loss": 0.3067,
+      "step": 630
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.9333947936654014e-05,
+      "loss": 0.2989,
+      "step": 635
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.932149707340452e-05,
+      "loss": 0.3148,
+      "step": 640
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.9308932510629766e-05,
+      "loss": 0.3096,
+      "step": 645
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.929625430706728e-05,
+      "loss": 0.311,
+      "step": 650
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.9283462521985856e-05,
+      "loss": 0.3012,
+      "step": 655
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.9270557215185255e-05,
+      "loss": 0.3161,
+      "step": 660
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.925753844699594e-05,
+      "loss": 0.285,
+      "step": 665
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.924440627827881e-05,
+      "loss": 0.2979,
+      "step": 670
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.923116077042487e-05,
+      "loss": 0.314,
+      "step": 675
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.921780198535497e-05,
+      "loss": 0.3089,
+      "step": 680
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.920432998551954e-05,
+      "loss": 0.2923,
+      "step": 685
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.919074483389823e-05,
+      "loss": 0.2936,
+      "step": 690
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.91770465939997e-05,
+      "loss": 0.3143,
+      "step": 695
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.916323532986127e-05,
+      "loss": 0.3053,
+      "step": 700
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.914931110604861e-05,
+      "loss": 0.2989,
+      "step": 705
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.91352739876555e-05,
+      "loss": 0.3132,
+      "step": 710
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.9121124040303446e-05,
+      "loss": 0.2958,
+      "step": 715
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.910686133014143e-05,
+      "loss": 0.3043,
+      "step": 720
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.90924859238456e-05,
+      "loss": 0.2997,
+      "step": 725
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.907799788861891e-05,
+      "loss": 0.3003,
+      "step": 730
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.906339729219087e-05,
+      "loss": 0.3002,
+      "step": 735
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.9048684202817174e-05,
+      "loss": 0.2945,
+      "step": 740
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.903385868927942e-05,
+      "loss": 0.3017,
+      "step": 745
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.901892082088476e-05,
+      "loss": 0.2843,
+      "step": 750
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.9003870667465606e-05,
+      "loss": 0.2878,
+      "step": 755
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.898870829937926e-05,
+      "loss": 0.3041,
+      "step": 760
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.897343378750764e-05,
+      "loss": 0.2983,
+      "step": 765
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.895804720325691e-05,
+      "loss": 0.2995,
+      "step": 770
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.894254861855714e-05,
+      "loss": 0.2874,
+      "step": 775
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.8926938105862e-05,
+      "loss": 0.3056,
+      "step": 780
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.891121573814842e-05,
+      "loss": 0.3153,
+      "step": 785
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.8895381588916225e-05,
+      "loss": 0.3026,
+      "step": 790
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.887943573218779e-05,
+      "loss": 0.2953,
+      "step": 795
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.886337824250772e-05,
+      "loss": 0.299,
+      "step": 800
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.884720919494249e-05,
+      "loss": 0.284,
+      "step": 805
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.883092866508009e-05,
+      "loss": 0.287,
+      "step": 810
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.881453672902966e-05,
+      "loss": 0.2969,
+      "step": 815
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.879803346342118e-05,
+      "loss": 0.2858,
+      "step": 820
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.878141894540503e-05,
+      "loss": 0.3071,
+      "step": 825
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.8764693252651746e-05,
+      "loss": 0.2957,
+      "step": 830
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.8747856463351526e-05,
+      "loss": 0.2901,
+      "step": 835
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.873090865621396e-05,
+      "loss": 0.2908,
+      "step": 840
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.871384991046763e-05,
+      "loss": 0.2958,
+      "step": 845
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.8696680305859744e-05,
+      "loss": 0.2905,
+      "step": 850
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.8679399922655734e-05,
+      "loss": 0.3062,
+      "step": 855
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.866200884163893e-05,
+      "loss": 0.2891,
+      "step": 860
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.8644507144110165e-05,
+      "loss": 0.2846,
+      "step": 865
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.862689491188736e-05,
+      "loss": 0.2921,
+      "step": 870
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.860917222730518e-05,
+      "loss": 0.2872,
+      "step": 875
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.859133917321467e-05,
+      "loss": 0.2969,
+      "step": 880
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.857339583298278e-05,
+      "loss": 0.288,
+      "step": 885
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.855534229049209e-05,
+      "loss": 0.2847,
+      "step": 890
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.8537178630140315e-05,
+      "loss": 0.2852,
+      "step": 895
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.851890493683998e-05,
+      "loss": 0.2859,
+      "step": 900
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.850052129601799e-05,
+      "loss": 0.2875,
+      "step": 905
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.8482027793615236e-05,
+      "loss": 0.2923,
+      "step": 910
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.846342451608622e-05,
+      "loss": 0.2778,
+      "step": 915
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.844471155039858e-05,
+      "loss": 0.2761,
+      "step": 920
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.842588898403277e-05,
+      "loss": 0.2865,
+      "step": 925
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.840695690498161e-05,
+      "loss": 0.2842,
+      "step": 930
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.838791540174984e-05,
+      "loss": 0.2807,
+      "step": 935
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.83687645633538e-05,
+      "loss": 0.2892,
+      "step": 940
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.8349504479320894e-05,
+      "loss": 0.2764,
+      "step": 945
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.8330135239689286e-05,
+      "loss": 0.2859,
+      "step": 950
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.8310656935007406e-05,
+      "loss": 0.2816,
+      "step": 955
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.8291069656333526e-05,
+      "loss": 0.2867,
+      "step": 960
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.8271373495235396e-05,
+      "loss": 0.276,
+      "step": 965
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.825156854378974e-05,
+      "loss": 0.2831,
+      "step": 970
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.82316548945819e-05,
+      "loss": 0.2703,
+      "step": 975
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.8211632640705315e-05,
+      "loss": 0.285,
+      "step": 980
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.819150187576118e-05,
+      "loss": 0.2683,
+      "step": 985
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.817126269385794e-05,
+      "loss": 0.2796,
+      "step": 990
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.815091518961089e-05,
+      "loss": 0.2905,
+      "step": 995
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.8130459458141694e-05,
+      "loss": 0.277,
+      "step": 1000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.810989559507798e-05,
+      "loss": 0.2857,
+      "step": 1005
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.8089223696552865e-05,
+      "loss": 0.2927,
+      "step": 1010
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.806844385920453e-05,
+      "loss": 0.2858,
+      "step": 1015
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.804755618017574e-05,
+      "loss": 0.289,
+      "step": 1020
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.802656075711342e-05,
+      "loss": 0.2788,
+      "step": 1025
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.800545768816817e-05,
+      "loss": 0.2823,
+      "step": 1030
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.7984247071993825e-05,
+      "loss": 0.2889,
+      "step": 1035
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.7962929007746984e-05,
+      "loss": 0.28,
+      "step": 1040
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.794150359508656e-05,
+      "loss": 0.2925,
+      "step": 1045
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.791997093417328e-05,
+      "loss": 0.278,
+      "step": 1050
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.7898331125669264e-05,
+      "loss": 0.2715,
+      "step": 1055
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.7876584270737545e-05,
+      "loss": 0.2685,
+      "step": 1060
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.7854730471041544e-05,
+      "loss": 0.2916,
+      "step": 1065
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.7832769828744655e-05,
+      "loss": 0.2863,
+      "step": 1070
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.781070244650976e-05,
+      "loss": 0.2823,
+      "step": 1075
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.77885284274987e-05,
+      "loss": 0.2773,
+      "step": 1080
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.7766247875371856e-05,
+      "loss": 0.272,
+      "step": 1085
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.774386089428763e-05,
+      "loss": 0.272,
+      "step": 1090
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.772136758890196e-05,
+      "loss": 0.2635,
+      "step": 1095
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.7698768064367827e-05,
+      "loss": 0.2879,
+      "step": 1100
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.767606242633479e-05,
+      "loss": 0.2689,
+      "step": 1105
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.765325078094846e-05,
+      "loss": 0.2698,
+      "step": 1110
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.763033323485003e-05,
+      "loss": 0.2824,
+      "step": 1115
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.7607309895175744e-05,
+      "loss": 0.2834,
+      "step": 1120
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.758418086955643e-05,
+      "loss": 0.2834,
+      "step": 1125
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.7560946266116976e-05,
+      "loss": 0.2635,
+      "step": 1130
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.753760619347585e-05,
+      "loss": 0.2818,
+      "step": 1135
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.751416076074454e-05,
+      "loss": 0.2675,
+      "step": 1140
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.74906100775271e-05,
+      "loss": 0.2679,
+      "step": 1145
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.746695425391961e-05,
+      "loss": 0.2698,
+      "step": 1150
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.744319340050969e-05,
+      "loss": 0.2666,
+      "step": 1155
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.74193276283759e-05,
+      "loss": 0.261,
+      "step": 1160
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.7395357049087334e-05,
+      "loss": 0.269,
+      "step": 1165
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.737128177470303e-05,
+      "loss": 0.2696,
+      "step": 1170
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.7347101917771456e-05,
+      "loss": 0.2835,
+      "step": 1175
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.7322817591329984e-05,
+      "loss": 0.2726,
+      "step": 1180
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.729842890890437e-05,
+      "loss": 0.282,
+      "step": 1185
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.727393598450823e-05,
+      "loss": 0.2672,
+      "step": 1190
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.724933893264248e-05,
+      "loss": 0.2835,
+      "step": 1195
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.722463786829483e-05,
+      "loss": 0.2723,
+      "step": 1200
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.7199832906939226e-05,
+      "loss": 0.273,
+      "step": 1205
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.717492416453532e-05,
+      "loss": 0.2683,
+      "step": 1210
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.714991175752792e-05,
+      "loss": 0.2807,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.712479580284646e-05,
+      "loss": 0.266,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.709957641790445e-05,
+      "loss": 0.2769,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.707425372059891e-05,
+      "loss": 0.2752,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.704882782930982e-05,
+      "loss": 0.2651,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.7023298862899604e-05,
+      "loss": 0.2729,
+      "step": 1240
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.699766694071253e-05,
+      "loss": 0.2618,
+      "step": 1245
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.697193218257417e-05,
+      "loss": 0.2657,
+      "step": 1250
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.6946094708790846e-05,
+      "loss": 0.2755,
+      "step": 1255
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.692015464014906e-05,
+      "loss": 0.2655,
+      "step": 1260
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.689411209791492e-05,
+      "loss": 0.2739,
+      "step": 1265
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.68679672038336e-05,
+      "loss": 0.2544,
+      "step": 1270
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.684172008012875e-05,
+      "loss": 0.2616,
+      "step": 1275
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.681537084950191e-05,
+      "loss": 0.2744,
+      "step": 1280
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.6788919635131976e-05,
+      "loss": 0.2809,
+      "step": 1285
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.6762366560674606e-05,
+      "loss": 0.2683,
+      "step": 1290
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.6735711750261626e-05,
+      "loss": 0.2809,
+      "step": 1295
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.670895532850046e-05,
+      "loss": 0.2619,
+      "step": 1300
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.668209742047357e-05,
+      "loss": 0.2642,
+      "step": 1305
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.665513815173783e-05,
+      "loss": 0.2613,
+      "step": 1310
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.662807764832398e-05,
+      "loss": 0.2682,
+      "step": 1315
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.6600916036736e-05,
+      "loss": 0.2649,
+      "step": 1320
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.6573653443950556e-05,
+      "loss": 0.2595,
+      "step": 1325
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.6546289997416364e-05,
+      "loss": 0.2627,
+      "step": 1330
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.6518825825053636e-05,
+      "loss": 0.2715,
+      "step": 1335
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.6491261055253457e-05,
+      "loss": 0.2542,
+      "step": 1340
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.646359581687718e-05,
+      "loss": 0.2718,
+      "step": 1345
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.643583023925585e-05,
+      "loss": 0.2556,
+      "step": 1350
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.640796445218958e-05,
+      "loss": 0.2667,
+      "step": 1355
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.6379998585946935e-05,
+      "loss": 0.2751,
+      "step": 1360
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.635193277126435e-05,
+      "loss": 0.2681,
+      "step": 1365
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.632376713934551e-05,
+      "loss": 0.2649,
+      "step": 1370
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.629550182186071e-05,
+      "loss": 0.2738,
+      "step": 1375
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.6267136950946266e-05,
+      "loss": 0.2679,
+      "step": 1380
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.623867265920392e-05,
+      "loss": 0.2626,
+      "step": 1385
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.6210109079700154e-05,
+      "loss": 0.2704,
+      "step": 1390
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.618144634596563e-05,
+      "loss": 0.2533,
+      "step": 1395
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.6152684591994524e-05,
+      "loss": 0.2722,
+      "step": 1400
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.612382395224394e-05,
+      "loss": 0.2622,
+      "step": 1405
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.609486456163325e-05,
+      "loss": 0.2563,
+      "step": 1410
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.606580655554346e-05,
+      "loss": 0.2581,
+      "step": 1415
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.60366500698166e-05,
+      "loss": 0.2772,
+      "step": 1420
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.600739524075508e-05,
+      "loss": 0.2646,
+      "step": 1425
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.5978042205121037e-05,
+      "loss": 0.2602,
+      "step": 1430
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.594859110013574e-05,
+      "loss": 0.2483,
+      "step": 1435
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.591904206347888e-05,
+      "loss": 0.261,
+      "step": 1440
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.588939523328799e-05,
+      "loss": 0.2707,
+      "step": 1445
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.585965074815777e-05,
+      "loss": 0.262,
+      "step": 1450
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.582980874713943e-05,
+      "loss": 0.2515,
+      "step": 1455
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.579986936974009e-05,
+      "loss": 0.2639,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.576983275592203e-05,
+      "loss": 0.2632,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.573969904610215e-05,
+      "loss": 0.2672,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.5709468381151225e-05,
+      "loss": 0.2563,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.567914090239332e-05,
+      "loss": 0.2576,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.5648716751605036e-05,
+      "loss": 0.2642,
+      "step": 1485
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.561819607101495e-05,
+      "loss": 0.2547,
+      "step": 1490
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.5587579003302886e-05,
+      "loss": 0.2573,
+      "step": 1495
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.555686569159926e-05,
+      "loss": 0.2619,
+      "step": 1500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.552605627948443e-05,
+      "loss": 0.267,
+      "step": 1505
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.5495150910987974e-05,
+      "loss": 0.2628,
+      "step": 1510
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.546414973058808e-05,
+      "loss": 0.2614,
+      "step": 1515
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.5433052883210855e-05,
+      "loss": 0.2693,
+      "step": 1520
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.540186051422961e-05,
+      "loss": 0.2485,
+      "step": 1525
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.537057276946422e-05,
+      "loss": 0.2514,
+      "step": 1530
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.5339189795180414e-05,
+      "loss": 0.2486,
+      "step": 1535
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.530771173808912e-05,
+      "loss": 0.2615,
+      "step": 1540
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.527613874534576e-05,
+      "loss": 0.2601,
+      "step": 1545
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.524447096454956e-05,
+      "loss": 0.2498,
+      "step": 1550
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.5212708543742884e-05,
+      "loss": 0.249,
+      "step": 1555
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.5180851631410504e-05,
+      "loss": 0.2447,
+      "step": 1560
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.5148900376478936e-05,
+      "loss": 0.2509,
+      "step": 1565
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.5116854928315736e-05,
+      "loss": 0.2569,
+      "step": 1570
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.50847154367288e-05,
+      "loss": 0.2573,
+      "step": 1575
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.505248205196565e-05,
+      "loss": 0.2638,
+      "step": 1580
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.5020154924712774e-05,
+      "loss": 0.259,
+      "step": 1585
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.4987734206094855e-05,
+      "loss": 0.2667,
+      "step": 1590
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.495522004767414e-05,
+      "loss": 0.2536,
+      "step": 1595
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.4922612601449665e-05,
+      "loss": 0.2443,
+      "step": 1600
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.488991201985659e-05,
+      "loss": 0.256,
+      "step": 1605
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.485711845576547e-05,
+      "loss": 0.2593,
+      "step": 1610
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.4824232062481505e-05,
+      "loss": 0.2478,
+      "step": 1615
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.479125299374392e-05,
+      "loss": 0.2591,
+      "step": 1620
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.475818140372514e-05,
+      "loss": 0.2613,
+      "step": 1625
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.472501744703012e-05,
+      "loss": 0.2624,
+      "step": 1630
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.469176127869562e-05,
+      "loss": 0.247,
+      "step": 1635
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.465841305418948e-05,
+      "loss": 0.2626,
+      "step": 1640
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.46249729294099e-05,
+      "loss": 0.2573,
+      "step": 1645
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.4591441060684666e-05,
+      "loss": 0.2502,
+      "step": 1650
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.455781760477049e-05,
+      "loss": 0.2472,
+      "step": 1655
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.4524102718852225e-05,
+      "loss": 0.25,
+      "step": 1660
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.449029656054213e-05,
+      "loss": 0.2523,
+      "step": 1665
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.44563992878792e-05,
+      "loss": 0.2638,
+      "step": 1670
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.442241105932831e-05,
+      "loss": 0.2405,
+      "step": 1675
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.43883320337796e-05,
+      "loss": 0.2517,
+      "step": 1680
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4354162370547626e-05,
+      "loss": 0.2444,
+      "step": 1685
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.43199022293707e-05,
+      "loss": 0.2322,
+      "step": 1690
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4285551770410084e-05,
+      "loss": 0.2563,
+      "step": 1695
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.425111115424927e-05,
+      "loss": 0.2408,
+      "step": 1700
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.421658054189321e-05,
+      "loss": 0.2447,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.418196009476759e-05,
+      "loss": 0.2506,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.4147249974718065e-05,
+      "loss": 0.2682,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.411245034400948e-05,
+      "loss": 0.2487,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.407756136532514e-05,
+      "loss": 0.2462,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.4042583201766044e-05,
+      "loss": 0.2431,
+      "step": 1730
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.400751601685009e-05,
+      "loss": 0.248,
+      "step": 1735
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.3972359974511386e-05,
+      "loss": 0.2458,
+      "step": 1740
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.3937115239099394e-05,
+      "loss": 0.2425,
+      "step": 1745
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.390178197537822e-05,
+      "loss": 0.2331,
+      "step": 1750
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.386636034852582e-05,
+      "loss": 0.2424,
+      "step": 1755
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.383085052413327e-05,
+      "loss": 0.2466,
+      "step": 1760
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.379525266820389e-05,
+      "loss": 0.2422,
+      "step": 1765
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.375956694715259e-05,
+      "loss": 0.2505,
+      "step": 1770
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.372379352780502e-05,
+      "loss": 0.2446,
+      "step": 1775
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.36879325773968e-05,
+      "loss": 0.2513,
+      "step": 1780
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.365198426357276e-05,
+      "loss": 0.2471,
+      "step": 1785
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.3615948754386125e-05,
+      "loss": 0.2402,
+      "step": 1790
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.357982621829776e-05,
+      "loss": 0.2516,
+      "step": 1795
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.354361682417536e-05,
+      "loss": 0.2507,
+      "step": 1800
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.3507320741292665e-05,
+      "loss": 0.2441,
+      "step": 1805
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.347093813932869e-05,
+      "loss": 0.2519,
+      "step": 1810
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.34344691883669e-05,
+      "loss": 0.2329,
+      "step": 1815
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.339791405889442e-05,
+      "loss": 0.2408,
+      "step": 1820
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.3361272921801275e-05,
+      "loss": 0.2475,
+      "step": 1825
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.332454594837955e-05,
+      "loss": 0.2408,
+      "step": 1830
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.328773331032259e-05,
+      "loss": 0.2385,
+      "step": 1835
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.3250835179724215e-05,
+      "loss": 0.2483,
+      "step": 1840
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.3213851729077926e-05,
+      "loss": 0.235,
+      "step": 1845
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.3176783131276076e-05,
+      "loss": 0.2452,
+      "step": 1850
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.3139629559609056e-05,
+      "loss": 0.2296,
+      "step": 1855
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.3102391187764515e-05,
+      "loss": 0.2309,
+      "step": 1860
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.3065068189826515e-05,
+      "loss": 0.25,
+      "step": 1865
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.3027660740274756e-05,
+      "loss": 0.2578,
+      "step": 1870
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.29901690139837e-05,
+      "loss": 0.2413,
+      "step": 1875
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.2952593186221834e-05,
+      "loss": 0.2394,
+      "step": 1880
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.291493343265078e-05,
+      "loss": 0.2416,
+      "step": 1885
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.28771899293245e-05,
+      "loss": 0.242,
+      "step": 1890
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.283936285268848e-05,
+      "loss": 0.2469,
+      "step": 1895
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.280145237957892e-05,
+      "loss": 0.2436,
+      "step": 1900
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.276345868722185e-05,
+      "loss": 0.2403,
+      "step": 1905
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.2725381953232357e-05,
+      "loss": 0.2484,
+      "step": 1910
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.268722235561374e-05,
+      "loss": 0.246,
+      "step": 1915
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.2648980072756664e-05,
+      "loss": 0.2427,
+      "step": 1920
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.2610655283438344e-05,
+      "loss": 0.2383,
+      "step": 1925
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.257224816682169e-05,
+      "loss": 0.2443,
+      "step": 1930
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.25337589024545e-05,
+      "loss": 0.2374,
+      "step": 1935
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.2495187670268566e-05,
+      "loss": 0.2351,
+      "step": 1940
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.245653465057892e-05,
+      "loss": 0.2407,
+      "step": 1945
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.2417800024082874e-05,
+      "loss": 0.2507,
+      "step": 1950
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.23789839718593e-05,
+      "loss": 0.2372,
+      "step": 1955
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.2340086675367687e-05,
+      "loss": 0.2483,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.2301108316447344e-05,
+      "loss": 0.2357,
+      "step": 1965
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.226204907731654e-05,
+      "loss": 0.2485,
+      "step": 1970
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.222290914057163e-05,
+      "loss": 0.2509,
+      "step": 1975
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.218368868918623e-05,
+      "loss": 0.2392,
+      "step": 1980
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.2144387906510354e-05,
+      "loss": 0.2504,
+      "step": 1985
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.2105006976269535e-05,
+      "loss": 0.2338,
+      "step": 1990
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.2065546082564006e-05,
+      "loss": 0.2407,
+      "step": 1995
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.202600540986781e-05,
+      "loss": 0.2413,
+      "step": 2000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.198638514302795e-05,
+      "loss": 0.2458,
+      "step": 2005
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.194668546726351e-05,
+      "loss": 0.2353,
+      "step": 2010
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.1906906568164805e-05,
+      "loss": 0.2459,
+      "step": 2015
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.1867048631692495e-05,
+      "loss": 0.2347,
+      "step": 2020
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.182711184417676e-05,
+      "loss": 0.2372,
+      "step": 2025
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.1787096392316375e-05,
+      "loss": 0.2316,
+      "step": 2030
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.174700246317784e-05,
+      "loss": 0.224,
+      "step": 2035
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.170683024419458e-05,
+      "loss": 0.2326,
+      "step": 2040
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.1666579923165964e-05,
+      "loss": 0.2362,
+      "step": 2045
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.162625168825648e-05,
+      "loss": 0.2293,
+      "step": 2050
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.15858457279949e-05,
+      "loss": 0.2453,
+      "step": 2055
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.15453622312733e-05,
+      "loss": 0.2412,
+      "step": 2060
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.1504801387346234e-05,
+      "loss": 0.233,
+      "step": 2065
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.146416338582988e-05,
+      "loss": 0.2472,
+      "step": 2070
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.142344841670107e-05,
+      "loss": 0.2368,
+      "step": 2075
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.138265667029648e-05,
+      "loss": 0.2258,
+      "step": 2080
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.134178833731169e-05,
+      "loss": 0.2328,
+      "step": 2085
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.130084360880033e-05,
+      "loss": 0.2308,
+      "step": 2090
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.1259822676173155e-05,
+      "loss": 0.2322,
+      "step": 2095
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.121872573119717e-05,
+      "loss": 0.2466,
+      "step": 2100
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.117755296599471e-05,
+      "loss": 0.2395,
+      "step": 2105
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.113630457304258e-05,
+      "loss": 0.2261,
+      "step": 2110
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.109498074517114e-05,
+      "loss": 0.2319,
+      "step": 2115
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.1053581675563356e-05,
+      "loss": 0.2403,
+      "step": 2120
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.101210755775399e-05,
+      "loss": 0.2276,
+      "step": 2125
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.097055858562861e-05,
+      "loss": 0.2339,
+      "step": 2130
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.0928934953422725e-05,
+      "loss": 0.2425,
+      "step": 2135
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.088723685572087e-05,
+      "loss": 0.2436,
+      "step": 2140
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.0845464487455706e-05,
+      "loss": 0.2258,
+      "step": 2145
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.080361804390708e-05,
+      "loss": 0.2437,
+      "step": 2150
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.076169772070113e-05,
+      "loss": 0.2352,
+      "step": 2155
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.071970371380941e-05,
+      "loss": 0.2364,
+      "step": 2160
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.067763621954787e-05,
+      "loss": 0.2236,
+      "step": 2165
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.063549543457606e-05,
+      "loss": 0.2335,
+      "step": 2170
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.059328155589612e-05,
+      "loss": 0.2333,
+      "step": 2175
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.0550994780851914e-05,
+      "loss": 0.2317,
+      "step": 2180
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.050863530712807e-05,
+      "loss": 0.224,
+      "step": 2185
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.0466203332749084e-05,
+      "loss": 0.2436,
+      "step": 2190
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.042369905607836e-05,
+      "loss": 0.2406,
+      "step": 2195
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.038112267581734e-05,
+      "loss": 0.2325,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.033847439100451e-05,
+      "loss": 0.2368,
+      "step": 2205
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.029575440101451e-05,
+      "loss": 0.2362,
+      "step": 2210
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.02529629055572e-05,
+      "loss": 0.2354,
+      "step": 2215
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.021010010467671e-05,
+      "loss": 0.2472,
+      "step": 2220
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.0167166198750505e-05,
+      "loss": 0.2294,
+      "step": 2225
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.012416138848846e-05,
+      "loss": 0.2385,
+      "step": 2230
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.008108587493193e-05,
+      "loss": 0.2363,
+      "step": 2235
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.003793985945279e-05,
+      "loss": 0.2307,
+      "step": 2240
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.99947235437525e-05,
+      "loss": 0.2277,
+      "step": 2245
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.9951437129861155e-05,
+      "loss": 0.2273,
+      "step": 2250
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.990808082013657e-05,
+      "loss": 0.2334,
+      "step": 2255
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.986465481726329e-05,
+      "loss": 0.2263,
+      "step": 2260
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.982115932425168e-05,
+      "loss": 0.2229,
+      "step": 2265
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.977759454443696e-05,
+      "loss": 0.2205,
+      "step": 2270
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.973396068147825e-05,
+      "loss": 0.2331,
+      "step": 2275
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.969025793935763e-05,
+      "loss": 0.2336,
+      "step": 2280
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.964648652237918e-05,
+      "loss": 0.2265,
+      "step": 2285
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.960264663516802e-05,
+      "loss": 0.2221,
+      "step": 2290
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.955873848266937e-05,
+      "loss": 0.2393,
+      "step": 2295
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.951476227014755e-05,
+      "loss": 0.2473,
+      "step": 2300
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.9470718203185095e-05,
+      "loss": 0.2279,
+      "step": 2305
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.942660648768173e-05,
+      "loss": 0.2256,
+      "step": 2310
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.93824273298534e-05,
+      "loss": 0.2244,
+      "step": 2315
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.9338180936231386e-05,
+      "loss": 0.2334,
+      "step": 2320
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.929386751366124e-05,
+      "loss": 0.225,
+      "step": 2325
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.924948726930189e-05,
+      "loss": 0.2233,
+      "step": 2330
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.9205040410624637e-05,
+      "loss": 0.2242,
+      "step": 2335
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.9160527145412205e-05,
+      "loss": 0.2397,
+      "step": 2340
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.911594768175775e-05,
+      "loss": 0.2288,
+      "step": 2345
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.90713022280639e-05,
+      "loss": 0.2244,
+      "step": 2350
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.902659099304176e-05,
+      "loss": 0.2253,
+      "step": 2355
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.898181418570998e-05,
+      "loss": 0.2283,
+      "step": 2360
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.893697201539373e-05,
+      "loss": 0.2382,
+      "step": 2365
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.889206469172378e-05,
+      "loss": 0.2326,
+      "step": 2370
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.8847092424635425e-05,
+      "loss": 0.2324,
+      "step": 2375
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.88020554243676e-05,
+      "loss": 0.2143,
+      "step": 2380
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.875695390146185e-05,
+      "loss": 0.2257,
+      "step": 2385
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.8711788066761346e-05,
+      "loss": 0.2255,
+      "step": 2390
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.866655813140992e-05,
+      "loss": 0.2269,
+      "step": 2395
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.862126430685107e-05,
+      "loss": 0.2277,
+      "step": 2400
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.857590680482695e-05,
+      "loss": 0.2223,
+      "step": 2405
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.8530485837377394e-05,
+      "loss": 0.2236,
+      "step": 2410
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.848500161683895e-05,
+      "loss": 0.2223,
+      "step": 2415
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.843945435584385e-05,
+      "loss": 0.2197,
+      "step": 2420
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.8393844267319036e-05,
+      "loss": 0.2376,
+      "step": 2425
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.834817156448516e-05,
+      "loss": 0.225,
+      "step": 2430
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.8302436460855576e-05,
+      "loss": 0.2218,
+      "step": 2435
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.8256639170235356e-05,
+      "loss": 0.2229,
+      "step": 2440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.821077990672031e-05,
+      "loss": 0.2172,
+      "step": 2445
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8164858884695945e-05,
+      "loss": 0.2257,
+      "step": 2450
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.811887631883647e-05,
+      "loss": 0.2277,
+      "step": 2455
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.807283242410381e-05,
+      "loss": 0.1779,
+      "step": 2460
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8026727415746613e-05,
+      "loss": 0.1799,
+      "step": 2465
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.798056150929919e-05,
+      "loss": 0.1758,
+      "step": 2470
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7934334920580574e-05,
+      "loss": 0.1828,
+      "step": 2475
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7888047865693465e-05,
+      "loss": 0.1708,
+      "step": 2480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.784170056102323e-05,
+      "loss": 0.1762,
+      "step": 2485
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.77952932232369e-05,
+      "loss": 0.1783,
+      "step": 2490
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7748826069282136e-05,
+      "loss": 0.1766,
+      "step": 2495
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770229931638628e-05,
+      "loss": 0.1816,
+      "step": 2500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765571318205523e-05,
+      "loss": 0.1778,
+      "step": 2505
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760906788407252e-05,
+      "loss": 0.179,
+      "step": 2510
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756236364049826e-05,
+      "loss": 0.169,
+      "step": 2515
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.751560066966811e-05,
+      "loss": 0.1656,
+      "step": 2520
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.7468779190192275e-05,
+      "loss": 0.1677,
+      "step": 2525
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.7421899420954496e-05,
+      "loss": 0.1789,
+      "step": 2530
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.7374961581111e-05,
+      "loss": 0.181,
+      "step": 2535
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.7327965890089485e-05,
+      "loss": 0.1825,
+      "step": 2540
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.728091256758808e-05,
+      "loss": 0.1738,
+      "step": 2545
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.7233801833574365e-05,
+      "loss": 0.1801,
+      "step": 2550
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.718663390828428e-05,
+      "loss": 0.1809,
+      "step": 2555
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.713940901222116e-05,
+      "loss": 0.1746,
+      "step": 2560
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.7092127366154614e-05,
+      "loss": 0.178,
+      "step": 2565
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.704478919111962e-05,
+      "loss": 0.1733,
+      "step": 2570
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.699739470841537e-05,
+      "loss": 0.1779,
+      "step": 2575
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.6949944139604295e-05,
+      "loss": 0.1817,
+      "step": 2580
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.690243770651105e-05,
+      "loss": 0.1788,
+      "step": 2585
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.68548756312214e-05,
+      "loss": 0.1783,
+      "step": 2590
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.680725813608127e-05,
+      "loss": 0.1672,
+      "step": 2595
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.6759585443695636e-05,
+      "loss": 0.1829,
+      "step": 2600
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.671185777692754e-05,
+      "loss": 0.1742,
+      "step": 2605
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.6664075358897e-05,
+      "loss": 0.1726,
+      "step": 2610
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.661623841298e-05,
+      "loss": 0.1878,
+      "step": 2615
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.6568347162807426e-05,
+      "loss": 0.1758,
+      "step": 2620
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.652040183226404e-05,
+      "loss": 0.1774,
+      "step": 2625
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.64724026454874e-05,
+      "loss": 0.1726,
+      "step": 2630
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.642434982686687e-05,
+      "loss": 0.1761,
+      "step": 2635
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.63762436010425e-05,
+      "loss": 0.1793,
+      "step": 2640
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.632808419290403e-05,
+      "loss": 0.1793,
+      "step": 2645
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.6279871827589806e-05,
+      "loss": 0.1716,
+      "step": 2650
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.623160673048576e-05,
+      "loss": 0.1871,
+      "step": 2655
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.618328912722433e-05,
+      "loss": 0.1772,
+      "step": 2660
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.6134919243683394e-05,
+      "loss": 0.176,
+      "step": 2665
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.608649730598528e-05,
+      "loss": 0.1771,
+      "step": 2670
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.6038023540495606e-05,
+      "loss": 0.1719,
+      "step": 2675
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.598949817382231e-05,
+      "loss": 0.1815,
+      "step": 2680
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.594092143281454e-05,
+      "loss": 0.1771,
+      "step": 2685
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.589229354456164e-05,
+      "loss": 0.1814,
+      "step": 2690
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.584361473639204e-05,
+      "loss": 0.1769,
+      "step": 2695
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.5794885235872205e-05,
+      "loss": 0.1659,
+      "step": 2700
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.57461052708056e-05,
+      "loss": 0.1813,
+      "step": 2705
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.5697275069231594e-05,
+      "loss": 0.1776,
+      "step": 2710
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.564839485942441e-05,
+      "loss": 0.1683,
+      "step": 2715
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.559946486989203e-05,
+      "loss": 0.1814,
+      "step": 2720
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.5550485329375175e-05,
+      "loss": 0.173,
+      "step": 2725
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.550145646684622e-05,
+      "loss": 0.1778,
+      "step": 2730
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.5452378511508065e-05,
+      "loss": 0.1749,
+      "step": 2735
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.5403251692793174e-05,
+      "loss": 0.1781,
+      "step": 2740
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.535407624036238e-05,
+      "loss": 0.1754,
+      "step": 2745
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.530485238410391e-05,
+      "loss": 0.174,
+      "step": 2750
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.525558035413227e-05,
+      "loss": 0.1748,
+      "step": 2755
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.520626038078715e-05,
+      "loss": 0.1823,
+      "step": 2760
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.51568926946324e-05,
+      "loss": 0.1736,
+      "step": 2765
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.510747752645488e-05,
+      "loss": 0.1728,
+      "step": 2770
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.5058015107263465e-05,
+      "loss": 0.1749,
+      "step": 2775
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.500850566828788e-05,
+      "loss": 0.1748,
+      "step": 2780
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.4958949440977694e-05,
+      "loss": 0.1789,
+      "step": 2785
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.490934665700119e-05,
+      "loss": 0.1687,
+      "step": 2790
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.48596975482443e-05,
+      "loss": 0.1802,
+      "step": 2795
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.4810002346809515e-05,
+      "loss": 0.1802,
+      "step": 2800
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.47602612850148e-05,
+      "loss": 0.1816,
+      "step": 2805
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.471047459539253e-05,
+      "loss": 0.1758,
+      "step": 2810
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.4660642510688354e-05,
+      "loss": 0.1739,
+      "step": 2815
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.461076526386014e-05,
+      "loss": 0.1732,
+      "step": 2820
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.456084308807693e-05,
+      "loss": 0.1756,
+      "step": 2825
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.451087621671773e-05,
+      "loss": 0.1757,
+      "step": 2830
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.4460864883370555e-05,
+      "loss": 0.166,
+      "step": 2835
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.4410809321831216e-05,
+      "loss": 0.1715,
+      "step": 2840
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.436070976610234e-05,
+      "loss": 0.1731,
+      "step": 2845
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.431056645039219e-05,
+      "loss": 0.1728,
+      "step": 2850
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.4260379609113594e-05,
+      "loss": 0.1691,
+      "step": 2855
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.421014947688288e-05,
+      "loss": 0.1687,
+      "step": 2860
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.415987628851874e-05,
+      "loss": 0.1823,
+      "step": 2865
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.411962689533804e-05,
+      "loss": 0.1785,
+      "step": 2870
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.406927679831862e-05,
+      "loss": 0.1734,
+      "step": 2875
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.401888430372543e-05,
+      "loss": 0.1746,
+      "step": 2880
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.396844964713618e-05,
+      "loss": 0.176,
+      "step": 2885
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.391797306432568e-05,
+      "loss": 0.1742,
+      "step": 2890
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.386745479126476e-05,
+      "loss": 0.181,
+      "step": 2895
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.381689506411912e-05,
+      "loss": 0.1756,
+      "step": 2900
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.3766294119248265e-05,
+      "loss": 0.1627,
+      "step": 2905
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.3715652193204395e-05,
+      "loss": 0.1859,
+      "step": 2910
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.366496952273127e-05,
+      "loss": 0.1814,
+      "step": 2915
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.361424634476313e-05,
+      "loss": 0.169,
+      "step": 2920
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.3563482896423624e-05,
+      "loss": 0.1716,
+      "step": 2925
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.351267941502457e-05,
+      "loss": 0.1774,
+      "step": 2930
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.3461836138065026e-05,
+      "loss": 0.1738,
+      "step": 2935
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.341095330323005e-05,
+      "loss": 0.1679,
+      "step": 2940
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.3360031148389605e-05,
+      "loss": 0.1722,
+      "step": 2945
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.330906991159751e-05,
+      "loss": 0.1737,
+      "step": 2950
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.3258069831090264e-05,
+      "loss": 0.1798,
+      "step": 2955
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.320703114528596e-05,
+      "loss": 0.1772,
+      "step": 2960
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.315595409278316e-05,
+      "loss": 0.1658,
+      "step": 2965
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.31048389123598e-05,
+      "loss": 0.1837,
+      "step": 2970
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.305368584297204e-05,
+      "loss": 0.1727,
+      "step": 2975
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.3002495123753175e-05,
+      "loss": 0.1696,
+      "step": 2980
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.29512669940125e-05,
+      "loss": 0.1797,
+      "step": 2985
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.2900001693234223e-05,
+      "loss": 0.1673,
+      "step": 2990
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.2848699461076286e-05,
+      "loss": 0.1693,
+      "step": 2995
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.279736053736931e-05,
+      "loss": 0.1748,
+      "step": 3000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.2745985162115425e-05,
+      "loss": 0.1744,
+      "step": 3005
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.269457357548718e-05,
+      "loss": 0.1772,
+      "step": 3010
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.264312601782639e-05,
+      "loss": 0.1779,
+      "step": 3015
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.2591642729643034e-05,
+      "loss": 0.1713,
+      "step": 3020
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.254012395161415e-05,
+      "loss": 0.1762,
+      "step": 3025
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.248856992458265e-05,
+      "loss": 0.1741,
+      "step": 3030
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.2436980889556264e-05,
+      "loss": 0.178,
+      "step": 3035
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.238535708770635e-05,
+      "loss": 0.1763,
+      "step": 3040
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.233369876036682e-05,
+      "loss": 0.1822,
+      "step": 3045
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.2282006149032964e-05,
+      "loss": 0.1724,
+      "step": 3050
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.223027949536036e-05,
+      "loss": 0.1726,
+      "step": 3055
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.2178519041163744e-05,
+      "loss": 0.1792,
+      "step": 3060
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.212672502841582e-05,
+      "loss": 0.1814,
+      "step": 3065
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.207489769924622e-05,
+      "loss": 0.1792,
+      "step": 3070
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.20230372959403e-05,
+      "loss": 0.1675,
+      "step": 3075
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.1971144060938034e-05,
+      "loss": 0.1832,
+      "step": 3080
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.1919218236832886e-05,
+      "loss": 0.1646,
+      "step": 3085
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.186726006637067e-05,
+      "loss": 0.1733,
+      "step": 3090
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.181526979244842e-05,
+      "loss": 0.1717,
+      "step": 3095
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.176324765811323e-05,
+      "loss": 0.1743,
+      "step": 3100
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.171119390656115e-05,
+      "loss": 0.1711,
+      "step": 3105
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.1659108781136046e-05,
+      "loss": 0.1661,
+      "step": 3110
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.160699252532843e-05,
+      "loss": 0.1714,
+      "step": 3115
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.1554845382774346e-05,
+      "loss": 0.1712,
+      "step": 3120
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.150266759725425e-05,
+      "loss": 0.1848,
+      "step": 3125
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.145045941269184e-05,
+      "loss": 0.1735,
+      "step": 3130
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.1398221073152914e-05,
+      "loss": 0.1747,
+      "step": 3135
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.134595282284426e-05,
+      "loss": 0.1708,
+      "step": 3140
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.1293654906112475e-05,
+      "loss": 0.163,
+      "step": 3145
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.1241327567442844e-05,
+      "loss": 0.1704,
+      "step": 3150
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.118897105145822e-05,
+      "loss": 0.1672,
+      "step": 3155
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.1136585602917816e-05,
+      "loss": 0.1688,
+      "step": 3160
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.108417146671615e-05,
+      "loss": 0.1756,
+      "step": 3165
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.103172888788178e-05,
+      "loss": 0.179,
+      "step": 3170
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.097925811157632e-05,
+      "loss": 0.174,
+      "step": 3175
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.092675938309313e-05,
+      "loss": 0.168,
+      "step": 3180
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.0874232947856266e-05,
+      "loss": 0.1713,
+      "step": 3185
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.082167905141933e-05,
+      "loss": 0.1711,
+      "step": 3190
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.076909793946426e-05,
+      "loss": 0.1715,
+      "step": 3195
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.071648985780026e-05,
+      "loss": 0.1713,
+      "step": 3200
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.066385505236259e-05,
+      "loss": 0.1802,
+      "step": 3205
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.0611193769211446e-05,
+      "loss": 0.1684,
+      "step": 3210
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.0558506254530824e-05,
+      "loss": 0.1808,
+      "step": 3215
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.0505792754627304e-05,
+      "loss": 0.1769,
+      "step": 3220
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.045305351592899e-05,
+      "loss": 0.176,
+      "step": 3225
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.0400288784984282e-05,
+      "loss": 0.1727,
+      "step": 3230
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.0347498808460766e-05,
+      "loss": 0.1725,
+      "step": 3235
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.029468383314404e-05,
+      "loss": 0.1674,
+      "step": 3240
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.0241844105936564e-05,
+      "loss": 0.1722,
+      "step": 3245
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.0188979873856527e-05,
+      "loss": 0.1686,
+      "step": 3250
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.0136091384036657e-05,
+      "loss": 0.1762,
+      "step": 3255
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.00831788837231e-05,
+      "loss": 0.1758,
+      "step": 3260
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.003024262027423e-05,
+      "loss": 0.1709,
+      "step": 3265
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.9977282841159515e-05,
+      "loss": 0.1726,
+      "step": 3270
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.9924299793958366e-05,
+      "loss": 0.172,
+      "step": 3275
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.9871293726358956e-05,
+      "loss": 0.1694,
+      "step": 3280
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.9818264886157094e-05,
+      "loss": 0.1749,
+      "step": 3285
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.9765213521255016e-05,
+      "loss": 0.173,
+      "step": 3290
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.971213987966029e-05,
+      "loss": 0.1727,
+      "step": 3295
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.96590442094846e-05,
+      "loss": 0.1682,
+      "step": 3300
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.960592675894263e-05,
+      "loss": 0.164,
+      "step": 3305
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.9552787776350864e-05,
+      "loss": 0.1781,
+      "step": 3310
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.9499627510126472e-05,
+      "loss": 0.1688,
+      "step": 3315
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.94464462087861e-05,
+      "loss": 0.177,
+      "step": 3320
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.9393244120944737e-05,
+      "loss": 0.1728,
+      "step": 3325
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.9340021495314528e-05,
+      "loss": 0.1708,
+      "step": 3330
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.9286778580703676e-05,
+      "loss": 0.1681,
+      "step": 3335
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.9233515626015185e-05,
+      "loss": 0.1754,
+      "step": 3340
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.9180232880245766e-05,
+      "loss": 0.1708,
+      "step": 3345
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.912693059248464e-05,
+      "loss": 0.1665,
+      "step": 3350
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.9073609011912407e-05,
+      "loss": 0.1779,
+      "step": 3355
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.9020268387799815e-05,
+      "loss": 0.1709,
+      "step": 3360
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.896690896950668e-05,
+      "loss": 0.1689,
+      "step": 3365
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.891353100648066e-05,
+      "loss": 0.1697,
+      "step": 3370
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.8860134748256107e-05,
+      "loss": 0.1729,
+      "step": 3375
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8806720444452907e-05,
+      "loss": 0.1619,
+      "step": 3380
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8753288344775286e-05,
+      "loss": 0.1741,
+      "step": 3385
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8699838699010707e-05,
+      "loss": 0.1679,
+      "step": 3390
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.86463717570286e-05,
+      "loss": 0.1637,
+      "step": 3395
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8592887768779303e-05,
+      "loss": 0.1595,
+      "step": 3400
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.8539386984292822e-05,
+      "loss": 0.174,
+      "step": 3405
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.8485869653677666e-05,
+      "loss": 0.1721,
+      "step": 3410
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.8432336027119732e-05,
+      "loss": 0.1783,
+      "step": 3415
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.837878635488107e-05,
+      "loss": 0.1724,
+      "step": 3420
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.8325220887298755e-05,
+      "loss": 0.173,
+      "step": 3425
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.8271639874783683e-05,
+      "loss": 0.1701,
+      "step": 3430
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.8218043567819447e-05,
+      "loss": 0.1719,
+      "step": 3435
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.8164432216961135e-05,
+      "loss": 0.1703,
+      "step": 3440
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.811080607283414e-05,
+      "loss": 0.1749,
+      "step": 3445
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.8057165386133038e-05,
+      "loss": 0.1657,
+      "step": 3450
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.800351040762037e-05,
+      "loss": 0.1715,
+      "step": 3455
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.7949841388125503e-05,
+      "loss": 0.1749,
+      "step": 3460
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.789615857854344e-05,
+      "loss": 0.171,
+      "step": 3465
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.7842462229833645e-05,
+      "loss": 0.1653,
+      "step": 3470
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.778875259301888e-05,
+      "loss": 0.1759,
+      "step": 3475
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7735029919184024e-05,
+      "loss": 0.1696,
+      "step": 3480
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7681294459474915e-05,
+      "loss": 0.1718,
+      "step": 3485
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7627546465097143e-05,
+      "loss": 0.1673,
+      "step": 3490
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7573786187314897e-05,
+      "loss": 0.1697,
+      "step": 3495
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.7520013877449826e-05,
+      "loss": 0.171,
+      "step": 3500
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.746622978687977e-05,
+      "loss": 0.1698,
+      "step": 3505
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.7412434167037694e-05,
+      "loss": 0.1658,
+      "step": 3510
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.7358627269410413e-05,
+      "loss": 0.1646,
+      "step": 3515
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.7304809345537507e-05,
+      "loss": 0.1712,
+      "step": 3520
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.7250980647010083e-05,
+      "loss": 0.1733,
+      "step": 3525
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.719714142546961e-05,
+      "loss": 0.1732,
+      "step": 3530
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.714329193260676e-05,
+      "loss": 0.1681,
+      "step": 3535
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.7089432420160234e-05,
+      "loss": 0.1697,
+      "step": 3540
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.7035563139915553e-05,
+      "loss": 0.1692,
+      "step": 3545
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.6981684343703906e-05,
+      "loss": 0.1656,
+      "step": 3550
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.6927796283400963e-05,
+      "loss": 0.1749,
+      "step": 3555
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.6873899210925736e-05,
+      "loss": 0.1621,
+      "step": 3560
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.681999337823931e-05,
+      "loss": 0.1686,
+      "step": 3565
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.6766079037343782e-05,
+      "loss": 0.17,
+      "step": 3570
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.671215644028097e-05,
+      "loss": 0.1684,
+      "step": 3575
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.665822583913133e-05,
+      "loss": 0.1738,
+      "step": 3580
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.6604287486012726e-05,
+      "loss": 0.165,
+      "step": 3585
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.6550341633079233e-05,
+      "loss": 0.1729,
+      "step": 3590
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.6496388532520032e-05,
+      "loss": 0.1673,
+      "step": 3595
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.6442428436558148e-05,
+      "loss": 0.1655,
+      "step": 3600
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.6388461597449333e-05,
+      "loss": 0.167,
+      "step": 3605
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.633448826748084e-05,
+      "loss": 0.1705,
+      "step": 3610
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.6280508698970286e-05,
+      "loss": 0.1599,
+      "step": 3615
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.622652314426445e-05,
+      "loss": 0.166,
+      "step": 3620
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.617253185573808e-05,
+      "loss": 0.1655,
+      "step": 3625
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.6118535085792745e-05,
+      "loss": 0.1669,
+      "step": 3630
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.606453308685563e-05,
+      "loss": 0.1673,
+      "step": 3635
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.601052611137837e-05,
+      "loss": 0.1696,
+      "step": 3640
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.595651441183586e-05,
+      "loss": 0.1731,
+      "step": 3645
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.5902498240725076e-05,
+      "loss": 0.1637,
+      "step": 3650
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.5848477850563918e-05,
+      "loss": 0.1714,
+      "step": 3655
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.5794453493889987e-05,
+      "loss": 0.1724,
+      "step": 3660
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.574042542325943e-05,
+      "loss": 0.1675,
+      "step": 3665
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.5686393891245763e-05,
+      "loss": 0.1671,
+      "step": 3670
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.5632359150438696e-05,
+      "loss": 0.1753,
+      "step": 3675
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.55783214534429e-05,
+      "loss": 0.1681,
+      "step": 3680
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.552428105287691e-05,
+      "loss": 0.1721,
+      "step": 3685
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.547023820137187e-05,
+      "loss": 0.1664,
+      "step": 3690
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.54161931515704e-05,
+      "loss": 0.1698,
+      "step": 3695
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5362146156125392e-05,
+      "loss": 0.1637,
+      "step": 3700
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.530809746769882e-05,
+      "loss": 0.1724,
+      "step": 3705
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.525404733896059e-05,
+      "loss": 0.1604,
+      "step": 3710
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.519999602258733e-05,
+      "loss": 0.1604,
+      "step": 3715
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5145943771261226e-05,
+      "loss": 0.1653,
+      "step": 3720
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.509189083766884e-05,
+      "loss": 0.1606,
+      "step": 3725
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.503783747449991e-05,
+      "loss": 0.1671,
+      "step": 3730
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.4983783934446188e-05,
+      "loss": 0.1683,
+      "step": 3735
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.492973047020026e-05,
+      "loss": 0.17,
+      "step": 3740
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.4875677334454344e-05,
+      "loss": 0.1648,
+      "step": 3745
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.4821624779899137e-05,
+      "loss": 0.167,
+      "step": 3750
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.47675730592226e-05,
+      "loss": 0.1645,
+      "step": 3755
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.4713522425108835e-05,
+      "loss": 0.1702,
+      "step": 3760
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.4659473130236797e-05,
+      "loss": 0.1695,
+      "step": 3765
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.4605425427279257e-05,
+      "loss": 0.165,
+      "step": 3770
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.4551379568901486e-05,
+      "loss": 0.1649,
+      "step": 3775
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.4497335807760162e-05,
+      "loss": 0.1626,
+      "step": 3780
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.4443294396502144e-05,
+      "loss": 0.1667,
+      "step": 3785
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.4389255587763322e-05,
+      "loss": 0.1602,
+      "step": 3790
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.43352196341674e-05,
+      "loss": 0.17,
+      "step": 3795
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.428118678832474e-05,
+      "loss": 0.1622,
+      "step": 3800
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.4227157302831192e-05,
+      "loss": 0.1656,
+      "step": 3805
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.4173131430266868e-05,
+      "loss": 0.1674,
+      "step": 3810
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.4119109423195023e-05,
+      "loss": 0.1618,
+      "step": 3815
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4065091534160813e-05,
+      "loss": 0.1648,
+      "step": 3820
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4011078015690165e-05,
+      "loss": 0.1671,
+      "step": 3825
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.395706912028855e-05,
+      "loss": 0.1623,
+      "step": 3830
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.3903065100439855e-05,
+      "loss": 0.1574,
+      "step": 3835
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.3849066208605142e-05,
+      "loss": 0.1609,
+      "step": 3840
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.3795072697221545e-05,
+      "loss": 0.1616,
+      "step": 3845
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.3741084818700988e-05,
+      "loss": 0.1665,
+      "step": 3850
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.368710282542912e-05,
+      "loss": 0.1584,
+      "step": 3855
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.363312696976403e-05,
+      "loss": 0.1671,
+      "step": 3860
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.357915750403515e-05,
+      "loss": 0.1586,
+      "step": 3865
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.3525194680542008e-05,
+      "loss": 0.1724,
+      "step": 3870
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.3471238751553114e-05,
+      "loss": 0.1614,
+      "step": 3875
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.341728996930471e-05,
+      "loss": 0.1643,
+      "step": 3880
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.336334858599967e-05,
+      "loss": 0.1622,
+      "step": 3885
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.3309414853806245e-05,
+      "loss": 0.16,
+      "step": 3890
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.3255489024856932e-05,
+      "loss": 0.1642,
+      "step": 3895
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.320157135124728e-05,
+      "loss": 0.1614,
+      "step": 3900
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.3147662085034713e-05,
+      "loss": 0.1683,
+      "step": 3905
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.3093761478237346e-05,
+      "loss": 0.1609,
+      "step": 3910
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.3039869782832832e-05,
+      "loss": 0.1656,
+      "step": 3915
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.298598725075713e-05,
+      "loss": 0.1734,
+      "step": 3920
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.2932114133903405e-05,
+      "loss": 0.1716,
+      "step": 3925
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.287825068412076e-05,
+      "loss": 0.1703,
+      "step": 3930
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.2824397153213156e-05,
+      "loss": 0.1582,
+      "step": 3935
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.2770553792938137e-05,
+      "loss": 0.17,
+      "step": 3940
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.2716720855005737e-05,
+      "loss": 0.1681,
+      "step": 3945
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.2662898591077237e-05,
+      "loss": 0.1652,
+      "step": 3950
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.260908725276405e-05,
+      "loss": 0.1625,
+      "step": 3955
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.255528709162649e-05,
+      "loss": 0.1679,
+      "step": 3960
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.2501498359172628e-05,
+      "loss": 0.1683,
+      "step": 3965
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.24477213068571e-05,
+      "loss": 0.1596,
+      "step": 3970
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.239395618607996e-05,
+      "loss": 0.1669,
+      "step": 3975
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.234020324818546e-05,
+      "loss": 0.1619,
+      "step": 3980
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.22864627444609e-05,
+      "loss": 0.1683,
+      "step": 3985
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.2232734926135475e-05,
+      "loss": 0.1591,
+      "step": 3990
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.217902004437904e-05,
+      "loss": 0.1644,
+      "step": 3995
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.2125318350301032e-05,
+      "loss": 0.1675,
+      "step": 4000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.2071630094949157e-05,
+      "loss": 0.164,
+      "step": 4005
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.201795552930838e-05,
+      "loss": 0.1627,
+      "step": 4010
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.1964294904299607e-05,
+      "loss": 0.167,
+      "step": 4015
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.1910648470778606e-05,
+      "loss": 0.1615,
+      "step": 4020
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.1857016479534786e-05,
+      "loss": 0.1699,
+      "step": 4025
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.1803399181290054e-05,
+      "loss": 0.1591,
+      "step": 4030
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.174979682669761e-05,
+      "loss": 0.1699,
+      "step": 4035
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.169620966634082e-05,
+      "loss": 0.1579,
+      "step": 4040
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.1642637950731996e-05,
+      "loss": 0.169,
+      "step": 4045
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.1589081930311263e-05,
+      "loss": 0.1615,
+      "step": 4050
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.1535541855445353e-05,
+      "loss": 0.1654,
+      "step": 4055
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.1482017976426502e-05,
+      "loss": 0.1618,
+      "step": 4060
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.1428510543471163e-05,
+      "loss": 0.1664,
+      "step": 4065
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.137501980671897e-05,
+      "loss": 0.1546,
+      "step": 4070
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.1321546016231457e-05,
+      "loss": 0.1619,
+      "step": 4075
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.1268089421990978e-05,
+      "loss": 0.1597,
+      "step": 4080
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.1214650273899462e-05,
+      "loss": 0.1613,
+      "step": 4085
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.1161228821777305e-05,
+      "loss": 0.1659,
+      "step": 4090
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.1107825315362153e-05,
+      "loss": 0.1637,
+      "step": 4095
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.1054440004307785e-05,
+      "loss": 0.1716,
+      "step": 4100
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.1001073138182896e-05,
+      "loss": 0.1698,
+      "step": 4105
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.0947724966469968e-05,
+      "loss": 0.1634,
+      "step": 4110
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.089439573856408e-05,
+      "loss": 0.1676,
+      "step": 4115
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.0841085703771766e-05,
+      "loss": 0.1686,
+      "step": 4120
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.0787795111309807e-05,
+      "loss": 0.1659,
+      "step": 4125
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.073452421030414e-05,
+      "loss": 0.1665,
+      "step": 4130
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.0681273249788587e-05,
+      "loss": 0.1655,
+      "step": 4135
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.062804247870382e-05,
+      "loss": 0.1634,
+      "step": 4140
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.0574832145896044e-05,
+      "loss": 0.1674,
+      "step": 4145
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.0521642500116013e-05,
+      "loss": 0.1603,
+      "step": 4150
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.046847379001769e-05,
+      "loss": 0.1638,
+      "step": 4155
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.0415326264157225e-05,
+      "loss": 0.1628,
+      "step": 4160
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.0362200170991696e-05,
+      "loss": 0.1618,
+      "step": 4165
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.0309095758878005e-05,
+      "loss": 0.1611,
+      "step": 4170
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.02560132760717e-05,
+      "loss": 0.1585,
+      "step": 4175
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.0202952970725797e-05,
+      "loss": 0.1735,
+      "step": 4180
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.0149915090889653e-05,
+      "loss": 0.16,
+      "step": 4185
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.0096899884507777e-05,
+      "loss": 0.153,
+      "step": 4190
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.0043907599418693e-05,
+      "loss": 0.1662,
+      "step": 4195
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.999093848335376e-05,
+      "loss": 0.1595,
+      "step": 4200
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.9937992783936036e-05,
+      "loss": 0.1672,
+      "step": 4205
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.9885070748679088e-05,
+      "loss": 0.1643,
+      "step": 4210
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.9832172624985906e-05,
+      "loss": 0.1633,
+      "step": 4215
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.9779298660147623e-05,
+      "loss": 0.1641,
+      "step": 4220
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.972644910134251e-05,
+      "loss": 0.1582,
+      "step": 4225
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.9673624195634676e-05,
+      "loss": 0.1629,
+      "step": 4230
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.9620824189973037e-05,
+      "loss": 0.162,
+      "step": 4235
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.9568049331190056e-05,
+      "loss": 0.1568,
+      "step": 4240
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.9515299866000678e-05,
+      "loss": 0.1613,
+      "step": 4245
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.9462576041001108e-05,
+      "loss": 0.1549,
+      "step": 4250
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.9409878102667707e-05,
+      "loss": 0.1648,
+      "step": 4255
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.9357206297355796e-05,
+      "loss": 0.1571,
+      "step": 4260
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.930456087129856e-05,
+      "loss": 0.1611,
+      "step": 4265
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.9251942070605837e-05,
+      "loss": 0.1613,
+      "step": 4270
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.919935014126301e-05,
+      "loss": 0.1646,
+      "step": 4275
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.9146785329129824e-05,
+      "loss": 0.1543,
+      "step": 4280
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.9094247879939296e-05,
+      "loss": 0.1605,
+      "step": 4285
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.9041738039296466e-05,
+      "loss": 0.1675,
+      "step": 4290
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.8989256052677363e-05,
+      "loss": 0.167,
+      "step": 4295
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.8936802165427764e-05,
+      "loss": 0.1615,
+      "step": 4300
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.8884376622762115e-05,
+      "loss": 0.1653,
+      "step": 4305
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.883197966976233e-05,
+      "loss": 0.1684,
+      "step": 4310
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.8779611551376686e-05,
+      "loss": 0.1627,
+      "step": 4315
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.872727251241865e-05,
+      "loss": 0.1616,
+      "step": 4320
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.8674962797565765e-05,
+      "loss": 0.1611,
+      "step": 4325
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.8622682651358468e-05,
+      "loss": 0.1624,
+      "step": 4330
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8570432318198984e-05,
+      "loss": 0.1624,
+      "step": 4335
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.851821204235015e-05,
+      "loss": 0.1632,
+      "step": 4340
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8466022067934308e-05,
+      "loss": 0.1629,
+      "step": 4345
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8413862638932124e-05,
+      "loss": 0.1568,
+      "step": 4350
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.836173399918148e-05,
+      "loss": 0.1565,
+      "step": 4355
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.830963639237633e-05,
+      "loss": 0.1638,
+      "step": 4360
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8257570062065534e-05,
+      "loss": 0.1573,
+      "step": 4365
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8205535251651767e-05,
+      "loss": 0.1576,
+      "step": 4370
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.815353220439031e-05,
+      "loss": 0.1525,
+      "step": 4375
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.810156116338802e-05,
+      "loss": 0.164,
+      "step": 4380
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.804962237160205e-05,
+      "loss": 0.1625,
+      "step": 4385
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.7997716071838862e-05,
+      "loss": 0.1647,
+      "step": 4390
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.7945842506752984e-05,
+      "loss": 0.1593,
+      "step": 4395
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.7894001918845926e-05,
+      "loss": 0.1628,
+      "step": 4400
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.784219455046503e-05,
+      "loss": 0.161,
+      "step": 4405
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.7790420643802353e-05,
+      "loss": 0.1681,
+      "step": 4410
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.7738680440893502e-05,
+      "loss": 0.1515,
+      "step": 4415
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.7686974183616544e-05,
+      "loss": 0.1624,
+      "step": 4420
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.7635302113690845e-05,
+      "loss": 0.1663,
+      "step": 4425
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.7583664472675952e-05,
+      "loss": 0.1546,
+      "step": 4430
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.753206150197046e-05,
+      "loss": 0.1523,
+      "step": 4435
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.74804934428109e-05,
+      "loss": 0.1617,
+      "step": 4440
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.7428960536270555e-05,
+      "loss": 0.1647,
+      "step": 4445
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.737746302325844e-05,
+      "loss": 0.1584,
+      "step": 4450
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.7326001144518043e-05,
+      "loss": 0.1537,
+      "step": 4455
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.7274575140626318e-05,
+      "loss": 0.1652,
+      "step": 4460
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.7223185251992475e-05,
+      "loss": 0.1638,
+      "step": 4465
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.717183171885691e-05,
+      "loss": 0.1518,
+      "step": 4470
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.7120514781290053e-05,
+      "loss": 0.156,
+      "step": 4475
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.7069234679191267e-05,
+      "loss": 0.1615,
+      "step": 4480
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.7017991652287693e-05,
+      "loss": 0.176,
+      "step": 4485
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.6966785940133175e-05,
+      "loss": 0.1587,
+      "step": 4490
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.6915617782107094e-05,
+      "loss": 0.15,
+      "step": 4495
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.6864487417413288e-05,
+      "loss": 0.1642,
+      "step": 4500
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.6813395085078897e-05,
+      "loss": 0.1651,
+      "step": 4505
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.676234102395331e-05,
+      "loss": 0.16,
+      "step": 4510
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.6711325472706925e-05,
+      "loss": 0.1504,
+      "step": 4515
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.66603486698302e-05,
+      "loss": 0.1576,
+      "step": 4520
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.6609410853632382e-05,
+      "loss": 0.1637,
+      "step": 4525
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.65585122622405e-05,
+      "loss": 0.1582,
+      "step": 4530
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.6507653133598193e-05,
+      "loss": 0.1582,
+      "step": 4535
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.6456833705464626e-05,
+      "loss": 0.1551,
+      "step": 4540
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.640605421541337e-05,
+      "loss": 0.1581,
+      "step": 4545
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.635531490083128e-05,
+      "loss": 0.1538,
+      "step": 4550
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.6304615998917418e-05,
+      "loss": 0.1596,
+      "step": 4555
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.6253957746681897e-05,
+      "loss": 0.1642,
+      "step": 4560
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.6203340380944822e-05,
+      "loss": 0.1597,
+      "step": 4565
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.6152764138335138e-05,
+      "loss": 0.1619,
+      "step": 4570
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.610222925528957e-05,
+      "loss": 0.1584,
+      "step": 4575
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.605173596805146e-05,
+      "loss": 0.1591,
+      "step": 4580
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6001284512669736e-05,
+      "loss": 0.161,
+      "step": 4585
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.5950875124997724e-05,
+      "loss": 0.1509,
+      "step": 4590
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.590050804069214e-05,
+      "loss": 0.1548,
+      "step": 4595
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.585018349521188e-05,
+      "loss": 0.1621,
+      "step": 4600
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.579990172381703e-05,
+      "loss": 0.1556,
+      "step": 4605
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.5749662961567674e-05,
+      "loss": 0.1534,
+      "step": 4610
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.569946744332286e-05,
+      "loss": 0.1607,
+      "step": 4615
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.564931540373946e-05,
+      "loss": 0.1596,
+      "step": 4620
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.559920707727111e-05,
+      "loss": 0.1561,
+      "step": 4625
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.5549142698167064e-05,
+      "loss": 0.1612,
+      "step": 4630
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.5499122500471154e-05,
+      "loss": 0.16,
+      "step": 4635
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.544914671802065e-05,
+      "loss": 0.1532,
+      "step": 4640
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.5399215584445216e-05,
+      "loss": 0.1643,
+      "step": 4645
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.5349329333165748e-05,
+      "loss": 0.1646,
+      "step": 4650
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.529948819739337e-05,
+      "loss": 0.1597,
+      "step": 4655
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.5249692410128241e-05,
+      "loss": 0.1569,
+      "step": 4660
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.5199942204158596e-05,
+      "loss": 0.1568,
+      "step": 4665
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.5150237812059503e-05,
+      "loss": 0.1574,
+      "step": 4670
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.5100579466191928e-05,
+      "loss": 0.1572,
+      "step": 4675
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.505096739870152e-05,
+      "loss": 0.1574,
+      "step": 4680
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.5001401841517632e-05,
+      "loss": 0.1523,
+      "step": 4685
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.4951883026352149e-05,
+      "loss": 0.1602,
+      "step": 4690
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.490241118469847e-05,
+      "loss": 0.1593,
+      "step": 4695
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.4852986547830378e-05,
+      "loss": 0.1555,
+      "step": 4700
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.4803609346801004e-05,
+      "loss": 0.1615,
+      "step": 4705
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.4754279812441698e-05,
+      "loss": 0.1543,
+      "step": 4710
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.4704998175361001e-05,
+      "loss": 0.1465,
+      "step": 4715
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.4655764665943511e-05,
+      "loss": 0.1541,
+      "step": 4720
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.4606579514348878e-05,
+      "loss": 0.1554,
+      "step": 4725
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.4557442950510641e-05,
+      "loss": 0.1597,
+      "step": 4730
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.4508355204135227e-05,
+      "loss": 0.1598,
+      "step": 4735
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.4459316504700837e-05,
+      "loss": 0.1607,
+      "step": 4740
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.4410327081456376e-05,
+      "loss": 0.162,
+      "step": 4745
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.4361387163420423e-05,
+      "loss": 0.1654,
+      "step": 4750
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.4312496979380093e-05,
+      "loss": 0.156,
+      "step": 4755
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.426365675789002e-05,
+      "loss": 0.157,
+      "step": 4760
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.4214866727271247e-05,
+      "loss": 0.1575,
+      "step": 4765
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.4166127115610223e-05,
+      "loss": 0.1586,
+      "step": 4770
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.411743815075765e-05,
+      "loss": 0.159,
+      "step": 4775
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.4068800060327514e-05,
+      "loss": 0.1533,
+      "step": 4780
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.4020213071695903e-05,
+      "loss": 0.1583,
+      "step": 4785
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.3971677412000084e-05,
+      "loss": 0.1533,
+      "step": 4790
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.3923193308137306e-05,
+      "loss": 0.1512,
+      "step": 4795
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.3874760986763858e-05,
+      "loss": 0.1567,
+      "step": 4800
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.3826380674293907e-05,
+      "loss": 0.1607,
+      "step": 4805
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.377805259689851e-05,
+      "loss": 0.161,
+      "step": 4810
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.3729776980504516e-05,
+      "loss": 0.1564,
+      "step": 4815
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.3681554050793558e-05,
+      "loss": 0.1624,
+      "step": 4820
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.363338403320093e-05,
+      "loss": 0.1556,
+      "step": 4825
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.3585267152914594e-05,
+      "loss": 0.1544,
+      "step": 4830
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.3537203634874074e-05,
+      "loss": 0.1563,
+      "step": 4835
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.3489193703769476e-05,
+      "loss": 0.1499,
+      "step": 4840
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.3441237584040351e-05,
+      "loss": 0.153,
+      "step": 4845
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.3393335499874742e-05,
+      "loss": 0.1562,
+      "step": 4850
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.334548767520801e-05,
+      "loss": 0.1564,
+      "step": 4855
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.329769433372193e-05,
+      "loss": 0.1599,
+      "step": 4860
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.324995569884353e-05,
+      "loss": 0.1555,
+      "step": 4865
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.3202271993744129e-05,
+      "loss": 0.1544,
+      "step": 4870
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.3154643441338227e-05,
+      "loss": 0.1614,
+      "step": 4875
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.3107070264282506e-05,
+      "loss": 0.1599,
+      "step": 4880
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.3059552684974763e-05,
+      "loss": 0.1555,
+      "step": 4885
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.301209092555292e-05,
+      "loss": 0.1555,
+      "step": 4890
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.2964685207893907e-05,
+      "loss": 0.1592,
+      "step": 4895
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.2917335753612686e-05,
+      "loss": 0.1532,
+      "step": 4900
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 7365,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "total_flos": 1.5165757447300841e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}