diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4722 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999201239676023,
+  "eval_steps": 500,
+  "global_step": 7824,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.385696040868455e-07,
+      "loss": 9.3159,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.277139208173691e-06,
+      "loss": 8.4375,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9157088122605362e-06,
+      "loss": 3.9338,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.554278416347382e-06,
+      "loss": 1.0118,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1928480204342275e-06,
+      "loss": 0.8395,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8314176245210725e-06,
+      "loss": 0.7496,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469987228607918e-06,
+      "loss": 0.7392,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.108556832694764e-06,
+      "loss": 0.7435,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.747126436781609e-06,
+      "loss": 0.7452,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.385696040868455e-06,
+      "loss": 0.7477,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.024265644955301e-06,
+      "loss": 0.7422,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.662835249042145e-06,
+      "loss": 0.7743,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.301404853128992e-06,
+      "loss": 0.7389,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.939974457215837e-06,
+      "loss": 0.7386,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.578544061302683e-06,
+      "loss": 0.7448,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0217113665389528e-05,
+      "loss": 0.733,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0855683269476373e-05,
+      "loss": 0.7533,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1494252873563218e-05,
+      "loss": 0.7424,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2132822477650065e-05,
+      "loss": 0.7343,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.277139208173691e-05,
+      "loss": 0.7555,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.3409961685823755e-05,
+      "loss": 0.7502,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.4048531289910602e-05,
+      "loss": 0.7448,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.4687100893997447e-05,
+      "loss": 0.7469,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.532567049808429e-05,
+      "loss": 0.7501,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.596424010217114e-05,
+      "loss": 0.7382,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.6602809706257983e-05,
+      "loss": 0.7283,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.7241379310344828e-05,
+      "loss": 0.7342,
+      "step": 270
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.7879948914431673e-05,
+      "loss": 0.7612,
+      "step": 280
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.72,
+      "step": 290
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9157088122605367e-05,
+      "loss": 0.7448,
+      "step": 300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.979565772669221e-05,
+      "loss": 0.7642,
+      "step": 310
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.0434227330779057e-05,
+      "loss": 0.7369,
+      "step": 320
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.10727969348659e-05,
+      "loss": 0.7464,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.1711366538952747e-05,
+      "loss": 0.7398,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.234993614303959e-05,
+      "loss": 0.7369,
+      "step": 350
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.2988505747126437e-05,
+      "loss": 0.722,
+      "step": 360
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.3627075351213285e-05,
+      "loss": 0.7598,
+      "step": 370
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.426564495530013e-05,
+      "loss": 0.7592,
+      "step": 380
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.4904214559386975e-05,
+      "loss": 0.7506,
+      "step": 390
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.554278416347382e-05,
+      "loss": 0.7526,
+      "step": 400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.618135376756067e-05,
+      "loss": 0.7527,
+      "step": 410
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.681992337164751e-05,
+      "loss": 0.7702,
+      "step": 420
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.745849297573436e-05,
+      "loss": 0.7481,
+      "step": 430
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.8097062579821203e-05,
+      "loss": 0.7456,
+      "step": 440
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.8735632183908045e-05,
+      "loss": 0.7528,
+      "step": 450
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.9374201787994893e-05,
+      "loss": 0.7353,
+      "step": 460
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.0012771392081738e-05,
+      "loss": 0.78,
+      "step": 470
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.065134099616858e-05,
+      "loss": 0.7394,
+      "step": 480
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.128991060025543e-05,
+      "loss": 0.7464,
+      "step": 490
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.192848020434228e-05,
+      "loss": 0.7401,
+      "step": 500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.256704980842912e-05,
+      "loss": 0.7561,
+      "step": 510
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.3205619412515967e-05,
+      "loss": 0.7509,
+      "step": 520
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.3844189016602815e-05,
+      "loss": 0.7687,
+      "step": 530
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.4482758620689657e-05,
+      "loss": 0.7568,
+      "step": 540
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.51213282247765e-05,
+      "loss": 0.759,
+      "step": 550
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.5759897828863347e-05,
+      "loss": 0.7655,
+      "step": 560
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.6398467432950195e-05,
+      "loss": 0.7691,
+      "step": 570
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.7619,
+      "step": 580
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.7675606641123885e-05,
+      "loss": 0.7558,
+      "step": 590
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.831417624521073e-05,
+      "loss": 0.7387,
+      "step": 600
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.8952745849297575e-05,
+      "loss": 0.7463,
+      "step": 610
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 3.959131545338442e-05,
+      "loss": 0.7437,
+      "step": 620
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.0229885057471265e-05,
+      "loss": 0.7572,
+      "step": 630
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.086845466155811e-05,
+      "loss": 0.7478,
+      "step": 640
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.1507024265644955e-05,
+      "loss": 0.7301,
+      "step": 650
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.21455938697318e-05,
+      "loss": 0.7689,
+      "step": 660
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.278416347381865e-05,
+      "loss": 0.774,
+      "step": 670
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.342273307790549e-05,
+      "loss": 0.791,
+      "step": 680
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.406130268199234e-05,
+      "loss": 0.7642,
+      "step": 690
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.469987228607918e-05,
+      "loss": 0.7622,
+      "step": 700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.5338441890166025e-05,
+      "loss": 0.7601,
+      "step": 710
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.597701149425287e-05,
+      "loss": 0.7694,
+      "step": 720
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.661558109833972e-05,
+      "loss": 0.7641,
+      "step": 730
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.725415070242657e-05,
+      "loss": 0.7724,
+      "step": 740
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.789272030651341e-05,
+      "loss": 0.7576,
+      "step": 750
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.853128991060026e-05,
+      "loss": 0.7803,
+      "step": 760
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.916985951468711e-05,
+      "loss": 0.7796,
+      "step": 770
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.980842911877395e-05,
+      "loss": 0.7445,
+      "step": 780
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.995029115182503e-05,
+      "loss": 0.7646,
+      "step": 790
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9879278511575064e-05,
+      "loss": 0.779,
+      "step": 800
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.98082658713251e-05,
+      "loss": 0.7788,
+      "step": 810
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.973725323107513e-05,
+      "loss": 0.763,
+      "step": 820
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.966624059082517e-05,
+      "loss": 0.7842,
+      "step": 830
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9595227950575206e-05,
+      "loss": 0.789,
+      "step": 840
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.952421531032524e-05,
+      "loss": 0.7753,
+      "step": 850
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.945320267007528e-05,
+      "loss": 0.7613,
+      "step": 860
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9382190029825306e-05,
+      "loss": 0.7671,
+      "step": 870
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.931117738957535e-05,
+      "loss": 0.7752,
+      "step": 880
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9240164749325384e-05,
+      "loss": 0.7732,
+      "step": 890
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.916915210907542e-05,
+      "loss": 0.7685,
+      "step": 900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9098139468825455e-05,
+      "loss": 0.7711,
+      "step": 910
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9027126828575484e-05,
+      "loss": 0.7659,
+      "step": 920
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8956114188325526e-05,
+      "loss": 0.7941,
+      "step": 930
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.888510154807556e-05,
+      "loss": 0.7545,
+      "step": 940
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.88140889078256e-05,
+      "loss": 0.7846,
+      "step": 950
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.874307626757563e-05,
+      "loss": 0.7642,
+      "step": 960
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.867206362732566e-05,
+      "loss": 0.7496,
+      "step": 970
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8601050987075704e-05,
+      "loss": 0.7763,
+      "step": 980
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.853003834682573e-05,
+      "loss": 0.7835,
+      "step": 990
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8459025706575775e-05,
+      "loss": 0.7607,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.838801306632581e-05,
+      "loss": 0.7646,
+      "step": 1010
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.831700042607584e-05,
+      "loss": 0.7992,
+      "step": 1020
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.824598778582588e-05,
+      "loss": 0.771,
+      "step": 1030
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.817497514557591e-05,
+      "loss": 0.7834,
+      "step": 1040
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.810396250532595e-05,
+      "loss": 0.7796,
+      "step": 1050
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.803294986507599e-05,
+      "loss": 0.7689,
+      "step": 1060
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.796193722482602e-05,
+      "loss": 0.7945,
+      "step": 1070
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.789092458457606e-05,
+      "loss": 0.7719,
+      "step": 1080
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.781991194432609e-05,
+      "loss": 0.744,
+      "step": 1090
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.774889930407613e-05,
+      "loss": 0.7855,
+      "step": 1100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7677886663826166e-05,
+      "loss": 0.7597,
+      "step": 1110
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7606874023576195e-05,
+      "loss": 0.784,
+      "step": 1120
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.753586138332624e-05,
+      "loss": 0.7669,
+      "step": 1130
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7464848743076266e-05,
+      "loss": 0.748,
+      "step": 1140
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.739383610282631e-05,
+      "loss": 0.7783,
+      "step": 1150
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7322823462576344e-05,
+      "loss": 0.7464,
+      "step": 1160
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.725181082232637e-05,
+      "loss": 0.7627,
+      "step": 1170
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7180798182076415e-05,
+      "loss": 0.7589,
+      "step": 1180
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7109785541826444e-05,
+      "loss": 0.7478,
+      "step": 1190
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7038772901576486e-05,
+      "loss": 0.7418,
+      "step": 1200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.696776026132652e-05,
+      "loss": 0.7598,
+      "step": 1210
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.689674762107655e-05,
+      "loss": 0.7802,
+      "step": 1220
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.682573498082659e-05,
+      "loss": 0.7706,
+      "step": 1230
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.675472234057662e-05,
+      "loss": 0.7475,
+      "step": 1240
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.6683709700326664e-05,
+      "loss": 0.7633,
+      "step": 1250
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.661269706007669e-05,
+      "loss": 0.7681,
+      "step": 1260
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.654168441982673e-05,
+      "loss": 0.7815,
+      "step": 1270
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.647067177957677e-05,
+      "loss": 0.7444,
+      "step": 1280
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.63996591393268e-05,
+      "loss": 0.8014,
+      "step": 1290
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.632864649907684e-05,
+      "loss": 0.7483,
+      "step": 1300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.625763385882687e-05,
+      "loss": 0.731,
+      "step": 1310
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.6186621218576906e-05,
+      "loss": 0.7478,
+      "step": 1320
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.611560857832695e-05,
+      "loss": 0.7595,
+      "step": 1330
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.604459593807698e-05,
+      "loss": 0.7513,
+      "step": 1340
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.597358329782702e-05,
+      "loss": 0.7762,
+      "step": 1350
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.590257065757705e-05,
+      "loss": 0.7686,
+      "step": 1360
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.5831558017327084e-05,
+      "loss": 0.7642,
+      "step": 1370
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.5760545377077126e-05,
+      "loss": 0.7674,
+      "step": 1380
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.5689532736827155e-05,
+      "loss": 0.7493,
+      "step": 1390
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.56185200965772e-05,
+      "loss": 0.7537,
+      "step": 1400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.5547507456327226e-05,
+      "loss": 0.7671,
+      "step": 1410
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.547649481607726e-05,
+      "loss": 0.7573,
+      "step": 1420
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.5405482175827304e-05,
+      "loss": 0.777,
+      "step": 1430
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.533446953557733e-05,
+      "loss": 0.7624,
+      "step": 1440
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.5263456895327375e-05,
+      "loss": 0.7643,
+      "step": 1450
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.5192444255077404e-05,
+      "loss": 0.7413,
+      "step": 1460
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.512143161482744e-05,
+      "loss": 0.7338,
+      "step": 1470
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.505041897457748e-05,
+      "loss": 0.7537,
+      "step": 1480
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.497940633432751e-05,
+      "loss": 0.7592,
+      "step": 1490
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.4908393694077546e-05,
+      "loss": 0.7529,
+      "step": 1500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.483738105382758e-05,
+      "loss": 0.7532,
+      "step": 1510
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.476636841357762e-05,
+      "loss": 0.7747,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.469535577332765e-05,
+      "loss": 0.7623,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.462434313307769e-05,
+      "loss": 0.7427,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.4553330492827724e-05,
+      "loss": 0.7517,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.448231785257776e-05,
+      "loss": 0.7723,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.4411305212327795e-05,
+      "loss": 0.7476,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.434029257207783e-05,
+      "loss": 0.7341,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.4269279931827866e-05,
+      "loss": 0.764,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.41982672915779e-05,
+      "loss": 0.7538,
+      "step": 1600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.412725465132794e-05,
+      "loss": 0.7351,
+      "step": 1610
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.405624201107797e-05,
+      "loss": 0.7314,
+      "step": 1620
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.398522937082801e-05,
+      "loss": 0.7629,
+      "step": 1630
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.3914216730578044e-05,
+      "loss": 0.7572,
+      "step": 1640
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.384320409032808e-05,
+      "loss": 0.73,
+      "step": 1650
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.3772191450078115e-05,
+      "loss": 0.7676,
+      "step": 1660
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.370117880982815e-05,
+      "loss": 0.7674,
+      "step": 1670
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.3630166169578186e-05,
+      "loss": 0.7647,
+      "step": 1680
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.355915352932822e-05,
+      "loss": 0.7528,
+      "step": 1690
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.348814088907826e-05,
+      "loss": 0.7442,
+      "step": 1700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.341712824882829e-05,
+      "loss": 0.7506,
+      "step": 1710
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.334611560857833e-05,
+      "loss": 0.7435,
+      "step": 1720
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.3275102968328364e-05,
+      "loss": 0.7419,
+      "step": 1730
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.32040903280784e-05,
+      "loss": 0.7401,
+      "step": 1740
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.3133077687828435e-05,
+      "loss": 0.7359,
+      "step": 1750
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.306206504757847e-05,
+      "loss": 0.7528,
+      "step": 1760
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.2991052407328506e-05,
+      "loss": 0.7629,
+      "step": 1770
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.292003976707854e-05,
+      "loss": 0.7386,
+      "step": 1780
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.284902712682858e-05,
+      "loss": 0.746,
+      "step": 1790
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.277801448657861e-05,
+      "loss": 0.7287,
+      "step": 1800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.270700184632865e-05,
+      "loss": 0.7499,
+      "step": 1810
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.2635989206078684e-05,
+      "loss": 0.7369,
+      "step": 1820
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.256497656582872e-05,
+      "loss": 0.7407,
+      "step": 1830
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.2493963925578755e-05,
+      "loss": 0.7536,
+      "step": 1840
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.242295128532879e-05,
+      "loss": 0.747,
+      "step": 1850
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.2351938645078826e-05,
+      "loss": 0.7301,
+      "step": 1860
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.228092600482886e-05,
+      "loss": 0.7436,
+      "step": 1870
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.22099133645789e-05,
+      "loss": 0.7572,
+      "step": 1880
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.213890072432893e-05,
+      "loss": 0.7345,
+      "step": 1890
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.206788808407897e-05,
+      "loss": 0.743,
+      "step": 1900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.1996875443829004e-05,
+      "loss": 0.7416,
+      "step": 1910
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.192586280357904e-05,
+      "loss": 0.7489,
+      "step": 1920
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.1854850163329075e-05,
+      "loss": 0.7215,
+      "step": 1930
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.178383752307911e-05,
+      "loss": 0.7523,
+      "step": 1940
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.1712824882829146e-05,
+      "loss": 0.7274,
+      "step": 1950
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.164181224257918e-05,
+      "loss": 0.7483,
+      "step": 1960
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.157079960232922e-05,
+      "loss": 0.7417,
+      "step": 1970
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.149978696207925e-05,
+      "loss": 0.741,
+      "step": 1980
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.142877432182929e-05,
+      "loss": 0.7498,
+      "step": 1990
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.1357761681579324e-05,
+      "loss": 0.7368,
+      "step": 2000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.128674904132936e-05,
+      "loss": 0.7318,
+      "step": 2010
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.1215736401079395e-05,
+      "loss": 0.7341,
+      "step": 2020
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.114472376082943e-05,
+      "loss": 0.7304,
+      "step": 2030
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.1073711120579466e-05,
+      "loss": 0.7182,
+      "step": 2040
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.10026984803295e-05,
+      "loss": 0.7311,
+      "step": 2050
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.093168584007954e-05,
+      "loss": 0.7248,
+      "step": 2060
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.0860673199829566e-05,
+      "loss": 0.7125,
+      "step": 2070
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.078966055957961e-05,
+      "loss": 0.7476,
+      "step": 2080
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.0718647919329644e-05,
+      "loss": 0.7643,
+      "step": 2090
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.064763527907968e-05,
+      "loss": 0.7373,
+      "step": 2100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.0576622638829715e-05,
+      "loss": 0.7433,
+      "step": 2110
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.0505609998579744e-05,
+      "loss": 0.7409,
+      "step": 2120
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.0434597358329786e-05,
+      "loss": 0.7417,
+      "step": 2130
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.036358471807982e-05,
+      "loss": 0.749,
+      "step": 2140
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.029257207782986e-05,
+      "loss": 0.7511,
+      "step": 2150
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.022155943757989e-05,
+      "loss": 0.7466,
+      "step": 2160
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.015054679732992e-05,
+      "loss": 0.7455,
+      "step": 2170
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.0079534157079964e-05,
+      "loss": 0.74,
+      "step": 2180
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.000852151683e-05,
+      "loss": 0.7205,
+      "step": 2190
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.9937508876580035e-05,
+      "loss": 0.7404,
+      "step": 2200
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.986649623633007e-05,
+      "loss": 0.7334,
+      "step": 2210
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.97954835960801e-05,
+      "loss": 0.7365,
+      "step": 2220
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.972447095583014e-05,
+      "loss": 0.7198,
+      "step": 2230
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.965345831558018e-05,
+      "loss": 0.7547,
+      "step": 2240
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.958244567533021e-05,
+      "loss": 0.7407,
+      "step": 2250
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.951143303508025e-05,
+      "loss": 0.7287,
+      "step": 2260
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.944042039483028e-05,
+      "loss": 0.7335,
+      "step": 2270
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.936940775458032e-05,
+      "loss": 0.7389,
+      "step": 2280
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.9298395114330355e-05,
+      "loss": 0.7366,
+      "step": 2290
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.922738247408039e-05,
+      "loss": 0.7554,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.9156369833830426e-05,
+      "loss": 0.7431,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.9085357193580455e-05,
+      "loss": 0.7115,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.90143445533305e-05,
+      "loss": 0.7131,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.894333191308053e-05,
+      "loss": 0.7364,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.887231927283057e-05,
+      "loss": 0.7224,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.8801306632580604e-05,
+      "loss": 0.7222,
+      "step": 2360
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.873029399233063e-05,
+      "loss": 0.7259,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.8659281352080675e-05,
+      "loss": 0.721,
+      "step": 2380
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.8588268711830704e-05,
+      "loss": 0.7396,
+      "step": 2390
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.8517256071580746e-05,
+      "loss": 0.7323,
+      "step": 2400
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.844624343133078e-05,
+      "loss": 0.7137,
+      "step": 2410
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.837523079108081e-05,
+      "loss": 0.7248,
+      "step": 2420
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.830421815083085e-05,
+      "loss": 0.7102,
+      "step": 2430
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.823320551058088e-05,
+      "loss": 0.7188,
+      "step": 2440
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.8162192870330924e-05,
+      "loss": 0.7301,
+      "step": 2450
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.809118023008096e-05,
+      "loss": 0.7042,
+      "step": 2460
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.802016758983099e-05,
+      "loss": 0.7114,
+      "step": 2470
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.794915494958103e-05,
+      "loss": 0.7383,
+      "step": 2480
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.787814230933106e-05,
+      "loss": 0.7193,
+      "step": 2490
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.7807129669081095e-05,
+      "loss": 0.7426,
+      "step": 2500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.773611702883114e-05,
+      "loss": 0.7068,
+      "step": 2510
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.7665104388581166e-05,
+      "loss": 0.7152,
+      "step": 2520
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.759409174833121e-05,
+      "loss": 0.7279,
+      "step": 2530
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.752307910808124e-05,
+      "loss": 0.7172,
+      "step": 2540
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.745206646783127e-05,
+      "loss": 0.7385,
+      "step": 2550
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.7381053827581315e-05,
+      "loss": 0.7195,
+      "step": 2560
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.7310041187331344e-05,
+      "loss": 0.7259,
+      "step": 2570
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.7239028547081386e-05,
+      "loss": 0.7472,
+      "step": 2580
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.7168015906831415e-05,
+      "loss": 0.7215,
+      "step": 2590
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.709700326658145e-05,
+      "loss": 0.7173,
+      "step": 2600
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.702599062633149e-05,
+      "loss": 0.7024,
+      "step": 2610
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.695497798608152e-05,
+      "loss": 0.7162,
+      "step": 2620
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.6883965345831564e-05,
+      "loss": 0.715,
+      "step": 2630
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.681295270558159e-05,
+      "loss": 0.7198,
+      "step": 2640
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.674194006533163e-05,
+      "loss": 0.7144,
+      "step": 2650
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.6670927425081664e-05,
+      "loss": 0.7269,
+      "step": 2660
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.65999147848317e-05,
+      "loss": 0.7196,
+      "step": 2670
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.652890214458174e-05,
+      "loss": 0.7251,
+      "step": 2680
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.645788950433177e-05,
+      "loss": 0.7044,
+      "step": 2690
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.6386876864081806e-05,
+      "loss": 0.7492,
+      "step": 2700
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.631586422383184e-05,
+      "loss": 0.7418,
+      "step": 2710
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.624485158358188e-05,
+      "loss": 0.7326,
+      "step": 2720
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.617383894333192e-05,
+      "loss": 0.7113,
+      "step": 2730
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.610282630308195e-05,
+      "loss": 0.715,
+      "step": 2740
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.6031813662831984e-05,
+      "loss": 0.7124,
+      "step": 2750
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.596080102258202e-05,
+      "loss": 0.7297,
+      "step": 2760
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.5889788382332055e-05,
+      "loss": 0.7335,
+      "step": 2770
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.58187757420821e-05,
+      "loss": 0.6997,
+      "step": 2780
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.5747763101832126e-05,
+      "loss": 0.7187,
+      "step": 2790
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.567675046158216e-05,
+      "loss": 0.7192,
+      "step": 2800
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.56057378213322e-05,
+      "loss": 0.6959,
+      "step": 2810
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.553472518108223e-05,
+      "loss": 0.7402,
+      "step": 2820
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.5463712540832275e-05,
+      "loss": 0.7132,
+      "step": 2830
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.5392699900582304e-05,
+      "loss": 0.7282,
+      "step": 2840
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.532168726033234e-05,
+      "loss": 0.7315,
+      "step": 2850
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.5250674620082375e-05,
+      "loss": 0.7325,
+      "step": 2860
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.517966197983241e-05,
+      "loss": 0.7135,
+      "step": 2870
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.510864933958245e-05,
+      "loss": 0.7064,
+      "step": 2880
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.503763669933248e-05,
+      "loss": 0.7245,
+      "step": 2890
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.496662405908252e-05,
+      "loss": 0.7101,
+      "step": 2900
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.489561141883255e-05,
+      "loss": 0.7231,
+      "step": 2910
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.482459877858259e-05,
+      "loss": 0.7259,
+      "step": 2920
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.4753586138332624e-05,
+      "loss": 0.7084,
+      "step": 2930
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.468257349808266e-05,
+      "loss": 0.7271,
+      "step": 2940
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.4611560857832695e-05,
+      "loss": 0.7395,
+      "step": 2950
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.454054821758273e-05,
+      "loss": 0.7062,
+      "step": 2960
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.4469535577332766e-05,
+      "loss": 0.7237,
+      "step": 2970
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.43985229370828e-05,
+      "loss": 0.7218,
+      "step": 2980
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.432751029683284e-05,
+      "loss": 0.7228,
+      "step": 2990
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.425649765658287e-05,
+      "loss": 0.7204,
+      "step": 3000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.418548501633291e-05,
+      "loss": 0.7217,
+      "step": 3010
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.4114472376082944e-05,
+      "loss": 0.7272,
+      "step": 3020
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.404345973583298e-05,
+      "loss": 0.6991,
+      "step": 3030
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.3972447095583015e-05,
+      "loss": 0.7208,
+      "step": 3040
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.390143445533305e-05,
+      "loss": 0.7424,
+      "step": 3050
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.3830421815083086e-05,
+      "loss": 0.7233,
+      "step": 3060
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.375940917483312e-05,
+      "loss": 0.7086,
+      "step": 3070
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.368839653458316e-05,
+      "loss": 0.7281,
+      "step": 3080
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.361738389433319e-05,
+      "loss": 0.7297,
+      "step": 3090
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.354637125408323e-05,
+      "loss": 0.7326,
+      "step": 3100
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.3475358613833264e-05,
+      "loss": 0.7141,
+      "step": 3110
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.34043459735833e-05,
+      "loss": 0.7073,
+      "step": 3120
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.7012,
+      "step": 3130
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.326232069308337e-05,
+      "loss": 0.7241,
+      "step": 3140
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.3191308052833406e-05,
+      "loss": 0.7123,
+      "step": 3150
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.312029541258344e-05,
+      "loss": 0.7181,
+      "step": 3160
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.304928277233348e-05,
+      "loss": 0.731,
+      "step": 3170
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.297827013208351e-05,
+      "loss": 0.7242,
+      "step": 3180
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.290725749183355e-05,
+      "loss": 0.7139,
+      "step": 3190
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.2836244851583584e-05,
+      "loss": 0.7049,
+      "step": 3200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.276523221133362e-05,
+      "loss": 0.7306,
+      "step": 3210
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.2694219571083655e-05,
+      "loss": 0.7236,
+      "step": 3220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.262320693083369e-05,
+      "loss": 0.6916,
+      "step": 3230
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.2552194290583726e-05,
+      "loss": 0.7358,
+      "step": 3240
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.248118165033376e-05,
+      "loss": 0.7286,
+      "step": 3250
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.24101690100838e-05,
+      "loss": 0.7039,
+      "step": 3260
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.233915636983383e-05,
+      "loss": 0.7218,
+      "step": 3270
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.226814372958387e-05,
+      "loss": 0.7268,
+      "step": 3280
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.2197131089333904e-05,
+      "loss": 0.7107,
+      "step": 3290
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.212611844908394e-05,
+      "loss": 0.7112,
+      "step": 3300
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.2055105808833975e-05,
+      "loss": 0.7175,
+      "step": 3310
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.198409316858401e-05,
+      "loss": 0.7226,
+      "step": 3320
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.1913080528334046e-05,
+      "loss": 0.7058,
+      "step": 3330
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.184206788808408e-05,
+      "loss": 0.7191,
+      "step": 3340
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.177105524783412e-05,
+      "loss": 0.6697,
+      "step": 3350
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.170004260758415e-05,
+      "loss": 0.6974,
+      "step": 3360
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.162902996733419e-05,
+      "loss": 0.7192,
+      "step": 3370
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.1558017327084224e-05,
+      "loss": 0.7083,
+      "step": 3380
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.148700468683426e-05,
+      "loss": 0.7262,
+      "step": 3390
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.1415992046584295e-05,
+      "loss": 0.7107,
+      "step": 3400
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.134497940633433e-05,
+      "loss": 0.717,
+      "step": 3410
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.1273966766084366e-05,
+      "loss": 0.708,
+      "step": 3420
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.12029541258344e-05,
+      "loss": 0.6808,
+      "step": 3430
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.113194148558444e-05,
+      "loss": 0.7085,
+      "step": 3440
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.106092884533447e-05,
+      "loss": 0.7232,
+      "step": 3450
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.098991620508451e-05,
+      "loss": 0.6909,
+      "step": 3460
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.091890356483454e-05,
+      "loss": 0.7017,
+      "step": 3470
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 3.084789092458458e-05,
+      "loss": 0.7178,
+      "step": 3480
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.0776878284334615e-05,
+      "loss": 0.7212,
+      "step": 3490
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.0705865644084644e-05,
+      "loss": 0.7242,
+      "step": 3500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.0634853003834686e-05,
+      "loss": 0.6952,
+      "step": 3510
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.0563840363584715e-05,
+      "loss": 0.6957,
+      "step": 3520
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.0492827723334754e-05,
+      "loss": 0.6831,
+      "step": 3530
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.0421815083084793e-05,
+      "loss": 0.6967,
+      "step": 3540
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.0350802442834825e-05,
+      "loss": 0.7292,
+      "step": 3550
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.0279789802584864e-05,
+      "loss": 0.6992,
+      "step": 3560
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.0208777162334896e-05,
+      "loss": 0.7154,
+      "step": 3570
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.013776452208493e-05,
+      "loss": 0.6801,
+      "step": 3580
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 3.006675188183497e-05,
+      "loss": 0.6921,
+      "step": 3590
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.9995739241585003e-05,
+      "loss": 0.6887,
+      "step": 3600
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.992472660133504e-05,
+      "loss": 0.6936,
+      "step": 3610
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.9853713961085074e-05,
+      "loss": 0.7014,
+      "step": 3620
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.978270132083511e-05,
+      "loss": 0.7101,
+      "step": 3630
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.9711688680585148e-05,
+      "loss": 0.7153,
+      "step": 3640
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.964067604033518e-05,
+      "loss": 0.7185,
+      "step": 3650
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.956966340008522e-05,
+      "loss": 0.6934,
+      "step": 3660
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.949865075983525e-05,
+      "loss": 0.6766,
+      "step": 3670
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.9427638119585287e-05,
+      "loss": 0.6801,
+      "step": 3680
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.9356625479335326e-05,
+      "loss": 0.7132,
+      "step": 3690
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.9285612839085358e-05,
+      "loss": 0.7118,
+      "step": 3700
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.9214600198835397e-05,
+      "loss": 0.6812,
+      "step": 3710
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.914358755858543e-05,
+      "loss": 0.6859,
+      "step": 3720
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.9072574918335465e-05,
+      "loss": 0.6849,
+      "step": 3730
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.9001562278085497e-05,
+      "loss": 0.6989,
+      "step": 3740
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8930549637835536e-05,
+      "loss": 0.6891,
+      "step": 3750
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8859536997585575e-05,
+      "loss": 0.7033,
+      "step": 3760
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8788524357335607e-05,
+      "loss": 0.7101,
+      "step": 3770
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8717511717085643e-05,
+      "loss": 0.73,
+      "step": 3780
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8646499076835675e-05,
+      "loss": 0.741,
+      "step": 3790
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8575486436585714e-05,
+      "loss": 0.6963,
+      "step": 3800
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8504473796335753e-05,
+      "loss": 0.7064,
+      "step": 3810
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8433461156085785e-05,
+      "loss": 0.7061,
+      "step": 3820
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.836244851583582e-05,
+      "loss": 0.6784,
+      "step": 3830
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8291435875585853e-05,
+      "loss": 0.6701,
+      "step": 3840
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.822042323533589e-05,
+      "loss": 0.7048,
+      "step": 3850
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.814941059508593e-05,
+      "loss": 0.7003,
+      "step": 3860
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8078397954835963e-05,
+      "loss": 0.7068,
+      "step": 3870
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8007385314585998e-05,
+      "loss": 0.6989,
+      "step": 3880
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.793637267433603e-05,
+      "loss": 0.7108,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.786536003408607e-05,
+      "loss": 0.6959,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.7794347393836108e-05,
+      "loss": 0.6825,
+      "step": 3910
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.772333475358614e-05,
+      "loss": 0.6964,
+      "step": 3920
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.7652322113336176e-05,
+      "loss": 0.6977,
+      "step": 3930
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.7581309473086208e-05,
+      "loss": 0.7007,
+      "step": 3940
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.7510296832836247e-05,
+      "loss": 0.6982,
+      "step": 3950
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7439284192586283e-05,
+      "loss": 0.6941,
+      "step": 3960
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7368271552336315e-05,
+      "loss": 0.7094,
+      "step": 3970
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7297258912086354e-05,
+      "loss": 0.7026,
+      "step": 3980
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7226246271836386e-05,
+      "loss": 0.702,
+      "step": 3990
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7155233631586425e-05,
+      "loss": 0.6997,
+      "step": 4000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.708422099133646e-05,
+      "loss": 0.6988,
+      "step": 4010
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7013208351086493e-05,
+      "loss": 0.7055,
+      "step": 4020
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.694219571083653e-05,
+      "loss": 0.6954,
+      "step": 4030
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.6871183070586564e-05,
+      "loss": 0.7034,
+      "step": 4040
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.6800170430336603e-05,
+      "loss": 0.6989,
+      "step": 4050
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.6729157790086635e-05,
+      "loss": 0.7173,
+      "step": 4060
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.665814514983667e-05,
+      "loss": 0.7075,
+      "step": 4070
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.658713250958671e-05,
+      "loss": 0.6919,
+      "step": 4080
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.651611986933674e-05,
+      "loss": 0.6843,
+      "step": 4090
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.644510722908678e-05,
+      "loss": 0.7114,
+      "step": 4100
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.6374094588836813e-05,
+      "loss": 0.7106,
+      "step": 4110
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.6303081948586848e-05,
+      "loss": 0.7068,
+      "step": 4120
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.6232069308336887e-05,
+      "loss": 0.6979,
+      "step": 4130
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.616105666808692e-05,
+      "loss": 0.692,
+      "step": 4140
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.6090044027836958e-05,
+      "loss": 0.6915,
+      "step": 4150
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.601903138758699e-05,
+      "loss": 0.687,
+      "step": 4160
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.5948018747337026e-05,
+      "loss": 0.6917,
+      "step": 4170
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.5877006107087065e-05,
+      "loss": 0.6857,
+      "step": 4180
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5805993466837097e-05,
+      "loss": 0.688,
+      "step": 4190
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5734980826587136e-05,
+      "loss": 0.6983,
+      "step": 4200
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5663968186337168e-05,
+      "loss": 0.6983,
+      "step": 4210
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5592955546087204e-05,
+      "loss": 0.7054,
+      "step": 4220
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5521942905837243e-05,
+      "loss": 0.6811,
+      "step": 4230
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5450930265587275e-05,
+      "loss": 0.6871,
+      "step": 4240
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5379917625337314e-05,
+      "loss": 0.7152,
+      "step": 4250
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.5308904985087346e-05,
+      "loss": 0.6943,
+      "step": 4260
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.523789234483738e-05,
+      "loss": 0.7022,
+      "step": 4270
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.516687970458742e-05,
+      "loss": 0.7081,
+      "step": 4280
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.5095867064337453e-05,
+      "loss": 0.6955,
+      "step": 4290
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.502485442408749e-05,
+      "loss": 0.6884,
+      "step": 4300
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.4953841783837527e-05,
+      "loss": 0.6842,
+      "step": 4310
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.488282914358756e-05,
+      "loss": 0.6955,
+      "step": 4320
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.4811816503337595e-05,
+      "loss": 0.679,
+      "step": 4330
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.474080386308763e-05,
+      "loss": 0.6774,
+      "step": 4340
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.4669791222837666e-05,
+      "loss": 0.666,
+      "step": 4350
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.45987785825877e-05,
+      "loss": 0.6718,
+      "step": 4360
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.4527765942337737e-05,
+      "loss": 0.7057,
+      "step": 4370
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.4456753302087773e-05,
+      "loss": 0.6815,
+      "step": 4380
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.4385740661837808e-05,
+      "loss": 0.711,
+      "step": 4390
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.4314728021587844e-05,
+      "loss": 0.6843,
+      "step": 4400
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.424371538133788e-05,
+      "loss": 0.7058,
+      "step": 4410
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.4172702741087915e-05,
+      "loss": 0.7073,
+      "step": 4420
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.410169010083795e-05,
+      "loss": 0.7122,
+      "step": 4430
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.4030677460587986e-05,
+      "loss": 0.6984,
+      "step": 4440
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.395966482033802e-05,
+      "loss": 0.7016,
+      "step": 4450
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.3888652180088057e-05,
+      "loss": 0.6844,
+      "step": 4460
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.3817639539838093e-05,
+      "loss": 0.6795,
+      "step": 4470
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.3746626899588128e-05,
+      "loss": 0.699,
+      "step": 4480
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.3675614259338164e-05,
+      "loss": 0.6874,
+      "step": 4490
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.36046016190882e-05,
+      "loss": 0.7105,
+      "step": 4500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.3533588978838235e-05,
+      "loss": 0.6791,
+      "step": 4510
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.3462576338588267e-05,
+      "loss": 0.7111,
+      "step": 4520
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.3391563698338306e-05,
+      "loss": 0.681,
+      "step": 4530
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.332055105808834e-05,
+      "loss": 0.6831,
+      "step": 4540
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.3249538417838377e-05,
+      "loss": 0.6878,
+      "step": 4550
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.3178525777588413e-05,
+      "loss": 0.6966,
+      "step": 4560
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.3107513137338445e-05,
+      "loss": 0.6963,
+      "step": 4570
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.3036500497088484e-05,
+      "loss": 0.6924,
+      "step": 4580
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.296548785683852e-05,
+      "loss": 0.6861,
+      "step": 4590
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.2894475216588555e-05,
+      "loss": 0.6896,
+      "step": 4600
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.282346257633859e-05,
+      "loss": 0.6947,
+      "step": 4610
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.2752449936088622e-05,
+      "loss": 0.678,
+      "step": 4620
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.268143729583866e-05,
+      "loss": 0.6809,
+      "step": 4630
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.2610424655588697e-05,
+      "loss": 0.6717,
+      "step": 4640
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.2539412015338732e-05,
+      "loss": 0.6835,
+      "step": 4650
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.2468399375088768e-05,
+      "loss": 0.6964,
+      "step": 4660
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.23973867348388e-05,
+      "loss": 0.7003,
+      "step": 4670
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.2326374094588836e-05,
+      "loss": 0.6567,
+      "step": 4680
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.2255361454338875e-05,
+      "loss": 0.7228,
+      "step": 4690
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.218434881408891e-05,
+      "loss": 0.6627,
+      "step": 4700
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.2113336173838946e-05,
+      "loss": 0.6879,
+      "step": 4710
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.2042323533588978e-05,
+      "loss": 0.686,
+      "step": 4720
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.1971310893339014e-05,
+      "loss": 0.6913,
+      "step": 4730
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.1900298253089052e-05,
+      "loss": 0.6763,
+      "step": 4740
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.1829285612839088e-05,
+      "loss": 0.6938,
+      "step": 4750
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.1758272972589124e-05,
+      "loss": 0.6794,
+      "step": 4760
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.1687260332339156e-05,
+      "loss": 0.686,
+      "step": 4770
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.161624769208919e-05,
+      "loss": 0.6871,
+      "step": 4780
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.1545235051839227e-05,
+      "loss": 0.6935,
+      "step": 4790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.1474222411589266e-05,
+      "loss": 0.6647,
+      "step": 4800
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.14032097713393e-05,
+      "loss": 0.6968,
+      "step": 4810
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.1332197131089334e-05,
+      "loss": 0.6798,
+      "step": 4820
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.126118449083937e-05,
+      "loss": 0.692,
+      "step": 4830
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.1190171850589405e-05,
+      "loss": 0.6994,
+      "step": 4840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.1119159210339444e-05,
+      "loss": 0.6742,
+      "step": 4850
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.104814657008948e-05,
+      "loss": 0.6821,
+      "step": 4860
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.097713392983951e-05,
+      "loss": 0.6936,
+      "step": 4870
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.0906121289589547e-05,
+      "loss": 0.6859,
+      "step": 4880
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.0835108649339582e-05,
+      "loss": 0.6948,
+      "step": 4890
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.076409600908962e-05,
+      "loss": 0.6807,
+      "step": 4900
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.0693083368839657e-05,
+      "loss": 0.6846,
+      "step": 4910
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.062207072858969e-05,
+      "loss": 0.681,
+      "step": 4920
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.0551058088339725e-05,
+      "loss": 0.6979,
+      "step": 4930
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.048004544808976e-05,
+      "loss": 0.6765,
+      "step": 4940
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.0409032807839796e-05,
+      "loss": 0.6879,
+      "step": 4950
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.033802016758983e-05,
+      "loss": 0.6818,
+      "step": 4960
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.0267007527339867e-05,
+      "loss": 0.678,
+      "step": 4970
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.0195994887089902e-05,
+      "loss": 0.6697,
+      "step": 4980
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.0124982246839938e-05,
+      "loss": 0.6896,
+      "step": 4990
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.0053969606589974e-05,
+      "loss": 0.6762,
+      "step": 5000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.998295696634001e-05,
+      "loss": 0.6778,
+      "step": 5010
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.9911944326090045e-05,
+      "loss": 0.686,
+      "step": 5020
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.984093168584008e-05,
+      "loss": 0.6818,
+      "step": 5030
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.9769919045590116e-05,
+      "loss": 0.6903,
+      "step": 5040
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.969890640534015e-05,
+      "loss": 0.6866,
+      "step": 5050
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9627893765090187e-05,
+      "loss": 0.7182,
+      "step": 5060
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9556881124840222e-05,
+      "loss": 0.6859,
+      "step": 5070
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9485868484590258e-05,
+      "loss": 0.6423,
+      "step": 5080
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9414855844340294e-05,
+      "loss": 0.7043,
+      "step": 5090
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.934384320409033e-05,
+      "loss": 0.6776,
+      "step": 5100
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9272830563840365e-05,
+      "loss": 0.6662,
+      "step": 5110
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.92018179235904e-05,
+      "loss": 0.6868,
+      "step": 5120
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9130805283340436e-05,
+      "loss": 0.683,
+      "step": 5130
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.905979264309047e-05,
+      "loss": 0.6825,
+      "step": 5140
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8988780002840507e-05,
+      "loss": 0.6694,
+      "step": 5150
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8917767362590542e-05,
+      "loss": 0.6972,
+      "step": 5160
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8846754722340578e-05,
+      "loss": 0.696,
+      "step": 5170
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8775742082090614e-05,
+      "loss": 0.6963,
+      "step": 5180
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.870472944184065e-05,
+      "loss": 0.6752,
+      "step": 5190
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8633716801590685e-05,
+      "loss": 0.6757,
+      "step": 5200
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.856270416134072e-05,
+      "loss": 0.6859,
+      "step": 5210
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8491691521090752e-05,
+      "loss": 0.6734,
+      "step": 5220
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.842067888084079e-05,
+      "loss": 0.6965,
+      "step": 5230
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8349666240590827e-05,
+      "loss": 0.686,
+      "step": 5240
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8278653600340862e-05,
+      "loss": 0.6699,
+      "step": 5250
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8207640960090898e-05,
+      "loss": 0.6782,
+      "step": 5260
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.813662831984093e-05,
+      "loss": 0.6734,
+      "step": 5270
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.806561567959097e-05,
+      "loss": 0.6753,
+      "step": 5280
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7994603039341005e-05,
+      "loss": 0.6875,
+      "step": 5290
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.792359039909104e-05,
+      "loss": 0.6677,
+      "step": 5300
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7852577758841076e-05,
+      "loss": 0.671,
+      "step": 5310
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7781565118591108e-05,
+      "loss": 0.6761,
+      "step": 5320
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7710552478341147e-05,
+      "loss": 0.679,
+      "step": 5330
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7639539838091182e-05,
+      "loss": 0.673,
+      "step": 5340
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7568527197841218e-05,
+      "loss": 0.6538,
+      "step": 5350
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7497514557591254e-05,
+      "loss": 0.673,
+      "step": 5360
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7426501917341286e-05,
+      "loss": 0.6775,
+      "step": 5370
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.735548927709132e-05,
+      "loss": 0.7055,
+      "step": 5380
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.728447663684136e-05,
+      "loss": 0.6855,
+      "step": 5390
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7213463996591396e-05,
+      "loss": 0.6852,
+      "step": 5400
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.714245135634143e-05,
+      "loss": 0.6856,
+      "step": 5410
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7071438716091463e-05,
+      "loss": 0.6907,
+      "step": 5420
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.70004260758415e-05,
+      "loss": 0.6834,
+      "step": 5430
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6929413435591538e-05,
+      "loss": 0.682,
+      "step": 5440
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6858400795341574e-05,
+      "loss": 0.7019,
+      "step": 5450
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6787388155091606e-05,
+      "loss": 0.6877,
+      "step": 5460
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.671637551484164e-05,
+      "loss": 0.6895,
+      "step": 5470
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6645362874591677e-05,
+      "loss": 0.6872,
+      "step": 5480
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6574350234341712e-05,
+      "loss": 0.699,
+      "step": 5490
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.650333759409175e-05,
+      "loss": 0.6867,
+      "step": 5500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.6432324953841783e-05,
+      "loss": 0.6678,
+      "step": 5510
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.636131231359182e-05,
+      "loss": 0.6796,
+      "step": 5520
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6290299673341855e-05,
+      "loss": 0.6605,
+      "step": 5530
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.621928703309189e-05,
+      "loss": 0.6697,
+      "step": 5540
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.614827439284193e-05,
+      "loss": 0.6754,
+      "step": 5550
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.607726175259196e-05,
+      "loss": 0.6861,
+      "step": 5560
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6006249112341997e-05,
+      "loss": 0.6769,
+      "step": 5570
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.5935236472092032e-05,
+      "loss": 0.6824,
+      "step": 5580
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.5864223831842068e-05,
+      "loss": 0.682,
+      "step": 5590
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.5793211191592107e-05,
+      "loss": 0.6625,
+      "step": 5600
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.572219855134214e-05,
+      "loss": 0.6823,
+      "step": 5610
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.5651185911092175e-05,
+      "loss": 0.6506,
+      "step": 5620
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.558017327084221e-05,
+      "loss": 0.6623,
+      "step": 5630
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.5509160630592246e-05,
+      "loss": 0.6825,
+      "step": 5640
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.543814799034228e-05,
+      "loss": 0.6693,
+      "step": 5650
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.5367135350092317e-05,
+      "loss": 0.6937,
+      "step": 5660
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.5296122709842352e-05,
+      "loss": 0.6885,
+      "step": 5670
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.5225110069592388e-05,
+      "loss": 0.6897,
+      "step": 5680
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.5154097429342423e-05,
+      "loss": 0.6652,
+      "step": 5690
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.5083084789092459e-05,
+      "loss": 0.665,
+      "step": 5700
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.5012072148842496e-05,
+      "loss": 0.6658,
+      "step": 5710
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.494105950859253e-05,
+      "loss": 0.6568,
+      "step": 5720
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4870046868342566e-05,
+      "loss": 0.6552,
+      "step": 5730
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4799034228092601e-05,
+      "loss": 0.6945,
+      "step": 5740
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4728021587842635e-05,
+      "loss": 0.6607,
+      "step": 5750
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.465700894759267e-05,
+      "loss": 0.6575,
+      "step": 5760
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4585996307342708e-05,
+      "loss": 0.6739,
+      "step": 5770
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4514983667092743e-05,
+      "loss": 0.6737,
+      "step": 5780
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4443971026842779e-05,
+      "loss": 0.6605,
+      "step": 5790
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4372958386592813e-05,
+      "loss": 0.6725,
+      "step": 5800
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4301945746342848e-05,
+      "loss": 0.661,
+      "step": 5810
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4230933106092886e-05,
+      "loss": 0.6811,
+      "step": 5820
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.4159920465842921e-05,
+      "loss": 0.6774,
+      "step": 5830
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.4088907825592957e-05,
+      "loss": 0.671,
+      "step": 5840
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.401789518534299e-05,
+      "loss": 0.6679,
+      "step": 5850
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3946882545093026e-05,
+      "loss": 0.6647,
+      "step": 5860
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3875869904843063e-05,
+      "loss": 0.6762,
+      "step": 5870
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3804857264593099e-05,
+      "loss": 0.6751,
+      "step": 5880
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3733844624343135e-05,
+      "loss": 0.6746,
+      "step": 5890
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3662831984093168e-05,
+      "loss": 0.6906,
+      "step": 5900
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3591819343843204e-05,
+      "loss": 0.6657,
+      "step": 5910
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.352080670359324e-05,
+      "loss": 0.6533,
+      "step": 5920
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3449794063343277e-05,
+      "loss": 0.6633,
+      "step": 5930
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3378781423093312e-05,
+      "loss": 0.6712,
+      "step": 5940
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3307768782843346e-05,
+      "loss": 0.6539,
+      "step": 5950
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3236756142593382e-05,
+      "loss": 0.6741,
+      "step": 5960
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3165743502343417e-05,
+      "loss": 0.6917,
+      "step": 5970
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3094730862093455e-05,
+      "loss": 0.6637,
+      "step": 5980
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.302371822184349e-05,
+      "loss": 0.66,
+      "step": 5990
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2952705581593524e-05,
+      "loss": 0.6441,
+      "step": 6000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.288169294134356e-05,
+      "loss": 0.6562,
+      "step": 6010
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2810680301093595e-05,
+      "loss": 0.6759,
+      "step": 6020
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2739667660843629e-05,
+      "loss": 0.6586,
+      "step": 6030
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2668655020593668e-05,
+      "loss": 0.6815,
+      "step": 6040
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2597642380343702e-05,
+      "loss": 0.6721,
+      "step": 6050
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2526629740093737e-05,
+      "loss": 0.6756,
+      "step": 6060
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2455617099843773e-05,
+      "loss": 0.6681,
+      "step": 6070
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2384604459593808e-05,
+      "loss": 0.6493,
+      "step": 6080
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2313591819343844e-05,
+      "loss": 0.6536,
+      "step": 6090
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.224257917909388e-05,
+      "loss": 0.6364,
+      "step": 6100
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2171566538843915e-05,
+      "loss": 0.6811,
+      "step": 6110
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.210055389859395e-05,
+      "loss": 0.6988,
+      "step": 6120
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2029541258343986e-05,
+      "loss": 0.668,
+      "step": 6130
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1958528618094022e-05,
+      "loss": 0.6277,
+      "step": 6140
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1887515977844056e-05,
+      "loss": 0.6787,
+      "step": 6150
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1816503337594093e-05,
+      "loss": 0.6281,
+      "step": 6160
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1745490697344128e-05,
+      "loss": 0.6876,
+      "step": 6170
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1674478057094164e-05,
+      "loss": 0.6683,
+      "step": 6180
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.16034654168442e-05,
+      "loss": 0.6853,
+      "step": 6190
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1532452776594233e-05,
+      "loss": 0.6537,
+      "step": 6200
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.146144013634427e-05,
+      "loss": 0.6657,
+      "step": 6210
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1390427496094304e-05,
+      "loss": 0.6724,
+      "step": 6220
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.131941485584434e-05,
+      "loss": 0.6592,
+      "step": 6230
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.1248402215594377e-05,
+      "loss": 0.648,
+      "step": 6240
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.1177389575344411e-05,
+      "loss": 0.6812,
+      "step": 6250
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.1106376935094448e-05,
+      "loss": 0.6671,
+      "step": 6260
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.1035364294844482e-05,
+      "loss": 0.6767,
+      "step": 6270
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0964351654594518e-05,
+      "loss": 0.6575,
+      "step": 6280
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0893339014344555e-05,
+      "loss": 0.6696,
+      "step": 6290
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.0822326374094589e-05,
+      "loss": 0.6668,
+      "step": 6300
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.0751313733844624e-05,
+      "loss": 0.6713,
+      "step": 6310
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.068030109359466e-05,
+      "loss": 0.6528,
+      "step": 6320
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.0609288453344696e-05,
+      "loss": 0.6673,
+      "step": 6330
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.0538275813094733e-05,
+      "loss": 0.6594,
+      "step": 6340
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.0467263172844767e-05,
+      "loss": 0.6674,
+      "step": 6350
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.0396250532594802e-05,
+      "loss": 0.6703,
+      "step": 6360
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.0325237892344838e-05,
+      "loss": 0.665,
+      "step": 6370
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.0254225252094873e-05,
+      "loss": 0.6737,
+      "step": 6380
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.0183212611844909e-05,
+      "loss": 0.6737,
+      "step": 6390
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.0112199971594944e-05,
+      "loss": 0.6542,
+      "step": 6400
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.004118733134498e-05,
+      "loss": 0.6634,
+      "step": 6410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.970174691095016e-06,
+      "loss": 0.6655,
+      "step": 6420
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.899162050845051e-06,
+      "loss": 0.6761,
+      "step": 6430
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.828149410595087e-06,
+      "loss": 0.6706,
+      "step": 6440
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.757136770345122e-06,
+      "loss": 0.6789,
+      "step": 6450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.686124130095158e-06,
+      "loss": 0.675,
+      "step": 6460
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.615111489845192e-06,
+      "loss": 0.6644,
+      "step": 6470
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.544098849595229e-06,
+      "loss": 0.6864,
+      "step": 6480
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.473086209345264e-06,
+      "loss": 0.6446,
+      "step": 6490
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.402073569095298e-06,
+      "loss": 0.6724,
+      "step": 6500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.331060928845336e-06,
+      "loss": 0.6699,
+      "step": 6510
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.26004828859537e-06,
+      "loss": 0.6459,
+      "step": 6520
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.189035648345407e-06,
+      "loss": 0.6658,
+      "step": 6530
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.118023008095442e-06,
+      "loss": 0.6634,
+      "step": 6540
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.047010367845476e-06,
+      "loss": 0.6499,
+      "step": 6550
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.975997727595513e-06,
+      "loss": 0.6583,
+      "step": 6560
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.904985087345547e-06,
+      "loss": 0.6714,
+      "step": 6570
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.833972447095583e-06,
+      "loss": 0.6636,
+      "step": 6580
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.76295980684562e-06,
+      "loss": 0.6533,
+      "step": 6590
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.691947166595654e-06,
+      "loss": 0.678,
+      "step": 6600
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.620934526345691e-06,
+      "loss": 0.6598,
+      "step": 6610
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.549921886095725e-06,
+      "loss": 0.6668,
+      "step": 6620
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.47890924584576e-06,
+      "loss": 0.6343,
+      "step": 6630
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.407896605595798e-06,
+      "loss": 0.6585,
+      "step": 6640
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.336883965345832e-06,
+      "loss": 0.6651,
+      "step": 6650
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.265871325095867e-06,
+      "loss": 0.6533,
+      "step": 6660
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.194858684845903e-06,
+      "loss": 0.653,
+      "step": 6670
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.123846044595938e-06,
+      "loss": 0.6585,
+      "step": 6680
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.052833404345974e-06,
+      "loss": 0.6585,
+      "step": 6690
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.98182076409601e-06,
+      "loss": 0.66,
+      "step": 6700
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.910808123846045e-06,
+      "loss": 0.6508,
+      "step": 6710
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.83979548359608e-06,
+      "loss": 0.6544,
+      "step": 6720
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.768782843346116e-06,
+      "loss": 0.6532,
+      "step": 6730
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.697770203096152e-06,
+      "loss": 0.6634,
+      "step": 6740
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.626757562846187e-06,
+      "loss": 0.6525,
+      "step": 6750
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.555744922596223e-06,
+      "loss": 0.646,
+      "step": 6760
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.4847322823462574e-06,
+      "loss": 0.6707,
+      "step": 6770
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.413719642096294e-06,
+      "loss": 0.6563,
+      "step": 6780
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.3427070018463285e-06,
+      "loss": 0.6429,
+      "step": 6790
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.271694361596365e-06,
+      "loss": 0.6718,
+      "step": 6800
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.2006817213464e-06,
+      "loss": 0.634,
+      "step": 6810
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.129669081096435e-06,
+      "loss": 0.6701,
+      "step": 6820
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 7.058656440846472e-06,
+      "loss": 0.6529,
+      "step": 6830
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.987643800596506e-06,
+      "loss": 0.6725,
+      "step": 6840
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.916631160346541e-06,
+      "loss": 0.6615,
+      "step": 6850
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.845618520096577e-06,
+      "loss": 0.6589,
+      "step": 6860
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.774605879846613e-06,
+      "loss": 0.642,
+      "step": 6870
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.703593239596649e-06,
+      "loss": 0.6523,
+      "step": 6880
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.632580599346684e-06,
+      "loss": 0.6571,
+      "step": 6890
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.561567959096719e-06,
+      "loss": 0.6631,
+      "step": 6900
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.490555318846755e-06,
+      "loss": 0.6816,
+      "step": 6910
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.419542678596791e-06,
+      "loss": 0.6856,
+      "step": 6920
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.3485300383468255e-06,
+      "loss": 0.6602,
+      "step": 6930
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.277517398096862e-06,
+      "loss": 0.6664,
+      "step": 6940
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.2065047578468966e-06,
+      "loss": 0.666,
+      "step": 6950
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.135492117596932e-06,
+      "loss": 0.6494,
+      "step": 6960
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.0644794773469685e-06,
+      "loss": 0.6365,
+      "step": 6970
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.993466837097004e-06,
+      "loss": 0.6745,
+      "step": 6980
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.922454196847039e-06,
+      "loss": 0.6621,
+      "step": 6990
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.851441556597074e-06,
+      "loss": 0.655,
+      "step": 7000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.78042891634711e-06,
+      "loss": 0.6534,
+      "step": 7010
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.709416276097146e-06,
+      "loss": 0.6776,
+      "step": 7020
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.638403635847181e-06,
+      "loss": 0.6504,
+      "step": 7030
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.5673909955972166e-06,
+      "loss": 0.6631,
+      "step": 7040
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.496378355347252e-06,
+      "loss": 0.6412,
+      "step": 7050
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.425365715097288e-06,
+      "loss": 0.6557,
+      "step": 7060
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.354353074847323e-06,
+      "loss": 0.6925,
+      "step": 7070
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.283340434597359e-06,
+      "loss": 0.6316,
+      "step": 7080
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.212327794347394e-06,
+      "loss": 0.6795,
+      "step": 7090
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.141315154097429e-06,
+      "loss": 0.6524,
+      "step": 7100
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.070302513847465e-06,
+      "loss": 0.6303,
+      "step": 7110
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.999289873597501e-06,
+      "loss": 0.6486,
+      "step": 7120
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.9282772333475366e-06,
+      "loss": 0.6484,
+      "step": 7130
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.857264593097571e-06,
+      "loss": 0.6456,
+      "step": 7140
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.786251952847607e-06,
+      "loss": 0.6651,
+      "step": 7150
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.715239312597642e-06,
+      "loss": 0.6643,
+      "step": 7160
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.644226672347679e-06,
+      "loss": 0.6694,
+      "step": 7170
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.5732140320977135e-06,
+      "loss": 0.6469,
+      "step": 7180
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.502201391847749e-06,
+      "loss": 0.6473,
+      "step": 7190
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.431188751597785e-06,
+      "loss": 0.6363,
+      "step": 7200
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.36017611134782e-06,
+      "loss": 0.678,
+      "step": 7210
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.289163471097856e-06,
+      "loss": 0.6461,
+      "step": 7220
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.218150830847891e-06,
+      "loss": 0.6539,
+      "step": 7230
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.147138190597927e-06,
+      "loss": 0.6617,
+      "step": 7240
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.076125550347962e-06,
+      "loss": 0.6626,
+      "step": 7250
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.005112910097997e-06,
+      "loss": 0.6496,
+      "step": 7260
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.9341002698480335e-06,
+      "loss": 0.6494,
+      "step": 7270
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.863087629598069e-06,
+      "loss": 0.6606,
+      "step": 7280
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.7920749893481046e-06,
+      "loss": 0.6816,
+      "step": 7290
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.7210623490981393e-06,
+      "loss": 0.6408,
+      "step": 7300
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.6500497088481753e-06,
+      "loss": 0.6575,
+      "step": 7310
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.579037068598211e-06,
+      "loss": 0.66,
+      "step": 7320
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.5080244283482464e-06,
+      "loss": 0.65,
+      "step": 7330
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.4370117880982815e-06,
+      "loss": 0.6722,
+      "step": 7340
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.365999147848317e-06,
+      "loss": 0.646,
+      "step": 7350
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.294986507598353e-06,
+      "loss": 0.6651,
+      "step": 7360
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2239738673483878e-06,
+      "loss": 0.6574,
+      "step": 7370
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.1529612270984233e-06,
+      "loss": 0.6609,
+      "step": 7380
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.0819485868484593e-06,
+      "loss": 0.6489,
+      "step": 7390
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.0109359465984944e-06,
+      "loss": 0.6537,
+      "step": 7400
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.9399233063485304e-06,
+      "loss": 0.6499,
+      "step": 7410
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.8689106660985655e-06,
+      "loss": 0.6691,
+      "step": 7420
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.797898025848601e-06,
+      "loss": 0.6451,
+      "step": 7430
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7268853855986366e-06,
+      "loss": 0.661,
+      "step": 7440
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.655872745348672e-06,
+      "loss": 0.6642,
+      "step": 7450
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.5848601050987078e-06,
+      "loss": 0.6638,
+      "step": 7460
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.5138474648487433e-06,
+      "loss": 0.6564,
+      "step": 7470
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.4428348245987784e-06,
+      "loss": 0.6448,
+      "step": 7480
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.3718221843488144e-06,
+      "loss": 0.6541,
+      "step": 7490
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.3008095440988495e-06,
+      "loss": 0.6631,
+      "step": 7500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.229796903848885e-06,
+      "loss": 0.674,
+      "step": 7510
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.1587842635989207e-06,
+      "loss": 0.6427,
+      "step": 7520
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.087771623348956e-06,
+      "loss": 0.6707,
+      "step": 7530
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.0167589830989918e-06,
+      "loss": 0.6325,
+      "step": 7540
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9457463428490273e-06,
+      "loss": 0.6388,
+      "step": 7550
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.8747337025990627e-06,
+      "loss": 0.647,
+      "step": 7560
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.8037210623490984e-06,
+      "loss": 0.6585,
+      "step": 7570
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.7327084220991338e-06,
+      "loss": 0.6278,
+      "step": 7580
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.6616957818491693e-06,
+      "loss": 0.64,
+      "step": 7590
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5906831415992047e-06,
+      "loss": 0.6687,
+      "step": 7600
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5196705013492402e-06,
+      "loss": 0.6673,
+      "step": 7610
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.4486578610992758e-06,
+      "loss": 0.6704,
+      "step": 7620
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.3776452208493111e-06,
+      "loss": 0.6629,
+      "step": 7630
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.3066325805993467e-06,
+      "loss": 0.6456,
+      "step": 7640
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2356199403493822e-06,
+      "loss": 0.6287,
+      "step": 7650
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.1646073000994178e-06,
+      "loss": 0.6684,
+      "step": 7660
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0935946598494533e-06,
+      "loss": 0.6393,
+      "step": 7670
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0225820195994887e-06,
+      "loss": 0.6369,
+      "step": 7680
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 9.515693793495242e-07,
+      "loss": 0.6481,
+      "step": 7690
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.805567390995598e-07,
+      "loss": 0.6594,
+      "step": 7700
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.095440988495952e-07,
+      "loss": 0.6508,
+      "step": 7710
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 7.385314585996308e-07,
+      "loss": 0.6445,
+      "step": 7720
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.675188183496662e-07,
+      "loss": 0.6492,
+      "step": 7730
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.965061780997018e-07,
+      "loss": 0.6332,
+      "step": 7740
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.254935378497374e-07,
+      "loss": 0.6739,
+      "step": 7750
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.544808975997728e-07,
+      "loss": 0.6653,
+      "step": 7760
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.8346825734980826e-07,
+      "loss": 0.6566,
+      "step": 7770
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.124556170998438e-07,
+      "loss": 0.661,
+      "step": 7780
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.414429768498793e-07,
+      "loss": 0.6686,
+      "step": 7790
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.704303365999148e-07,
+      "loss": 0.6837,
+      "step": 7800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.941769634995029e-08,
+      "loss": 0.6575,
+      "step": 7810
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.84050560999858e-08,
+      "loss": 0.6388,
+      "step": 7820
+    },
+    {
+      "epoch": 1.0,
+      "step": 7824,
+      "total_flos": 1.2493204182427238e+18,
+      "train_loss": 0.730289878990265,
+      "train_runtime": 56725.1631,
+      "train_samples_per_second": 17.656,
+      "train_steps_per_second": 0.138
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 7824,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 1.2493204182427238e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}