diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,15413 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.2618031500583697,
+  "eval_steps": 500,
+  "global_step": 21978,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005741297150402847,
+      "grad_norm": 47.5,
+      "learning_rate": 8.604206500956023e-07,
+      "loss": 12.2289,
+      "step": 10
+    },
+    {
+      "epoch": 0.0011482594300805695,
+      "grad_norm": 41.0,
+      "learning_rate": 1.8164435946462718e-06,
+      "loss": 12.1447,
+      "step": 20
+    },
+    {
+      "epoch": 0.0017223891451208543,
+      "grad_norm": 38.75,
+      "learning_rate": 2.772466539196941e-06,
+      "loss": 12.0576,
+      "step": 30
+    },
+    {
+      "epoch": 0.002296518860161139,
+      "grad_norm": 75.0,
+      "learning_rate": 3.7284894837476104e-06,
+      "loss": 11.9311,
+      "step": 40
+    },
+    {
+      "epoch": 0.0028706485752014238,
+      "grad_norm": 38.25,
+      "learning_rate": 4.68451242829828e-06,
+      "loss": 11.6058,
+      "step": 50
+    },
+    {
+      "epoch": 0.0034447782902417086,
+      "grad_norm": 35.0,
+      "learning_rate": 5.6405353728489485e-06,
+      "loss": 10.9243,
+      "step": 60
+    },
+    {
+      "epoch": 0.0040189080052819934,
+      "grad_norm": 59.25,
+      "learning_rate": 6.596558317399617e-06,
+      "loss": 10.1567,
+      "step": 70
+    },
+    {
+      "epoch": 0.004593037720322278,
+      "grad_norm": 52.0,
+      "learning_rate": 7.552581261950287e-06,
+      "loss": 9.6776,
+      "step": 80
+    },
+    {
+      "epoch": 0.005167167435362563,
+      "grad_norm": 18.75,
+      "learning_rate": 8.508604206500955e-06,
+      "loss": 9.3481,
+      "step": 90
+    },
+    {
+      "epoch": 0.0057412971504028475,
+      "grad_norm": 19.25,
+      "learning_rate": 9.464627151051626e-06,
+      "loss": 9.0556,
+      "step": 100
+    },
+    {
+      "epoch": 0.006315426865443133,
+      "grad_norm": 70.0,
+      "learning_rate": 1.0420650095602295e-05,
+      "loss": 8.7948,
+      "step": 110
+    },
+    {
+      "epoch": 0.006889556580483417,
+      "grad_norm": 54.0,
+      "learning_rate": 1.1376673040152965e-05,
+      "loss": 8.6711,
+      "step": 120
+    },
+    {
+      "epoch": 0.007463686295523702,
+      "grad_norm": 22.25,
+      "learning_rate": 1.2332695984703634e-05,
+      "loss": 8.5725,
+      "step": 130
+    },
+    {
+      "epoch": 0.008037816010563987,
+      "grad_norm": 5.03125,
+      "learning_rate": 1.3288718929254305e-05,
+      "loss": 8.3944,
+      "step": 140
+    },
+    {
+      "epoch": 0.008611945725604272,
+      "grad_norm": 4.84375,
+      "learning_rate": 1.4244741873804973e-05,
+      "loss": 8.1984,
+      "step": 150
+    },
+    {
+      "epoch": 0.009186075440644556,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.5200764818355642e-05,
+      "loss": 8.0769,
+      "step": 160
+    },
+    {
+      "epoch": 0.009760205155684841,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.615678776290631e-05,
+      "loss": 8.0288,
+      "step": 170
+    },
+    {
+      "epoch": 0.010334334870725126,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.7112810707456982e-05,
+      "loss": 7.9832,
+      "step": 180
+    },
+    {
+      "epoch": 0.01090846458576541,
+      "grad_norm": 5.71875,
+      "learning_rate": 1.806883365200765e-05,
+      "loss": 7.956,
+      "step": 190
+    },
+    {
+      "epoch": 0.011482594300805695,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.902485659655832e-05,
+      "loss": 7.9347,
+      "step": 200
+    },
+    {
+      "epoch": 0.01205672401584598,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.9980879541108987e-05,
+      "loss": 7.9107,
+      "step": 210
+    },
+    {
+      "epoch": 0.012630853730886266,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.0936902485659657e-05,
+      "loss": 7.8908,
+      "step": 220
+    },
+    {
+      "epoch": 0.01320498344592655,
+      "grad_norm": 2.375,
+      "learning_rate": 2.1892925430210324e-05,
+      "loss": 7.8553,
+      "step": 230
+    },
+    {
+      "epoch": 0.013779113160966834,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.2848948374760995e-05,
+      "loss": 7.8601,
+      "step": 240
+    },
+    {
+      "epoch": 0.01435324287600712,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.3804971319311666e-05,
+      "loss": 7.8253,
+      "step": 250
+    },
+    {
+      "epoch": 0.014927372591047403,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.4760994263862333e-05,
+      "loss": 7.8335,
+      "step": 260
+    },
+    {
+      "epoch": 0.015501502306087689,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.5717017208413003e-05,
+      "loss": 7.8141,
+      "step": 270
+    },
+    {
+      "epoch": 0.016075632021127974,
+      "grad_norm": 2.03125,
+      "learning_rate": 2.6673040152963674e-05,
+      "loss": 7.8022,
+      "step": 280
+    },
+    {
+      "epoch": 0.016649761736168257,
+      "grad_norm": 4.15625,
+      "learning_rate": 2.762906309751434e-05,
+      "loss": 7.7881,
+      "step": 290
+    },
+    {
+      "epoch": 0.017223891451208544,
+      "grad_norm": 1.96875,
+      "learning_rate": 2.858508604206501e-05,
+      "loss": 7.7425,
+      "step": 300
+    },
+    {
+      "epoch": 0.017798021166248828,
+      "grad_norm": 1.953125,
+      "learning_rate": 2.954110898661568e-05,
+      "loss": 7.7538,
+      "step": 310
+    },
+    {
+      "epoch": 0.01837215088128911,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.049713193116635e-05,
+      "loss": 7.7481,
+      "step": 320
+    },
+    {
+      "epoch": 0.0189462805963294,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.145315487571702e-05,
+      "loss": 7.7378,
+      "step": 330
+    },
+    {
+      "epoch": 0.019520410311369682,
+      "grad_norm": 1.875,
+      "learning_rate": 3.240917782026769e-05,
+      "loss": 7.7184,
+      "step": 340
+    },
+    {
+      "epoch": 0.020094540026409966,
+      "grad_norm": 2.25,
+      "learning_rate": 3.3365200764818354e-05,
+      "loss": 7.6801,
+      "step": 350
+    },
+    {
+      "epoch": 0.020668669741450253,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.432122370936903e-05,
+      "loss": 7.6844,
+      "step": 360
+    },
+    {
+      "epoch": 0.021242799456490536,
+      "grad_norm": 1.875,
+      "learning_rate": 3.5277246653919695e-05,
+      "loss": 7.6768,
+      "step": 370
+    },
+    {
+      "epoch": 0.02181692917153082,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.623326959847036e-05,
+      "loss": 7.6567,
+      "step": 380
+    },
+    {
+      "epoch": 0.022391058886571107,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.7189292543021036e-05,
+      "loss": 7.6527,
+      "step": 390
+    },
+    {
+      "epoch": 0.02296518860161139,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.8145315487571704e-05,
+      "loss": 7.6469,
+      "step": 400
+    },
+    {
+      "epoch": 0.023539318316651677,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.910133843212238e-05,
+      "loss": 7.6354,
+      "step": 410
+    },
+    {
+      "epoch": 0.02411344803169196,
+      "grad_norm": 1.8828125,
+      "learning_rate": 4.0057361376673045e-05,
+      "loss": 7.6377,
+      "step": 420
+    },
+    {
+      "epoch": 0.024687577746732244,
+      "grad_norm": 3.0,
+      "learning_rate": 4.101338432122371e-05,
+      "loss": 7.6348,
+      "step": 430
+    },
+    {
+      "epoch": 0.02526170746177253,
+      "grad_norm": 1.9296875,
+      "learning_rate": 4.196940726577438e-05,
+      "loss": 7.6112,
+      "step": 440
+    },
+    {
+      "epoch": 0.025835837176812815,
+      "grad_norm": 1.875,
+      "learning_rate": 4.292543021032505e-05,
+      "loss": 7.6331,
+      "step": 450
+    },
+    {
+      "epoch": 0.0264099668918531,
+      "grad_norm": 2.0,
+      "learning_rate": 4.388145315487572e-05,
+      "loss": 7.5849,
+      "step": 460
+    },
+    {
+      "epoch": 0.026984096606893385,
+      "grad_norm": 2.046875,
+      "learning_rate": 4.483747609942639e-05,
+      "loss": 7.5836,
+      "step": 470
+    },
+    {
+      "epoch": 0.02755822632193367,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.5793499043977055e-05,
+      "loss": 7.5977,
+      "step": 480
+    },
+    {
+      "epoch": 0.028132356036973952,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.674952198852773e-05,
+      "loss": 7.5915,
+      "step": 490
+    },
+    {
+      "epoch": 0.02870648575201424,
+      "grad_norm": 1.9921875,
+      "learning_rate": 4.7705544933078396e-05,
+      "loss": 7.5873,
+      "step": 500
+    },
+    {
+      "epoch": 0.029280615467054523,
+      "grad_norm": 1.90625,
+      "learning_rate": 4.866156787762906e-05,
+      "loss": 7.5587,
+      "step": 510
+    },
+    {
+      "epoch": 0.029854745182094806,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.961759082217973e-05,
+      "loss": 7.5544,
+      "step": 520
+    },
+    {
+      "epoch": 0.030428874897135093,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.05736137667304e-05,
+      "loss": 7.5507,
+      "step": 530
+    },
+    {
+      "epoch": 0.031003004612175377,
+      "grad_norm": 1.96875,
+      "learning_rate": 5.152963671128107e-05,
+      "loss": 7.5432,
+      "step": 540
+    },
+    {
+      "epoch": 0.031577134327215664,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.2485659655831745e-05,
+      "loss": 7.5419,
+      "step": 550
+    },
+    {
+      "epoch": 0.03215126404225595,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.344168260038241e-05,
+      "loss": 7.5418,
+      "step": 560
+    },
+    {
+      "epoch": 0.03272539375729623,
+      "grad_norm": 2.25,
+      "learning_rate": 5.4397705544933086e-05,
+      "loss": 7.5354,
+      "step": 570
+    },
+    {
+      "epoch": 0.033299523472336515,
+      "grad_norm": 2.0,
+      "learning_rate": 5.535372848948375e-05,
+      "loss": 7.5175,
+      "step": 580
+    },
+    {
+      "epoch": 0.0338736531873768,
+      "grad_norm": 2.25,
+      "learning_rate": 5.630975143403442e-05,
+      "loss": 7.5083,
+      "step": 590
+    },
+    {
+      "epoch": 0.03444778290241709,
+      "grad_norm": 2.25,
+      "learning_rate": 5.726577437858509e-05,
+      "loss": 7.509,
+      "step": 600
+    },
+    {
+      "epoch": 0.03502191261745737,
+      "grad_norm": 2.0,
+      "learning_rate": 5.822179732313576e-05,
+      "loss": 7.4949,
+      "step": 610
+    },
+    {
+      "epoch": 0.035596042332497656,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.917782026768642e-05,
+      "loss": 7.5187,
+      "step": 620
+    },
+    {
+      "epoch": 0.03617017204753794,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.0133843212237096e-05,
+      "loss": 7.471,
+      "step": 630
+    },
+    {
+      "epoch": 0.03674430176257822,
+      "grad_norm": 2.125,
+      "learning_rate": 6.108986615678777e-05,
+      "loss": 7.4908,
+      "step": 640
+    },
+    {
+      "epoch": 0.03731843147761851,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.204588910133844e-05,
+      "loss": 7.4648,
+      "step": 650
+    },
+    {
+      "epoch": 0.0378925611926588,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.30019120458891e-05,
+      "loss": 7.4751,
+      "step": 660
+    },
+    {
+      "epoch": 0.03846669090769908,
+      "grad_norm": 1.9453125,
+      "learning_rate": 6.395793499043978e-05,
+      "loss": 7.451,
+      "step": 670
+    },
+    {
+      "epoch": 0.039040820622739364,
+      "grad_norm": 2.25,
+      "learning_rate": 6.491395793499044e-05,
+      "loss": 7.4656,
+      "step": 680
+    },
+    {
+      "epoch": 0.03961495033777965,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.586998087954111e-05,
+      "loss": 7.4733,
+      "step": 690
+    },
+    {
+      "epoch": 0.04018908005281993,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.682600382409177e-05,
+      "loss": 7.4538,
+      "step": 700
+    },
+    {
+      "epoch": 0.04076320976786022,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.778202676864245e-05,
+      "loss": 7.4687,
+      "step": 710
+    },
+    {
+      "epoch": 0.041337339482900505,
+      "grad_norm": 1.9140625,
+      "learning_rate": 6.873804971319312e-05,
+      "loss": 7.4557,
+      "step": 720
+    },
+    {
+      "epoch": 0.04191146919794079,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.96940726577438e-05,
+      "loss": 7.4609,
+      "step": 730
+    },
+    {
+      "epoch": 0.04248559891298107,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.065009560229447e-05,
+      "loss": 7.4635,
+      "step": 740
+    },
+    {
+      "epoch": 0.043059728628021356,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.160611854684513e-05,
+      "loss": 7.4266,
+      "step": 750
+    },
+    {
+      "epoch": 0.04363385834306164,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.256214149139579e-05,
+      "loss": 7.4294,
+      "step": 760
+    },
+    {
+      "epoch": 0.04420798805810193,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.351816443594646e-05,
+      "loss": 7.4687,
+      "step": 770
+    },
+    {
+      "epoch": 0.04478211777314221,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.447418738049714e-05,
+      "loss": 7.441,
+      "step": 780
+    },
+    {
+      "epoch": 0.0453562474881825,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.54302103250478e-05,
+      "loss": 7.429,
+      "step": 790
+    },
+    {
+      "epoch": 0.04593037720322278,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.638623326959847e-05,
+      "loss": 7.4286,
+      "step": 800
+    },
+    {
+      "epoch": 0.046504506918263064,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.734225621414915e-05,
+      "loss": 7.433,
+      "step": 810
+    },
+    {
+      "epoch": 0.047078636633303354,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.829827915869982e-05,
+      "loss": 7.4379,
+      "step": 820
+    },
+    {
+      "epoch": 0.04765276634834364,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.925430210325048e-05,
+      "loss": 7.4205,
+      "step": 830
+    },
+    {
+      "epoch": 0.04822689606338392,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.021032504780115e-05,
+      "loss": 7.4257,
+      "step": 840
+    },
+    {
+      "epoch": 0.048801025778424205,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.116634799235181e-05,
+      "loss": 7.4371,
+      "step": 850
+    },
+    {
+      "epoch": 0.04937515549346449,
+      "grad_norm": 2.0,
+      "learning_rate": 8.212237093690249e-05,
+      "loss": 7.4293,
+      "step": 860
+    },
+    {
+      "epoch": 0.04994928520850477,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.307839388145315e-05,
+      "loss": 7.4071,
+      "step": 870
+    },
+    {
+      "epoch": 0.05052341492354506,
+      "grad_norm": 2.125,
+      "learning_rate": 8.403441682600382e-05,
+      "loss": 7.4027,
+      "step": 880
+    },
+    {
+      "epoch": 0.051097544638585346,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.49904397705545e-05,
+      "loss": 7.4302,
+      "step": 890
+    },
+    {
+      "epoch": 0.05167167435362563,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.594646271510517e-05,
+      "loss": 7.4265,
+      "step": 900
+    },
+    {
+      "epoch": 0.05224580406866591,
+      "grad_norm": 1.890625,
+      "learning_rate": 8.690248565965584e-05,
+      "loss": 7.4176,
+      "step": 910
+    },
+    {
+      "epoch": 0.0528199337837062,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.78585086042065e-05,
+      "loss": 7.4276,
+      "step": 920
+    },
+    {
+      "epoch": 0.05339406349874648,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.881453154875718e-05,
+      "loss": 7.3996,
+      "step": 930
+    },
+    {
+      "epoch": 0.05396819321378677,
+      "grad_norm": 2.125,
+      "learning_rate": 8.977055449330784e-05,
+      "loss": 7.4079,
+      "step": 940
+    },
+    {
+      "epoch": 0.054542322928827054,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.072657743785851e-05,
+      "loss": 7.3955,
+      "step": 950
+    },
+    {
+      "epoch": 0.05511645264386734,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.168260038240917e-05,
+      "loss": 7.4203,
+      "step": 960
+    },
+    {
+      "epoch": 0.05569058235890762,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.263862332695985e-05,
+      "loss": 7.39,
+      "step": 970
+    },
+    {
+      "epoch": 0.056264712073947905,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.359464627151052e-05,
+      "loss": 7.4234,
+      "step": 980
+    },
+    {
+      "epoch": 0.056838841788988195,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.45506692160612e-05,
+      "loss": 7.3976,
+      "step": 990
+    },
+    {
+      "epoch": 0.05741297150402848,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.550669216061186e-05,
+      "loss": 7.4,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05798710121906876,
+      "grad_norm": 2.25,
+      "learning_rate": 9.646271510516253e-05,
+      "loss": 7.3905,
+      "step": 1010
+    },
+    {
+      "epoch": 0.058561230934109046,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.74187380497132e-05,
+      "loss": 7.3955,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05913536064914933,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.837476099426386e-05,
+      "loss": 7.3945,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05970949036418961,
+      "grad_norm": 1.90625,
+      "learning_rate": 9.933078393881452e-05,
+      "loss": 7.3953,
+      "step": 1040
+    },
+    {
+      "epoch": 0.0602836200792299,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.999999805483122e-05,
+      "loss": 7.3903,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06085774979427019,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.99999634740572e-05,
+      "loss": 7.3952,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06143187950931047,
+      "grad_norm": 2.25,
+      "learning_rate": 9.999988566734478e-05,
+      "loss": 7.3853,
+      "step": 1070
+    },
+    {
+      "epoch": 0.062006009224350754,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.999976463476122e-05,
+      "loss": 7.3878,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06258013893939104,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.999960037641117e-05,
+      "loss": 7.3885,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06315426865443133,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.999939289243663e-05,
+      "loss": 7.377,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06372839836947161,
+      "grad_norm": 1.8671875,
+      "learning_rate": 9.999914218301699e-05,
+      "loss": 7.3959,
+      "step": 1110
+    },
+    {
+      "epoch": 0.0643025280845119,
+      "grad_norm": 1.875,
+      "learning_rate": 9.999884824836898e-05,
+      "loss": 7.3735,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06487665779955218,
+      "grad_norm": 1.8984375,
+      "learning_rate": 9.99985110887467e-05,
+      "loss": 7.3862,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06545078751459246,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.999813070444166e-05,
+      "loss": 7.3695,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06602491722963275,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.999770709578267e-05,
+      "loss": 7.3761,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06659904694467303,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.999724026313598e-05,
+      "loss": 7.355,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06717317665971331,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.999673020690516e-05,
+      "loss": 7.3493,
+      "step": 1170
+    },
+    {
+      "epoch": 0.0677473063747536,
+      "grad_norm": 1.84375,
+      "learning_rate": 9.999617692753119e-05,
+      "loss": 7.3738,
+      "step": 1180
+    },
+    {
+      "epoch": 0.0683214360897939,
+      "grad_norm": 1.8671875,
+      "learning_rate": 9.999558042549236e-05,
+      "loss": 7.3737,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06889556580483418,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.999494070130435e-05,
+      "loss": 7.3867,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06946969551987446,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.999425775552025e-05,
+      "loss": 7.3736,
+      "step": 1210
+    },
+    {
+      "epoch": 0.07004382523491474,
+      "grad_norm": 1.828125,
+      "learning_rate": 9.999353158873045e-05,
+      "loss": 7.3807,
+      "step": 1220
+    },
+    {
+      "epoch": 0.07061795494995503,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.999276220156276e-05,
+      "loss": 7.3413,
+      "step": 1230
+    },
+    {
+      "epoch": 0.07119208466499531,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.99919495946823e-05,
+      "loss": 7.3401,
+      "step": 1240
+    },
+    {
+      "epoch": 0.0717662143800356,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.999109376879163e-05,
+      "loss": 7.3607,
+      "step": 1250
+    },
+    {
+      "epoch": 0.07234034409507588,
+      "grad_norm": 1.875,
+      "learning_rate": 9.999019472463057e-05,
+      "loss": 7.353,
+      "step": 1260
+    },
+    {
+      "epoch": 0.07291447381011616,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.998925246297641e-05,
+      "loss": 7.3633,
+      "step": 1270
+    },
+    {
+      "epoch": 0.07348860352515645,
+      "grad_norm": 1.8984375,
+      "learning_rate": 9.998826698464372e-05,
+      "loss": 7.3313,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07406273324019673,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.998723829048449e-05,
+      "loss": 7.371,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07463686295523703,
+      "grad_norm": 1.828125,
+      "learning_rate": 9.998616638138802e-05,
+      "loss": 7.3106,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07521099267027731,
+      "grad_norm": 1.9375,
+      "learning_rate": 9.998505125828104e-05,
+      "loss": 7.3412,
+      "step": 1310
+    },
+    {
+      "epoch": 0.0757851223853176,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.998389292212755e-05,
+      "loss": 7.3174,
+      "step": 1320
+    },
+    {
+      "epoch": 0.07635925210035788,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.998269137392897e-05,
+      "loss": 7.34,
+      "step": 1330
+    },
+    {
+      "epoch": 0.07693338181539816,
+      "grad_norm": 1.8671875,
+      "learning_rate": 9.998144661472406e-05,
+      "loss": 7.3135,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07750751153043844,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.998015864558895e-05,
+      "loss": 7.3592,
+      "step": 1350
+    },
+    {
+      "epoch": 0.07808164124547873,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.99788274676371e-05,
+      "loss": 7.3103,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07865577096051901,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.997745308201935e-05,
+      "loss": 7.3127,
+      "step": 1370
+    },
+    {
+      "epoch": 0.0792299006755593,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.997603548992387e-05,
+      "loss": 7.3279,
+      "step": 1380
+    },
+    {
+      "epoch": 0.07980403039059958,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.99745746925762e-05,
+      "loss": 7.3282,
+      "step": 1390
+    },
+    {
+      "epoch": 0.08037816010563986,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.997307069123925e-05,
+      "loss": 7.3304,
+      "step": 1400
+    },
+    {
+      "epoch": 0.08095228982068015,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.997152348721324e-05,
+      "loss": 7.3042,
+      "step": 1410
+    },
+    {
+      "epoch": 0.08152641953572044,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.996993308183575e-05,
+      "loss": 7.3112,
+      "step": 1420
+    },
+    {
+      "epoch": 0.08210054925076073,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.996829947648172e-05,
+      "loss": 7.3347,
+      "step": 1430
+    },
+    {
+      "epoch": 0.08267467896580101,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.996662267256344e-05,
+      "loss": 7.3229,
+      "step": 1440
+    },
+    {
+      "epoch": 0.0832488086808413,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.996490267153053e-05,
+      "loss": 7.3004,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08382293839588158,
+      "grad_norm": 1.875,
+      "learning_rate": 9.996313947486999e-05,
+      "loss": 7.3087,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08439706811092186,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.996133308410609e-05,
+      "loss": 7.3292,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08497119782596214,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.995948350080055e-05,
+      "loss": 7.314,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08554532754100243,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.995759072655231e-05,
+      "loss": 7.2709,
+      "step": 1490
+    },
+    {
+      "epoch": 0.08611945725604271,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.995565476299777e-05,
+      "loss": 7.314,
+      "step": 1500
+    },
+    {
+      "epoch": 0.086693586971083,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.995367561181057e-05,
+      "loss": 7.3032,
+      "step": 1510
+    },
+    {
+      "epoch": 0.08726771668612328,
+      "grad_norm": 1.828125,
+      "learning_rate": 9.99516532747017e-05,
+      "loss": 7.3275,
+      "step": 1520
+    },
+    {
+      "epoch": 0.08784184640116358,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.994958775341958e-05,
+      "loss": 7.3007,
+      "step": 1530
+    },
+    {
+      "epoch": 0.08841597611620386,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.994747904974983e-05,
+      "loss": 7.3164,
+      "step": 1540
+    },
+    {
+      "epoch": 0.08899010583124414,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.994532716551551e-05,
+      "loss": 7.2968,
+      "step": 1550
+    },
+    {
+      "epoch": 0.08956423554628443,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.994313210257694e-05,
+      "loss": 7.2734,
+      "step": 1560
+    },
+    {
+      "epoch": 0.09013836526132471,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.994089386283181e-05,
+      "loss": 7.3032,
+      "step": 1570
+    },
+    {
+      "epoch": 0.090712494976365,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.99386124482151e-05,
+      "loss": 7.295,
+      "step": 1580
+    },
+    {
+      "epoch": 0.09128662469140528,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.993628786069914e-05,
+      "loss": 7.2955,
+      "step": 1590
+    },
+    {
+      "epoch": 0.09186075440644556,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.993392010229361e-05,
+      "loss": 7.2999,
+      "step": 1600
+    },
+    {
+      "epoch": 0.09243488412148584,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.993150917504545e-05,
+      "loss": 7.2974,
+      "step": 1610
+    },
+    {
+      "epoch": 0.09300901383652613,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.992905508103897e-05,
+      "loss": 7.3055,
+      "step": 1620
+    },
+    {
+      "epoch": 0.09358314355156641,
+      "grad_norm": 1.875,
+      "learning_rate": 9.992655782239577e-05,
+      "loss": 7.2906,
+      "step": 1630
+    },
+    {
+      "epoch": 0.09415727326660671,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.992401740127481e-05,
+      "loss": 7.2949,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09473140298164699,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.992143381987229e-05,
+      "loss": 7.2723,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09530553269668728,
+      "grad_norm": 1.75,
+      "learning_rate": 9.99188070804218e-05,
+      "loss": 7.3072,
+      "step": 1660
+    },
+    {
+      "epoch": 0.09587966241172756,
+      "grad_norm": 1.8984375,
+      "learning_rate": 9.991613718519419e-05,
+      "loss": 7.2643,
+      "step": 1670
+    },
+    {
+      "epoch": 0.09645379212676784,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.991342413649764e-05,
+      "loss": 7.2772,
+      "step": 1680
+    },
+    {
+      "epoch": 0.09702792184180813,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.991066793667763e-05,
+      "loss": 7.2988,
+      "step": 1690
+    },
+    {
+      "epoch": 0.09760205155684841,
+      "grad_norm": 1.8359375,
+      "learning_rate": 9.990786858811695e-05,
+      "loss": 7.291,
+      "step": 1700
+    },
+    {
+      "epoch": 0.0981761812718887,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.99050260932357e-05,
+      "loss": 7.281,
+      "step": 1710
+    },
+    {
+      "epoch": 0.09875031098692898,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.990214045449127e-05,
+      "loss": 7.2702,
+      "step": 1720
+    },
+    {
+      "epoch": 0.09932444070196926,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.989921167437833e-05,
+      "loss": 7.2667,
+      "step": 1730
+    },
+    {
+      "epoch": 0.09989857041700954,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.989623975542888e-05,
+      "loss": 7.2564,
+      "step": 1740
+    },
+    {
+      "epoch": 0.10047270013204983,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.989322470021221e-05,
+      "loss": 7.2945,
+      "step": 1750
+    },
+    {
+      "epoch": 0.10104682984709012,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.98901665113349e-05,
+      "loss": 7.2481,
+      "step": 1760
+    },
+    {
+      "epoch": 0.10162095956213041,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.98870651914408e-05,
+      "loss": 7.2643,
+      "step": 1770
+    },
+    {
+      "epoch": 0.10219508927717069,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.988392074321105e-05,
+      "loss": 7.2704,
+      "step": 1780
+    },
+    {
+      "epoch": 0.10276921899221098,
+      "grad_norm": 1.75,
+      "learning_rate": 9.98807331693641e-05,
+      "loss": 7.2702,
+      "step": 1790
+    },
+    {
+      "epoch": 0.10334334870725126,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.987750247265568e-05,
+      "loss": 7.2794,
+      "step": 1800
+    },
+    {
+      "epoch": 0.10391747842229154,
+      "grad_norm": 1.875,
+      "learning_rate": 9.987422865587878e-05,
+      "loss": 7.2656,
+      "step": 1810
+    },
+    {
+      "epoch": 0.10449160813733183,
+      "grad_norm": 1.7890625,
+      "learning_rate": 9.987091172186367e-05,
+      "loss": 7.2393,
+      "step": 1820
+    },
+    {
+      "epoch": 0.10506573785237211,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.986755167347791e-05,
+      "loss": 7.2547,
+      "step": 1830
+    },
+    {
+      "epoch": 0.1056398675674124,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.986414851362633e-05,
+      "loss": 7.2572,
+      "step": 1840
+    },
+    {
+      "epoch": 0.10621399728245268,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.986070224525101e-05,
+      "loss": 7.2767,
+      "step": 1850
+    },
+    {
+      "epoch": 0.10678812699749296,
+      "grad_norm": 1.8046875,
+      "learning_rate": 9.985721287133136e-05,
+      "loss": 7.2495,
+      "step": 1860
+    },
+    {
+      "epoch": 0.10736225671253326,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.985368039488397e-05,
+      "loss": 7.2741,
+      "step": 1870
+    },
+    {
+      "epoch": 0.10793638642757354,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.985010481896274e-05,
+      "loss": 7.267,
+      "step": 1880
+    },
+    {
+      "epoch": 0.10851051614261382,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.984648614665884e-05,
+      "loss": 7.2617,
+      "step": 1890
+    },
+    {
+      "epoch": 0.10908464585765411,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.984282438110067e-05,
+      "loss": 7.2747,
+      "step": 1900
+    },
+    {
+      "epoch": 0.10965877557269439,
+      "grad_norm": 1.7890625,
+      "learning_rate": 9.983911952545391e-05,
+      "loss": 7.2657,
+      "step": 1910
+    },
+    {
+      "epoch": 0.11023290528773468,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.983537158292145e-05,
+      "loss": 7.2559,
+      "step": 1920
+    },
+    {
+      "epoch": 0.11080703500277496,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.98315805567435e-05,
+      "loss": 7.271,
+      "step": 1930
+    },
+    {
+      "epoch": 0.11138116471781524,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.982774645019746e-05,
+      "loss": 7.2564,
+      "step": 1940
+    },
+    {
+      "epoch": 0.11195529443285553,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.982386926659798e-05,
+      "loss": 7.2654,
+      "step": 1950
+    },
+    {
+      "epoch": 0.11252942414789581,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.981994900929694e-05,
+      "loss": 7.2521,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1131035538629361,
+      "grad_norm": 1.75,
+      "learning_rate": 9.981598568168354e-05,
+      "loss": 7.2378,
+      "step": 1970
+    },
+    {
+      "epoch": 0.11367768357797639,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.98119792871841e-05,
+      "loss": 7.2558,
+      "step": 1980
+    },
+    {
+      "epoch": 0.11425181329301667,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.980792982926224e-05,
+      "loss": 7.2372,
+      "step": 1990
+    },
+    {
+      "epoch": 0.11482594300805696,
+      "grad_norm": 1.625,
+      "learning_rate": 9.98038373114188e-05,
+      "loss": 7.2435,
+      "step": 2000
+    },
+    {
+      "epoch": 0.11540007272309724,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.979970173719186e-05,
+      "loss": 7.2298,
+      "step": 2010
+    },
+    {
+      "epoch": 0.11597420243813752,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.979552311015666e-05,
+      "loss": 7.2634,
+      "step": 2020
+    },
+    {
+      "epoch": 0.11654833215317781,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.979130143392575e-05,
+      "loss": 7.2516,
+      "step": 2030
+    },
+    {
+      "epoch": 0.11712246186821809,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.978703671214881e-05,
+      "loss": 7.2689,
+      "step": 2040
+    },
+    {
+      "epoch": 0.11769659158325838,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.97827289485128e-05,
+      "loss": 7.2549,
+      "step": 2050
+    },
+    {
+      "epoch": 0.11827072129829866,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.977837814674186e-05,
+      "loss": 7.2468,
+      "step": 2060
+    },
+    {
+      "epoch": 0.11884485101333894,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.977398431059734e-05,
+      "loss": 7.2569,
+      "step": 2070
+    },
+    {
+      "epoch": 0.11941898072837923,
+      "grad_norm": 1.7890625,
+      "learning_rate": 9.97695474438778e-05,
+      "loss": 7.2475,
+      "step": 2080
+    },
+    {
+      "epoch": 0.11999311044341952,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.976506755041898e-05,
+      "loss": 7.2458,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1205672401584598,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.976054463409388e-05,
+      "loss": 7.2387,
+      "step": 2100
+    },
+    {
+      "epoch": 0.12114136987350009,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.97559786988126e-05,
+      "loss": 7.2162,
+      "step": 2110
+    },
+    {
+      "epoch": 0.12171549958854037,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.97513697485225e-05,
+      "loss": 7.2453,
+      "step": 2120
+    },
+    {
+      "epoch": 0.12228962930358066,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.97467177872081e-05,
+      "loss": 7.2229,
+      "step": 2130
+    },
+    {
+      "epoch": 0.12286375901862094,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.974202281889114e-05,
+      "loss": 7.2433,
+      "step": 2140
+    },
+    {
+      "epoch": 0.12343788873366122,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.973728484763047e-05,
+      "loss": 7.232,
+      "step": 2150
+    },
+    {
+      "epoch": 0.12401201844870151,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.973250387752217e-05,
+      "loss": 7.2322,
+      "step": 2160
+    },
+    {
+      "epoch": 0.12458614816374179,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.97276799126995e-05,
+      "loss": 7.2194,
+      "step": 2170
+    },
+    {
+      "epoch": 0.1251602778787821,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.972281295733286e-05,
+      "loss": 7.2392,
+      "step": 2180
+    },
+    {
+      "epoch": 0.12573440759382237,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.971790301562981e-05,
+      "loss": 7.2556,
+      "step": 2190
+    },
+    {
+      "epoch": 0.12630853730886266,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.971295009183512e-05,
+      "loss": 7.2522,
+      "step": 2200
+    },
+    {
+      "epoch": 0.12688266702390294,
+      "grad_norm": 1.625,
+      "learning_rate": 9.970795419023065e-05,
+      "loss": 7.2535,
+      "step": 2210
+    },
+    {
+      "epoch": 0.12745679673894322,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.970291531513551e-05,
+      "loss": 7.2208,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1280309264539835,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.969783347090585e-05,
+      "loss": 7.2127,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1286050561690238,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.969270866193506e-05,
+      "loss": 7.2056,
+      "step": 2240
+    },
+    {
+      "epoch": 0.12917918588406407,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.968754089265362e-05,
+      "loss": 7.231,
+      "step": 2250
+    },
+    {
+      "epoch": 0.12975331559910436,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.968233016752917e-05,
+      "loss": 7.2163,
+      "step": 2260
+    },
+    {
+      "epoch": 0.13032744531414464,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.967707649106648e-05,
+      "loss": 7.2007,
+      "step": 2270
+    },
+    {
+      "epoch": 0.13090157502918492,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.967177986780746e-05,
+      "loss": 7.2608,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1314757047442252,
+      "grad_norm": 1.7890625,
+      "learning_rate": 9.966644030233114e-05,
+      "loss": 7.222,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1320498344592655,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.966105779925367e-05,
+      "loss": 7.2187,
+      "step": 2300
+    },
+    {
+      "epoch": 0.13262396417430578,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.965563236322836e-05,
+      "loss": 7.2438,
+      "step": 2310
+    },
+    {
+      "epoch": 0.13319809388934606,
+      "grad_norm": 1.7890625,
+      "learning_rate": 9.965016399894556e-05,
+      "loss": 7.1962,
+      "step": 2320
+    },
+    {
+      "epoch": 0.13377222360438634,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.964465271113282e-05,
+      "loss": 7.222,
+      "step": 2330
+    },
+    {
+      "epoch": 0.13434635331942663,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.963909850455473e-05,
+      "loss": 7.1955,
+      "step": 2340
+    },
+    {
+      "epoch": 0.1349204830344669,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.963350138401299e-05,
+      "loss": 7.2496,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1354946127495072,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.962786135434648e-05,
+      "loss": 7.2438,
+      "step": 2360
+    },
+    {
+      "epoch": 0.13606874246454748,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.962217842043106e-05,
+      "loss": 7.2093,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1366428721795878,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.961645258717976e-05,
+      "loss": 7.1862,
+      "step": 2380
+    },
+    {
+      "epoch": 0.13721700189462807,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.961068385954265e-05,
+      "loss": 7.2067,
+      "step": 2390
+    },
+    {
+      "epoch": 0.13779113160966835,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.960487224250694e-05,
+      "loss": 7.2353,
+      "step": 2400
+    },
+    {
+      "epoch": 0.13836526132470864,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.959901774109687e-05,
+      "loss": 7.1877,
+      "step": 2410
+    },
+    {
+      "epoch": 0.13893939103974892,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.959312036037379e-05,
+      "loss": 7.2055,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1395135207547892,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.958718010543607e-05,
+      "loss": 7.2765,
+      "step": 2430
+    },
+    {
+      "epoch": 0.1400876504698295,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.958119698141917e-05,
+      "loss": 7.2015,
+      "step": 2440
+    },
+    {
+      "epoch": 0.14066178018486977,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.957517099349564e-05,
+      "loss": 7.2288,
+      "step": 2450
+    },
+    {
+      "epoch": 0.14123590989991006,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.956910214687507e-05,
+      "loss": 7.251,
+      "step": 2460
+    },
+    {
+      "epoch": 0.14181003961495034,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.956299044680409e-05,
+      "loss": 7.2318,
+      "step": 2470
+    },
+    {
+      "epoch": 0.14238416932999062,
+      "grad_norm": 1.890625,
+      "learning_rate": 9.955683589856634e-05,
+      "loss": 7.2268,
+      "step": 2480
+    },
+    {
+      "epoch": 0.1429582990450309,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.955063850748263e-05,
+      "loss": 7.2451,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1435324287600712,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.954439827891065e-05,
+      "loss": 7.2232,
+      "step": 2500
+    },
+    {
+      "epoch": 0.14410655847511147,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.953811521824522e-05,
+      "loss": 7.2054,
+      "step": 2510
+    },
+    {
+      "epoch": 0.14468068819015176,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.953178933091818e-05,
+      "loss": 7.2127,
+      "step": 2520
+    },
+    {
+      "epoch": 0.14525481790519204,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.952542062239838e-05,
+      "loss": 7.2002,
+      "step": 2530
+    },
+    {
+      "epoch": 0.14582894762023232,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.951900909819169e-05,
+      "loss": 7.235,
+      "step": 2540
+    },
+    {
+      "epoch": 0.1464030773352726,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.9512554763841e-05,
+      "loss": 7.2237,
+      "step": 2550
+    },
+    {
+      "epoch": 0.1469772070503129,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.95060576249262e-05,
+      "loss": 7.2328,
+      "step": 2560
+    },
+    {
+      "epoch": 0.14755133676535317,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.949951768706419e-05,
+      "loss": 7.2167,
+      "step": 2570
+    },
+    {
+      "epoch": 0.14812546648039346,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.949293495590888e-05,
+      "loss": 7.1892,
+      "step": 2580
+    },
+    {
+      "epoch": 0.14869959619543374,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.948630943715118e-05,
+      "loss": 7.1999,
+      "step": 2590
+    },
+    {
+      "epoch": 0.14927372591047405,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.947964113651896e-05,
+      "loss": 7.2392,
+      "step": 2600
+    },
+    {
+      "epoch": 0.14984785562551434,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.947293005977709e-05,
+      "loss": 7.1989,
+      "step": 2610
+    },
+    {
+      "epoch": 0.15042198534055462,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.946617621272745e-05,
+      "loss": 7.1833,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1509961150555949,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.945937960120886e-05,
+      "loss": 7.2117,
+      "step": 2630
+    },
+    {
+      "epoch": 0.1515702447706352,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.945254023109712e-05,
+      "loss": 7.1973,
+      "step": 2640
+    },
+    {
+      "epoch": 0.15214437448567547,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.9445658108305e-05,
+      "loss": 7.1816,
+      "step": 2650
+    },
+    {
+      "epoch": 0.15271850420071575,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.943873323878221e-05,
+      "loss": 7.1913,
+      "step": 2660
+    },
+    {
+      "epoch": 0.15329263391575604,
+      "grad_norm": 1.625,
+      "learning_rate": 9.943176562851548e-05,
+      "loss": 7.2218,
+      "step": 2670
+    },
+    {
+      "epoch": 0.15386676363079632,
+      "grad_norm": 1.75,
+      "learning_rate": 9.942475528352842e-05,
+      "loss": 7.2328,
+      "step": 2680
+    },
+    {
+      "epoch": 0.1544408933458366,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.941770220988158e-05,
+      "loss": 7.1863,
+      "step": 2690
+    },
+    {
+      "epoch": 0.1550150230608769,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.941060641367253e-05,
+      "loss": 7.2014,
+      "step": 2700
+    },
+    {
+      "epoch": 0.15558915277591717,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.940346790103569e-05,
+      "loss": 7.2158,
+      "step": 2710
+    },
+    {
+      "epoch": 0.15616328249095746,
+      "grad_norm": 1.7890625,
+      "learning_rate": 9.939628667814244e-05,
+      "loss": 7.1862,
+      "step": 2720
+    },
+    {
+      "epoch": 0.15673741220599774,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.938906275120112e-05,
+      "loss": 7.169,
+      "step": 2730
+    },
+    {
+      "epoch": 0.15731154192103802,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.938179612645693e-05,
+      "loss": 7.2047,
+      "step": 2740
+    },
+    {
+      "epoch": 0.1578856716360783,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.9374486810192e-05,
+      "loss": 7.2038,
+      "step": 2750
+    },
+    {
+      "epoch": 0.1584598013511186,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.93671348087254e-05,
+      "loss": 7.2069,
+      "step": 2760
+    },
+    {
+      "epoch": 0.15903393106615887,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.935974012841305e-05,
+      "loss": 7.2183,
+      "step": 2770
+    },
+    {
+      "epoch": 0.15960806078119916,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.935230277564782e-05,
+      "loss": 7.2102,
+      "step": 2780
+    },
+    {
+      "epoch": 0.16018219049623944,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.934482275685943e-05,
+      "loss": 7.2052,
+      "step": 2790
+    },
+    {
+      "epoch": 0.16075632021127972,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.933730007851451e-05,
+      "loss": 7.1924,
+      "step": 2800
+    },
+    {
+      "epoch": 0.16133044992632,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.932973474711655e-05,
+      "loss": 7.2124,
+      "step": 2810
+    },
+    {
+      "epoch": 0.1619045796413603,
+      "grad_norm": 1.625,
+      "learning_rate": 9.932212676920595e-05,
+      "loss": 7.2017,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1624787093564006,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.931447615135994e-05,
+      "loss": 7.2266,
+      "step": 2830
+    },
+    {
+      "epoch": 0.16305283907144089,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.93067829001926e-05,
+      "loss": 7.1716,
+      "step": 2840
+    },
+    {
+      "epoch": 0.16362696878648117,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.929904702235495e-05,
+      "loss": 7.1671,
+      "step": 2850
+    },
+    {
+      "epoch": 0.16420109850152145,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.929126852453477e-05,
+      "loss": 7.1841,
+      "step": 2860
+    },
+    {
+      "epoch": 0.16477522821656174,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.928344741345672e-05,
+      "loss": 7.2104,
+      "step": 2870
+    },
+    {
+      "epoch": 0.16534935793160202,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.92755836958823e-05,
+      "loss": 7.1866,
+      "step": 2880
+    },
+    {
+      "epoch": 0.1659234876466423,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.926767737860988e-05,
+      "loss": 7.1634,
+      "step": 2890
+    },
+    {
+      "epoch": 0.1664976173616826,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.92597284684746e-05,
+      "loss": 7.2023,
+      "step": 2900
+    },
+    {
+      "epoch": 0.16707174707672287,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.925173697234844e-05,
+      "loss": 7.1846,
+      "step": 2910
+    },
+    {
+      "epoch": 0.16764587679176315,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.924370289714022e-05,
+      "loss": 7.1909,
+      "step": 2920
+    },
+    {
+      "epoch": 0.16822000650680344,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.923562624979555e-05,
+      "loss": 7.1852,
+      "step": 2930
+    },
+    {
+      "epoch": 0.16879413622184372,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.922750703729684e-05,
+      "loss": 7.1982,
+      "step": 2940
+    },
+    {
+      "epoch": 0.169368265936884,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.921934526666332e-05,
+      "loss": 7.1717,
+      "step": 2950
+    },
+    {
+      "epoch": 0.1699423956519243,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.921114094495099e-05,
+      "loss": 7.1865,
+      "step": 2960
+    },
+    {
+      "epoch": 0.17051652536696457,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.920289407925263e-05,
+      "loss": 7.194,
+      "step": 2970
+    },
+    {
+      "epoch": 0.17109065508200486,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.919460467669785e-05,
+      "loss": 7.1622,
+      "step": 2980
+    },
+    {
+      "epoch": 0.17166478479704514,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.918627274445297e-05,
+      "loss": 7.1727,
+      "step": 2990
+    },
+    {
+      "epoch": 0.17223891451208542,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.917789828972113e-05,
+      "loss": 7.2052,
+      "step": 3000
+    },
+    {
+      "epoch": 0.1728130442271257,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.916948131974217e-05,
+      "loss": 7.2079,
+      "step": 3010
+    },
+    {
+      "epoch": 0.173387173942166,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.916102184179279e-05,
+      "loss": 7.1754,
+      "step": 3020
+    },
+    {
+      "epoch": 0.17396130365720627,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.91525198631863e-05,
+      "loss": 7.1914,
+      "step": 3030
+    },
+    {
+      "epoch": 0.17453543337224656,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.914397539127289e-05,
+      "loss": 7.1679,
+      "step": 3040
+    },
+    {
+      "epoch": 0.17510956308728687,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.913538843343936e-05,
+      "loss": 7.181,
+      "step": 3050
+    },
+    {
+      "epoch": 0.17568369280232715,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.912675899710934e-05,
+      "loss": 7.1929,
+      "step": 3060
+    },
+    {
+      "epoch": 0.17625782251736744,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.911808708974315e-05,
+      "loss": 7.174,
+      "step": 3070
+    },
+    {
+      "epoch": 0.17683195223240772,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.91093727188378e-05,
+      "loss": 7.1991,
+      "step": 3080
+    },
+    {
+      "epoch": 0.177406081947448,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.910061589192705e-05,
+      "loss": 7.1869,
+      "step": 3090
+    },
+    {
+      "epoch": 0.17798021166248829,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.909181661658134e-05,
+      "loss": 7.1955,
+      "step": 3100
+    },
+    {
+      "epoch": 0.17855434137752857,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.908297490040778e-05,
+      "loss": 7.1625,
+      "step": 3110
+    },
+    {
+      "epoch": 0.17912847109256885,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.907409075105028e-05,
+      "loss": 7.201,
+      "step": 3120
+    },
+    {
+      "epoch": 0.17970260080760914,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.90651641761893e-05,
+      "loss": 7.167,
+      "step": 3130
+    },
+    {
+      "epoch": 0.18027673052264942,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.905619518354205e-05,
+      "loss": 7.1732,
+      "step": 3140
+    },
+    {
+      "epoch": 0.1808508602376897,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.904718378086242e-05,
+      "loss": 7.175,
+      "step": 3150
+    },
+    {
+      "epoch": 0.18142498995273,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.903812997594092e-05,
+      "loss": 7.169,
+      "step": 3160
+    },
+    {
+      "epoch": 0.18199911966777027,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.902903377660473e-05,
+      "loss": 7.1731,
+      "step": 3170
+    },
+    {
+      "epoch": 0.18257324938281055,
+      "grad_norm": 1.921875,
+      "learning_rate": 9.901989519071773e-05,
+      "loss": 7.1691,
+      "step": 3180
+    },
+    {
+      "epoch": 0.18314737909785084,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.901071422618036e-05,
+      "loss": 7.1727,
+      "step": 3190
+    },
+    {
+      "epoch": 0.18372150881289112,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.900149089092978e-05,
+      "loss": 7.1871,
+      "step": 3200
+    },
+    {
+      "epoch": 0.1842956385279314,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.899222519293971e-05,
+      "loss": 7.1621,
+      "step": 3210
+    },
+    {
+      "epoch": 0.1848697682429717,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.898291714022055e-05,
+      "loss": 7.1787,
+      "step": 3220
+    },
+    {
+      "epoch": 0.18544389795801197,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.897356674081928e-05,
+      "loss": 7.1843,
+      "step": 3230
+    },
+    {
+      "epoch": 0.18601802767305226,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.89641740028195e-05,
+      "loss": 7.1441,
+      "step": 3240
+    },
+    {
+      "epoch": 0.18659215738809254,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.895473893434142e-05,
+      "loss": 7.1718,
+      "step": 3250
+    },
+    {
+      "epoch": 0.18716628710313282,
+      "grad_norm": 1.75,
+      "learning_rate": 9.894526154354185e-05,
+      "loss": 7.177,
+      "step": 3260
+    },
+    {
+      "epoch": 0.1877404168181731,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.893574183861417e-05,
+      "loss": 7.1814,
+      "step": 3270
+    },
+    {
+      "epoch": 0.18831454653321342,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.892617982778833e-05,
+      "loss": 7.166,
+      "step": 3280
+    },
+    {
+      "epoch": 0.1888886762482537,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.89165755193309e-05,
+      "loss": 7.1755,
+      "step": 3290
+    },
+    {
+      "epoch": 0.18946280596329398,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.890692892154498e-05,
+      "loss": 7.1605,
+      "step": 3300
+    },
+    {
+      "epoch": 0.19003693567833427,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.889724004277023e-05,
+      "loss": 7.1718,
+      "step": 3310
+    },
+    {
+      "epoch": 0.19061106539337455,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.88875088913829e-05,
+      "loss": 7.174,
+      "step": 3320
+    },
+    {
+      "epoch": 0.19118519510841483,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.887773547579574e-05,
+      "loss": 7.1784,
+      "step": 3330
+    },
+    {
+      "epoch": 0.19175932482345512,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.886791980445806e-05,
+      "loss": 7.1844,
+      "step": 3340
+    },
+    {
+      "epoch": 0.1923334545384954,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.885806188585571e-05,
+      "loss": 7.1523,
+      "step": 3350
+    },
+    {
+      "epoch": 0.19290758425353569,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.884816172851104e-05,
+      "loss": 7.1403,
+      "step": 3360
+    },
+    {
+      "epoch": 0.19348171396857597,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.883821934098292e-05,
+      "loss": 7.1545,
+      "step": 3370
+    },
+    {
+      "epoch": 0.19405584368361625,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.882823473186675e-05,
+      "loss": 7.1925,
+      "step": 3380
+    },
+    {
+      "epoch": 0.19462997339865654,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.881820790979443e-05,
+      "loss": 7.1765,
+      "step": 3390
+    },
+    {
+      "epoch": 0.19520410311369682,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.880813888343431e-05,
+      "loss": 7.1598,
+      "step": 3400
+    },
+    {
+      "epoch": 0.1957782328287371,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.87980276614913e-05,
+      "loss": 7.1651,
+      "step": 3410
+    },
+    {
+      "epoch": 0.1963523625437774,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.87878742527067e-05,
+      "loss": 7.1966,
+      "step": 3420
+    },
+    {
+      "epoch": 0.19692649225881767,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.877767866585837e-05,
+      "loss": 7.1851,
+      "step": 3430
+    },
+    {
+      "epoch": 0.19750062197385795,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.876744090976056e-05,
+      "loss": 7.1835,
+      "step": 3440
+    },
+    {
+      "epoch": 0.19807475168889824,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.875716099326404e-05,
+      "loss": 7.2052,
+      "step": 3450
+    },
+    {
+      "epoch": 0.19864888140393852,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.874683892525598e-05,
+      "loss": 7.1426,
+      "step": 3460
+    },
+    {
+      "epoch": 0.1992230111189788,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.873647471466e-05,
+      "loss": 7.1271,
+      "step": 3470
+    },
+    {
+      "epoch": 0.1997971408340191,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.872606837043617e-05,
+      "loss": 7.1351,
+      "step": 3480
+    },
+    {
+      "epoch": 0.20037127054905937,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.871561990158097e-05,
+      "loss": 7.1292,
+      "step": 3490
+    },
+    {
+      "epoch": 0.20094540026409966,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.870512931712734e-05,
+      "loss": 7.1518,
+      "step": 3500
+    },
+    {
+      "epoch": 0.20151952997913997,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.869459662614455e-05,
+      "loss": 7.1549,
+      "step": 3510
+    },
+    {
+      "epoch": 0.20209365969418025,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.868402183773833e-05,
+      "loss": 7.1718,
+      "step": 3520
+    },
+    {
+      "epoch": 0.20266778940922053,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.867340496105079e-05,
+      "loss": 7.1645,
+      "step": 3530
+    },
+    {
+      "epoch": 0.20324191912426082,
+      "grad_norm": 1.625,
+      "learning_rate": 9.866274600526043e-05,
+      "loss": 7.1599,
+      "step": 3540
+    },
+    {
+      "epoch": 0.2038160488393011,
+      "grad_norm": 1.75,
+      "learning_rate": 9.865204497958211e-05,
+      "loss": 7.1633,
+      "step": 3550
+    },
+    {
+      "epoch": 0.20439017855434138,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.864130189326709e-05,
+      "loss": 7.1397,
+      "step": 3560
+    },
+    {
+      "epoch": 0.20496430826938167,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.863051675560297e-05,
+      "loss": 7.1739,
+      "step": 3570
+    },
+    {
+      "epoch": 0.20553843798442195,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.861968957591372e-05,
+      "loss": 7.157,
+      "step": 3580
+    },
+    {
+      "epoch": 0.20611256769946223,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.860882036355962e-05,
+      "loss": 7.1219,
+      "step": 3590
+    },
+    {
+      "epoch": 0.20668669741450252,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.859790912793737e-05,
+      "loss": 7.1577,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2072608271295428,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.858695587847987e-05,
+      "loss": 7.1883,
+      "step": 3610
+    },
+    {
+      "epoch": 0.20783495684458309,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.857596062465648e-05,
+      "loss": 7.1774,
+      "step": 3620
+    },
+    {
+      "epoch": 0.20840908655962337,
+      "grad_norm": 1.625,
+      "learning_rate": 9.856492337597276e-05,
+      "loss": 7.1515,
+      "step": 3630
+    },
+    {
+      "epoch": 0.20898321627466365,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.855384414197067e-05,
+      "loss": 7.1831,
+      "step": 3640
+    },
+    {
+      "epoch": 0.20955734598970394,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.854272293222841e-05,
+      "loss": 7.1699,
+      "step": 3650
+    },
+    {
+      "epoch": 0.21013147570474422,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.853155975636045e-05,
+      "loss": 7.1675,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2107056054197845,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.852035462401764e-05,
+      "loss": 7.1572,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2112797351348248,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.850910754488697e-05,
+      "loss": 7.1519,
+      "step": 3680
+    },
+    {
+      "epoch": 0.21185386484986507,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.849781852869176e-05,
+      "loss": 7.1442,
+      "step": 3690
+    },
+    {
+      "epoch": 0.21242799456490535,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.848648758519161e-05,
+      "loss": 7.1557,
+      "step": 3700
+    },
+    {
+      "epoch": 0.21300212427994564,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.847511472418235e-05,
+      "loss": 7.1443,
+      "step": 3710
+    },
+    {
+      "epoch": 0.21357625399498592,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.846369995549601e-05,
+      "loss": 7.1367,
+      "step": 3720
+    },
+    {
+      "epoch": 0.21415038371002623,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.84522432890009e-05,
+      "loss": 7.1849,
+      "step": 3730
+    },
+    {
+      "epoch": 0.21472451342506652,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.844074473460152e-05,
+      "loss": 7.1564,
+      "step": 3740
+    },
+    {
+      "epoch": 0.2152986431401068,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.842920430223858e-05,
+      "loss": 7.1428,
+      "step": 3750
+    },
+    {
+      "epoch": 0.21587277285514708,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.841762200188904e-05,
+      "loss": 7.1381,
+      "step": 3760
+    },
+    {
+      "epoch": 0.21644690257018737,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.840599784356601e-05,
+      "loss": 7.1563,
+      "step": 3770
+    },
+    {
+      "epoch": 0.21702103228522765,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.839433183731879e-05,
+      "loss": 7.1821,
+      "step": 3780
+    },
+    {
+      "epoch": 0.21759516200026793,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.838262399323288e-05,
+      "loss": 7.1049,
+      "step": 3790
+    },
+    {
+      "epoch": 0.21816929171530822,
+      "grad_norm": 1.625,
+      "learning_rate": 9.837087432142993e-05,
+      "loss": 7.1677,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2187434214303485,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.835908283206777e-05,
+      "loss": 7.1602,
+      "step": 3810
+    },
+    {
+      "epoch": 0.21931755114538878,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.834724953534036e-05,
+      "loss": 7.158,
+      "step": 3820
+    },
+    {
+      "epoch": 0.21989168086042907,
+      "grad_norm": 1.75,
+      "learning_rate": 9.833537444147781e-05,
+      "loss": 7.1471,
+      "step": 3830
+    },
+    {
+      "epoch": 0.22046581057546935,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.832345756074639e-05,
+      "loss": 7.1652,
+      "step": 3840
+    },
+    {
+      "epoch": 0.22103994029050963,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.831149890344846e-05,
+      "loss": 7.1406,
+      "step": 3850
+    },
+    {
+      "epoch": 0.22161407000554992,
+      "grad_norm": 1.625,
+      "learning_rate": 9.82994984799225e-05,
+      "loss": 7.1428,
+      "step": 3860
+    },
+    {
+      "epoch": 0.2221881997205902,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.828745630054314e-05,
+      "loss": 7.1159,
+      "step": 3870
+    },
+    {
+      "epoch": 0.22276232943563049,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.827537237572107e-05,
+      "loss": 7.1448,
+      "step": 3880
+    },
+    {
+      "epoch": 0.22333645915067077,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.826324671590303e-05,
+      "loss": 7.132,
+      "step": 3890
+    },
+    {
+      "epoch": 0.22391058886571105,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.825107933157196e-05,
+      "loss": 7.1557,
+      "step": 3900
+    },
+    {
+      "epoch": 0.22448471858075134,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.823887023324675e-05,
+      "loss": 7.1258,
+      "step": 3910
+    },
+    {
+      "epoch": 0.22505884829579162,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.822661943148243e-05,
+      "loss": 7.1438,
+      "step": 3920
+    },
+    {
+      "epoch": 0.2256329780108319,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.821432693687004e-05,
+      "loss": 7.1421,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2262071077258722,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.820199276003667e-05,
+      "loss": 7.1424,
+      "step": 3940
+    },
+    {
+      "epoch": 0.22678123744091247,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.818961691164548e-05,
+      "loss": 7.1514,
+      "step": 3950
+    },
+    {
+      "epoch": 0.22735536715595278,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.817719940239563e-05,
+      "loss": 7.128,
+      "step": 3960
+    },
+    {
+      "epoch": 0.22792949687099306,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.816474024302228e-05,
+      "loss": 7.1122,
+      "step": 3970
+    },
+    {
+      "epoch": 0.22850362658603335,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.815223944429662e-05,
+      "loss": 7.1329,
+      "step": 3980
+    },
+    {
+      "epoch": 0.22907775630107363,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.813969701702583e-05,
+      "loss": 7.1356,
+      "step": 3990
+    },
+    {
+      "epoch": 0.22965188601611392,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.81271129720531e-05,
+      "loss": 7.1245,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2302260157311542,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.811448732025757e-05,
+      "loss": 7.1262,
+      "step": 4010
+    },
+    {
+      "epoch": 0.23080014544619448,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.810182007255435e-05,
+      "loss": 7.1314,
+      "step": 4020
+    },
+    {
+      "epoch": 0.23137427516123477,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.808911123989452e-05,
+      "loss": 7.134,
+      "step": 4030
+    },
+    {
+      "epoch": 0.23194840487627505,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.807636083326515e-05,
+      "loss": 7.1211,
+      "step": 4040
+    },
+    {
+      "epoch": 0.23252253459131533,
+      "grad_norm": 1.625,
+      "learning_rate": 9.806356886368917e-05,
+      "loss": 7.1423,
+      "step": 4050
+    },
+    {
+      "epoch": 0.23309666430635562,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.80507353422255e-05,
+      "loss": 7.1196,
+      "step": 4060
+    },
+    {
+      "epoch": 0.2336707940213959,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.803786027996899e-05,
+      "loss": 7.1225,
+      "step": 4070
+    },
+    {
+      "epoch": 0.23424492373643618,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.802494368805035e-05,
+      "loss": 7.1493,
+      "step": 4080
+    },
+    {
+      "epoch": 0.23481905345147647,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.801198557763623e-05,
+      "loss": 7.1441,
+      "step": 4090
+    },
+    {
+      "epoch": 0.23539318316651675,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.799898595992919e-05,
+      "loss": 7.1074,
+      "step": 4100
+    },
+    {
+      "epoch": 0.23596731288155703,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.798594484616762e-05,
+      "loss": 7.1399,
+      "step": 4110
+    },
+    {
+      "epoch": 0.23654144259659732,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.797286224762584e-05,
+      "loss": 7.1218,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2371155723116376,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.7959738175614e-05,
+      "loss": 7.1224,
+      "step": 4130
+    },
+    {
+      "epoch": 0.23768970202667788,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.794657264147811e-05,
+      "loss": 7.1262,
+      "step": 4140
+    },
+    {
+      "epoch": 0.23826383174171817,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.793336565660005e-05,
+      "loss": 7.1277,
+      "step": 4150
+    },
+    {
+      "epoch": 0.23883796145675845,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.792011723239751e-05,
+      "loss": 7.1333,
+      "step": 4160
+    },
+    {
+      "epoch": 0.23941209117179874,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.790682738032397e-05,
+      "loss": 7.1199,
+      "step": 4170
+    },
+    {
+      "epoch": 0.23998622088683905,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.789349611186882e-05,
+      "loss": 7.1562,
+      "step": 4180
+    },
+    {
+      "epoch": 0.24056035060187933,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.788012343855716e-05,
+      "loss": 7.1346,
+      "step": 4190
+    },
+    {
+      "epoch": 0.2411344803169196,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.786670937194996e-05,
+      "loss": 7.1284,
+      "step": 4200
+    },
+    {
+      "epoch": 0.2417086100319599,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.785325392364391e-05,
+      "loss": 7.1572,
+      "step": 4210
+    },
+    {
+      "epoch": 0.24228273974700018,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.783975710527154e-05,
+      "loss": 7.1039,
+      "step": 4220
+    },
+    {
+      "epoch": 0.24285686946204046,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.782621892850106e-05,
+      "loss": 7.108,
+      "step": 4230
+    },
+    {
+      "epoch": 0.24343099917708075,
+      "grad_norm": 1.625,
+      "learning_rate": 9.781263940503653e-05,
+      "loss": 7.1546,
+      "step": 4240
+    },
+    {
+      "epoch": 0.24400512889212103,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.77990185466177e-05,
+      "loss": 7.1348,
+      "step": 4250
+    },
+    {
+      "epoch": 0.24457925860716132,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.778535636502005e-05,
+      "loss": 7.1317,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2451533883222016,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.777165287205484e-05,
+      "loss": 7.1193,
+      "step": 4270
+    },
+    {
+      "epoch": 0.24572751803724188,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.775790807956894e-05,
+      "loss": 7.1458,
+      "step": 4280
+    },
+    {
+      "epoch": 0.24630164775228217,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.774412199944507e-05,
+      "loss": 7.0719,
+      "step": 4290
+    },
+    {
+      "epoch": 0.24687577746732245,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.773029464360151e-05,
+      "loss": 7.1046,
+      "step": 4300
+    },
+    {
+      "epoch": 0.24744990718236273,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.771642602399229e-05,
+      "loss": 7.1422,
+      "step": 4310
+    },
+    {
+      "epoch": 0.24802403689740302,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.77025161526071e-05,
+      "loss": 7.0738,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2485981666124433,
+      "grad_norm": 1.5078125,
+      "learning_rate": 9.76885650414713e-05,
+      "loss": 7.0958,
+      "step": 4330
+    },
+    {
+      "epoch": 0.24917229632748358,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.76745727026459e-05,
+      "loss": 7.1058,
+      "step": 4340
+    },
+    {
+      "epoch": 0.24974642604252387,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.766053914822754e-05,
+      "loss": 7.1442,
+      "step": 4350
+    },
+    {
+      "epoch": 0.2503205557575642,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.764646439034849e-05,
+      "loss": 7.1123,
+      "step": 4360
+    },
+    {
+      "epoch": 0.25089468547260446,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.763234844117666e-05,
+      "loss": 7.1233,
+      "step": 4370
+    },
+    {
+      "epoch": 0.25146881518764475,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.761819131291557e-05,
+      "loss": 7.1272,
+      "step": 4380
+    },
+    {
+      "epoch": 0.25204294490268503,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.760399301780433e-05,
+      "loss": 7.0923,
+      "step": 4390
+    },
+    {
+      "epoch": 0.2526170746177253,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.758975356811763e-05,
+      "loss": 7.1102,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2531912043327656,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.757547297616576e-05,
+      "loss": 7.1003,
+      "step": 4410
+    },
+    {
+      "epoch": 0.2537653340478059,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.756115125429457e-05,
+      "loss": 7.1449,
+      "step": 4420
+    },
+    {
+      "epoch": 0.25433946376284616,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.754678841488545e-05,
+      "loss": 7.1108,
+      "step": 4430
+    },
+    {
+      "epoch": 0.25491359347788645,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.753238447035537e-05,
+      "loss": 7.1287,
+      "step": 4440
+    },
+    {
+      "epoch": 0.25548772319292673,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.751793943315683e-05,
+      "loss": 7.0953,
+      "step": 4450
+    },
+    {
+      "epoch": 0.256061852907967,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.750345331577781e-05,
+      "loss": 7.1042,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2566359826230073,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.748892613074188e-05,
+      "loss": 7.1171,
+      "step": 4470
+    },
+    {
+      "epoch": 0.2572101123380476,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.747435789060804e-05,
+      "loss": 7.1258,
+      "step": 4480
+    },
+    {
+      "epoch": 0.25778424205308786,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.745974860797084e-05,
+      "loss": 7.116,
+      "step": 4490
+    },
+    {
+      "epoch": 0.25835837176812815,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.744509829546027e-05,
+      "loss": 7.1304,
+      "step": 4500
+    },
+    {
+      "epoch": 0.25893250148316843,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.743040696574182e-05,
+      "loss": 7.1077,
+      "step": 4510
+    },
+    {
+      "epoch": 0.2595066311982087,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.741567463151642e-05,
+      "loss": 7.1422,
+      "step": 4520
+    },
+    {
+      "epoch": 0.260080760913249,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.740090130552046e-05,
+      "loss": 7.1365,
+      "step": 4530
+    },
+    {
+      "epoch": 0.2606548906282893,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.73860870005258e-05,
+      "loss": 7.0826,
+      "step": 4540
+    },
+    {
+      "epoch": 0.26122902034332957,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.737123172933964e-05,
+      "loss": 7.1016,
+      "step": 4550
+    },
+    {
+      "epoch": 0.26180315005836985,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.735633550480469e-05,
+      "loss": 7.1019,
+      "step": 4560
+    },
+    {
+      "epoch": 0.26237727977341013,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.7341398339799e-05,
+      "loss": 7.0796,
+      "step": 4570
+    },
+    {
+      "epoch": 0.2629514094884504,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.732642024723605e-05,
+      "loss": 7.1059,
+      "step": 4580
+    },
+    {
+      "epoch": 0.2635255392034907,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.731140124006471e-05,
+      "loss": 7.1171,
+      "step": 4590
+    },
+    {
+      "epoch": 0.264099668918531,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.729634133126917e-05,
+      "loss": 7.1195,
+      "step": 4600
+    },
+    {
+      "epoch": 0.26467379863357127,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.728124053386905e-05,
+      "loss": 7.1011,
+      "step": 4610
+    },
+    {
+      "epoch": 0.26524792834861155,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.726609886091925e-05,
+      "loss": 7.1241,
+      "step": 4620
+    },
+    {
+      "epoch": 0.26582205806365183,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.725091632551002e-05,
+      "loss": 7.1232,
+      "step": 4630
+    },
+    {
+      "epoch": 0.2663961877786921,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.723569294076702e-05,
+      "loss": 7.1256,
+      "step": 4640
+    },
+    {
+      "epoch": 0.2669703174937324,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.722042871985112e-05,
+      "loss": 7.1163,
+      "step": 4650
+    },
+    {
+      "epoch": 0.2675444472087727,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.720512367595854e-05,
+      "loss": 7.1269,
+      "step": 4660
+    },
+    {
+      "epoch": 0.26811857692381297,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.718977782232079e-05,
+      "loss": 7.1275,
+      "step": 4670
+    },
+    {
+      "epoch": 0.26869270663885325,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.717439117220468e-05,
+      "loss": 7.0919,
+      "step": 4680
+    },
+    {
+      "epoch": 0.26926683635389353,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.715896373891222e-05,
+      "loss": 7.1171,
+      "step": 4690
+    },
+    {
+      "epoch": 0.2698409660689338,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.714349553578077e-05,
+      "loss": 7.0927,
+      "step": 4700
+    },
+    {
+      "epoch": 0.2704150957839741,
+      "grad_norm": 1.625,
+      "learning_rate": 9.712798657618287e-05,
+      "loss": 7.1006,
+      "step": 4710
+    },
+    {
+      "epoch": 0.2709892254990144,
+      "grad_norm": 1.625,
+      "learning_rate": 9.711243687352632e-05,
+      "loss": 7.0835,
+      "step": 4720
+    },
+    {
+      "epoch": 0.27156335521405467,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.709684644125413e-05,
+      "loss": 7.11,
+      "step": 4730
+    },
+    {
+      "epoch": 0.27213748492909495,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.708121529284455e-05,
+      "loss": 7.1328,
+      "step": 4740
+    },
+    {
+      "epoch": 0.2727116146441353,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.706554344181101e-05,
+      "loss": 7.084,
+      "step": 4750
+    },
+    {
+      "epoch": 0.2732857443591756,
+      "grad_norm": 1.625,
+      "learning_rate": 9.70498309017021e-05,
+      "loss": 7.1197,
+      "step": 4760
+    },
+    {
+      "epoch": 0.27385987407421586,
+      "grad_norm": 1.625,
+      "learning_rate": 9.703407768610164e-05,
+      "loss": 7.1251,
+      "step": 4770
+    },
+    {
+      "epoch": 0.27443400378925614,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.70182838086286e-05,
+      "loss": 7.0986,
+      "step": 4780
+    },
+    {
+      "epoch": 0.2750081335042964,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.700244928293708e-05,
+      "loss": 7.1065,
+      "step": 4790
+    },
+    {
+      "epoch": 0.2755822632193367,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.698657412271634e-05,
+      "loss": 7.1378,
+      "step": 4800
+    },
+    {
+      "epoch": 0.276156392934377,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.697065834169075e-05,
+      "loss": 7.096,
+      "step": 4810
+    },
+    {
+      "epoch": 0.2767305226494173,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.695470195361982e-05,
+      "loss": 7.105,
+      "step": 4820
+    },
+    {
+      "epoch": 0.27730465236445756,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.693870497229816e-05,
+      "loss": 7.1321,
+      "step": 4830
+    },
+    {
+      "epoch": 0.27787878207949784,
+      "grad_norm": 1.625,
+      "learning_rate": 9.692266741155547e-05,
+      "loss": 7.104,
+      "step": 4840
+    },
+    {
+      "epoch": 0.2784529117945381,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.690658928525653e-05,
+      "loss": 7.1017,
+      "step": 4850
+    },
+    {
+      "epoch": 0.2790270415095784,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.689047060730119e-05,
+      "loss": 7.1238,
+      "step": 4860
+    },
+    {
+      "epoch": 0.2796011712246187,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.687431139162437e-05,
+      "loss": 7.0816,
+      "step": 4870
+    },
+    {
+      "epoch": 0.280175300939659,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.6858111652196e-05,
+      "loss": 7.0878,
+      "step": 4880
+    },
+    {
+      "epoch": 0.28074943065469926,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.684187140302113e-05,
+      "loss": 7.0907,
+      "step": 4890
+    },
+    {
+      "epoch": 0.28132356036973954,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.682559065813969e-05,
+      "loss": 7.1139,
+      "step": 4900
+    },
+    {
+      "epoch": 0.28189769008477983,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.680926943162674e-05,
+      "loss": 7.1432,
+      "step": 4910
+    },
+    {
+      "epoch": 0.2824718197998201,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.67929077375923e-05,
+      "loss": 7.1069,
+      "step": 4920
+    },
+    {
+      "epoch": 0.2830459495148604,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.677650559018137e-05,
+      "loss": 7.1006,
+      "step": 4930
+    },
+    {
+      "epoch": 0.2836200792299007,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.676006300357392e-05,
+      "loss": 7.1214,
+      "step": 4940
+    },
+    {
+      "epoch": 0.28419420894494096,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.674357999198489e-05,
+      "loss": 7.1161,
+      "step": 4950
+    },
+    {
+      "epoch": 0.28476833865998125,
+      "grad_norm": 1.625,
+      "learning_rate": 9.672705656966417e-05,
+      "loss": 7.0781,
+      "step": 4960
+    },
+    {
+      "epoch": 0.28534246837502153,
+      "grad_norm": 1.625,
+      "learning_rate": 9.671049275089654e-05,
+      "loss": 7.0649,
+      "step": 4970
+    },
+    {
+      "epoch": 0.2859165980900618,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.669388855000178e-05,
+      "loss": 7.0862,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2864907278051021,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.667724398133455e-05,
+      "loss": 7.1193,
+      "step": 4990
+    },
+    {
+      "epoch": 0.2870648575201424,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.666055905928437e-05,
+      "loss": 7.1017,
+      "step": 5000
+    },
+    {
+      "epoch": 0.28763898723518266,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.66438337982757e-05,
+      "loss": 7.0887,
+      "step": 5010
+    },
+    {
+      "epoch": 0.28821311695022295,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.662706821276787e-05,
+      "loss": 7.135,
+      "step": 5020
+    },
+    {
+      "epoch": 0.28878724666526323,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.6610262317255e-05,
+      "loss": 7.1089,
+      "step": 5030
+    },
+    {
+      "epoch": 0.2893613763803035,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.659341612626618e-05,
+      "loss": 7.0617,
+      "step": 5040
+    },
+    {
+      "epoch": 0.2899355060953438,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.657652965436521e-05,
+      "loss": 7.1023,
+      "step": 5050
+    },
+    {
+      "epoch": 0.2905096358103841,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.655960291615081e-05,
+      "loss": 7.0837,
+      "step": 5060
+    },
+    {
+      "epoch": 0.29108376552542436,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.654263592625645e-05,
+      "loss": 7.0944,
+      "step": 5070
+    },
+    {
+      "epoch": 0.29165789524046465,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.652562869935045e-05,
+      "loss": 7.0902,
+      "step": 5080
+    },
+    {
+      "epoch": 0.29223202495550493,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.650858125013584e-05,
+      "loss": 7.0896,
+      "step": 5090
+    },
+    {
+      "epoch": 0.2928061546705452,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.649149359335053e-05,
+      "loss": 7.0806,
+      "step": 5100
+    },
+    {
+      "epoch": 0.2933802843855855,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.647436574376708e-05,
+      "loss": 7.1028,
+      "step": 5110
+    },
+    {
+      "epoch": 0.2939544141006258,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.645719771619288e-05,
+      "loss": 7.0584,
+      "step": 5120
+    },
+    {
+      "epoch": 0.29452854381566607,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.643998952547002e-05,
+      "loss": 7.0907,
+      "step": 5130
+    },
+    {
+      "epoch": 0.29510267353070635,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.642274118647529e-05,
+      "loss": 7.1007,
+      "step": 5140
+    },
+    {
+      "epoch": 0.29567680324574663,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.640545271412024e-05,
+      "loss": 7.1318,
+      "step": 5150
+    },
+    {
+      "epoch": 0.2962509329607869,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.638812412335108e-05,
+      "loss": 7.1222,
+      "step": 5160
+    },
+    {
+      "epoch": 0.2968250626758272,
+      "grad_norm": 1.625,
+      "learning_rate": 9.63707554291487e-05,
+      "loss": 7.0565,
+      "step": 5170
+    },
+    {
+      "epoch": 0.2973991923908675,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.63533466465287e-05,
+      "loss": 7.0986,
+      "step": 5180
+    },
+    {
+      "epoch": 0.29797332210590777,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.633589779054131e-05,
+      "loss": 7.0853,
+      "step": 5190
+    },
+    {
+      "epoch": 0.2985474518209481,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.631840887627138e-05,
+      "loss": 7.1219,
+      "step": 5200
+    },
+    {
+      "epoch": 0.2991215815359884,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.630087991883843e-05,
+      "loss": 7.0556,
+      "step": 5210
+    },
+    {
+      "epoch": 0.2996957112510287,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.628331093339657e-05,
+      "loss": 7.0984,
+      "step": 5220
+    },
+    {
+      "epoch": 0.30026984096606896,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.626570193513456e-05,
+      "loss": 7.0877,
+      "step": 5230
+    },
+    {
+      "epoch": 0.30084397068110924,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.624805293927568e-05,
+      "loss": 7.1067,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3014181003961495,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.623036396107785e-05,
+      "loss": 7.1021,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3019922301111898,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.621263501583356e-05,
+      "loss": 7.1186,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3025663598262301,
+      "grad_norm": 1.5,
+      "learning_rate": 9.619486611886976e-05,
+      "loss": 7.0824,
+      "step": 5270
+    },
+    {
+      "epoch": 0.3031404895412704,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.617705728554807e-05,
+      "loss": 7.1194,
+      "step": 5280
+    },
+    {
+      "epoch": 0.30371461925631066,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.615920853126456e-05,
+      "loss": 7.1096,
+      "step": 5290
+    },
+    {
+      "epoch": 0.30428874897135094,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.61413198714498e-05,
+      "loss": 7.1142,
+      "step": 5300
+    },
+    {
+      "epoch": 0.3048628786863912,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.612339132156889e-05,
+      "loss": 7.0574,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3054370084014315,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.610542289712143e-05,
+      "loss": 7.1176,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3060111381164718,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.608741461364145e-05,
+      "loss": 7.09,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3065852678315121,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.60693664866975e-05,
+      "loss": 7.1117,
+      "step": 5340
+    },
+    {
+      "epoch": 0.30715939754655236,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.605127853189246e-05,
+      "loss": 7.0935,
+      "step": 5350
+    },
+    {
+      "epoch": 0.30773352726159264,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.603315076486378e-05,
+      "loss": 7.0575,
+      "step": 5360
+    },
+    {
+      "epoch": 0.3083076569766329,
+      "grad_norm": 1.625,
+      "learning_rate": 9.601498320128324e-05,
+      "loss": 7.0817,
+      "step": 5370
+    },
+    {
+      "epoch": 0.3088817866916732,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.599677585685707e-05,
+      "loss": 7.0864,
+      "step": 5380
+    },
+    {
+      "epoch": 0.3094559164067135,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.597852874732585e-05,
+      "loss": 7.1067,
+      "step": 5390
+    },
+    {
+      "epoch": 0.3100300461217538,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.596024188846459e-05,
+      "loss": 7.1235,
+      "step": 5400
+    },
+    {
+      "epoch": 0.31060417583679406,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.59419152960826e-05,
+      "loss": 7.1188,
+      "step": 5410
+    },
+    {
+      "epoch": 0.31117830555183434,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.59235489860236e-05,
+      "loss": 7.1008,
+      "step": 5420
+    },
+    {
+      "epoch": 0.31175243526687463,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.590514297416561e-05,
+      "loss": 7.0965,
+      "step": 5430
+    },
+    {
+      "epoch": 0.3123265649819149,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.588669727642099e-05,
+      "loss": 7.1121,
+      "step": 5440
+    },
+    {
+      "epoch": 0.3129006946969552,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.586821190873639e-05,
+      "loss": 7.1177,
+      "step": 5450
+    },
+    {
+      "epoch": 0.3134748244119955,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.584968688709279e-05,
+      "loss": 7.1023,
+      "step": 5460
+    },
+    {
+      "epoch": 0.31404895412703576,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.58311222275054e-05,
+      "loss": 7.1225,
+      "step": 5470
+    },
+    {
+      "epoch": 0.31462308384207605,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.581251794602377e-05,
+      "loss": 7.0741,
+      "step": 5480
+    },
+    {
+      "epoch": 0.31519721355711633,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.579387405873164e-05,
+      "loss": 7.1023,
+      "step": 5490
+    },
+    {
+      "epoch": 0.3157713432721566,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.5775190581747e-05,
+      "loss": 7.1073,
+      "step": 5500
+    },
+    {
+      "epoch": 0.3163454729871969,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.57564675312221e-05,
+      "loss": 7.0511,
+      "step": 5510
+    },
+    {
+      "epoch": 0.3169196027022372,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.573770492334338e-05,
+      "loss": 7.0816,
+      "step": 5520
+    },
+    {
+      "epoch": 0.31749373241727746,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.571890277433144e-05,
+      "loss": 7.1016,
+      "step": 5530
+    },
+    {
+      "epoch": 0.31806786213231775,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.570006110044116e-05,
+      "loss": 7.0907,
+      "step": 5540
+    },
+    {
+      "epoch": 0.31864199184735803,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.568117991796148e-05,
+      "loss": 7.078,
+      "step": 5550
+    },
+    {
+      "epoch": 0.3192161215623983,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.56622592432156e-05,
+      "loss": 7.0899,
+      "step": 5560
+    },
+    {
+      "epoch": 0.3197902512774386,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.564329909256078e-05,
+      "loss": 7.1083,
+      "step": 5570
+    },
+    {
+      "epoch": 0.3203643809924789,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.562429948238842e-05,
+      "loss": 7.0546,
+      "step": 5580
+    },
+    {
+      "epoch": 0.32093851070751916,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.56052604291241e-05,
+      "loss": 7.0755,
+      "step": 5590
+    },
+    {
+      "epoch": 0.32151264042255945,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.55861819492274e-05,
+      "loss": 7.0788,
+      "step": 5600
+    },
+    {
+      "epoch": 0.32208677013759973,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.556706405919208e-05,
+      "loss": 7.0758,
+      "step": 5610
+    },
+    {
+      "epoch": 0.32266089985264,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.55479067755459e-05,
+      "loss": 7.0838,
+      "step": 5620
+    },
+    {
+      "epoch": 0.3232350295676803,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.552871011485071e-05,
+      "loss": 7.0924,
+      "step": 5630
+    },
+    {
+      "epoch": 0.3238091592827206,
+      "grad_norm": 1.625,
+      "learning_rate": 9.550947409370239e-05,
+      "loss": 7.0698,
+      "step": 5640
+    },
+    {
+      "epoch": 0.3243832889977609,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.549019872873087e-05,
+      "loss": 7.0464,
+      "step": 5650
+    },
+    {
+      "epoch": 0.3249574187128012,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.547088403660005e-05,
+      "loss": 7.0699,
+      "step": 5660
+    },
+    {
+      "epoch": 0.3255315484278415,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.545153003400789e-05,
+      "loss": 7.0841,
+      "step": 5670
+    },
+    {
+      "epoch": 0.32610567814288177,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.543213673768627e-05,
+      "loss": 7.0842,
+      "step": 5680
+    },
+    {
+      "epoch": 0.32667980785792206,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.541270416440109e-05,
+      "loss": 7.0989,
+      "step": 5690
+    },
+    {
+      "epoch": 0.32725393757296234,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.539323233095219e-05,
+      "loss": 7.0961,
+      "step": 5700
+    },
+    {
+      "epoch": 0.3278280672880026,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.537372125417333e-05,
+      "loss": 7.0765,
+      "step": 5710
+    },
+    {
+      "epoch": 0.3284021970030429,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.535417095093222e-05,
+      "loss": 7.0622,
+      "step": 5720
+    },
+    {
+      "epoch": 0.3289763267180832,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.533458143813048e-05,
+      "loss": 7.0771,
+      "step": 5730
+    },
+    {
+      "epoch": 0.3295504564331235,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.531495273270363e-05,
+      "loss": 7.07,
+      "step": 5740
+    },
+    {
+      "epoch": 0.33012458614816376,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.529528485162105e-05,
+      "loss": 7.0796,
+      "step": 5750
+    },
+    {
+      "epoch": 0.33069871586320404,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.527557781188602e-05,
+      "loss": 7.1156,
+      "step": 5760
+    },
+    {
+      "epoch": 0.3312728455782443,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.525583163053566e-05,
+      "loss": 7.0224,
+      "step": 5770
+    },
+    {
+      "epoch": 0.3318469752932846,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.523604632464092e-05,
+      "loss": 7.0701,
+      "step": 5780
+    },
+    {
+      "epoch": 0.3324211050083249,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.521622191130656e-05,
+      "loss": 7.0623,
+      "step": 5790
+    },
+    {
+      "epoch": 0.3329952347233652,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.51963584076712e-05,
+      "loss": 7.1004,
+      "step": 5800
+    },
+    {
+      "epoch": 0.33356936443840546,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.517645583090722e-05,
+      "loss": 7.0978,
+      "step": 5810
+    },
+    {
+      "epoch": 0.33414349415344574,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.515651419822077e-05,
+      "loss": 7.0835,
+      "step": 5820
+    },
+    {
+      "epoch": 0.334717623868486,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.513653352685179e-05,
+      "loss": 7.0948,
+      "step": 5830
+    },
+    {
+      "epoch": 0.3352917535835263,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.511651383407395e-05,
+      "loss": 7.0186,
+      "step": 5840
+    },
+    {
+      "epoch": 0.3358658832985666,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.509645513719467e-05,
+      "loss": 7.0853,
+      "step": 5850
+    },
+    {
+      "epoch": 0.3364400130136069,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.507635745355509e-05,
+      "loss": 7.0859,
+      "step": 5860
+    },
+    {
+      "epoch": 0.33701414272864716,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.505622080053003e-05,
+      "loss": 7.0422,
+      "step": 5870
+    },
+    {
+      "epoch": 0.33758827244368744,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.503604519552803e-05,
+      "loss": 7.0574,
+      "step": 5880
+    },
+    {
+      "epoch": 0.3381624021587277,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.501583065599127e-05,
+      "loss": 7.062,
+      "step": 5890
+    },
+    {
+      "epoch": 0.338736531873768,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.499557719939564e-05,
+      "loss": 7.0634,
+      "step": 5900
+    },
+    {
+      "epoch": 0.3393106615888083,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.497528484325062e-05,
+      "loss": 7.0739,
+      "step": 5910
+    },
+    {
+      "epoch": 0.3398847913038486,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.495495360509937e-05,
+      "loss": 7.0782,
+      "step": 5920
+    },
+    {
+      "epoch": 0.34045892101888886,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.493458350251862e-05,
+      "loss": 7.0697,
+      "step": 5930
+    },
+    {
+      "epoch": 0.34103305073392914,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.491417455311875e-05,
+      "loss": 7.0745,
+      "step": 5940
+    },
+    {
+      "epoch": 0.3416071804489694,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.489372677454365e-05,
+      "loss": 7.0831,
+      "step": 5950
+    },
+    {
+      "epoch": 0.3421813101640097,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.487324018447086e-05,
+      "loss": 7.0725,
+      "step": 5960
+    },
+    {
+      "epoch": 0.34275543987905,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.48527148006114e-05,
+      "loss": 7.103,
+      "step": 5970
+    },
+    {
+      "epoch": 0.3433295695940903,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.48321506407099e-05,
+      "loss": 7.0543,
+      "step": 5980
+    },
+    {
+      "epoch": 0.34390369930913056,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.481154772254444e-05,
+      "loss": 7.1007,
+      "step": 5990
+    },
+    {
+      "epoch": 0.34447782902417085,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.479090606392664e-05,
+      "loss": 7.0926,
+      "step": 6000
+    },
+    {
+      "epoch": 0.34505195873921113,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.477022568270166e-05,
+      "loss": 7.0797,
+      "step": 6010
+    },
+    {
+      "epoch": 0.3456260884542514,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.474950659674804e-05,
+      "loss": 7.0643,
+      "step": 6020
+    },
+    {
+      "epoch": 0.3462002181692917,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.472874882397786e-05,
+      "loss": 7.0753,
+      "step": 6030
+    },
+    {
+      "epoch": 0.346774347884332,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.470795238233662e-05,
+      "loss": 7.0945,
+      "step": 6040
+    },
+    {
+      "epoch": 0.34734847759937226,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.468711728980323e-05,
+      "loss": 7.055,
+      "step": 6050
+    },
+    {
+      "epoch": 0.34792260731441255,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.466624356439004e-05,
+      "loss": 7.0759,
+      "step": 6060
+    },
+    {
+      "epoch": 0.34849673702945283,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.46453312241428e-05,
+      "loss": 7.0866,
+      "step": 6070
+    },
+    {
+      "epoch": 0.3490708667444931,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.462438028714061e-05,
+      "loss": 7.0849,
+      "step": 6080
+    },
+    {
+      "epoch": 0.3496449964595334,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.460339077149597e-05,
+      "loss": 7.101,
+      "step": 6090
+    },
+    {
+      "epoch": 0.35021912617457374,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.458236269535476e-05,
+      "loss": 7.0593,
+      "step": 6100
+    },
+    {
+      "epoch": 0.350793255889614,
+      "grad_norm": 1.625,
+      "learning_rate": 9.45612960768961e-05,
+      "loss": 7.0364,
+      "step": 6110
+    },
+    {
+      "epoch": 0.3513673856046543,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.454019093433253e-05,
+      "loss": 7.0575,
+      "step": 6120
+    },
+    {
+      "epoch": 0.3519415153196946,
+      "grad_norm": 1.5,
+      "learning_rate": 9.451904728590983e-05,
+      "loss": 7.1114,
+      "step": 6130
+    },
+    {
+      "epoch": 0.35251564503473487,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.449786514990713e-05,
+      "loss": 7.0604,
+      "step": 6140
+    },
+    {
+      "epoch": 0.35308977474977515,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.447664454463677e-05,
+      "loss": 7.0818,
+      "step": 6150
+    },
+    {
+      "epoch": 0.35366390446481544,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.445538548844436e-05,
+      "loss": 7.0759,
+      "step": 6160
+    },
+    {
+      "epoch": 0.3542380341798557,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.443408799970884e-05,
+      "loss": 7.0768,
+      "step": 6170
+    },
+    {
+      "epoch": 0.354812163894896,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.441275209684219e-05,
+      "loss": 7.0732,
+      "step": 6180
+    },
+    {
+      "epoch": 0.3553862936099363,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.43913777982898e-05,
+      "loss": 7.0547,
+      "step": 6190
+    },
+    {
+      "epoch": 0.35596042332497657,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.436996512253013e-05,
+      "loss": 7.0371,
+      "step": 6200
+    },
+    {
+      "epoch": 0.35653455304001685,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.434851408807487e-05,
+      "loss": 7.0853,
+      "step": 6210
+    },
+    {
+      "epoch": 0.35710868275505714,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.432702471346884e-05,
+      "loss": 7.0577,
+      "step": 6220
+    },
+    {
+      "epoch": 0.3576828124700974,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.430549701729004e-05,
+      "loss": 7.066,
+      "step": 6230
+    },
+    {
+      "epoch": 0.3582569421851377,
+      "grad_norm": 1.625,
+      "learning_rate": 9.428393101814954e-05,
+      "loss": 7.0415,
+      "step": 6240
+    },
+    {
+      "epoch": 0.358831071900178,
+      "grad_norm": 1.625,
+      "learning_rate": 9.426232673469162e-05,
+      "loss": 7.0349,
+      "step": 6250
+    },
+    {
+      "epoch": 0.3594052016152183,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.424068418559356e-05,
+      "loss": 7.0604,
+      "step": 6260
+    },
+    {
+      "epoch": 0.35997933133025856,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.421900338956578e-05,
+      "loss": 7.0384,
+      "step": 6270
+    },
+    {
+      "epoch": 0.36055346104529884,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.419728436535176e-05,
+      "loss": 7.0879,
+      "step": 6280
+    },
+    {
+      "epoch": 0.3611275907603391,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.4175527131728e-05,
+      "loss": 7.0746,
+      "step": 6290
+    },
+    {
+      "epoch": 0.3617017204753794,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.415373170750404e-05,
+      "loss": 7.0225,
+      "step": 6300
+    },
+    {
+      "epoch": 0.3622758501904197,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.413189811152247e-05,
+      "loss": 7.0722,
+      "step": 6310
+    },
+    {
+      "epoch": 0.36284997990546,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.411002636265886e-05,
+      "loss": 7.0454,
+      "step": 6320
+    },
+    {
+      "epoch": 0.36342410962050026,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.408811647982176e-05,
+      "loss": 7.0908,
+      "step": 6330
+    },
+    {
+      "epoch": 0.36399823933554054,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.406616848195266e-05,
+      "loss": 7.0544,
+      "step": 6340
+    },
+    {
+      "epoch": 0.3645723690505808,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.404418238802606e-05,
+      "loss": 7.0017,
+      "step": 6350
+    },
+    {
+      "epoch": 0.3651464987656211,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.402215821704935e-05,
+      "loss": 7.0659,
+      "step": 6360
+    },
+    {
+      "epoch": 0.3657206284806614,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.400009598806287e-05,
+      "loss": 7.0969,
+      "step": 6370
+    },
+    {
+      "epoch": 0.3662947581957017,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.397799572013982e-05,
+      "loss": 7.028,
+      "step": 6380
+    },
+    {
+      "epoch": 0.36686888791074196,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.395585743238633e-05,
+      "loss": 7.0821,
+      "step": 6390
+    },
+    {
+      "epoch": 0.36744301762578224,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.393368114394136e-05,
+      "loss": 7.0553,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3680171473408225,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.391146687397676e-05,
+      "loss": 7.0522,
+      "step": 6410
+    },
+    {
+      "epoch": 0.3685912770558628,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.388921464169719e-05,
+      "loss": 7.017,
+      "step": 6420
+    },
+    {
+      "epoch": 0.3691654067709031,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.386692446634016e-05,
+      "loss": 7.0541,
+      "step": 6430
+    },
+    {
+      "epoch": 0.3697395364859434,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.38445963671759e-05,
+      "loss": 7.069,
+      "step": 6440
+    },
+    {
+      "epoch": 0.37031366620098366,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.382223036350755e-05,
+      "loss": 7.0844,
+      "step": 6450
+    },
+    {
+      "epoch": 0.37088779591602394,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.379982647467091e-05,
+      "loss": 7.0528,
+      "step": 6460
+    },
+    {
+      "epoch": 0.3714619256310642,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.37773847200346e-05,
+      "loss": 7.0136,
+      "step": 6470
+    },
+    {
+      "epoch": 0.3720360553461045,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.375490511899994e-05,
+      "loss": 7.0743,
+      "step": 6480
+    },
+    {
+      "epoch": 0.3726101850611448,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.373238769100098e-05,
+      "loss": 7.0748,
+      "step": 6490
+    },
+    {
+      "epoch": 0.3731843147761851,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.370983245550449e-05,
+      "loss": 7.0398,
+      "step": 6500
+    },
+    {
+      "epoch": 0.37375844449122536,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.36872394320099e-05,
+      "loss": 7.0709,
+      "step": 6510
+    },
+    {
+      "epoch": 0.37433257420626564,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.36646086400493e-05,
+      "loss": 7.0296,
+      "step": 6520
+    },
+    {
+      "epoch": 0.37490670392130593,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.36419400991875e-05,
+      "loss": 7.0657,
+      "step": 6530
+    },
+    {
+      "epoch": 0.3754808336363462,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.361923382902182e-05,
+      "loss": 7.0604,
+      "step": 6540
+    },
+    {
+      "epoch": 0.37605496335138655,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.359648984918232e-05,
+      "loss": 7.0412,
+      "step": 6550
+    },
+    {
+      "epoch": 0.37662909306642683,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.35737081793316e-05,
+      "loss": 7.0582,
+      "step": 6560
+    },
+    {
+      "epoch": 0.3772032227814671,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.355088883916485e-05,
+      "loss": 7.0994,
+      "step": 6570
+    },
+    {
+      "epoch": 0.3777773524965074,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.352803184840983e-05,
+      "loss": 7.0822,
+      "step": 6580
+    },
+    {
+      "epoch": 0.3783514822115477,
+      "grad_norm": 1.484375,
+      "learning_rate": 9.350513722682687e-05,
+      "loss": 7.0721,
+      "step": 6590
+    },
+    {
+      "epoch": 0.37892561192658797,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.34822049942088e-05,
+      "loss": 7.0822,
+      "step": 6600
+    },
+    {
+      "epoch": 0.37949974164162825,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.3459235170381e-05,
+      "loss": 7.0448,
+      "step": 6610
+    },
+    {
+      "epoch": 0.38007387135666854,
+      "grad_norm": 1.625,
+      "learning_rate": 9.343622777520129e-05,
+      "loss": 7.037,
+      "step": 6620
+    },
+    {
+      "epoch": 0.3806480010717088,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.341318282856004e-05,
+      "loss": 7.0463,
+      "step": 6630
+    },
+    {
+      "epoch": 0.3812221307867491,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.339010035038005e-05,
+      "loss": 7.0267,
+      "step": 6640
+    },
+    {
+      "epoch": 0.3817962605017894,
+      "grad_norm": 1.625,
+      "learning_rate": 9.336698036061657e-05,
+      "loss": 7.0255,
+      "step": 6650
+    },
+    {
+      "epoch": 0.38237039021682967,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.334382287925726e-05,
+      "loss": 7.0543,
+      "step": 6660
+    },
+    {
+      "epoch": 0.38294451993186995,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.332062792632223e-05,
+      "loss": 7.0655,
+      "step": 6670
+    },
+    {
+      "epoch": 0.38351864964691024,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.329739552186396e-05,
+      "loss": 7.012,
+      "step": 6680
+    },
+    {
+      "epoch": 0.3840927793619505,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.327412568596735e-05,
+      "loss": 7.0628,
+      "step": 6690
+    },
+    {
+      "epoch": 0.3846669090769908,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.325081843874954e-05,
+      "loss": 7.0552,
+      "step": 6700
+    },
+    {
+      "epoch": 0.3852410387920311,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.322747380036019e-05,
+      "loss": 7.0691,
+      "step": 6710
+    },
+    {
+      "epoch": 0.38581516850707137,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.320409179098113e-05,
+      "loss": 7.0627,
+      "step": 6720
+    },
+    {
+      "epoch": 0.38638929822211165,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.31806724308266e-05,
+      "loss": 7.0231,
+      "step": 6730
+    },
+    {
+      "epoch": 0.38696342793715194,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.315721574014307e-05,
+      "loss": 7.0489,
+      "step": 6740
+    },
+    {
+      "epoch": 0.3875375576521922,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.31337217392093e-05,
+      "loss": 7.0758,
+      "step": 6750
+    },
+    {
+      "epoch": 0.3881116873672325,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.311019044833631e-05,
+      "loss": 7.0796,
+      "step": 6760
+    },
+    {
+      "epoch": 0.3886858170822728,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.308662188786738e-05,
+      "loss": 7.0804,
+      "step": 6770
+    },
+    {
+      "epoch": 0.38925994679731307,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.306301607817797e-05,
+      "loss": 7.0535,
+      "step": 6780
+    },
+    {
+      "epoch": 0.38983407651235336,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.303937303967578e-05,
+      "loss": 7.0569,
+      "step": 6790
+    },
+    {
+      "epoch": 0.39040820622739364,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.301569279280063e-05,
+      "loss": 7.0705,
+      "step": 6800
+    },
+    {
+      "epoch": 0.3909823359424339,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.29919753580246e-05,
+      "loss": 7.0721,
+      "step": 6810
+    },
+    {
+      "epoch": 0.3915564656574742,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.296822075585185e-05,
+      "loss": 7.1056,
+      "step": 6820
+    },
+    {
+      "epoch": 0.3921305953725145,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.29444290068187e-05,
+      "loss": 7.0806,
+      "step": 6830
+    },
+    {
+      "epoch": 0.3927047250875548,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.292060013149357e-05,
+      "loss": 6.9807,
+      "step": 6840
+    },
+    {
+      "epoch": 0.39327885480259506,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.289673415047701e-05,
+      "loss": 7.0552,
+      "step": 6850
+    },
+    {
+      "epoch": 0.39385298451763534,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.287283108440159e-05,
+      "loss": 7.0737,
+      "step": 6860
+    },
+    {
+      "epoch": 0.3944271142326756,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.2848890953932e-05,
+      "loss": 7.0574,
+      "step": 6870
+    },
+    {
+      "epoch": 0.3950012439477159,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.282491377976494e-05,
+      "loss": 7.0671,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3955753736627562,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.280089958262912e-05,
+      "loss": 6.9983,
+      "step": 6890
+    },
+    {
+      "epoch": 0.3961495033777965,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.277684838328532e-05,
+      "loss": 7.0531,
+      "step": 6900
+    },
+    {
+      "epoch": 0.39672363309283676,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.275276020252624e-05,
+      "loss": 7.0274,
+      "step": 6910
+    },
+    {
+      "epoch": 0.39729776280787704,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.272863506117659e-05,
+      "loss": 7.0561,
+      "step": 6920
+    },
+    {
+      "epoch": 0.3978718925229173,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.270447298009301e-05,
+      "loss": 7.0595,
+      "step": 6930
+    },
+    {
+      "epoch": 0.3984460222379576,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.26802739801641e-05,
+      "loss": 7.0799,
+      "step": 6940
+    },
+    {
+      "epoch": 0.3990201519529979,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.265603808231038e-05,
+      "loss": 7.0158,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3995942816680382,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.263176530748422e-05,
+      "loss": 7.0851,
+      "step": 6960
+    },
+    {
+      "epoch": 0.40016841138307846,
+      "grad_norm": 1.75,
+      "learning_rate": 9.260745567666992e-05,
+      "loss": 7.0673,
+      "step": 6970
+    },
+    {
+      "epoch": 0.40074254109811874,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.258310921088363e-05,
+      "loss": 7.0224,
+      "step": 6980
+    },
+    {
+      "epoch": 0.401316670813159,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.255872593117334e-05,
+      "loss": 7.0413,
+      "step": 6990
+    },
+    {
+      "epoch": 0.4018908005281993,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.253430585861887e-05,
+      "loss": 7.0657,
+      "step": 7000
+    },
+    {
+      "epoch": 0.40246493024323965,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.250984901433185e-05,
+      "loss": 6.9849,
+      "step": 7010
+    },
+    {
+      "epoch": 0.40303905995827993,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.248535541945569e-05,
+      "loss": 7.05,
+      "step": 7020
+    },
+    {
+      "epoch": 0.4036131896733202,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.246082509516558e-05,
+      "loss": 7.0623,
+      "step": 7030
+    },
+    {
+      "epoch": 0.4041873193883605,
+      "grad_norm": 1.625,
+      "learning_rate": 9.243625806266845e-05,
+      "loss": 7.0092,
+      "step": 7040
+    },
+    {
+      "epoch": 0.4047614491034008,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.2411654343203e-05,
+      "loss": 7.0477,
+      "step": 7050
+    },
+    {
+      "epoch": 0.40533557881844107,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.238701395803962e-05,
+      "loss": 7.0425,
+      "step": 7060
+    },
+    {
+      "epoch": 0.40590970853348135,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.236233692848035e-05,
+      "loss": 7.0353,
+      "step": 7070
+    },
+    {
+      "epoch": 0.40648383824852163,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.233762327585905e-05,
+      "loss": 7.0285,
+      "step": 7080
+    },
+    {
+      "epoch": 0.4070579679635619,
+      "grad_norm": 1.625,
+      "learning_rate": 9.231287302154107e-05,
+      "loss": 7.0415,
+      "step": 7090
+    },
+    {
+      "epoch": 0.4076320976786022,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.228808618692353e-05,
+      "loss": 7.0454,
+      "step": 7100
+    },
+    {
+      "epoch": 0.4082062273936425,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.226326279343512e-05,
+      "loss": 7.0603,
+      "step": 7110
+    },
+    {
+      "epoch": 0.40878035710868277,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.223840286253613e-05,
+      "loss": 7.0359,
+      "step": 7120
+    },
+    {
+      "epoch": 0.40935448682372305,
+      "grad_norm": 1.625,
+      "learning_rate": 9.221350641571848e-05,
+      "loss": 7.0449,
+      "step": 7130
+    },
+    {
+      "epoch": 0.40992861653876334,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.21885734745056e-05,
+      "loss": 7.0498,
+      "step": 7140
+    },
+    {
+      "epoch": 0.4105027462538036,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.216360406045254e-05,
+      "loss": 7.0622,
+      "step": 7150
+    },
+    {
+      "epoch": 0.4110768759688439,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.213859819514581e-05,
+      "loss": 7.0269,
+      "step": 7160
+    },
+    {
+      "epoch": 0.4116510056838842,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.211355590020348e-05,
+      "loss": 7.061,
+      "step": 7170
+    },
+    {
+      "epoch": 0.41222513539892447,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.208847719727509e-05,
+      "loss": 7.0471,
+      "step": 7180
+    },
+    {
+      "epoch": 0.41279926511396475,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.206336210804167e-05,
+      "loss": 7.0315,
+      "step": 7190
+    },
+    {
+      "epoch": 0.41337339482900504,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.203821065421571e-05,
+      "loss": 7.0379,
+      "step": 7200
+    },
+    {
+      "epoch": 0.4139475245440453,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.201302285754114e-05,
+      "loss": 7.0435,
+      "step": 7210
+    },
+    {
+      "epoch": 0.4145216542590856,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.198779873979326e-05,
+      "loss": 7.016,
+      "step": 7220
+    },
+    {
+      "epoch": 0.4150957839741259,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.196253832277883e-05,
+      "loss": 7.0658,
+      "step": 7230
+    },
+    {
+      "epoch": 0.41566991368916617,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.193724162833598e-05,
+      "loss": 7.0362,
+      "step": 7240
+    },
+    {
+      "epoch": 0.41624404340420645,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.191190867833419e-05,
+      "loss": 7.0765,
+      "step": 7250
+    },
+    {
+      "epoch": 0.41681817311924674,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.188653949467427e-05,
+      "loss": 7.0459,
+      "step": 7260
+    },
+    {
+      "epoch": 0.417392302834287,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.186113409928838e-05,
+      "loss": 7.0007,
+      "step": 7270
+    },
+    {
+      "epoch": 0.4179664325493273,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.183569251413999e-05,
+      "loss": 7.0554,
+      "step": 7280
+    },
+    {
+      "epoch": 0.4185405622643676,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.181021476122385e-05,
+      "loss": 7.0235,
+      "step": 7290
+    },
+    {
+      "epoch": 0.41911469197940787,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.178470086256593e-05,
+      "loss": 7.0363,
+      "step": 7300
+    },
+    {
+      "epoch": 0.41968882169444816,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.175915084022353e-05,
+      "loss": 7.0607,
+      "step": 7310
+    },
+    {
+      "epoch": 0.42026295140948844,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.173356471628511e-05,
+      "loss": 7.0638,
+      "step": 7320
+    },
+    {
+      "epoch": 0.4208370811245287,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.17079425128704e-05,
+      "loss": 7.0519,
+      "step": 7330
+    },
+    {
+      "epoch": 0.421411210839569,
+      "grad_norm": 1.625,
+      "learning_rate": 9.168228425213028e-05,
+      "loss": 7.0528,
+      "step": 7340
+    },
+    {
+      "epoch": 0.4219853405546093,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.165658995624681e-05,
+      "loss": 6.9998,
+      "step": 7350
+    },
+    {
+      "epoch": 0.4225594702696496,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.163085964743321e-05,
+      "loss": 7.0553,
+      "step": 7360
+    },
+    {
+      "epoch": 0.42313359998468986,
+      "grad_norm": 1.5,
+      "learning_rate": 9.160509334793384e-05,
+      "loss": 7.087,
+      "step": 7370
+    },
+    {
+      "epoch": 0.42370772969973014,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.157929108002414e-05,
+      "loss": 7.0492,
+      "step": 7380
+    },
+    {
+      "epoch": 0.4242818594147704,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.155345286601069e-05,
+      "loss": 7.0492,
+      "step": 7390
+    },
+    {
+      "epoch": 0.4248559891298107,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.152757872823113e-05,
+      "loss": 7.0173,
+      "step": 7400
+    },
+    {
+      "epoch": 0.425430118844851,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.150166868905414e-05,
+      "loss": 7.0533,
+      "step": 7410
+    },
+    {
+      "epoch": 0.4260042485598913,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.147572277087948e-05,
+      "loss": 7.03,
+      "step": 7420
+    },
+    {
+      "epoch": 0.42657837827493156,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.144974099613787e-05,
+      "loss": 7.0407,
+      "step": 7430
+    },
+    {
+      "epoch": 0.42715250798997184,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.142372338729108e-05,
+      "loss": 7.0591,
+      "step": 7440
+    },
+    {
+      "epoch": 0.4277266377050121,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.139766996683181e-05,
+      "loss": 7.0383,
+      "step": 7450
+    },
+    {
+      "epoch": 0.42830076742005246,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.137158075728377e-05,
+      "loss": 7.0378,
+      "step": 7460
+    },
+    {
+      "epoch": 0.42887489713509275,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.134545578120157e-05,
+      "loss": 7.0556,
+      "step": 7470
+    },
+    {
+      "epoch": 0.42944902685013303,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.131929506117078e-05,
+      "loss": 7.0647,
+      "step": 7480
+    },
+    {
+      "epoch": 0.4300231565651733,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.129309861980783e-05,
+      "loss": 7.0471,
+      "step": 7490
+    },
+    {
+      "epoch": 0.4305972862802136,
+      "grad_norm": 1.625,
+      "learning_rate": 9.126686647976008e-05,
+      "loss": 7.0375,
+      "step": 7500
+    },
+    {
+      "epoch": 0.4311714159952539,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.124059866370571e-05,
+      "loss": 7.0423,
+      "step": 7510
+    },
+    {
+      "epoch": 0.43174554571029417,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.121429519435374e-05,
+      "loss": 7.0172,
+      "step": 7520
+    },
+    {
+      "epoch": 0.43231967542533445,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.118795609444403e-05,
+      "loss": 7.0096,
+      "step": 7530
+    },
+    {
+      "epoch": 0.43289380514037473,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.116158138674729e-05,
+      "loss": 7.0544,
+      "step": 7540
+    },
+    {
+      "epoch": 0.433467934855415,
+      "grad_norm": 1.625,
+      "learning_rate": 9.11351710940649e-05,
+      "loss": 7.0353,
+      "step": 7550
+    },
+    {
+      "epoch": 0.4340420645704553,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.110872523922911e-05,
+      "loss": 7.0482,
+      "step": 7560
+    },
+    {
+      "epoch": 0.4346161942854956,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.108224384510286e-05,
+      "loss": 7.0514,
+      "step": 7570
+    },
+    {
+      "epoch": 0.43519032400053587,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.105572693457985e-05,
+      "loss": 7.0284,
+      "step": 7580
+    },
+    {
+      "epoch": 0.43576445371557615,
+      "grad_norm": 1.8046875,
+      "learning_rate": 9.102917453058444e-05,
+      "loss": 7.0148,
+      "step": 7590
+    },
+    {
+      "epoch": 0.43633858343061643,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.100258665607171e-05,
+      "loss": 7.0613,
+      "step": 7600
+    },
+    {
+      "epoch": 0.4369127131456567,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.097596333402738e-05,
+      "loss": 7.0254,
+      "step": 7610
+    },
+    {
+      "epoch": 0.437486842860697,
+      "grad_norm": 1.625,
+      "learning_rate": 9.094930458746784e-05,
+      "loss": 7.0785,
+      "step": 7620
+    },
+    {
+      "epoch": 0.4380609725757373,
+      "grad_norm": 1.4765625,
+      "learning_rate": 9.09226104394401e-05,
+      "loss": 7.0231,
+      "step": 7630
+    },
+    {
+      "epoch": 0.43863510229077757,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.089588091302176e-05,
+      "loss": 7.0644,
+      "step": 7640
+    },
+    {
+      "epoch": 0.43920923200581785,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.086911603132103e-05,
+      "loss": 7.0538,
+      "step": 7650
+    },
+    {
+      "epoch": 0.43978336172085813,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.084231581747662e-05,
+      "loss": 7.0289,
+      "step": 7660
+    },
+    {
+      "epoch": 0.4403574914358984,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.081548029465789e-05,
+      "loss": 7.0333,
+      "step": 7670
+    },
+    {
+      "epoch": 0.4409316211509387,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.078860948606464e-05,
+      "loss": 7.0796,
+      "step": 7680
+    },
+    {
+      "epoch": 0.441505750865979,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.076170341492722e-05,
+      "loss": 7.0289,
+      "step": 7690
+    },
+    {
+      "epoch": 0.44207988058101927,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.073476210450646e-05,
+      "loss": 7.0234,
+      "step": 7700
+    },
+    {
+      "epoch": 0.44265401029605955,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.070778557809362e-05,
+      "loss": 7.0468,
+      "step": 7710
+    },
+    {
+      "epoch": 0.44322814001109984,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.068077385901043e-05,
+      "loss": 7.048,
+      "step": 7720
+    },
+    {
+      "epoch": 0.4438022697261401,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.065372697060908e-05,
+      "loss": 7.0085,
+      "step": 7730
+    },
+    {
+      "epoch": 0.4443763994411804,
+      "grad_norm": 1.53125,
+      "learning_rate": 9.062664493627208e-05,
+      "loss": 7.0594,
+      "step": 7740
+    },
+    {
+      "epoch": 0.4449505291562207,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.059952777941241e-05,
+      "loss": 7.0759,
+      "step": 7750
+    },
+    {
+      "epoch": 0.44552465887126097,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.057237552347337e-05,
+      "loss": 7.0201,
+      "step": 7760
+    },
+    {
+      "epoch": 0.44609878858630125,
+      "grad_norm": 1.75,
+      "learning_rate": 9.054518819192862e-05,
+      "loss": 7.029,
+      "step": 7770
+    },
+    {
+      "epoch": 0.44667291830134154,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.051796580828212e-05,
+      "loss": 7.0293,
+      "step": 7780
+    },
+    {
+      "epoch": 0.4472470480163818,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.049070839606813e-05,
+      "loss": 6.9941,
+      "step": 7790
+    },
+    {
+      "epoch": 0.4478211777314221,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.046341597885126e-05,
+      "loss": 7.0569,
+      "step": 7800
+    },
+    {
+      "epoch": 0.4483953074464624,
+      "grad_norm": 1.625,
+      "learning_rate": 9.043608858022631e-05,
+      "loss": 7.0543,
+      "step": 7810
+    },
+    {
+      "epoch": 0.44896943716150267,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.040872622381834e-05,
+      "loss": 7.0131,
+      "step": 7820
+    },
+    {
+      "epoch": 0.44954356687654295,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.038132893328264e-05,
+      "loss": 7.0364,
+      "step": 7830
+    },
+    {
+      "epoch": 0.45011769659158324,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.035389673230472e-05,
+      "loss": 7.0294,
+      "step": 7840
+    },
+    {
+      "epoch": 0.4506918263066235,
+      "grad_norm": 1.75,
+      "learning_rate": 9.032642964460023e-05,
+      "loss": 7.046,
+      "step": 7850
+    },
+    {
+      "epoch": 0.4512659560216638,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.0298927693915e-05,
+      "loss": 7.0224,
+      "step": 7860
+    },
+    {
+      "epoch": 0.4518400857367041,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.027139090402499e-05,
+      "loss": 7.0183,
+      "step": 7870
+    },
+    {
+      "epoch": 0.4524142154517444,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.024381929873631e-05,
+      "loss": 7.037,
+      "step": 7880
+    },
+    {
+      "epoch": 0.45298834516678466,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.021621290188516e-05,
+      "loss": 7.0357,
+      "step": 7890
+    },
+    {
+      "epoch": 0.45356247488182494,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.018857173733776e-05,
+      "loss": 7.0316,
+      "step": 7900
+    },
+    {
+      "epoch": 0.4541366045968653,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.016089582899047e-05,
+      "loss": 7.0234,
+      "step": 7910
+    },
+    {
+      "epoch": 0.45471073431190556,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.013318520076964e-05,
+      "loss": 7.0453,
+      "step": 7920
+    },
+    {
+      "epoch": 0.45528486402694585,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.010543987663165e-05,
+      "loss": 7.0319,
+      "step": 7930
+    },
+    {
+      "epoch": 0.45585899374198613,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.007765988056284e-05,
+      "loss": 7.0268,
+      "step": 7940
+    },
+    {
+      "epoch": 0.4564331234570264,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.00498452365796e-05,
+      "loss": 6.972,
+      "step": 7950
+    },
+    {
+      "epoch": 0.4570072531720667,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.002199596872821e-05,
+      "loss": 7.062,
+      "step": 7960
+    },
+    {
+      "epoch": 0.457581382887107,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.99941121010849e-05,
+      "loss": 7.0066,
+      "step": 7970
+    },
+    {
+      "epoch": 0.45815551260214726,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.996619365775583e-05,
+      "loss": 7.0132,
+      "step": 7980
+    },
+    {
+      "epoch": 0.45872964231718755,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.993824066287699e-05,
+      "loss": 7.0405,
+      "step": 7990
+    },
+    {
+      "epoch": 0.45930377203222783,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.991025314061434e-05,
+      "loss": 7.0209,
+      "step": 8000
+    },
+    {
+      "epoch": 0.4598779017472681,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.988223111516363e-05,
+      "loss": 7.0209,
+      "step": 8010
+    },
+    {
+      "epoch": 0.4604520314623084,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.98541746107504e-05,
+      "loss": 7.0216,
+      "step": 8020
+    },
+    {
+      "epoch": 0.4610261611773487,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.982608365163009e-05,
+      "loss": 7.0717,
+      "step": 8030
+    },
+    {
+      "epoch": 0.46160029089238896,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.979795826208785e-05,
+      "loss": 7.0264,
+      "step": 8040
+    },
+    {
+      "epoch": 0.46217442060742925,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.976979846643865e-05,
+      "loss": 7.0146,
+      "step": 8050
+    },
+    {
+      "epoch": 0.46274855032246953,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.974160428902716e-05,
+      "loss": 7.012,
+      "step": 8060
+    },
+    {
+      "epoch": 0.4633226800375098,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.97133757542278e-05,
+      "loss": 7.0154,
+      "step": 8070
+    },
+    {
+      "epoch": 0.4638968097525501,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.968511288644468e-05,
+      "loss": 6.9963,
+      "step": 8080
+    },
+    {
+      "epoch": 0.4644709394675904,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.96568157101116e-05,
+      "loss": 7.0217,
+      "step": 8090
+    },
+    {
+      "epoch": 0.46504506918263067,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.962848424969201e-05,
+      "loss": 7.0337,
+      "step": 8100
+    },
+    {
+      "epoch": 0.46561919889767095,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.960011852967904e-05,
+      "loss": 7.0316,
+      "step": 8110
+    },
+    {
+      "epoch": 0.46619332861271123,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.957171857459538e-05,
+      "loss": 7.0226,
+      "step": 8120
+    },
+    {
+      "epoch": 0.4667674583277515,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.954328440899334e-05,
+      "loss": 7.0425,
+      "step": 8130
+    },
+    {
+      "epoch": 0.4673415880427918,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.95148160574548e-05,
+      "loss": 7.0096,
+      "step": 8140
+    },
+    {
+      "epoch": 0.4679157177578321,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.948631354459123e-05,
+      "loss": 7.0395,
+      "step": 8150
+    },
+    {
+      "epoch": 0.46848984747287237,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.945777689504357e-05,
+      "loss": 7.0209,
+      "step": 8160
+    },
+    {
+      "epoch": 0.46906397718791265,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.942920613348235e-05,
+      "loss": 7.0099,
+      "step": 8170
+    },
+    {
+      "epoch": 0.46963810690295293,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.940060128460752e-05,
+      "loss": 7.0199,
+      "step": 8180
+    },
+    {
+      "epoch": 0.4702122366179932,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.937196237314853e-05,
+      "loss": 7.0487,
+      "step": 8190
+    },
+    {
+      "epoch": 0.4707863663330335,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.934328942386427e-05,
+      "loss": 7.0168,
+      "step": 8200
+    },
+    {
+      "epoch": 0.4713604960480738,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.931458246154307e-05,
+      "loss": 7.0126,
+      "step": 8210
+    },
+    {
+      "epoch": 0.47193462576311407,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.928584151100265e-05,
+      "loss": 7.0564,
+      "step": 8220
+    },
+    {
+      "epoch": 0.47250875547815435,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.925706659709014e-05,
+      "loss": 7.0058,
+      "step": 8230
+    },
+    {
+      "epoch": 0.47308288519319464,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.922825774468198e-05,
+      "loss": 7.0144,
+      "step": 8240
+    },
+    {
+      "epoch": 0.4736570149082349,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.919941497868398e-05,
+      "loss": 7.0205,
+      "step": 8250
+    },
+    {
+      "epoch": 0.4742311446232752,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.917053832403131e-05,
+      "loss": 7.0391,
+      "step": 8260
+    },
+    {
+      "epoch": 0.4748052743383155,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.914162780568836e-05,
+      "loss": 7.004,
+      "step": 8270
+    },
+    {
+      "epoch": 0.47537940405335577,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.911268344864885e-05,
+      "loss": 6.9943,
+      "step": 8280
+    },
+    {
+      "epoch": 0.47595353376839605,
+      "grad_norm": 1.625,
+      "learning_rate": 8.908370527793573e-05,
+      "loss": 7.0023,
+      "step": 8290
+    },
+    {
+      "epoch": 0.47652766348343634,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.905469331860121e-05,
+      "loss": 6.9991,
+      "step": 8300
+    },
+    {
+      "epoch": 0.4771017931984766,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.902564759572667e-05,
+      "loss": 7.0314,
+      "step": 8310
+    },
+    {
+      "epoch": 0.4776759229135169,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.899656813442273e-05,
+      "loss": 7.0244,
+      "step": 8320
+    },
+    {
+      "epoch": 0.4782500526285572,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.89674549598291e-05,
+      "loss": 6.9878,
+      "step": 8330
+    },
+    {
+      "epoch": 0.47882418234359747,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.893830809711472e-05,
+      "loss": 7.0248,
+      "step": 8340
+    },
+    {
+      "epoch": 0.47939831205863775,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.89091275714776e-05,
+      "loss": 7.0032,
+      "step": 8350
+    },
+    {
+      "epoch": 0.4799724417736781,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.88799134081449e-05,
+      "loss": 7.0167,
+      "step": 8360
+    },
+    {
+      "epoch": 0.4805465714887184,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.88506656323728e-05,
+      "loss": 7.0079,
+      "step": 8370
+    },
+    {
+      "epoch": 0.48112070120375866,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.88213842694466e-05,
+      "loss": 7.0017,
+      "step": 8380
+    },
+    {
+      "epoch": 0.48169483091879894,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.879206934468056e-05,
+      "loss": 7.0265,
+      "step": 8390
+    },
+    {
+      "epoch": 0.4822689606338392,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.876272088341804e-05,
+      "loss": 6.9952,
+      "step": 8400
+    },
+    {
+      "epoch": 0.4828430903488795,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.873333891103135e-05,
+      "loss": 6.9986,
+      "step": 8410
+    },
+    {
+      "epoch": 0.4834172200639198,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.870392345292175e-05,
+      "loss": 6.9903,
+      "step": 8420
+    },
+    {
+      "epoch": 0.4839913497789601,
+      "grad_norm": 1.625,
+      "learning_rate": 8.867447453451952e-05,
+      "loss": 7.0363,
+      "step": 8430
+    },
+    {
+      "epoch": 0.48456547949400036,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.864499218128377e-05,
+      "loss": 7.0042,
+      "step": 8440
+    },
+    {
+      "epoch": 0.48513960920904065,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.86154764187026e-05,
+      "loss": 7.0164,
+      "step": 8450
+    },
+    {
+      "epoch": 0.48571373892408093,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.858592727229295e-05,
+      "loss": 7.0584,
+      "step": 8460
+    },
+    {
+      "epoch": 0.4862878686391212,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.855634476760061e-05,
+      "loss": 7.0018,
+      "step": 8470
+    },
+    {
+      "epoch": 0.4868619983541615,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.852672893020027e-05,
+      "loss": 6.9839,
+      "step": 8480
+    },
+    {
+      "epoch": 0.4874361280692018,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.849707978569537e-05,
+      "loss": 7.004,
+      "step": 8490
+    },
+    {
+      "epoch": 0.48801025778424206,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.846739735971817e-05,
+      "loss": 7.0146,
+      "step": 8500
+    },
+    {
+      "epoch": 0.48858438749928235,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.843768167792971e-05,
+      "loss": 7.0398,
+      "step": 8510
+    },
+    {
+      "epoch": 0.48915851721432263,
+      "grad_norm": 1.625,
+      "learning_rate": 8.840793276601977e-05,
+      "loss": 7.0074,
+      "step": 8520
+    },
+    {
+      "epoch": 0.4897326469293629,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.837815064970687e-05,
+      "loss": 6.9928,
+      "step": 8530
+    },
+    {
+      "epoch": 0.4903067766444032,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.834833535473822e-05,
+      "loss": 7.0171,
+      "step": 8540
+    },
+    {
+      "epoch": 0.4908809063594435,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.831848690688972e-05,
+      "loss": 7.0176,
+      "step": 8550
+    },
+    {
+      "epoch": 0.49145503607448376,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.828860533196593e-05,
+      "loss": 7.0408,
+      "step": 8560
+    },
+    {
+      "epoch": 0.49202916578952405,
+      "grad_norm": 1.625,
+      "learning_rate": 8.825869065580006e-05,
+      "loss": 7.0392,
+      "step": 8570
+    },
+    {
+      "epoch": 0.49260329550456433,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.822874290425391e-05,
+      "loss": 7.0104,
+      "step": 8580
+    },
+    {
+      "epoch": 0.4931774252196046,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.819876210321792e-05,
+      "loss": 7.001,
+      "step": 8590
+    },
+    {
+      "epoch": 0.4937515549346449,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.816874827861103e-05,
+      "loss": 7.0417,
+      "step": 8600
+    },
+    {
+      "epoch": 0.4943256846496852,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.813870145638083e-05,
+      "loss": 7.0276,
+      "step": 8610
+    },
+    {
+      "epoch": 0.49489981436472547,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.810862166250335e-05,
+      "loss": 7.0135,
+      "step": 8620
+    },
+    {
+      "epoch": 0.49547394407976575,
+      "grad_norm": 1.5,
+      "learning_rate": 8.807850892298315e-05,
+      "loss": 7.0043,
+      "step": 8630
+    },
+    {
+      "epoch": 0.49604807379480603,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.804836326385328e-05,
+      "loss": 7.0027,
+      "step": 8640
+    },
+    {
+      "epoch": 0.4966222035098463,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.801818471117528e-05,
+      "loss": 7.0387,
+      "step": 8650
+    },
+    {
+      "epoch": 0.4971963332248866,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.798797329103905e-05,
+      "loss": 7.0251,
+      "step": 8660
+    },
+    {
+      "epoch": 0.4977704629399269,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.795772902956297e-05,
+      "loss": 7.0148,
+      "step": 8670
+    },
+    {
+      "epoch": 0.49834459265496717,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.792745195289378e-05,
+      "loss": 7.0388,
+      "step": 8680
+    },
+    {
+      "epoch": 0.49891872237000745,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.789714208720661e-05,
+      "loss": 7.0125,
+      "step": 8690
+    },
+    {
+      "epoch": 0.49949285208504773,
+      "grad_norm": 1.625,
+      "learning_rate": 8.786679945870491e-05,
+      "loss": 7.0172,
+      "step": 8700
+    },
+    {
+      "epoch": 0.500066981800088,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.78364240936205e-05,
+      "loss": 7.0287,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5006411115151284,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.780601601821345e-05,
+      "loss": 7.0163,
+      "step": 8720
+    },
+    {
+      "epoch": 0.5012152412301686,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.777557525877216e-05,
+      "loss": 7.0036,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5017893709452089,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.774510184161322e-05,
+      "loss": 6.9887,
+      "step": 8740
+    },
+    {
+      "epoch": 0.5023635006602492,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.77145957930815e-05,
+      "loss": 7.0021,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5029376303752895,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.768405713955009e-05,
+      "loss": 7.0171,
+      "step": 8760
+    },
+    {
+      "epoch": 0.5035117600903297,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.765348590742021e-05,
+      "loss": 7.0025,
+      "step": 8770
+    },
+    {
+      "epoch": 0.5040858898053701,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.762288212312133e-05,
+      "loss": 7.0257,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5046600195204103,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.759224581311098e-05,
+      "loss": 7.0121,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5052341492354506,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.756157700387487e-05,
+      "loss": 7.0393,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5058082789504909,
+      "grad_norm": 1.625,
+      "learning_rate": 8.753087572192675e-05,
+      "loss": 7.0138,
+      "step": 8810
+    },
+    {
+      "epoch": 0.5063824086655312,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.750014199380852e-05,
+      "loss": 7.0661,
+      "step": 8820
+    },
+    {
+      "epoch": 0.5069565383805714,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.746937584609003e-05,
+      "loss": 7.0297,
+      "step": 8830
+    },
+    {
+      "epoch": 0.5075306680956118,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.743857730536925e-05,
+      "loss": 7.0149,
+      "step": 8840
+    },
+    {
+      "epoch": 0.508104797810652,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.74077463982721e-05,
+      "loss": 7.0319,
+      "step": 8850
+    },
+    {
+      "epoch": 0.5086789275256923,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.737688315145251e-05,
+      "loss": 7.0174,
+      "step": 8860
+    },
+    {
+      "epoch": 0.5092530572407326,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.734598759159234e-05,
+      "loss": 7.0097,
+      "step": 8870
+    },
+    {
+      "epoch": 0.5098271869557729,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.731505974540139e-05,
+      "loss": 6.9883,
+      "step": 8880
+    },
+    {
+      "epoch": 0.5104013166708131,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.728409963961744e-05,
+      "loss": 7.0213,
+      "step": 8890
+    },
+    {
+      "epoch": 0.5109754463858535,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.725310730100602e-05,
+      "loss": 7.0088,
+      "step": 8900
+    },
+    {
+      "epoch": 0.5115495761008937,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.722208275636068e-05,
+      "loss": 6.9966,
+      "step": 8910
+    },
+    {
+      "epoch": 0.512123705815934,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.71910260325027e-05,
+      "loss": 7.0113,
+      "step": 8920
+    },
+    {
+      "epoch": 0.5126978355309743,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.715993715628122e-05,
+      "loss": 7.0126,
+      "step": 8930
+    },
+    {
+      "epoch": 0.5132719652460146,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.71288161545732e-05,
+      "loss": 7.0036,
+      "step": 8940
+    },
+    {
+      "epoch": 0.5138460949610548,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.709766305428334e-05,
+      "loss": 7.004,
+      "step": 8950
+    },
+    {
+      "epoch": 0.5144202246760952,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.70664778823441e-05,
+      "loss": 7.0153,
+      "step": 8960
+    },
+    {
+      "epoch": 0.5149943543911354,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.703526066571565e-05,
+      "loss": 7.0101,
+      "step": 8970
+    },
+    {
+      "epoch": 0.5155684841061757,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.70040114313859e-05,
+      "loss": 7.0252,
+      "step": 8980
+    },
+    {
+      "epoch": 0.516142613821216,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.697273020637042e-05,
+      "loss": 6.9998,
+      "step": 8990
+    },
+    {
+      "epoch": 0.5167167435362563,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.694141701771241e-05,
+      "loss": 7.0078,
+      "step": 9000
+    },
+    {
+      "epoch": 0.5172908732512965,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.691007189248276e-05,
+      "loss": 7.0193,
+      "step": 9010
+    },
+    {
+      "epoch": 0.5178650029663369,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.687869485777993e-05,
+      "loss": 7.0169,
+      "step": 9020
+    },
+    {
+      "epoch": 0.5184391326813771,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.684728594072995e-05,
+      "loss": 6.9945,
+      "step": 9030
+    },
+    {
+      "epoch": 0.5190132623964174,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.681584516848648e-05,
+      "loss": 7.002,
+      "step": 9040
+    },
+    {
+      "epoch": 0.5195873921114578,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.678437256823065e-05,
+      "loss": 7.0061,
+      "step": 9050
+    },
+    {
+      "epoch": 0.520161521826498,
+      "grad_norm": 1.46875,
+      "learning_rate": 8.675286816717114e-05,
+      "loss": 7.0683,
+      "step": 9060
+    },
+    {
+      "epoch": 0.5207356515415383,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.67213319925441e-05,
+      "loss": 6.9938,
+      "step": 9070
+    },
+    {
+      "epoch": 0.5213097812565786,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.66897640716132e-05,
+      "loss": 7.0302,
+      "step": 9080
+    },
+    {
+      "epoch": 0.5218839109716189,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.66581644316695e-05,
+      "loss": 6.9934,
+      "step": 9090
+    },
+    {
+      "epoch": 0.5224580406866591,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.66265331000315e-05,
+      "loss": 7.0016,
+      "step": 9100
+    },
+    {
+      "epoch": 0.5230321704016995,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.659487010404511e-05,
+      "loss": 7.0131,
+      "step": 9110
+    },
+    {
+      "epoch": 0.5236063001167397,
+      "grad_norm": 1.8203125,
+      "learning_rate": 8.656317547108356e-05,
+      "loss": 6.9904,
+      "step": 9120
+    },
+    {
+      "epoch": 0.52418042983178,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.653144922854755e-05,
+      "loss": 7.0,
+      "step": 9130
+    },
+    {
+      "epoch": 0.5247545595468203,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.649969140386497e-05,
+      "loss": 7.0017,
+      "step": 9140
+    },
+    {
+      "epoch": 0.5253286892618606,
+      "grad_norm": 1.625,
+      "learning_rate": 8.646790202449114e-05,
+      "loss": 6.9942,
+      "step": 9150
+    },
+    {
+      "epoch": 0.5259028189769008,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.64360811179085e-05,
+      "loss": 6.9673,
+      "step": 9160
+    },
+    {
+      "epoch": 0.5264769486919412,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.640422871162693e-05,
+      "loss": 7.0162,
+      "step": 9170
+    },
+    {
+      "epoch": 0.5270510784069814,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.637234483318342e-05,
+      "loss": 7.0132,
+      "step": 9180
+    },
+    {
+      "epoch": 0.5276252081220217,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.634042951014219e-05,
+      "loss": 6.9909,
+      "step": 9190
+    },
+    {
+      "epoch": 0.528199337837062,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.630848277009465e-05,
+      "loss": 6.9867,
+      "step": 9200
+    },
+    {
+      "epoch": 0.5287734675521023,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.627650464065942e-05,
+      "loss": 7.0013,
+      "step": 9210
+    },
+    {
+      "epoch": 0.5293475972671425,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.624449514948216e-05,
+      "loss": 7.0366,
+      "step": 9220
+    },
+    {
+      "epoch": 0.5299217269821829,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.621245432423575e-05,
+      "loss": 7.042,
+      "step": 9230
+    },
+    {
+      "epoch": 0.5304958566972231,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.618038219262006e-05,
+      "loss": 7.017,
+      "step": 9240
+    },
+    {
+      "epoch": 0.5310699864122634,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.614827878236209e-05,
+      "loss": 6.9998,
+      "step": 9250
+    },
+    {
+      "epoch": 0.5316441161273037,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.611614412121584e-05,
+      "loss": 6.9938,
+      "step": 9260
+    },
+    {
+      "epoch": 0.532218245842344,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.608397823696239e-05,
+      "loss": 7.0051,
+      "step": 9270
+    },
+    {
+      "epoch": 0.5327923755573842,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.605178115740975e-05,
+      "loss": 6.9972,
+      "step": 9280
+    },
+    {
+      "epoch": 0.5333665052724246,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.60195529103929e-05,
+      "loss": 7.0035,
+      "step": 9290
+    },
+    {
+      "epoch": 0.5339406349874648,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.598729352377381e-05,
+      "loss": 7.0136,
+      "step": 9300
+    },
+    {
+      "epoch": 0.5345147647025051,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.595500302544133e-05,
+      "loss": 7.0247,
+      "step": 9310
+    },
+    {
+      "epoch": 0.5350888944175454,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.592268144331124e-05,
+      "loss": 6.9363,
+      "step": 9320
+    },
+    {
+      "epoch": 0.5356630241325857,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.589032880532615e-05,
+      "loss": 6.9205,
+      "step": 9330
+    },
+    {
+      "epoch": 0.5362371538476259,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.585794513945557e-05,
+      "loss": 6.9942,
+      "step": 9340
+    },
+    {
+      "epoch": 0.5368112835626663,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.582553047369579e-05,
+      "loss": 7.0191,
+      "step": 9350
+    },
+    {
+      "epoch": 0.5373854132777065,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.579308483606991e-05,
+      "loss": 6.9831,
+      "step": 9360
+    },
+    {
+      "epoch": 0.5379595429927468,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.576060825462784e-05,
+      "loss": 6.9771,
+      "step": 9370
+    },
+    {
+      "epoch": 0.5385336727077871,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.57281007574462e-05,
+      "loss": 7.0119,
+      "step": 9380
+    },
+    {
+      "epoch": 0.5391078024228274,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.569556237262834e-05,
+      "loss": 6.9919,
+      "step": 9390
+    },
+    {
+      "epoch": 0.5396819321378676,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.566299312830433e-05,
+      "loss": 7.0048,
+      "step": 9400
+    },
+    {
+      "epoch": 0.540256061852908,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.563039305263095e-05,
+      "loss": 7.0365,
+      "step": 9410
+    },
+    {
+      "epoch": 0.5408301915679482,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.559776217379154e-05,
+      "loss": 7.0129,
+      "step": 9420
+    },
+    {
+      "epoch": 0.5414043212829885,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.556510051999616e-05,
+      "loss": 6.9662,
+      "step": 9430
+    },
+    {
+      "epoch": 0.5419784509980288,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.553240811948144e-05,
+      "loss": 6.9831,
+      "step": 9440
+    },
+    {
+      "epoch": 0.5425525807130691,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.54996850005106e-05,
+      "loss": 7.0207,
+      "step": 9450
+    },
+    {
+      "epoch": 0.5431267104281093,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.54669311913734e-05,
+      "loss": 6.9951,
+      "step": 9460
+    },
+    {
+      "epoch": 0.5437008401431497,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.543414672038615e-05,
+      "loss": 7.0068,
+      "step": 9470
+    },
+    {
+      "epoch": 0.5442749698581899,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.540133161589165e-05,
+      "loss": 6.9715,
+      "step": 9480
+    },
+    {
+      "epoch": 0.5448490995732302,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.536848590625923e-05,
+      "loss": 6.9811,
+      "step": 9490
+    },
+    {
+      "epoch": 0.5454232292882706,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.53356096198846e-05,
+      "loss": 6.9818,
+      "step": 9500
+    },
+    {
+      "epoch": 0.5459973590033108,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.530270278518997e-05,
+      "loss": 7.032,
+      "step": 9510
+    },
+    {
+      "epoch": 0.5465714887183512,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.52697654306239e-05,
+      "loss": 7.0155,
+      "step": 9520
+    },
+    {
+      "epoch": 0.5471456184333914,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.523679758466144e-05,
+      "loss": 7.0373,
+      "step": 9530
+    },
+    {
+      "epoch": 0.5477197481484317,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.520379927580386e-05,
+      "loss": 7.0093,
+      "step": 9540
+    },
+    {
+      "epoch": 0.548293877863472,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.51707705325789e-05,
+      "loss": 7.0006,
+      "step": 9550
+    },
+    {
+      "epoch": 0.5488680075785123,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.513771138354052e-05,
+      "loss": 6.9994,
+      "step": 9560
+    },
+    {
+      "epoch": 0.5494421372935525,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.5104621857269e-05,
+      "loss": 7.004,
+      "step": 9570
+    },
+    {
+      "epoch": 0.5500162670085929,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.507150198237087e-05,
+      "loss": 7.0039,
+      "step": 9580
+    },
+    {
+      "epoch": 0.5505903967236331,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.503835178747892e-05,
+      "loss": 6.9992,
+      "step": 9590
+    },
+    {
+      "epoch": 0.5511645264386734,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.500517130125212e-05,
+      "loss": 7.0309,
+      "step": 9600
+    },
+    {
+      "epoch": 0.5517386561537136,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.497196055237565e-05,
+      "loss": 6.9839,
+      "step": 9610
+    },
+    {
+      "epoch": 0.552312785868754,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.493871956956083e-05,
+      "loss": 6.9932,
+      "step": 9620
+    },
+    {
+      "epoch": 0.5528869155837942,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.490544838154518e-05,
+      "loss": 7.0001,
+      "step": 9630
+    },
+    {
+      "epoch": 0.5534610452988346,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.487214701709225e-05,
+      "loss": 7.0033,
+      "step": 9640
+    },
+    {
+      "epoch": 0.5540351750138748,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.48388155049917e-05,
+      "loss": 7.0179,
+      "step": 9650
+    },
+    {
+      "epoch": 0.5546093047289151,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.480545387405933e-05,
+      "loss": 7.0158,
+      "step": 9660
+    },
+    {
+      "epoch": 0.5551834344439553,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.477206215313687e-05,
+      "loss": 6.971,
+      "step": 9670
+    },
+    {
+      "epoch": 0.5557575641589957,
+      "grad_norm": 1.625,
+      "learning_rate": 8.473864037109212e-05,
+      "loss": 6.9604,
+      "step": 9680
+    },
+    {
+      "epoch": 0.5563316938740359,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.470518855681886e-05,
+      "loss": 7.0133,
+      "step": 9690
+    },
+    {
+      "epoch": 0.5569058235890763,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.467170673923684e-05,
+      "loss": 6.9575,
+      "step": 9700
+    },
+    {
+      "epoch": 0.5574799533041165,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.463819494729173e-05,
+      "loss": 6.9639,
+      "step": 9710
+    },
+    {
+      "epoch": 0.5580540830191568,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.460465320995513e-05,
+      "loss": 6.995,
+      "step": 9720
+    },
+    {
+      "epoch": 0.558628212734197,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.45710815562245e-05,
+      "loss": 7.0093,
+      "step": 9730
+    },
+    {
+      "epoch": 0.5592023424492374,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.453748001512322e-05,
+      "loss": 6.9967,
+      "step": 9740
+    },
+    {
+      "epoch": 0.5597764721642776,
+      "grad_norm": 1.625,
+      "learning_rate": 8.450384861570047e-05,
+      "loss": 7.0163,
+      "step": 9750
+    },
+    {
+      "epoch": 0.560350601879318,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.447018738703122e-05,
+      "loss": 6.9971,
+      "step": 9760
+    },
+    {
+      "epoch": 0.5609247315943582,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.443649635821629e-05,
+      "loss": 6.9673,
+      "step": 9770
+    },
+    {
+      "epoch": 0.5614988613093985,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.44027755583822e-05,
+      "loss": 6.9829,
+      "step": 9780
+    },
+    {
+      "epoch": 0.5620729910244388,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.436902501668124e-05,
+      "loss": 6.9677,
+      "step": 9790
+    },
+    {
+      "epoch": 0.5626471207394791,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.433524476229142e-05,
+      "loss": 7.0134,
+      "step": 9800
+    },
+    {
+      "epoch": 0.5632212504545193,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.430143482441643e-05,
+      "loss": 7.0174,
+      "step": 9810
+    },
+    {
+      "epoch": 0.5637953801695597,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.42675952322856e-05,
+      "loss": 7.0116,
+      "step": 9820
+    },
+    {
+      "epoch": 0.5643695098845999,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.423372601515391e-05,
+      "loss": 6.9594,
+      "step": 9830
+    },
+    {
+      "epoch": 0.5649436395996402,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.419982720230199e-05,
+      "loss": 6.9933,
+      "step": 9840
+    },
+    {
+      "epoch": 0.5655177693146805,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.416589882303598e-05,
+      "loss": 6.9948,
+      "step": 9850
+    },
+    {
+      "epoch": 0.5660918990297208,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.413194090668766e-05,
+      "loss": 6.9934,
+      "step": 9860
+    },
+    {
+      "epoch": 0.566666028744761,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.409795348261427e-05,
+      "loss": 6.9736,
+      "step": 9870
+    },
+    {
+      "epoch": 0.5672401584598014,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.40639365801986e-05,
+      "loss": 6.9948,
+      "step": 9880
+    },
+    {
+      "epoch": 0.5678142881748416,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.402989022884896e-05,
+      "loss": 6.9928,
+      "step": 9890
+    },
+    {
+      "epoch": 0.5683884178898819,
+      "grad_norm": 1.625,
+      "learning_rate": 8.399581445799905e-05,
+      "loss": 6.9894,
+      "step": 9900
+    },
+    {
+      "epoch": 0.5689625476049222,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.396170929710805e-05,
+      "loss": 6.9837,
+      "step": 9910
+    },
+    {
+      "epoch": 0.5695366773199625,
+      "grad_norm": 944.0,
+      "learning_rate": 8.392757477566051e-05,
+      "loss": 7.0183,
+      "step": 9920
+    },
+    {
+      "epoch": 0.5701108070350027,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.389341092316642e-05,
+      "loss": 6.9961,
+      "step": 9930
+    },
+    {
+      "epoch": 0.5706849367500431,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.385921776916106e-05,
+      "loss": 7.0278,
+      "step": 9940
+    },
+    {
+      "epoch": 0.5712590664650834,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.382499534320509e-05,
+      "loss": 7.014,
+      "step": 9950
+    },
+    {
+      "epoch": 0.5718331961801236,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.379074367488446e-05,
+      "loss": 6.9726,
+      "step": 9960
+    },
+    {
+      "epoch": 0.572407325895164,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.375646279381042e-05,
+      "loss": 7.0086,
+      "step": 9970
+    },
+    {
+      "epoch": 0.5729814556102042,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.372215272961943e-05,
+      "loss": 6.9934,
+      "step": 9980
+    },
+    {
+      "epoch": 0.5735555853252445,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.368781351197321e-05,
+      "loss": 7.0058,
+      "step": 9990
+    },
+    {
+      "epoch": 0.5741297150402848,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.36534451705587e-05,
+      "loss": 6.9995,
+      "step": 10000
+    },
+    {
+      "epoch": 0.5747038447553251,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.361904773508798e-05,
+      "loss": 7.0011,
+      "step": 10010
+    },
+    {
+      "epoch": 0.5752779744703653,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.358462123529829e-05,
+      "loss": 6.9823,
+      "step": 10020
+    },
+    {
+      "epoch": 0.5758521041854057,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.355016570095204e-05,
+      "loss": 7.0104,
+      "step": 10030
+    },
+    {
+      "epoch": 0.5764262339004459,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.351568116183667e-05,
+      "loss": 7.0075,
+      "step": 10040
+    },
+    {
+      "epoch": 0.5770003636154862,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.348116764776475e-05,
+      "loss": 6.9775,
+      "step": 10050
+    },
+    {
+      "epoch": 0.5775744933305265,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.344662518857388e-05,
+      "loss": 6.9956,
+      "step": 10060
+    },
+    {
+      "epoch": 0.5781486230455668,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.34120538141267e-05,
+      "loss": 7.0129,
+      "step": 10070
+    },
+    {
+      "epoch": 0.578722752760607,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.337745355431083e-05,
+      "loss": 6.985,
+      "step": 10080
+    },
+    {
+      "epoch": 0.5792968824756474,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.334282443903886e-05,
+      "loss": 7.025,
+      "step": 10090
+    },
+    {
+      "epoch": 0.5798710121906876,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.330816649824833e-05,
+      "loss": 6.9988,
+      "step": 10100
+    },
+    {
+      "epoch": 0.5804451419057279,
+      "grad_norm": 1.4921875,
+      "learning_rate": 8.32734797619017e-05,
+      "loss": 6.9752,
+      "step": 10110
+    },
+    {
+      "epoch": 0.5810192716207682,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.323876425998633e-05,
+      "loss": 6.9748,
+      "step": 10120
+    },
+    {
+      "epoch": 0.5815934013358085,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.320402002251446e-05,
+      "loss": 7.0053,
+      "step": 10130
+    },
+    {
+      "epoch": 0.5821675310508487,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.316924707952312e-05,
+      "loss": 6.9957,
+      "step": 10140
+    },
+    {
+      "epoch": 0.5827416607658891,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.313444546107423e-05,
+      "loss": 6.9575,
+      "step": 10150
+    },
+    {
+      "epoch": 0.5833157904809293,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.309961519725444e-05,
+      "loss": 6.9751,
+      "step": 10160
+    },
+    {
+      "epoch": 0.5838899201959696,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.30647563181752e-05,
+      "loss": 6.9962,
+      "step": 10170
+    },
+    {
+      "epoch": 0.5844640499110099,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.30298688539727e-05,
+      "loss": 6.9946,
+      "step": 10180
+    },
+    {
+      "epoch": 0.5850381796260502,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.29949528348078e-05,
+      "loss": 6.9892,
+      "step": 10190
+    },
+    {
+      "epoch": 0.5856123093410904,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.296000829086611e-05,
+      "loss": 6.9865,
+      "step": 10200
+    },
+    {
+      "epoch": 0.5861864390561308,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.292503525235785e-05,
+      "loss": 6.9942,
+      "step": 10210
+    },
+    {
+      "epoch": 0.586760568771171,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.289003374951786e-05,
+      "loss": 6.988,
+      "step": 10220
+    },
+    {
+      "epoch": 0.5873346984862113,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.285500381260567e-05,
+      "loss": 6.9678,
+      "step": 10230
+    },
+    {
+      "epoch": 0.5879088282012516,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.28199454719053e-05,
+      "loss": 6.9993,
+      "step": 10240
+    },
+    {
+      "epoch": 0.5884829579162919,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.27848587577254e-05,
+      "loss": 6.9852,
+      "step": 10250
+    },
+    {
+      "epoch": 0.5890570876313321,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.274974370039909e-05,
+      "loss": 6.9722,
+      "step": 10260
+    },
+    {
+      "epoch": 0.5896312173463725,
+      "grad_norm": 1.546875,
+      "learning_rate": 8.271460033028401e-05,
+      "loss": 7.0244,
+      "step": 10270
+    },
+    {
+      "epoch": 0.5902053470614127,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.267942867776233e-05,
+      "loss": 7.0165,
+      "step": 10280
+    },
+    {
+      "epoch": 0.590779476776453,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.264422877324059e-05,
+      "loss": 6.9987,
+      "step": 10290
+    },
+    {
+      "epoch": 0.5913536064914933,
+      "grad_norm": 1.625,
+      "learning_rate": 8.260900064714978e-05,
+      "loss": 6.9803,
+      "step": 10300
+    },
+    {
+      "epoch": 0.5919277362065336,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.257374432994532e-05,
+      "loss": 6.9853,
+      "step": 10310
+    },
+    {
+      "epoch": 0.5925018659215738,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.253845985210697e-05,
+      "loss": 6.9921,
+      "step": 10320
+    },
+    {
+      "epoch": 0.5930759956366142,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.250314724413888e-05,
+      "loss": 7.0052,
+      "step": 10330
+    },
+    {
+      "epoch": 0.5936501253516544,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.246780653656942e-05,
+      "loss": 6.9785,
+      "step": 10340
+    },
+    {
+      "epoch": 0.5942242550666947,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.243243775995138e-05,
+      "loss": 7.0001,
+      "step": 10350
+    },
+    {
+      "epoch": 0.594798384781735,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.239704094486171e-05,
+      "loss": 6.9794,
+      "step": 10360
+    },
+    {
+      "epoch": 0.5953725144967753,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.236161612190167e-05,
+      "loss": 6.9948,
+      "step": 10370
+    },
+    {
+      "epoch": 0.5959466442118155,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.232616332169669e-05,
+      "loss": 6.9464,
+      "step": 10380
+    },
+    {
+      "epoch": 0.5965207739268559,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.229068257489643e-05,
+      "loss": 6.9786,
+      "step": 10390
+    },
+    {
+      "epoch": 0.5970949036418962,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.225517391217464e-05,
+      "loss": 7.0033,
+      "step": 10400
+    },
+    {
+      "epoch": 0.5976690333569364,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.221963736422929e-05,
+      "loss": 6.974,
+      "step": 10410
+    },
+    {
+      "epoch": 0.5982431630719768,
+      "grad_norm": 1.625,
+      "learning_rate": 8.218407296178238e-05,
+      "loss": 6.9626,
+      "step": 10420
+    },
+    {
+      "epoch": 0.598817292787017,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.214848073558006e-05,
+      "loss": 6.9609,
+      "step": 10430
+    },
+    {
+      "epoch": 0.5993914225020573,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.211286071639246e-05,
+      "loss": 6.9778,
+      "step": 10440
+    },
+    {
+      "epoch": 0.5999655522170976,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.207721293501383e-05,
+      "loss": 6.9841,
+      "step": 10450
+    },
+    {
+      "epoch": 0.6005396819321379,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.20415374222623e-05,
+      "loss": 6.9789,
+      "step": 10460
+    },
+    {
+      "epoch": 0.6011138116471781,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.200583420898012e-05,
+      "loss": 6.9939,
+      "step": 10470
+    },
+    {
+      "epoch": 0.6016879413622185,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.197010332603336e-05,
+      "loss": 6.9697,
+      "step": 10480
+    },
+    {
+      "epoch": 0.6022620710772587,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.193434480431206e-05,
+      "loss": 6.975,
+      "step": 10490
+    },
+    {
+      "epoch": 0.602836200792299,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.189855867473018e-05,
+      "loss": 7.0106,
+      "step": 10500
+    },
+    {
+      "epoch": 0.6034103305073393,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.186274496822552e-05,
+      "loss": 6.9875,
+      "step": 10510
+    },
+    {
+      "epoch": 0.6039844602223796,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.182690371575971e-05,
+      "loss": 6.9691,
+      "step": 10520
+    },
+    {
+      "epoch": 0.6045585899374198,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.179103494831821e-05,
+      "loss": 6.9972,
+      "step": 10530
+    },
+    {
+      "epoch": 0.6051327196524602,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.175513869691027e-05,
+      "loss": 6.9925,
+      "step": 10540
+    },
+    {
+      "epoch": 0.6057068493675004,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.171921499256891e-05,
+      "loss": 6.9612,
+      "step": 10550
+    },
+    {
+      "epoch": 0.6062809790825407,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.168326386635083e-05,
+      "loss": 6.9736,
+      "step": 10560
+    },
+    {
+      "epoch": 0.606855108797581,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.164728534933653e-05,
+      "loss": 6.9866,
+      "step": 10570
+    },
+    {
+      "epoch": 0.6074292385126213,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.161127947263007e-05,
+      "loss": 7.0039,
+      "step": 10580
+    },
+    {
+      "epoch": 0.6080033682276615,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.15752462673593e-05,
+      "loss": 7.0022,
+      "step": 10590
+    },
+    {
+      "epoch": 0.6085774979427019,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.153918576467558e-05,
+      "loss": 7.0146,
+      "step": 10600
+    },
+    {
+      "epoch": 0.6091516276577421,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.150309799575394e-05,
+      "loss": 6.998,
+      "step": 10610
+    },
+    {
+      "epoch": 0.6097257573727825,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.146698299179291e-05,
+      "loss": 6.9919,
+      "step": 10620
+    },
+    {
+      "epoch": 0.6102998870878227,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.143084078401467e-05,
+      "loss": 6.9656,
+      "step": 10630
+    },
+    {
+      "epoch": 0.610874016802863,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.139467140366483e-05,
+      "loss": 6.9782,
+      "step": 10640
+    },
+    {
+      "epoch": 0.6114481465179032,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.135847488201251e-05,
+      "loss": 6.9906,
+      "step": 10650
+    },
+    {
+      "epoch": 0.6120222762329436,
+      "grad_norm": 1.5,
+      "learning_rate": 8.132225125035032e-05,
+      "loss": 6.9975,
+      "step": 10660
+    },
+    {
+      "epoch": 0.6125964059479838,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.128600053999431e-05,
+      "loss": 6.9186,
+      "step": 10670
+    },
+    {
+      "epoch": 0.6131705356630242,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.124972278228389e-05,
+      "loss": 6.9996,
+      "step": 10680
+    },
+    {
+      "epoch": 0.6137446653780644,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.121341800858189e-05,
+      "loss": 6.9948,
+      "step": 10690
+    },
+    {
+      "epoch": 0.6143187950931047,
+      "grad_norm": 1.625,
+      "learning_rate": 8.117708625027451e-05,
+      "loss": 6.972,
+      "step": 10700
+    },
+    {
+      "epoch": 0.614892924808145,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.114072753877125e-05,
+      "loss": 7.0159,
+      "step": 10710
+    },
+    {
+      "epoch": 0.6154670545231853,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.110434190550493e-05,
+      "loss": 6.9762,
+      "step": 10720
+    },
+    {
+      "epoch": 0.6160411842382255,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.106792938193162e-05,
+      "loss": 6.9775,
+      "step": 10730
+    },
+    {
+      "epoch": 0.6166153139532659,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.103148999953065e-05,
+      "loss": 6.9968,
+      "step": 10740
+    },
+    {
+      "epoch": 0.6171894436683061,
+      "grad_norm": 1.625,
+      "learning_rate": 8.099502378980459e-05,
+      "loss": 6.9856,
+      "step": 10750
+    },
+    {
+      "epoch": 0.6177635733833464,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.095853078427918e-05,
+      "loss": 7.0284,
+      "step": 10760
+    },
+    {
+      "epoch": 0.6183377030983866,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.092201101450332e-05,
+      "loss": 6.9898,
+      "step": 10770
+    },
+    {
+      "epoch": 0.618911832813427,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.088546451204909e-05,
+      "loss": 6.9919,
+      "step": 10780
+    },
+    {
+      "epoch": 0.6194859625284672,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.084889130851163e-05,
+      "loss": 7.0162,
+      "step": 10790
+    },
+    {
+      "epoch": 0.6200600922435076,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.081229143550917e-05,
+      "loss": 6.9979,
+      "step": 10800
+    },
+    {
+      "epoch": 0.6206342219585478,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.077566492468302e-05,
+      "loss": 6.9826,
+      "step": 10810
+    },
+    {
+      "epoch": 0.6212083516735881,
+      "grad_norm": 1.515625,
+      "learning_rate": 8.073901180769752e-05,
+      "loss": 6.9901,
+      "step": 10820
+    },
+    {
+      "epoch": 0.6217824813886283,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.070233211623999e-05,
+      "loss": 6.9824,
+      "step": 10830
+    },
+    {
+      "epoch": 0.6223566111036687,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.066562588202073e-05,
+      "loss": 6.9176,
+      "step": 10840
+    },
+    {
+      "epoch": 0.622930740818709,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.062889313677302e-05,
+      "loss": 6.9748,
+      "step": 10850
+    },
+    {
+      "epoch": 0.6235048705337493,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.059213391225301e-05,
+      "loss": 7.001,
+      "step": 10860
+    },
+    {
+      "epoch": 0.6240790002487896,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.055534824023976e-05,
+      "loss": 6.9788,
+      "step": 10870
+    },
+    {
+      "epoch": 0.6246531299638298,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.05185361525352e-05,
+      "loss": 6.9838,
+      "step": 10880
+    },
+    {
+      "epoch": 0.6252272596788702,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.04816976809641e-05,
+      "loss": 6.9763,
+      "step": 10890
+    },
+    {
+      "epoch": 0.6258013893939104,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.044483285737401e-05,
+      "loss": 6.9989,
+      "step": 10900
+    },
+    {
+      "epoch": 0.6263755191089507,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.040794171363531e-05,
+      "loss": 7.0087,
+      "step": 10910
+    },
+    {
+      "epoch": 0.626949648823991,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.037102428164112e-05,
+      "loss": 6.9784,
+      "step": 10920
+    },
+    {
+      "epoch": 0.6275237785390313,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.033408059330725e-05,
+      "loss": 6.9524,
+      "step": 10930
+    },
+    {
+      "epoch": 0.6280979082540715,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.029711068057224e-05,
+      "loss": 6.9713,
+      "step": 10940
+    },
+    {
+      "epoch": 0.6286720379691119,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.02601145753973e-05,
+      "loss": 6.9852,
+      "step": 10950
+    },
+    {
+      "epoch": 0.6292461676841521,
+      "grad_norm": 1.625,
+      "learning_rate": 8.022309230976628e-05,
+      "loss": 7.0048,
+      "step": 10960
+    },
+    {
+      "epoch": 0.6298202973991924,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.018604391568564e-05,
+      "loss": 6.982,
+      "step": 10970
+    },
+    {
+      "epoch": 0.6303944271142327,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.014896942518446e-05,
+      "loss": 6.9608,
+      "step": 10980
+    },
+    {
+      "epoch": 0.630968556829273,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.011186887031434e-05,
+      "loss": 6.9604,
+      "step": 10990
+    },
+    {
+      "epoch": 0.6315426865443132,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.007474228314942e-05,
+      "loss": 6.9843,
+      "step": 11000
+    },
+    {
+      "epoch": 0.6321168162593536,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.003758969578636e-05,
+      "loss": 6.988,
+      "step": 11010
+    },
+    {
+      "epoch": 0.6326909459743938,
+      "grad_norm": 1.625,
+      "learning_rate": 8.000041114034431e-05,
+      "loss": 6.9857,
+      "step": 11020
+    },
+    {
+      "epoch": 0.6332650756894341,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.996320664896483e-05,
+      "loss": 6.9709,
+      "step": 11030
+    },
+    {
+      "epoch": 0.6338392054044744,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.992597625381195e-05,
+      "loss": 6.9667,
+      "step": 11040
+    },
+    {
+      "epoch": 0.6344133351195147,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.988871998707204e-05,
+      "loss": 6.9538,
+      "step": 11050
+    },
+    {
+      "epoch": 0.6349874648345549,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.985143788095389e-05,
+      "loss": 6.9634,
+      "step": 11060
+    },
+    {
+      "epoch": 0.6355615945495953,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.981412996768858e-05,
+      "loss": 6.9727,
+      "step": 11070
+    },
+    {
+      "epoch": 0.6361357242646355,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.977679627952953e-05,
+      "loss": 7.0057,
+      "step": 11080
+    },
+    {
+      "epoch": 0.6367098539796758,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.973943684875245e-05,
+      "loss": 6.9734,
+      "step": 11090
+    },
+    {
+      "epoch": 0.6372839836947161,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.970205170765528e-05,
+      "loss": 6.9587,
+      "step": 11100
+    },
+    {
+      "epoch": 0.6378581134097564,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.966464088855822e-05,
+      "loss": 6.9155,
+      "step": 11110
+    },
+    {
+      "epoch": 0.6384322431247966,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.962720442380364e-05,
+      "loss": 6.961,
+      "step": 11120
+    },
+    {
+      "epoch": 0.639006372839837,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.958974234575607e-05,
+      "loss": 6.9626,
+      "step": 11130
+    },
+    {
+      "epoch": 0.6395805025548772,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.955225468680223e-05,
+      "loss": 6.9431,
+      "step": 11140
+    },
+    {
+      "epoch": 0.6401546322699175,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.951474147935091e-05,
+      "loss": 6.9833,
+      "step": 11150
+    },
+    {
+      "epoch": 0.6407287619849578,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.947720275583301e-05,
+      "loss": 6.9595,
+      "step": 11160
+    },
+    {
+      "epoch": 0.6413028916999981,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.943963854870149e-05,
+      "loss": 6.9866,
+      "step": 11170
+    },
+    {
+      "epoch": 0.6418770214150383,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.940204889043135e-05,
+      "loss": 6.9559,
+      "step": 11180
+    },
+    {
+      "epoch": 0.6424511511300787,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.936443381351954e-05,
+      "loss": 6.9778,
+      "step": 11190
+    },
+    {
+      "epoch": 0.6430252808451189,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.932679335048506e-05,
+      "loss": 6.9584,
+      "step": 11200
+    },
+    {
+      "epoch": 0.6435994105601592,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.92891275338688e-05,
+      "loss": 6.9656,
+      "step": 11210
+    },
+    {
+      "epoch": 0.6441735402751995,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.92514363962336e-05,
+      "loss": 6.9864,
+      "step": 11220
+    },
+    {
+      "epoch": 0.6447476699902398,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.921371997016416e-05,
+      "loss": 6.9844,
+      "step": 11230
+    },
+    {
+      "epoch": 0.64532179970528,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.91759782882671e-05,
+      "loss": 6.9804,
+      "step": 11240
+    },
+    {
+      "epoch": 0.6458959294203204,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.913821138317079e-05,
+      "loss": 6.9777,
+      "step": 11250
+    },
+    {
+      "epoch": 0.6464700591353606,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.91004192875255e-05,
+      "loss": 6.974,
+      "step": 11260
+    },
+    {
+      "epoch": 0.6470441888504009,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.906260203400319e-05,
+      "loss": 6.9603,
+      "step": 11270
+    },
+    {
+      "epoch": 0.6476183185654412,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.902475965529763e-05,
+      "loss": 6.9737,
+      "step": 11280
+    },
+    {
+      "epoch": 0.6481924482804815,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.898689218412427e-05,
+      "loss": 6.9711,
+      "step": 11290
+    },
+    {
+      "epoch": 0.6487665779955218,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.894899965322031e-05,
+      "loss": 6.956,
+      "step": 11300
+    },
+    {
+      "epoch": 0.6493407077105621,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.891108209534455e-05,
+      "loss": 6.9562,
+      "step": 11310
+    },
+    {
+      "epoch": 0.6499148374256024,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.887313954327745e-05,
+      "loss": 7.0049,
+      "step": 11320
+    },
+    {
+      "epoch": 0.6504889671406426,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.88351720298211e-05,
+      "loss": 7.0021,
+      "step": 11330
+    },
+    {
+      "epoch": 0.651063096855683,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.879717958779915e-05,
+      "loss": 6.9747,
+      "step": 11340
+    },
+    {
+      "epoch": 0.6516372265707232,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.87591622500568e-05,
+      "loss": 6.9826,
+      "step": 11350
+    },
+    {
+      "epoch": 0.6522113562857635,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.872112004946075e-05,
+      "loss": 6.9759,
+      "step": 11360
+    },
+    {
+      "epoch": 0.6527854860008038,
+      "grad_norm": 1.625,
+      "learning_rate": 7.868305301889927e-05,
+      "loss": 6.9862,
+      "step": 11370
+    },
+    {
+      "epoch": 0.6533596157158441,
+      "grad_norm": 1.5,
+      "learning_rate": 7.864496119128202e-05,
+      "loss": 6.9906,
+      "step": 11380
+    },
+    {
+      "epoch": 0.6539337454308843,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.860684459954011e-05,
+      "loss": 6.978,
+      "step": 11390
+    },
+    {
+      "epoch": 0.6545078751459247,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.856870327662611e-05,
+      "loss": 6.9735,
+      "step": 11400
+    },
+    {
+      "epoch": 0.6550820048609649,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.853053725551389e-05,
+      "loss": 6.9732,
+      "step": 11410
+    },
+    {
+      "epoch": 0.6556561345760052,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.849234656919875e-05,
+      "loss": 7.0027,
+      "step": 11420
+    },
+    {
+      "epoch": 0.6562302642910455,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.845413125069727e-05,
+      "loss": 6.9497,
+      "step": 11430
+    },
+    {
+      "epoch": 0.6568043940060858,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.841589133304732e-05,
+      "loss": 6.9555,
+      "step": 11440
+    },
+    {
+      "epoch": 0.657378523721126,
+      "grad_norm": 1.4921875,
+      "learning_rate": 7.837762684930806e-05,
+      "loss": 6.9795,
+      "step": 11450
+    },
+    {
+      "epoch": 0.6579526534361664,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.833933783255988e-05,
+      "loss": 6.9998,
+      "step": 11460
+    },
+    {
+      "epoch": 0.6585267831512066,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.83010243159044e-05,
+      "loss": 7.0157,
+      "step": 11470
+    },
+    {
+      "epoch": 0.659100912866247,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.826268633246435e-05,
+      "loss": 6.9558,
+      "step": 11480
+    },
+    {
+      "epoch": 0.6596750425812872,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.822432391538371e-05,
+      "loss": 6.9717,
+      "step": 11490
+    },
+    {
+      "epoch": 0.6602491722963275,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.818593709782749e-05,
+      "loss": 6.9613,
+      "step": 11500
+    },
+    {
+      "epoch": 0.6608233020113677,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.814752591298186e-05,
+      "loss": 6.9635,
+      "step": 11510
+    },
+    {
+      "epoch": 0.6613974317264081,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.810909039405402e-05,
+      "loss": 7.0205,
+      "step": 11520
+    },
+    {
+      "epoch": 0.6619715614414483,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.807063057427226e-05,
+      "loss": 6.9883,
+      "step": 11530
+    },
+    {
+      "epoch": 0.6625456911564886,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.803214648688581e-05,
+      "loss": 6.9535,
+      "step": 11540
+    },
+    {
+      "epoch": 0.6631198208715289,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.799363816516491e-05,
+      "loss": 6.944,
+      "step": 11550
+    },
+    {
+      "epoch": 0.6636939505865692,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.795510564240076e-05,
+      "loss": 6.9636,
+      "step": 11560
+    },
+    {
+      "epoch": 0.6642680803016094,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.791654895190548e-05,
+      "loss": 7.007,
+      "step": 11570
+    },
+    {
+      "epoch": 0.6648422100166498,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.787796812701204e-05,
+      "loss": 6.9788,
+      "step": 11580
+    },
+    {
+      "epoch": 0.66541633973169,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.783936320107437e-05,
+      "loss": 6.9309,
+      "step": 11590
+    },
+    {
+      "epoch": 0.6659904694467303,
+      "grad_norm": 1.625,
+      "learning_rate": 7.780073420746712e-05,
+      "loss": 6.9335,
+      "step": 11600
+    },
+    {
+      "epoch": 0.6665645991617706,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.776208117958585e-05,
+      "loss": 6.9752,
+      "step": 11610
+    },
+    {
+      "epoch": 0.6671387288768109,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.772340415084681e-05,
+      "loss": 6.9545,
+      "step": 11620
+    },
+    {
+      "epoch": 0.6677128585918511,
+      "grad_norm": 1.8125,
+      "learning_rate": 7.768470315468707e-05,
+      "loss": 6.9765,
+      "step": 11630
+    },
+    {
+      "epoch": 0.6682869883068915,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.76459782245644e-05,
+      "loss": 6.9908,
+      "step": 11640
+    },
+    {
+      "epoch": 0.6688611180219317,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.760722939395724e-05,
+      "loss": 6.9714,
+      "step": 11650
+    },
+    {
+      "epoch": 0.669435247736972,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.756845669636469e-05,
+      "loss": 6.9427,
+      "step": 11660
+    },
+    {
+      "epoch": 0.6700093774520123,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.752966016530652e-05,
+      "loss": 6.9655,
+      "step": 11670
+    },
+    {
+      "epoch": 0.6705835071670526,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.749083983432308e-05,
+      "loss": 7.0029,
+      "step": 11680
+    },
+    {
+      "epoch": 0.6711576368820928,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.74519957369753e-05,
+      "loss": 6.9661,
+      "step": 11690
+    },
+    {
+      "epoch": 0.6717317665971332,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.741312790684465e-05,
+      "loss": 6.9679,
+      "step": 11700
+    },
+    {
+      "epoch": 0.6723058963121734,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.737423637753313e-05,
+      "loss": 6.9563,
+      "step": 11710
+    },
+    {
+      "epoch": 0.6728800260272138,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.73353211826632e-05,
+      "loss": 6.9657,
+      "step": 11720
+    },
+    {
+      "epoch": 0.673454155742254,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.729638235587783e-05,
+      "loss": 6.9761,
+      "step": 11730
+    },
+    {
+      "epoch": 0.6740282854572943,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.72574199308404e-05,
+      "loss": 6.9707,
+      "step": 11740
+    },
+    {
+      "epoch": 0.6746024151723347,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.721843394123465e-05,
+      "loss": 6.9245,
+      "step": 11750
+    },
+    {
+      "epoch": 0.6751765448873749,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.717942442076473e-05,
+      "loss": 6.9743,
+      "step": 11760
+    },
+    {
+      "epoch": 0.6757506746024152,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.714039140315514e-05,
+      "loss": 6.9635,
+      "step": 11770
+    },
+    {
+      "epoch": 0.6763248043174555,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.710133492215066e-05,
+      "loss": 6.9753,
+      "step": 11780
+    },
+    {
+      "epoch": 0.6768989340324958,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.706225501151641e-05,
+      "loss": 6.9818,
+      "step": 11790
+    },
+    {
+      "epoch": 0.677473063747536,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.702315170503769e-05,
+      "loss": 6.9832,
+      "step": 11800
+    },
+    {
+      "epoch": 0.6780471934625764,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.69840250365201e-05,
+      "loss": 7.0003,
+      "step": 11810
+    },
+    {
+      "epoch": 0.6786213231776166,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.69448750397894e-05,
+      "loss": 6.9929,
+      "step": 11820
+    },
+    {
+      "epoch": 0.6791954528926569,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.690570174869149e-05,
+      "loss": 6.97,
+      "step": 11830
+    },
+    {
+      "epoch": 0.6797695826076972,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.686650519709249e-05,
+      "loss": 6.9556,
+      "step": 11840
+    },
+    {
+      "epoch": 0.6803437123227375,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.682728541887854e-05,
+      "loss": 6.9593,
+      "step": 11850
+    },
+    {
+      "epoch": 0.6809178420377777,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.678804244795593e-05,
+      "loss": 6.9603,
+      "step": 11860
+    },
+    {
+      "epoch": 0.6814919717528181,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.674877631825093e-05,
+      "loss": 6.9669,
+      "step": 11870
+    },
+    {
+      "epoch": 0.6820661014678583,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.670948706370988e-05,
+      "loss": 6.9686,
+      "step": 11880
+    },
+    {
+      "epoch": 0.6826402311828986,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.667017471829914e-05,
+      "loss": 6.9484,
+      "step": 11890
+    },
+    {
+      "epoch": 0.6832143608979389,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.663083931600497e-05,
+      "loss": 6.944,
+      "step": 11900
+    },
+    {
+      "epoch": 0.6837884906129792,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.659148089083357e-05,
+      "loss": 6.9765,
+      "step": 11910
+    },
+    {
+      "epoch": 0.6843626203280194,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.65520994768111e-05,
+      "loss": 7.0007,
+      "step": 11920
+    },
+    {
+      "epoch": 0.6849367500430598,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.651269510798353e-05,
+      "loss": 6.9465,
+      "step": 11930
+    },
+    {
+      "epoch": 0.6855108797581,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.64732678184167e-05,
+      "loss": 6.9472,
+      "step": 11940
+    },
+    {
+      "epoch": 0.6860850094731403,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.64338176421963e-05,
+      "loss": 6.9246,
+      "step": 11950
+    },
+    {
+      "epoch": 0.6866591391881806,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.639434461342773e-05,
+      "loss": 6.9619,
+      "step": 11960
+    },
+    {
+      "epoch": 0.6872332689032209,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.63548487662362e-05,
+      "loss": 6.9673,
+      "step": 11970
+    },
+    {
+      "epoch": 0.6878073986182611,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.631533013476665e-05,
+      "loss": 6.9765,
+      "step": 11980
+    },
+    {
+      "epoch": 0.6883815283333015,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.627578875318372e-05,
+      "loss": 6.9593,
+      "step": 11990
+    },
+    {
+      "epoch": 0.6889556580483417,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.623622465567166e-05,
+      "loss": 6.9462,
+      "step": 12000
+    },
+    {
+      "epoch": 0.689529787763382,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.619663787643441e-05,
+      "loss": 6.9357,
+      "step": 12010
+    },
+    {
+      "epoch": 0.6901039174784223,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.615702844969553e-05,
+      "loss": 6.9577,
+      "step": 12020
+    },
+    {
+      "epoch": 0.6906780471934626,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.611739640969813e-05,
+      "loss": 6.9309,
+      "step": 12030
+    },
+    {
+      "epoch": 0.6912521769085028,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.607774179070485e-05,
+      "loss": 6.9118,
+      "step": 12040
+    },
+    {
+      "epoch": 0.6918263066235432,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.603806462699792e-05,
+      "loss": 6.976,
+      "step": 12050
+    },
+    {
+      "epoch": 0.6924004363385834,
+      "grad_norm": 1.9375,
+      "learning_rate": 7.599836495287898e-05,
+      "loss": 6.9466,
+      "step": 12060
+    },
+    {
+      "epoch": 0.6929745660536237,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.59586428026692e-05,
+      "loss": 6.9853,
+      "step": 12070
+    },
+    {
+      "epoch": 0.693548695768664,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.591889821070913e-05,
+      "loss": 6.9509,
+      "step": 12080
+    },
+    {
+      "epoch": 0.6941228254837043,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.587913121135875e-05,
+      "loss": 6.9566,
+      "step": 12090
+    },
+    {
+      "epoch": 0.6946969551987445,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.583934183899738e-05,
+      "loss": 6.9765,
+      "step": 12100
+    },
+    {
+      "epoch": 0.6952710849137849,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.579953012802374e-05,
+      "loss": 6.9199,
+      "step": 12110
+    },
+    {
+      "epoch": 0.6958452146288251,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.57596961128558e-05,
+      "loss": 6.9372,
+      "step": 12120
+    },
+    {
+      "epoch": 0.6964193443438654,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.571983982793086e-05,
+      "loss": 6.9888,
+      "step": 12130
+    },
+    {
+      "epoch": 0.6969934740589057,
+      "grad_norm": 1.75,
+      "learning_rate": 7.567996130770543e-05,
+      "loss": 6.9627,
+      "step": 12140
+    },
+    {
+      "epoch": 0.697567603773946,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.564006058665525e-05,
+      "loss": 6.9265,
+      "step": 12150
+    },
+    {
+      "epoch": 0.6981417334889862,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.560013769927532e-05,
+      "loss": 6.9392,
+      "step": 12160
+    },
+    {
+      "epoch": 0.6987158632040266,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.556019268007972e-05,
+      "loss": 6.9563,
+      "step": 12170
+    },
+    {
+      "epoch": 0.6992899929190668,
+      "grad_norm": 1.625,
+      "learning_rate": 7.55202255636017e-05,
+      "loss": 6.9561,
+      "step": 12180
+    },
+    {
+      "epoch": 0.6998641226341071,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.548023638439359e-05,
+      "loss": 6.9558,
+      "step": 12190
+    },
+    {
+      "epoch": 0.7004382523491475,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.544022517702684e-05,
+      "loss": 6.9474,
+      "step": 12200
+    },
+    {
+      "epoch": 0.7010123820641877,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.54001919760919e-05,
+      "loss": 6.973,
+      "step": 12210
+    },
+    {
+      "epoch": 0.701586511779228,
+      "grad_norm": 1.625,
+      "learning_rate": 7.536013681619822e-05,
+      "loss": 6.9548,
+      "step": 12220
+    },
+    {
+      "epoch": 0.7021606414942683,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.532005973197431e-05,
+      "loss": 6.9928,
+      "step": 12230
+    },
+    {
+      "epoch": 0.7027347712093086,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.527996075806757e-05,
+      "loss": 6.9436,
+      "step": 12240
+    },
+    {
+      "epoch": 0.7033089009243488,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.523983992914435e-05,
+      "loss": 6.9541,
+      "step": 12250
+    },
+    {
+      "epoch": 0.7038830306393892,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.519969727988984e-05,
+      "loss": 6.9461,
+      "step": 12260
+    },
+    {
+      "epoch": 0.7044571603544294,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.51595328450082e-05,
+      "loss": 6.9568,
+      "step": 12270
+    },
+    {
+      "epoch": 0.7050312900694697,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.511934665922232e-05,
+      "loss": 7.0033,
+      "step": 12280
+    },
+    {
+      "epoch": 0.70560541978451,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.507913875727397e-05,
+      "loss": 6.9849,
+      "step": 12290
+    },
+    {
+      "epoch": 0.7061795494995503,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.503890917392361e-05,
+      "loss": 6.9881,
+      "step": 12300
+    },
+    {
+      "epoch": 0.7067536792145905,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.499865794395057e-05,
+      "loss": 6.9328,
+      "step": 12310
+    },
+    {
+      "epoch": 0.7073278089296309,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.495838510215276e-05,
+      "loss": 6.9288,
+      "step": 12320
+    },
+    {
+      "epoch": 0.7079019386446711,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.491809068334685e-05,
+      "loss": 6.9334,
+      "step": 12330
+    },
+    {
+      "epoch": 0.7084760683597114,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.487777472236815e-05,
+      "loss": 6.9809,
+      "step": 12340
+    },
+    {
+      "epoch": 0.7090501980747517,
+      "grad_norm": 1.75,
+      "learning_rate": 7.48374372540706e-05,
+      "loss": 6.9187,
+      "step": 12350
+    },
+    {
+      "epoch": 0.709624327789792,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.47970783133267e-05,
+      "loss": 6.9398,
+      "step": 12360
+    },
+    {
+      "epoch": 0.7101984575048322,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.475669793502755e-05,
+      "loss": 6.9453,
+      "step": 12370
+    },
+    {
+      "epoch": 0.7107725872198726,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.471629615408278e-05,
+      "loss": 6.9427,
+      "step": 12380
+    },
+    {
+      "epoch": 0.7113467169349128,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.467587300542049e-05,
+      "loss": 6.9699,
+      "step": 12390
+    },
+    {
+      "epoch": 0.7119208466499531,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.463542852398728e-05,
+      "loss": 6.9523,
+      "step": 12400
+    },
+    {
+      "epoch": 0.7124949763649934,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.459496274474822e-05,
+      "loss": 6.9642,
+      "step": 12410
+    },
+    {
+      "epoch": 0.7130691060800337,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.455447570268673e-05,
+      "loss": 6.9485,
+      "step": 12420
+    },
+    {
+      "epoch": 0.7136432357950739,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.451396743280465e-05,
+      "loss": 6.922,
+      "step": 12430
+    },
+    {
+      "epoch": 0.7142173655101143,
+      "grad_norm": 1.625,
+      "learning_rate": 7.447343797012218e-05,
+      "loss": 6.9302,
+      "step": 12440
+    },
+    {
+      "epoch": 0.7147914952251545,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.443288734967782e-05,
+      "loss": 6.9553,
+      "step": 12450
+    },
+    {
+      "epoch": 0.7153656249401948,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.439231560652834e-05,
+      "loss": 6.9574,
+      "step": 12460
+    },
+    {
+      "epoch": 0.7159397546552351,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.435172277574885e-05,
+      "loss": 6.9633,
+      "step": 12470
+    },
+    {
+      "epoch": 0.7165138843702754,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.431110889243259e-05,
+      "loss": 6.9022,
+      "step": 12480
+    },
+    {
+      "epoch": 0.7170880140853156,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.427047399169108e-05,
+      "loss": 6.9603,
+      "step": 12490
+    },
+    {
+      "epoch": 0.717662143800356,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.422981810865397e-05,
+      "loss": 6.9143,
+      "step": 12500
+    },
+    {
+      "epoch": 0.7182362735153962,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.418914127846906e-05,
+      "loss": 6.917,
+      "step": 12510
+    },
+    {
+      "epoch": 0.7188104032304365,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.414844353630226e-05,
+      "loss": 6.9381,
+      "step": 12520
+    },
+    {
+      "epoch": 0.7193845329454768,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.410772491733756e-05,
+      "loss": 6.9498,
+      "step": 12530
+    },
+    {
+      "epoch": 0.7199586626605171,
+      "grad_norm": 1.4375,
+      "learning_rate": 7.406698545677698e-05,
+      "loss": 6.9534,
+      "step": 12540
+    },
+    {
+      "epoch": 0.7205327923755573,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.40262251898406e-05,
+      "loss": 6.9452,
+      "step": 12550
+    },
+    {
+      "epoch": 0.7211069220905977,
+      "grad_norm": 1.46875,
+      "learning_rate": 7.398544415176645e-05,
+      "loss": 6.9431,
+      "step": 12560
+    },
+    {
+      "epoch": 0.7216810518056379,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.394464237781053e-05,
+      "loss": 6.9655,
+      "step": 12570
+    },
+    {
+      "epoch": 0.7222551815206782,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.390381990324674e-05,
+      "loss": 6.9516,
+      "step": 12580
+    },
+    {
+      "epoch": 0.7228293112357185,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.386297676336696e-05,
+      "loss": 6.9533,
+      "step": 12590
+    },
+    {
+      "epoch": 0.7234034409507588,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.382211299348081e-05,
+      "loss": 6.9604,
+      "step": 12600
+    },
+    {
+      "epoch": 0.723977570665799,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.378122862891585e-05,
+      "loss": 6.9449,
+      "step": 12610
+    },
+    {
+      "epoch": 0.7245517003808394,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.37403237050174e-05,
+      "loss": 6.9143,
+      "step": 12620
+    },
+    {
+      "epoch": 0.7251258300958796,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.369939825714856e-05,
+      "loss": 6.9502,
+      "step": 12630
+    },
+    {
+      "epoch": 0.72569995981092,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.365845232069019e-05,
+      "loss": 6.9382,
+      "step": 12640
+    },
+    {
+      "epoch": 0.7262740895259603,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.36174859310408e-05,
+      "loss": 6.9101,
+      "step": 12650
+    },
+    {
+      "epoch": 0.7268482192410005,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.357649912361668e-05,
+      "loss": 6.9055,
+      "step": 12660
+    },
+    {
+      "epoch": 0.7274223489560409,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.353549193385168e-05,
+      "loss": 6.9336,
+      "step": 12670
+    },
+    {
+      "epoch": 0.7279964786710811,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.349446439719734e-05,
+      "loss": 6.9425,
+      "step": 12680
+    },
+    {
+      "epoch": 0.7285706083861214,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.345341654912274e-05,
+      "loss": 6.9389,
+      "step": 12690
+    },
+    {
+      "epoch": 0.7291447381011616,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.341234842511456e-05,
+      "loss": 6.9289,
+      "step": 12700
+    },
+    {
+      "epoch": 0.729718867816202,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.337126006067699e-05,
+      "loss": 6.9337,
+      "step": 12710
+    },
+    {
+      "epoch": 0.7302929975312422,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.333015149133169e-05,
+      "loss": 6.9392,
+      "step": 12720
+    },
+    {
+      "epoch": 0.7308671272462826,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.328902275261785e-05,
+      "loss": 6.9751,
+      "step": 12730
+    },
+    {
+      "epoch": 0.7314412569613228,
+      "grad_norm": 1.5078125,
+      "learning_rate": 7.324787388009204e-05,
+      "loss": 6.9742,
+      "step": 12740
+    },
+    {
+      "epoch": 0.7320153866763631,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.320670490932827e-05,
+      "loss": 6.9776,
+      "step": 12750
+    },
+    {
+      "epoch": 0.7325895163914034,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.31655158759179e-05,
+      "loss": 6.9609,
+      "step": 12760
+    },
+    {
+      "epoch": 0.7331636461064437,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.312430681546966e-05,
+      "loss": 6.9548,
+      "step": 12770
+    },
+    {
+      "epoch": 0.7337377758214839,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.308307776360959e-05,
+      "loss": 6.9416,
+      "step": 12780
+    },
+    {
+      "epoch": 0.7343119055365243,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.3041828755981e-05,
+      "loss": 6.9397,
+      "step": 12790
+    },
+    {
+      "epoch": 0.7348860352515645,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.300055982824443e-05,
+      "loss": 6.938,
+      "step": 12800
+    },
+    {
+      "epoch": 0.7354601649666048,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.295927101607771e-05,
+      "loss": 6.9473,
+      "step": 12810
+    },
+    {
+      "epoch": 0.736034294681645,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.29179623551758e-05,
+      "loss": 6.9411,
+      "step": 12820
+    },
+    {
+      "epoch": 0.7366084243966854,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.287663388125083e-05,
+      "loss": 6.9525,
+      "step": 12830
+    },
+    {
+      "epoch": 0.7371825541117256,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.283528563003208e-05,
+      "loss": 6.9232,
+      "step": 12840
+    },
+    {
+      "epoch": 0.737756683826766,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.27939176372659e-05,
+      "loss": 6.9248,
+      "step": 12850
+    },
+    {
+      "epoch": 0.7383308135418062,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.275252993871576e-05,
+      "loss": 6.9764,
+      "step": 12860
+    },
+    {
+      "epoch": 0.7389049432568465,
+      "grad_norm": 1.625,
+      "learning_rate": 7.27111225701621e-05,
+      "loss": 6.9333,
+      "step": 12870
+    },
+    {
+      "epoch": 0.7394790729718868,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.266969556740239e-05,
+      "loss": 6.9455,
+      "step": 12880
+    },
+    {
+      "epoch": 0.7400532026869271,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.262824896625107e-05,
+      "loss": 6.919,
+      "step": 12890
+    },
+    {
+      "epoch": 0.7406273324019673,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.258678280253954e-05,
+      "loss": 6.9078,
+      "step": 12900
+    },
+    {
+      "epoch": 0.7412014621170077,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.254529711211612e-05,
+      "loss": 6.9439,
+      "step": 12910
+    },
+    {
+      "epoch": 0.7417755918320479,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.2503791930846e-05,
+      "loss": 6.9729,
+      "step": 12920
+    },
+    {
+      "epoch": 0.7423497215470882,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.246226729461117e-05,
+      "loss": 6.9428,
+      "step": 12930
+    },
+    {
+      "epoch": 0.7429238512621285,
+      "grad_norm": 1.625,
+      "learning_rate": 7.242072323931051e-05,
+      "loss": 6.9696,
+      "step": 12940
+    },
+    {
+      "epoch": 0.7434979809771688,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.237915980085966e-05,
+      "loss": 6.9893,
+      "step": 12950
+    },
+    {
+      "epoch": 0.744072110692209,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.233757701519103e-05,
+      "loss": 6.9275,
+      "step": 12960
+    },
+    {
+      "epoch": 0.7446462404072494,
+      "grad_norm": 1.5234375,
+      "learning_rate": 7.229597491825374e-05,
+      "loss": 6.9448,
+      "step": 12970
+    },
+    {
+      "epoch": 0.7452203701222896,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.22543535460136e-05,
+      "loss": 6.9222,
+      "step": 12980
+    },
+    {
+      "epoch": 0.7457944998373299,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.221271293445308e-05,
+      "loss": 6.9139,
+      "step": 12990
+    },
+    {
+      "epoch": 0.7463686295523702,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.217105311957135e-05,
+      "loss": 6.9575,
+      "step": 13000
+    },
+    {
+      "epoch": 0.7469427592674105,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.212937413738408e-05,
+      "loss": 6.9712,
+      "step": 13010
+    },
+    {
+      "epoch": 0.7475168889824507,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.208767602392354e-05,
+      "loss": 6.9427,
+      "step": 13020
+    },
+    {
+      "epoch": 0.7480910186974911,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.204595881523862e-05,
+      "loss": 6.8905,
+      "step": 13030
+    },
+    {
+      "epoch": 0.7486651484125313,
+      "grad_norm": 1.625,
+      "learning_rate": 7.200422254739463e-05,
+      "loss": 6.9344,
+      "step": 13040
+    },
+    {
+      "epoch": 0.7492392781275716,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.196246725647338e-05,
+      "loss": 6.9794,
+      "step": 13050
+    },
+    {
+      "epoch": 0.7498134078426119,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.19206929785731e-05,
+      "loss": 6.9557,
+      "step": 13060
+    },
+    {
+      "epoch": 0.7503875375576522,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.187889974980852e-05,
+      "loss": 6.9585,
+      "step": 13070
+    },
+    {
+      "epoch": 0.7509616672726924,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.183708760631064e-05,
+      "loss": 6.9335,
+      "step": 13080
+    },
+    {
+      "epoch": 0.7515357969877328,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.179525658422693e-05,
+      "loss": 6.9735,
+      "step": 13090
+    },
+    {
+      "epoch": 0.7521099267027731,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.175340671972108e-05,
+      "loss": 6.9229,
+      "step": 13100
+    },
+    {
+      "epoch": 0.7526840564178133,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.17115380489731e-05,
+      "loss": 6.952,
+      "step": 13110
+    },
+    {
+      "epoch": 0.7532581861328537,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.166965060817929e-05,
+      "loss": 6.942,
+      "step": 13120
+    },
+    {
+      "epoch": 0.7538323158478939,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.162774443355218e-05,
+      "loss": 6.9544,
+      "step": 13130
+    },
+    {
+      "epoch": 0.7544064455629342,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.158581956132042e-05,
+      "loss": 6.9458,
+      "step": 13140
+    },
+    {
+      "epoch": 0.7549805752779745,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.154387602772889e-05,
+      "loss": 6.9283,
+      "step": 13150
+    },
+    {
+      "epoch": 0.7555547049930148,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.150191386903861e-05,
+      "loss": 6.9262,
+      "step": 13160
+    },
+    {
+      "epoch": 0.756128834708055,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.145993312152666e-05,
+      "loss": 6.9766,
+      "step": 13170
+    },
+    {
+      "epoch": 0.7567029644230954,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.141793382148621e-05,
+      "loss": 6.9395,
+      "step": 13180
+    },
+    {
+      "epoch": 0.7572770941381356,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.137591600522649e-05,
+      "loss": 6.9313,
+      "step": 13190
+    },
+    {
+      "epoch": 0.7578512238531759,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.133387970907268e-05,
+      "loss": 6.9541,
+      "step": 13200
+    },
+    {
+      "epoch": 0.7584253535682162,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.129182496936602e-05,
+      "loss": 6.9298,
+      "step": 13210
+    },
+    {
+      "epoch": 0.7589994832832565,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.124975182246361e-05,
+      "loss": 6.9438,
+      "step": 13220
+    },
+    {
+      "epoch": 0.7595736129982967,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.120766030473854e-05,
+      "loss": 6.9108,
+      "step": 13230
+    },
+    {
+      "epoch": 0.7601477427133371,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.116555045257969e-05,
+      "loss": 6.9529,
+      "step": 13240
+    },
+    {
+      "epoch": 0.7607218724283773,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.11234223023919e-05,
+      "loss": 6.9413,
+      "step": 13250
+    },
+    {
+      "epoch": 0.7612960021434176,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.108127589059573e-05,
+      "loss": 6.9513,
+      "step": 13260
+    },
+    {
+      "epoch": 0.7618701318584579,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.103911125362762e-05,
+      "loss": 6.9433,
+      "step": 13270
+    },
+    {
+      "epoch": 0.7624442615734982,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.099692842793964e-05,
+      "loss": 6.9542,
+      "step": 13280
+    },
+    {
+      "epoch": 0.7630183912885384,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.095472744999973e-05,
+      "loss": 6.9232,
+      "step": 13290
+    },
+    {
+      "epoch": 0.7635925210035788,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.091250835629143e-05,
+      "loss": 6.8932,
+      "step": 13300
+    },
+    {
+      "epoch": 0.764166650718619,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.087027118331397e-05,
+      "loss": 6.9483,
+      "step": 13310
+    },
+    {
+      "epoch": 0.7647407804336593,
+      "grad_norm": 1.625,
+      "learning_rate": 7.082801596758219e-05,
+      "loss": 6.9503,
+      "step": 13320
+    },
+    {
+      "epoch": 0.7653149101486996,
+      "grad_norm": 1.546875,
+      "learning_rate": 7.078574274562657e-05,
+      "loss": 6.9326,
+      "step": 13330
+    },
+    {
+      "epoch": 0.7658890398637399,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.07434515539931e-05,
+      "loss": 6.9463,
+      "step": 13340
+    },
+    {
+      "epoch": 0.7664631695787801,
+      "grad_norm": 1.625,
+      "learning_rate": 7.070114242924337e-05,
+      "loss": 6.9759,
+      "step": 13350
+    },
+    {
+      "epoch": 0.7670372992938205,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.06588154079544e-05,
+      "loss": 6.903,
+      "step": 13360
+    },
+    {
+      "epoch": 0.7676114290088607,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.061647052671873e-05,
+      "loss": 6.9557,
+      "step": 13370
+    },
+    {
+      "epoch": 0.768185558723901,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.057410782214438e-05,
+      "loss": 6.9652,
+      "step": 13380
+    },
+    {
+      "epoch": 0.7687596884389413,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.053172733085466e-05,
+      "loss": 6.9292,
+      "step": 13390
+    },
+    {
+      "epoch": 0.7693338181539816,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.048932908948839e-05,
+      "loss": 6.9163,
+      "step": 13400
+    },
+    {
+      "epoch": 0.7699079478690218,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.04469131346996e-05,
+      "loss": 6.9463,
+      "step": 13410
+    },
+    {
+      "epoch": 0.7704820775840622,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.040447950315779e-05,
+      "loss": 6.9376,
+      "step": 13420
+    },
+    {
+      "epoch": 0.7710562072991024,
+      "grad_norm": 1.53125,
+      "learning_rate": 7.03620282315476e-05,
+      "loss": 6.9563,
+      "step": 13430
+    },
+    {
+      "epoch": 0.7716303370141427,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.031955935656899e-05,
+      "loss": 6.938,
+      "step": 13440
+    },
+    {
+      "epoch": 0.772204466729183,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.027707291493711e-05,
+      "loss": 6.9266,
+      "step": 13450
+    },
+    {
+      "epoch": 0.7727785964442233,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.023456894338235e-05,
+      "loss": 6.9203,
+      "step": 13460
+    },
+    {
+      "epoch": 0.7733527261592635,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.01920474786502e-05,
+      "loss": 6.8945,
+      "step": 13470
+    },
+    {
+      "epoch": 0.7739268558743039,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.01495085575013e-05,
+      "loss": 6.9758,
+      "step": 13480
+    },
+    {
+      "epoch": 0.7745009855893441,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.010695221671135e-05,
+      "loss": 6.9387,
+      "step": 13490
+    },
+    {
+      "epoch": 0.7750751153043844,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.006437849307115e-05,
+      "loss": 6.9299,
+      "step": 13500
+    },
+    {
+      "epoch": 0.7756492450194247,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.00217874233865e-05,
+      "loss": 6.9433,
+      "step": 13510
+    },
+    {
+      "epoch": 0.776223374734465,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.997917904447823e-05,
+      "loss": 6.9512,
+      "step": 13520
+    },
+    {
+      "epoch": 0.7767975044495052,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.993655339318208e-05,
+      "loss": 6.9717,
+      "step": 13530
+    },
+    {
+      "epoch": 0.7773716341645456,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.989391050634877e-05,
+      "loss": 6.9223,
+      "step": 13540
+    },
+    {
+      "epoch": 0.7779457638795859,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.985125042084388e-05,
+      "loss": 6.9604,
+      "step": 13550
+    },
+    {
+      "epoch": 0.7785198935946261,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.980857317354792e-05,
+      "loss": 6.9727,
+      "step": 13560
+    },
+    {
+      "epoch": 0.7790940233096665,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.976587880135617e-05,
+      "loss": 6.9454,
+      "step": 13570
+    },
+    {
+      "epoch": 0.7796681530247067,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.972316734117874e-05,
+      "loss": 6.9135,
+      "step": 13580
+    },
+    {
+      "epoch": 0.780242282739747,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.968043882994054e-05,
+      "loss": 6.9029,
+      "step": 13590
+    },
+    {
+      "epoch": 0.7808164124547873,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.963769330458117e-05,
+      "loss": 6.9762,
+      "step": 13600
+    },
+    {
+      "epoch": 0.7813905421698276,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.959493080205499e-05,
+      "loss": 6.9443,
+      "step": 13610
+    },
+    {
+      "epoch": 0.7819646718848678,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.9552151359331e-05,
+      "loss": 6.9077,
+      "step": 13620
+    },
+    {
+      "epoch": 0.7825388015999082,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.950935501339284e-05,
+      "loss": 6.9295,
+      "step": 13630
+    },
+    {
+      "epoch": 0.7831129313149484,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.946654180123883e-05,
+      "loss": 6.9423,
+      "step": 13640
+    },
+    {
+      "epoch": 0.7836870610299888,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.942371175988178e-05,
+      "loss": 6.9018,
+      "step": 13650
+    },
+    {
+      "epoch": 0.784261190745029,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.93808649263491e-05,
+      "loss": 6.9466,
+      "step": 13660
+    },
+    {
+      "epoch": 0.7848353204600693,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.933800133768274e-05,
+      "loss": 6.9222,
+      "step": 13670
+    },
+    {
+      "epoch": 0.7854094501751095,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.929512103093905e-05,
+      "loss": 6.892,
+      "step": 13680
+    },
+    {
+      "epoch": 0.7859835798901499,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.925222404318892e-05,
+      "loss": 6.9525,
+      "step": 13690
+    },
+    {
+      "epoch": 0.7865577096051901,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.920931041151764e-05,
+      "loss": 6.9537,
+      "step": 13700
+    },
+    {
+      "epoch": 0.7871318393202305,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.916638017302484e-05,
+      "loss": 6.9071,
+      "step": 13710
+    },
+    {
+      "epoch": 0.7877059690352707,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.912343336482456e-05,
+      "loss": 6.9564,
+      "step": 13720
+    },
+    {
+      "epoch": 0.788280098750311,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.908047002404517e-05,
+      "loss": 6.937,
+      "step": 13730
+    },
+    {
+      "epoch": 0.7888542284653512,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.903749018782928e-05,
+      "loss": 6.983,
+      "step": 13740
+    },
+    {
+      "epoch": 0.7894283581803916,
+      "grad_norm": 1.625,
+      "learning_rate": 6.899449389333382e-05,
+      "loss": 6.9388,
+      "step": 13750
+    },
+    {
+      "epoch": 0.7900024878954318,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.89514811777299e-05,
+      "loss": 6.9097,
+      "step": 13760
+    },
+    {
+      "epoch": 0.7905766176104722,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.890845207820286e-05,
+      "loss": 6.9503,
+      "step": 13770
+    },
+    {
+      "epoch": 0.7911507473255124,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.886540663195218e-05,
+      "loss": 6.9628,
+      "step": 13780
+    },
+    {
+      "epoch": 0.7917248770405527,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.882234487619149e-05,
+      "loss": 6.921,
+      "step": 13790
+    },
+    {
+      "epoch": 0.792299006755593,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.877926684814853e-05,
+      "loss": 6.901,
+      "step": 13800
+    },
+    {
+      "epoch": 0.7928731364706333,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.873617258506504e-05,
+      "loss": 6.9464,
+      "step": 13810
+    },
+    {
+      "epoch": 0.7934472661856735,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.86930621241969e-05,
+      "loss": 6.9005,
+      "step": 13820
+    },
+    {
+      "epoch": 0.7940213959007139,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.864993550281393e-05,
+      "loss": 6.9325,
+      "step": 13830
+    },
+    {
+      "epoch": 0.7945955256157541,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.86067927581999e-05,
+      "loss": 6.9075,
+      "step": 13840
+    },
+    {
+      "epoch": 0.7951696553307944,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.856363392765257e-05,
+      "loss": 6.8993,
+      "step": 13850
+    },
+    {
+      "epoch": 0.7957437850458347,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.85204590484836e-05,
+      "loss": 6.9175,
+      "step": 13860
+    },
+    {
+      "epoch": 0.796317914760875,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.84772681580185e-05,
+      "loss": 6.9469,
+      "step": 13870
+    },
+    {
+      "epoch": 0.7968920444759152,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.843406129359661e-05,
+      "loss": 6.9306,
+      "step": 13880
+    },
+    {
+      "epoch": 0.7974661741909556,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.839083849257113e-05,
+      "loss": 6.925,
+      "step": 13890
+    },
+    {
+      "epoch": 0.7980403039059958,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.8347599792309e-05,
+      "loss": 6.9039,
+      "step": 13900
+    },
+    {
+      "epoch": 0.7986144336210361,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.830434523019091e-05,
+      "loss": 6.9328,
+      "step": 13910
+    },
+    {
+      "epoch": 0.7991885633360764,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.826107484361129e-05,
+      "loss": 6.9284,
+      "step": 13920
+    },
+    {
+      "epoch": 0.7997626930511167,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.821778866997822e-05,
+      "loss": 6.9301,
+      "step": 13930
+    },
+    {
+      "epoch": 0.8003368227661569,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.817448674671341e-05,
+      "loss": 6.9278,
+      "step": 13940
+    },
+    {
+      "epoch": 0.8009109524811973,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.813116911125225e-05,
+      "loss": 6.9447,
+      "step": 13950
+    },
+    {
+      "epoch": 0.8014850821962375,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.808783580104365e-05,
+      "loss": 6.9523,
+      "step": 13960
+    },
+    {
+      "epoch": 0.8020592119112778,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.804448685355011e-05,
+      "loss": 6.9536,
+      "step": 13970
+    },
+    {
+      "epoch": 0.802633341626318,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.800112230624764e-05,
+      "loss": 6.9397,
+      "step": 13980
+    },
+    {
+      "epoch": 0.8032074713413584,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.795774219662569e-05,
+      "loss": 6.9319,
+      "step": 13990
+    },
+    {
+      "epoch": 0.8037816010563986,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.791434656218729e-05,
+      "loss": 7.0099,
+      "step": 14000
+    },
+    {
+      "epoch": 0.804355730771439,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.787093544044873e-05,
+      "loss": 6.9383,
+      "step": 14010
+    },
+    {
+      "epoch": 0.8049298604864793,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.782750886893981e-05,
+      "loss": 6.952,
+      "step": 14020
+    },
+    {
+      "epoch": 0.8055039902015195,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.778406688520362e-05,
+      "loss": 6.9498,
+      "step": 14030
+    },
+    {
+      "epoch": 0.8060781199165599,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.774060952679661e-05,
+      "loss": 6.9687,
+      "step": 14040
+    },
+    {
+      "epoch": 0.8066522496316001,
+      "grad_norm": 1.484375,
+      "learning_rate": 6.769713683128851e-05,
+      "loss": 6.9473,
+      "step": 14050
+    },
+    {
+      "epoch": 0.8072263793466404,
+      "grad_norm": 1.625,
+      "learning_rate": 6.76536488362623e-05,
+      "loss": 6.924,
+      "step": 14060
+    },
+    {
+      "epoch": 0.8078005090616807,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.761014557931421e-05,
+      "loss": 6.9409,
+      "step": 14070
+    },
+    {
+      "epoch": 0.808374638776721,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.756662709805363e-05,
+      "loss": 6.9435,
+      "step": 14080
+    },
+    {
+      "epoch": 0.8089487684917612,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.752309343010316e-05,
+      "loss": 6.9117,
+      "step": 14090
+    },
+    {
+      "epoch": 0.8095228982068016,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.747954461309847e-05,
+      "loss": 6.9805,
+      "step": 14100
+    },
+    {
+      "epoch": 0.8100970279218418,
+      "grad_norm": 1.625,
+      "learning_rate": 6.743598068468837e-05,
+      "loss": 6.9137,
+      "step": 14110
+    },
+    {
+      "epoch": 0.8106711576368821,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.739240168253471e-05,
+      "loss": 6.9724,
+      "step": 14120
+    },
+    {
+      "epoch": 0.8112452873519224,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.734880764431242e-05,
+      "loss": 6.8846,
+      "step": 14130
+    },
+    {
+      "epoch": 0.8118194170669627,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.730519860770935e-05,
+      "loss": 6.9561,
+      "step": 14140
+    },
+    {
+      "epoch": 0.8123935467820029,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.726157461042637e-05,
+      "loss": 6.9365,
+      "step": 14150
+    },
+    {
+      "epoch": 0.8129676764970433,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.721793569017727e-05,
+      "loss": 6.8861,
+      "step": 14160
+    },
+    {
+      "epoch": 0.8135418062120835,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.717428188468875e-05,
+      "loss": 6.9391,
+      "step": 14170
+    },
+    {
+      "epoch": 0.8141159359271238,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.713061323170038e-05,
+      "loss": 6.9643,
+      "step": 14180
+    },
+    {
+      "epoch": 0.8146900656421641,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.708692976896454e-05,
+      "loss": 6.9478,
+      "step": 14190
+    },
+    {
+      "epoch": 0.8152641953572044,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.704323153424643e-05,
+      "loss": 6.8882,
+      "step": 14200
+    },
+    {
+      "epoch": 0.8158383250722446,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.699951856532405e-05,
+      "loss": 6.9549,
+      "step": 14210
+    },
+    {
+      "epoch": 0.816412454787285,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.695579089998808e-05,
+      "loss": 6.9198,
+      "step": 14220
+    },
+    {
+      "epoch": 0.8169865845023252,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.691204857604195e-05,
+      "loss": 6.9332,
+      "step": 14230
+    },
+    {
+      "epoch": 0.8175607142173655,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.686829163130173e-05,
+      "loss": 6.968,
+      "step": 14240
+    },
+    {
+      "epoch": 0.8181348439324058,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.682452010359616e-05,
+      "loss": 6.9519,
+      "step": 14250
+    },
+    {
+      "epoch": 0.8187089736474461,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.678073403076658e-05,
+      "loss": 6.9242,
+      "step": 14260
+    },
+    {
+      "epoch": 0.8192831033624863,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.673693345066691e-05,
+      "loss": 6.961,
+      "step": 14270
+    },
+    {
+      "epoch": 0.8198572330775267,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.669311840116357e-05,
+      "loss": 6.9111,
+      "step": 14280
+    },
+    {
+      "epoch": 0.8204313627925669,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.664928892013553e-05,
+      "loss": 6.9245,
+      "step": 14290
+    },
+    {
+      "epoch": 0.8210054925076072,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.660544504547423e-05,
+      "loss": 6.9198,
+      "step": 14300
+    },
+    {
+      "epoch": 0.8215796222226475,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.656158681508357e-05,
+      "loss": 6.9587,
+      "step": 14310
+    },
+    {
+      "epoch": 0.8221537519376878,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.651771426687983e-05,
+      "loss": 6.9579,
+      "step": 14320
+    },
+    {
+      "epoch": 0.822727881652728,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.647382743879166e-05,
+      "loss": 6.9472,
+      "step": 14330
+    },
+    {
+      "epoch": 0.8233020113677684,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.642992636876007e-05,
+      "loss": 6.9644,
+      "step": 14340
+    },
+    {
+      "epoch": 0.8238761410828086,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.638601109473842e-05,
+      "loss": 6.9212,
+      "step": 14350
+    },
+    {
+      "epoch": 0.8244502707978489,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.634208165469231e-05,
+      "loss": 6.9237,
+      "step": 14360
+    },
+    {
+      "epoch": 0.8250244005128892,
+      "grad_norm": 1.625,
+      "learning_rate": 6.629813808659958e-05,
+      "loss": 6.9283,
+      "step": 14370
+    },
+    {
+      "epoch": 0.8255985302279295,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.625418042845028e-05,
+      "loss": 6.9112,
+      "step": 14380
+    },
+    {
+      "epoch": 0.8261726599429697,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.621020871824668e-05,
+      "loss": 6.9496,
+      "step": 14390
+    },
+    {
+      "epoch": 0.8267467896580101,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.616622299400319e-05,
+      "loss": 6.9288,
+      "step": 14400
+    },
+    {
+      "epoch": 0.8273209193730503,
+      "grad_norm": 1.5,
+      "learning_rate": 6.612222329374631e-05,
+      "loss": 6.923,
+      "step": 14410
+    },
+    {
+      "epoch": 0.8278950490880906,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.607820965551462e-05,
+      "loss": 6.9171,
+      "step": 14420
+    },
+    {
+      "epoch": 0.8284691788031309,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.603418211735876e-05,
+      "loss": 6.9524,
+      "step": 14430
+    },
+    {
+      "epoch": 0.8290433085181712,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.599014071734145e-05,
+      "loss": 6.9356,
+      "step": 14440
+    },
+    {
+      "epoch": 0.8296174382332114,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.594608549353725e-05,
+      "loss": 6.8981,
+      "step": 14450
+    },
+    {
+      "epoch": 0.8301915679482518,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.59020164840328e-05,
+      "loss": 6.9348,
+      "step": 14460
+    },
+    {
+      "epoch": 0.8307656976632921,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.585793372692663e-05,
+      "loss": 6.9244,
+      "step": 14470
+    },
+    {
+      "epoch": 0.8313398273783323,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.581383726032912e-05,
+      "loss": 6.9678,
+      "step": 14480
+    },
+    {
+      "epoch": 0.8319139570933727,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.57697271223625e-05,
+      "loss": 6.9098,
+      "step": 14490
+    },
+    {
+      "epoch": 0.8324880868084129,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.572560335116087e-05,
+      "loss": 6.9442,
+      "step": 14500
+    },
+    {
+      "epoch": 0.8330622165234532,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.568146598487007e-05,
+      "loss": 6.9094,
+      "step": 14510
+    },
+    {
+      "epoch": 0.8336363462384935,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.563731506164772e-05,
+      "loss": 6.9335,
+      "step": 14520
+    },
+    {
+      "epoch": 0.8342104759535338,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.559315061966314e-05,
+      "loss": 6.9398,
+      "step": 14530
+    },
+    {
+      "epoch": 0.834784605668574,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.554897269709735e-05,
+      "loss": 6.9264,
+      "step": 14540
+    },
+    {
+      "epoch": 0.8353587353836144,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.550478133214304e-05,
+      "loss": 6.9198,
+      "step": 14550
+    },
+    {
+      "epoch": 0.8359328650986546,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.546057656300447e-05,
+      "loss": 6.9202,
+      "step": 14560
+    },
+    {
+      "epoch": 0.836506994813695,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.541635842789752e-05,
+      "loss": 6.9135,
+      "step": 14570
+    },
+    {
+      "epoch": 0.8370811245287352,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.537212696504968e-05,
+      "loss": 6.9019,
+      "step": 14580
+    },
+    {
+      "epoch": 0.8376552542437755,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.532788221269985e-05,
+      "loss": 6.891,
+      "step": 14590
+    },
+    {
+      "epoch": 0.8382293839588157,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.528362420909848e-05,
+      "loss": 6.9124,
+      "step": 14600
+    },
+    {
+      "epoch": 0.8388035136738561,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.52393529925075e-05,
+      "loss": 6.9593,
+      "step": 14610
+    },
+    {
+      "epoch": 0.8393776433888963,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.519506860120024e-05,
+      "loss": 6.9323,
+      "step": 14620
+    },
+    {
+      "epoch": 0.8399517731039366,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.515077107346139e-05,
+      "loss": 6.9259,
+      "step": 14630
+    },
+    {
+      "epoch": 0.8405259028189769,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.5106460447587e-05,
+      "loss": 6.9423,
+      "step": 14640
+    },
+    {
+      "epoch": 0.8411000325340172,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.506213676188453e-05,
+      "loss": 6.9264,
+      "step": 14650
+    },
+    {
+      "epoch": 0.8416741622490574,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.501780005467262e-05,
+      "loss": 6.9357,
+      "step": 14660
+    },
+    {
+      "epoch": 0.8422482919640978,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.497345036428124e-05,
+      "loss": 6.9022,
+      "step": 14670
+    },
+    {
+      "epoch": 0.842822421679138,
+      "grad_norm": 1.625,
+      "learning_rate": 6.492908772905154e-05,
+      "loss": 6.9009,
+      "step": 14680
+    },
+    {
+      "epoch": 0.8433965513941784,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.488471218733588e-05,
+      "loss": 6.8893,
+      "step": 14690
+    },
+    {
+      "epoch": 0.8439706811092186,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.484032377749777e-05,
+      "loss": 6.9118,
+      "step": 14700
+    },
+    {
+      "epoch": 0.8445448108242589,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.479592253791187e-05,
+      "loss": 6.8947,
+      "step": 14710
+    },
+    {
+      "epoch": 0.8451189405392991,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.47515085069639e-05,
+      "loss": 6.9224,
+      "step": 14720
+    },
+    {
+      "epoch": 0.8456930702543395,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.470708172305065e-05,
+      "loss": 6.9317,
+      "step": 14730
+    },
+    {
+      "epoch": 0.8462671999693797,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.466264222457997e-05,
+      "loss": 6.9228,
+      "step": 14740
+    },
+    {
+      "epoch": 0.84684132968442,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.46181900499706e-05,
+      "loss": 6.9008,
+      "step": 14750
+    },
+    {
+      "epoch": 0.8474154593994603,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.457372523765238e-05,
+      "loss": 6.928,
+      "step": 14760
+    },
+    {
+      "epoch": 0.8479895891145006,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.452924782606595e-05,
+      "loss": 6.9059,
+      "step": 14770
+    },
+    {
+      "epoch": 0.8485637188295408,
+      "grad_norm": 1.5078125,
+      "learning_rate": 6.448475785366291e-05,
+      "loss": 6.9193,
+      "step": 14780
+    },
+    {
+      "epoch": 0.8491378485445812,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.444025535890573e-05,
+      "loss": 6.904,
+      "step": 14790
+    },
+    {
+      "epoch": 0.8497119782596214,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.439574038026766e-05,
+      "loss": 6.9381,
+      "step": 14800
+    },
+    {
+      "epoch": 0.8502861079746618,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.435121295623276e-05,
+      "loss": 6.9774,
+      "step": 14810
+    },
+    {
+      "epoch": 0.850860237689702,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.430667312529585e-05,
+      "loss": 6.9291,
+      "step": 14820
+    },
+    {
+      "epoch": 0.8514343674047423,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.426212092596248e-05,
+      "loss": 6.9351,
+      "step": 14830
+    },
+    {
+      "epoch": 0.8520084971197825,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.421755639674889e-05,
+      "loss": 6.9359,
+      "step": 14840
+    },
+    {
+      "epoch": 0.8525826268348229,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.4172979576182e-05,
+      "loss": 6.8811,
+      "step": 14850
+    },
+    {
+      "epoch": 0.8531567565498631,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.412839050279929e-05,
+      "loss": 6.9069,
+      "step": 14860
+    },
+    {
+      "epoch": 0.8537308862649035,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.408378921514894e-05,
+      "loss": 6.936,
+      "step": 14870
+    },
+    {
+      "epoch": 0.8543050159799437,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.403917575178959e-05,
+      "loss": 6.9129,
+      "step": 14880
+    },
+    {
+      "epoch": 0.854879145694984,
+      "grad_norm": 1.46875,
+      "learning_rate": 6.399455015129043e-05,
+      "loss": 6.912,
+      "step": 14890
+    },
+    {
+      "epoch": 0.8554532754100242,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.394991245223121e-05,
+      "loss": 6.9321,
+      "step": 14900
+    },
+    {
+      "epoch": 0.8560274051250646,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.390526269320202e-05,
+      "loss": 6.8809,
+      "step": 14910
+    },
+    {
+      "epoch": 0.8566015348401049,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.38606009128035e-05,
+      "loss": 6.9148,
+      "step": 14920
+    },
+    {
+      "epoch": 0.8571756645551452,
+      "grad_norm": 1.625,
+      "learning_rate": 6.38159271496466e-05,
+      "loss": 6.9322,
+      "step": 14930
+    },
+    {
+      "epoch": 0.8577497942701855,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.377124144235265e-05,
+      "loss": 6.9338,
+      "step": 14940
+    },
+    {
+      "epoch": 0.8583239239852257,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.372654382955334e-05,
+      "loss": 6.8938,
+      "step": 14950
+    },
+    {
+      "epoch": 0.8588980537002661,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.368183434989058e-05,
+      "loss": 6.9226,
+      "step": 14960
+    },
+    {
+      "epoch": 0.8594721834153063,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.363711304201661e-05,
+      "loss": 6.9437,
+      "step": 14970
+    },
+    {
+      "epoch": 0.8600463131303466,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.359237994459388e-05,
+      "loss": 6.9135,
+      "step": 14980
+    },
+    {
+      "epoch": 0.8606204428453869,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.354763509629498e-05,
+      "loss": 6.9327,
+      "step": 14990
+    },
+    {
+      "epoch": 0.8611945725604272,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.350287853580273e-05,
+      "loss": 6.8962,
+      "step": 15000
+    },
+    {
+      "epoch": 0.8617687022754674,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.345811030181005e-05,
+      "loss": 6.9345,
+      "step": 15010
+    },
+    {
+      "epoch": 0.8623428319905078,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.341333043301993e-05,
+      "loss": 6.9071,
+      "step": 15020
+    },
+    {
+      "epoch": 0.862916961705548,
+      "grad_norm": 1.890625,
+      "learning_rate": 6.336853896814543e-05,
+      "loss": 6.9159,
+      "step": 15030
+    },
+    {
+      "epoch": 0.8634910914205883,
+      "grad_norm": 1.625,
+      "learning_rate": 6.332373594590964e-05,
+      "loss": 6.9337,
+      "step": 15040
+    },
+    {
+      "epoch": 0.8640652211356286,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.327892140504567e-05,
+      "loss": 6.9085,
+      "step": 15050
+    },
+    {
+      "epoch": 0.8646393508506689,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.323409538429656e-05,
+      "loss": 6.9236,
+      "step": 15060
+    },
+    {
+      "epoch": 0.8652134805657091,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.318925792241523e-05,
+      "loss": 6.9405,
+      "step": 15070
+    },
+    {
+      "epoch": 0.8657876102807495,
+      "grad_norm": 1.625,
+      "learning_rate": 6.314440905816457e-05,
+      "loss": 6.9579,
+      "step": 15080
+    },
+    {
+      "epoch": 0.8663617399957897,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.30995488303173e-05,
+      "loss": 6.9024,
+      "step": 15090
+    },
+    {
+      "epoch": 0.86693586971083,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.305467727765592e-05,
+      "loss": 6.9078,
+      "step": 15100
+    },
+    {
+      "epoch": 0.8675099994258703,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.30097944389728e-05,
+      "loss": 6.9246,
+      "step": 15110
+    },
+    {
+      "epoch": 0.8680841291409106,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.296490035306999e-05,
+      "loss": 6.9051,
+      "step": 15120
+    },
+    {
+      "epoch": 0.8686582588559508,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.291999505875932e-05,
+      "loss": 6.9181,
+      "step": 15130
+    },
+    {
+      "epoch": 0.8692323885709912,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.287507859486228e-05,
+      "loss": 6.9262,
+      "step": 15140
+    },
+    {
+      "epoch": 0.8698065182860314,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.283015100021002e-05,
+      "loss": 6.888,
+      "step": 15150
+    },
+    {
+      "epoch": 0.8703806480010717,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.278521231364334e-05,
+      "loss": 6.9464,
+      "step": 15160
+    },
+    {
+      "epoch": 0.870954777716112,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.274026257401258e-05,
+      "loss": 6.9193,
+      "step": 15170
+    },
+    {
+      "epoch": 0.8715289074311523,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.269530182017766e-05,
+      "loss": 6.8934,
+      "step": 15180
+    },
+    {
+      "epoch": 0.8721030371461925,
+      "grad_norm": 1.5,
+      "learning_rate": 6.265033009100805e-05,
+      "loss": 6.9851,
+      "step": 15190
+    },
+    {
+      "epoch": 0.8726771668612329,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.260534742538267e-05,
+      "loss": 6.9182,
+      "step": 15200
+    },
+    {
+      "epoch": 0.8732512965762731,
+      "grad_norm": 1.625,
+      "learning_rate": 6.256035386218989e-05,
+      "loss": 6.9386,
+      "step": 15210
+    },
+    {
+      "epoch": 0.8738254262913134,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.251534944032754e-05,
+      "loss": 6.9668,
+      "step": 15220
+    },
+    {
+      "epoch": 0.8743995560063537,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.247033419870281e-05,
+      "loss": 6.9103,
+      "step": 15230
+    },
+    {
+      "epoch": 0.874973685721394,
+      "grad_norm": 1.625,
+      "learning_rate": 6.242530817623225e-05,
+      "loss": 6.8926,
+      "step": 15240
+    },
+    {
+      "epoch": 0.8755478154364342,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.238027141184171e-05,
+      "loss": 6.9194,
+      "step": 15250
+    },
+    {
+      "epoch": 0.8761219451514746,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.23352239444664e-05,
+      "loss": 6.8804,
+      "step": 15260
+    },
+    {
+      "epoch": 0.8766960748665148,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.229016581305067e-05,
+      "loss": 6.9138,
+      "step": 15270
+    },
+    {
+      "epoch": 0.8772702045815551,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.224509705654818e-05,
+      "loss": 6.8961,
+      "step": 15280
+    },
+    {
+      "epoch": 0.8778443342965954,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.220001771392173e-05,
+      "loss": 6.9099,
+      "step": 15290
+    },
+    {
+      "epoch": 0.8784184640116357,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.21549278241433e-05,
+      "loss": 6.9214,
+      "step": 15300
+    },
+    {
+      "epoch": 0.8789925937266759,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.210982742619395e-05,
+      "loss": 6.8928,
+      "step": 15310
+    },
+    {
+      "epoch": 0.8795667234417163,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.206471655906388e-05,
+      "loss": 6.9285,
+      "step": 15320
+    },
+    {
+      "epoch": 0.8801408531567565,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.20195952617523e-05,
+      "loss": 6.9484,
+      "step": 15330
+    },
+    {
+      "epoch": 0.8807149828717968,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.197446357326745e-05,
+      "loss": 6.9297,
+      "step": 15340
+    },
+    {
+      "epoch": 0.8812891125868371,
+      "grad_norm": 1.625,
+      "learning_rate": 6.192932153262653e-05,
+      "loss": 6.8879,
+      "step": 15350
+    },
+    {
+      "epoch": 0.8818632423018774,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.188416917885572e-05,
+      "loss": 6.8707,
+      "step": 15360
+    },
+    {
+      "epoch": 0.8824373720169177,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.183900655099013e-05,
+      "loss": 6.9316,
+      "step": 15370
+    },
+    {
+      "epoch": 0.883011501731958,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.179383368807369e-05,
+      "loss": 6.9382,
+      "step": 15380
+    },
+    {
+      "epoch": 0.8835856314469983,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.174865062915924e-05,
+      "loss": 6.9445,
+      "step": 15390
+    },
+    {
+      "epoch": 0.8841597611620385,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.170345741330839e-05,
+      "loss": 6.8567,
+      "step": 15400
+    },
+    {
+      "epoch": 0.8847338908770789,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.165825407959158e-05,
+      "loss": 6.9408,
+      "step": 15410
+    },
+    {
+      "epoch": 0.8853080205921191,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.161304066708796e-05,
+      "loss": 6.9133,
+      "step": 15420
+    },
+    {
+      "epoch": 0.8858821503071594,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.156781721488538e-05,
+      "loss": 6.8569,
+      "step": 15430
+    },
+    {
+      "epoch": 0.8864562800221997,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.152258376208042e-05,
+      "loss": 6.8869,
+      "step": 15440
+    },
+    {
+      "epoch": 0.88703040973724,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.147734034777828e-05,
+      "loss": 6.9566,
+      "step": 15450
+    },
+    {
+      "epoch": 0.8876045394522802,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.143208701109274e-05,
+      "loss": 6.8694,
+      "step": 15460
+    },
+    {
+      "epoch": 0.8881786691673206,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.13868237911462e-05,
+      "loss": 6.9282,
+      "step": 15470
+    },
+    {
+      "epoch": 0.8887527988823608,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.13415507270696e-05,
+      "loss": 6.9229,
+      "step": 15480
+    },
+    {
+      "epoch": 0.8893269285974011,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.129626785800237e-05,
+      "loss": 6.9499,
+      "step": 15490
+    },
+    {
+      "epoch": 0.8899010583124414,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.125097522309245e-05,
+      "loss": 6.9157,
+      "step": 15500
+    },
+    {
+      "epoch": 0.8904751880274817,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.120567286149618e-05,
+      "loss": 6.9177,
+      "step": 15510
+    },
+    {
+      "epoch": 0.8910493177425219,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.116036081237834e-05,
+      "loss": 6.8992,
+      "step": 15520
+    },
+    {
+      "epoch": 0.8916234474575623,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.111503911491207e-05,
+      "loss": 6.8907,
+      "step": 15530
+    },
+    {
+      "epoch": 0.8921975771726025,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.106970780827886e-05,
+      "loss": 6.8982,
+      "step": 15540
+    },
+    {
+      "epoch": 0.8927717068876428,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.102436693166852e-05,
+      "loss": 6.9049,
+      "step": 15550
+    },
+    {
+      "epoch": 0.8933458366026831,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.0979016524279077e-05,
+      "loss": 6.9312,
+      "step": 15560
+    },
+    {
+      "epoch": 0.8939199663177234,
+      "grad_norm": 1.515625,
+      "learning_rate": 6.093365662531686e-05,
+      "loss": 6.8921,
+      "step": 15570
+    },
+    {
+      "epoch": 0.8944940960327636,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.0888287273996404e-05,
+      "loss": 6.9304,
+      "step": 15580
+    },
+    {
+      "epoch": 0.895068225747804,
+      "grad_norm": 1.53125,
+      "learning_rate": 6.084290850954036e-05,
+      "loss": 6.921,
+      "step": 15590
+    },
+    {
+      "epoch": 0.8956423554628442,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.0797520371179585e-05,
+      "loss": 6.9617,
+      "step": 15600
+    },
+    {
+      "epoch": 0.8962164851778845,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.0752122898152955e-05,
+      "loss": 6.9234,
+      "step": 15610
+    },
+    {
+      "epoch": 0.8967906148929248,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.070671612970751e-05,
+      "loss": 6.9161,
+      "step": 15620
+    },
+    {
+      "epoch": 0.8973647446079651,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.066130010509827e-05,
+      "loss": 6.9208,
+      "step": 15630
+    },
+    {
+      "epoch": 0.8979388743230053,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.061587486358826e-05,
+      "loss": 6.9033,
+      "step": 15640
+    },
+    {
+      "epoch": 0.8985130040380457,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.057044044444848e-05,
+      "loss": 6.9014,
+      "step": 15650
+    },
+    {
+      "epoch": 0.8990871337530859,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.0524996886957896e-05,
+      "loss": 6.8919,
+      "step": 15660
+    },
+    {
+      "epoch": 0.8996612634681262,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.047954423040332e-05,
+      "loss": 6.9082,
+      "step": 15670
+    },
+    {
+      "epoch": 0.9002353931831665,
+      "grad_norm": 1.625,
+      "learning_rate": 6.043408251407945e-05,
+      "loss": 6.9212,
+      "step": 15680
+    },
+    {
+      "epoch": 0.9008095228982068,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.038861177728884e-05,
+      "loss": 6.9136,
+      "step": 15690
+    },
+    {
+      "epoch": 0.901383652613247,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.03431320593418e-05,
+      "loss": 6.9018,
+      "step": 15700
+    },
+    {
+      "epoch": 0.9019577823282874,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.029764339955646e-05,
+      "loss": 6.9162,
+      "step": 15710
+    },
+    {
+      "epoch": 0.9025319120433276,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.0252145837258636e-05,
+      "loss": 6.9063,
+      "step": 15720
+    },
+    {
+      "epoch": 0.903106041758368,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.020663941178184e-05,
+      "loss": 6.8981,
+      "step": 15730
+    },
+    {
+      "epoch": 0.9036801714734082,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.016112416246726e-05,
+      "loss": 6.8945,
+      "step": 15740
+    },
+    {
+      "epoch": 0.9042543011884485,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.011560012866375e-05,
+      "loss": 6.9036,
+      "step": 15750
+    },
+    {
+      "epoch": 0.9048284309034887,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.0070067349727675e-05,
+      "loss": 6.9011,
+      "step": 15760
+    },
+    {
+      "epoch": 0.9054025606185291,
+      "grad_norm": 1.625,
+      "learning_rate": 6.002452586502303e-05,
+      "loss": 6.9039,
+      "step": 15770
+    },
+    {
+      "epoch": 0.9059766903335693,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.9978975713921294e-05,
+      "loss": 6.9252,
+      "step": 15780
+    },
+    {
+      "epoch": 0.9065508200486097,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.9933416935801466e-05,
+      "loss": 6.9102,
+      "step": 15790
+    },
+    {
+      "epoch": 0.9071249497636499,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.9887849570050016e-05,
+      "loss": 6.9262,
+      "step": 15800
+    },
+    {
+      "epoch": 0.9076990794786902,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.984227365606078e-05,
+      "loss": 6.877,
+      "step": 15810
+    },
+    {
+      "epoch": 0.9082732091937306,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.979668923323503e-05,
+      "loss": 6.9007,
+      "step": 15820
+    },
+    {
+      "epoch": 0.9088473389087708,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.9751096340981373e-05,
+      "loss": 6.9126,
+      "step": 15830
+    },
+    {
+      "epoch": 0.9094214686238111,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.970549501871578e-05,
+      "loss": 6.8955,
+      "step": 15840
+    },
+    {
+      "epoch": 0.9099955983388514,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.9659885305861476e-05,
+      "loss": 6.9163,
+      "step": 15850
+    },
+    {
+      "epoch": 0.9105697280538917,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.961426724184892e-05,
+      "loss": 6.9319,
+      "step": 15860
+    },
+    {
+      "epoch": 0.9111438577689319,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.956864086611581e-05,
+      "loss": 6.9329,
+      "step": 15870
+    },
+    {
+      "epoch": 0.9117179874839723,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.952300621810707e-05,
+      "loss": 6.9002,
+      "step": 15880
+    },
+    {
+      "epoch": 0.9122921171990125,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.9477363337274696e-05,
+      "loss": 6.8949,
+      "step": 15890
+    },
+    {
+      "epoch": 0.9128662469140528,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.9431712263077864e-05,
+      "loss": 6.9267,
+      "step": 15900
+    },
+    {
+      "epoch": 0.913440376629093,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.938605303498282e-05,
+      "loss": 6.9459,
+      "step": 15910
+    },
+    {
+      "epoch": 0.9140145063441334,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.934038569246283e-05,
+      "loss": 6.9093,
+      "step": 15920
+    },
+    {
+      "epoch": 0.9145886360591736,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.929471027499822e-05,
+      "loss": 6.8726,
+      "step": 15930
+    },
+    {
+      "epoch": 0.915162765774214,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.924902682207627e-05,
+      "loss": 6.8642,
+      "step": 15940
+    },
+    {
+      "epoch": 0.9157368954892542,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.9203335373191195e-05,
+      "loss": 6.9166,
+      "step": 15950
+    },
+    {
+      "epoch": 0.9163110252042945,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.915763596784415e-05,
+      "loss": 6.9005,
+      "step": 15960
+    },
+    {
+      "epoch": 0.9168851549193348,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.911192864554317e-05,
+      "loss": 6.912,
+      "step": 15970
+    },
+    {
+      "epoch": 0.9174592846343751,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.90662134458031e-05,
+      "loss": 6.9025,
+      "step": 15980
+    },
+    {
+      "epoch": 0.9180334143494153,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.902049040814563e-05,
+      "loss": 6.9095,
+      "step": 15990
+    },
+    {
+      "epoch": 0.9186075440644557,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.89747595720992e-05,
+      "loss": 6.8551,
+      "step": 16000
+    },
+    {
+      "epoch": 0.9191816737794959,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.8929020977199034e-05,
+      "loss": 6.9176,
+      "step": 16010
+    },
+    {
+      "epoch": 0.9197558034945362,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.888327466298701e-05,
+      "loss": 6.9277,
+      "step": 16020
+    },
+    {
+      "epoch": 0.9203299332095765,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.88375206690117e-05,
+      "loss": 6.8961,
+      "step": 16030
+    },
+    {
+      "epoch": 0.9209040629246168,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.879175903482833e-05,
+      "loss": 6.9116,
+      "step": 16040
+    },
+    {
+      "epoch": 0.921478192639657,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.874598979999873e-05,
+      "loss": 6.8972,
+      "step": 16050
+    },
+    {
+      "epoch": 0.9220523223546974,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.870021300409128e-05,
+      "loss": 6.9153,
+      "step": 16060
+    },
+    {
+      "epoch": 0.9226264520697376,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.8654428686680905e-05,
+      "loss": 6.8592,
+      "step": 16070
+    },
+    {
+      "epoch": 0.9232005817847779,
+      "grad_norm": 1.625,
+      "learning_rate": 5.860863688734903e-05,
+      "loss": 6.9135,
+      "step": 16080
+    },
+    {
+      "epoch": 0.9237747114998182,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.8562837645683575e-05,
+      "loss": 6.9168,
+      "step": 16090
+    },
+    {
+      "epoch": 0.9243488412148585,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.851703100127886e-05,
+      "loss": 6.8931,
+      "step": 16100
+    },
+    {
+      "epoch": 0.9249229709298987,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.8471216993735626e-05,
+      "loss": 6.9196,
+      "step": 16110
+    },
+    {
+      "epoch": 0.9254971006449391,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.842539566266095e-05,
+      "loss": 6.8941,
+      "step": 16120
+    },
+    {
+      "epoch": 0.9260712303599793,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.837956704766829e-05,
+      "loss": 6.9126,
+      "step": 16130
+    },
+    {
+      "epoch": 0.9266453600750196,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.833373118837734e-05,
+      "loss": 6.9539,
+      "step": 16140
+    },
+    {
+      "epoch": 0.9272194897900599,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.8287888124414126e-05,
+      "loss": 6.8976,
+      "step": 16150
+    },
+    {
+      "epoch": 0.9277936195051002,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.824203789541085e-05,
+      "loss": 6.8622,
+      "step": 16160
+    },
+    {
+      "epoch": 0.9283677492201404,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.819618054100591e-05,
+      "loss": 6.9183,
+      "step": 16170
+    },
+    {
+      "epoch": 0.9289418789351808,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.8150316100843895e-05,
+      "loss": 6.9262,
+      "step": 16180
+    },
+    {
+      "epoch": 0.929516008650221,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.81044446145755e-05,
+      "loss": 6.9072,
+      "step": 16190
+    },
+    {
+      "epoch": 0.9300901383652613,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.8058566121857514e-05,
+      "loss": 6.8663,
+      "step": 16200
+    },
+    {
+      "epoch": 0.9306642680803016,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.8012680662352795e-05,
+      "loss": 6.9084,
+      "step": 16210
+    },
+    {
+      "epoch": 0.9312383977953419,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.796678827573018e-05,
+      "loss": 6.8419,
+      "step": 16220
+    },
+    {
+      "epoch": 0.9318125275103821,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.792088900166457e-05,
+      "loss": 6.8997,
+      "step": 16230
+    },
+    {
+      "epoch": 0.9323866572254225,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.7874982879836746e-05,
+      "loss": 6.8838,
+      "step": 16240
+    },
+    {
+      "epoch": 0.9329607869404627,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.7829069949933464e-05,
+      "loss": 6.9244,
+      "step": 16250
+    },
+    {
+      "epoch": 0.933534916655503,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.778315025164731e-05,
+      "loss": 6.9197,
+      "step": 16260
+    },
+    {
+      "epoch": 0.9341090463705434,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.77372238246768e-05,
+      "loss": 6.8825,
+      "step": 16270
+    },
+    {
+      "epoch": 0.9346831760855836,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.769129070872619e-05,
+      "loss": 6.9021,
+      "step": 16280
+    },
+    {
+      "epoch": 0.9352573058006239,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.7645350943505547e-05,
+      "loss": 6.9201,
+      "step": 16290
+    },
+    {
+      "epoch": 0.9358314355156642,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.759940456873071e-05,
+      "loss": 6.8906,
+      "step": 16300
+    },
+    {
+      "epoch": 0.9364055652307045,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.755345162412318e-05,
+      "loss": 6.9528,
+      "step": 16310
+    },
+    {
+      "epoch": 0.9369796949457447,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.7507492149410204e-05,
+      "loss": 6.8871,
+      "step": 16320
+    },
+    {
+      "epoch": 0.9375538246607851,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.746152618432462e-05,
+      "loss": 6.9257,
+      "step": 16330
+    },
+    {
+      "epoch": 0.9381279543758253,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.7415553768604904e-05,
+      "loss": 6.9333,
+      "step": 16340
+    },
+    {
+      "epoch": 0.9387020840908656,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.736957494199509e-05,
+      "loss": 6.9548,
+      "step": 16350
+    },
+    {
+      "epoch": 0.9392762138059059,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.7323589744244765e-05,
+      "loss": 6.9107,
+      "step": 16360
+    },
+    {
+      "epoch": 0.9398503435209462,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.727759821510904e-05,
+      "loss": 6.9093,
+      "step": 16370
+    },
+    {
+      "epoch": 0.9404244732359864,
+      "grad_norm": 1.5,
+      "learning_rate": 5.72316003943485e-05,
+      "loss": 6.9069,
+      "step": 16380
+    },
+    {
+      "epoch": 0.9409986029510268,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.7185596321729106e-05,
+      "loss": 6.9084,
+      "step": 16390
+    },
+    {
+      "epoch": 0.941572732666067,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.71395860370223e-05,
+      "loss": 6.8798,
+      "step": 16400
+    },
+    {
+      "epoch": 0.9421468623811073,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.7093569580004855e-05,
+      "loss": 6.9022,
+      "step": 16410
+    },
+    {
+      "epoch": 0.9427209920961476,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.704754699045891e-05,
+      "loss": 6.9142,
+      "step": 16420
+    },
+    {
+      "epoch": 0.9432951218111879,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.700151830817187e-05,
+      "loss": 6.9078,
+      "step": 16430
+    },
+    {
+      "epoch": 0.9438692515262281,
+      "grad_norm": 1.625,
+      "learning_rate": 5.695548357293642e-05,
+      "loss": 6.966,
+      "step": 16440
+    },
+    {
+      "epoch": 0.9444433812412685,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.690944282455049e-05,
+      "loss": 6.8658,
+      "step": 16450
+    },
+    {
+      "epoch": 0.9450175109563087,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.6863396102817214e-05,
+      "loss": 6.8941,
+      "step": 16460
+    },
+    {
+      "epoch": 0.945591640671349,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.681734344754486e-05,
+      "loss": 6.895,
+      "step": 16470
+    },
+    {
+      "epoch": 0.9461657703863893,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.677128489854684e-05,
+      "loss": 6.897,
+      "step": 16480
+    },
+    {
+      "epoch": 0.9467399001014296,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.672522049564165e-05,
+      "loss": 6.872,
+      "step": 16490
+    },
+    {
+      "epoch": 0.9473140298164698,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.6679150278652895e-05,
+      "loss": 6.9132,
+      "step": 16500
+    },
+    {
+      "epoch": 0.9478881595315102,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.6633074287409135e-05,
+      "loss": 6.9661,
+      "step": 16510
+    },
+    {
+      "epoch": 0.9484622892465504,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.6586992561744e-05,
+      "loss": 6.9056,
+      "step": 16520
+    },
+    {
+      "epoch": 0.9490364189615907,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.654090514149598e-05,
+      "loss": 6.9056,
+      "step": 16530
+    },
+    {
+      "epoch": 0.949610548676631,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.649481206650859e-05,
+      "loss": 6.8932,
+      "step": 16540
+    },
+    {
+      "epoch": 0.9501846783916713,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.6448713376630194e-05,
+      "loss": 6.8895,
+      "step": 16550
+    },
+    {
+      "epoch": 0.9507588081067115,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.640260911171397e-05,
+      "loss": 6.9123,
+      "step": 16560
+    },
+    {
+      "epoch": 0.9513329378217519,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.635649931161794e-05,
+      "loss": 6.8596,
+      "step": 16570
+    },
+    {
+      "epoch": 0.9519070675367921,
+      "grad_norm": 1.796875,
+      "learning_rate": 5.6310384016204965e-05,
+      "loss": 6.899,
+      "step": 16580
+    },
+    {
+      "epoch": 0.9524811972518324,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.6264263265342586e-05,
+      "loss": 6.9051,
+      "step": 16590
+    },
+    {
+      "epoch": 0.9530553269668727,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.62181370989031e-05,
+      "loss": 6.8821,
+      "step": 16600
+    },
+    {
+      "epoch": 0.953629456681913,
+      "grad_norm": 1.625,
+      "learning_rate": 5.617200555676344e-05,
+      "loss": 6.8929,
+      "step": 16610
+    },
+    {
+      "epoch": 0.9542035863969532,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.612586867880525e-05,
+      "loss": 6.8953,
+      "step": 16620
+    },
+    {
+      "epoch": 0.9547777161119936,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.607972650491476e-05,
+      "loss": 6.9298,
+      "step": 16630
+    },
+    {
+      "epoch": 0.9553518458270338,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.6033579074982766e-05,
+      "loss": 6.8771,
+      "step": 16640
+    },
+    {
+      "epoch": 0.9559259755420741,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.598742642890461e-05,
+      "loss": 6.9059,
+      "step": 16650
+    },
+    {
+      "epoch": 0.9565001052571144,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.5941268606580146e-05,
+      "loss": 6.9305,
+      "step": 16660
+    },
+    {
+      "epoch": 0.9570742349721547,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.5895105647913716e-05,
+      "loss": 6.9076,
+      "step": 16670
+    },
+    {
+      "epoch": 0.9576483646871949,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.58489375928141e-05,
+      "loss": 6.9448,
+      "step": 16680
+    },
+    {
+      "epoch": 0.9582224944022353,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.580276448119447e-05,
+      "loss": 6.8704,
+      "step": 16690
+    },
+    {
+      "epoch": 0.9587966241172755,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.5756586352972374e-05,
+      "loss": 6.9062,
+      "step": 16700
+    },
+    {
+      "epoch": 0.9593707538323158,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.571040324806969e-05,
+      "loss": 6.9017,
+      "step": 16710
+    },
+    {
+      "epoch": 0.9599448835473562,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.566421520641263e-05,
+      "loss": 6.8891,
+      "step": 16720
+    },
+    {
+      "epoch": 0.9605190132623964,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.561802226793165e-05,
+      "loss": 6.9133,
+      "step": 16730
+    },
+    {
+      "epoch": 0.9610931429774368,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.557182447256142e-05,
+      "loss": 6.865,
+      "step": 16740
+    },
+    {
+      "epoch": 0.961667272692477,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.552562186024084e-05,
+      "loss": 6.8862,
+      "step": 16750
+    },
+    {
+      "epoch": 0.9622414024075173,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.547941447091297e-05,
+      "loss": 6.9085,
+      "step": 16760
+    },
+    {
+      "epoch": 0.9628155321225575,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.5433202344525e-05,
+      "loss": 6.8982,
+      "step": 16770
+    },
+    {
+      "epoch": 0.9633896618375979,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.53869855210282e-05,
+      "loss": 6.9161,
+      "step": 16780
+    },
+    {
+      "epoch": 0.9639637915526381,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.53407640403779e-05,
+      "loss": 6.8866,
+      "step": 16790
+    },
+    {
+      "epoch": 0.9645379212676785,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.52945379425335e-05,
+      "loss": 6.9069,
+      "step": 16800
+    },
+    {
+      "epoch": 0.9651120509827187,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.5248307267458334e-05,
+      "loss": 6.9002,
+      "step": 16810
+    },
+    {
+      "epoch": 0.965686180697759,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.5202072055119715e-05,
+      "loss": 6.8604,
+      "step": 16820
+    },
+    {
+      "epoch": 0.9662603104127993,
+      "grad_norm": 1.5,
+      "learning_rate": 5.5155832345488875e-05,
+      "loss": 6.8671,
+      "step": 16830
+    },
+    {
+      "epoch": 0.9668344401278396,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.510958817854097e-05,
+      "loss": 6.9308,
+      "step": 16840
+    },
+    {
+      "epoch": 0.9674085698428798,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.506333959425497e-05,
+      "loss": 6.9129,
+      "step": 16850
+    },
+    {
+      "epoch": 0.9679826995579202,
+      "grad_norm": 1.625,
+      "learning_rate": 5.501708663261366e-05,
+      "loss": 6.8769,
+      "step": 16860
+    },
+    {
+      "epoch": 0.9685568292729604,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.49708293336036e-05,
+      "loss": 6.8734,
+      "step": 16870
+    },
+    {
+      "epoch": 0.9691309589880007,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.492456773721517e-05,
+      "loss": 6.9219,
+      "step": 16880
+    },
+    {
+      "epoch": 0.969705088703041,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.4878301883442396e-05,
+      "loss": 6.8936,
+      "step": 16890
+    },
+    {
+      "epoch": 0.9702792184180813,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.483203181228301e-05,
+      "loss": 6.8887,
+      "step": 16900
+    },
+    {
+      "epoch": 0.9708533481331215,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.4785757563738396e-05,
+      "loss": 6.9107,
+      "step": 16910
+    },
+    {
+      "epoch": 0.9714274778481619,
+      "grad_norm": 1.625,
+      "learning_rate": 5.4739479177813516e-05,
+      "loss": 6.9171,
+      "step": 16920
+    },
+    {
+      "epoch": 0.9720016075632021,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.469319669451692e-05,
+      "loss": 6.9316,
+      "step": 16930
+    },
+    {
+      "epoch": 0.9725757372782424,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.4646910153860764e-05,
+      "loss": 6.8953,
+      "step": 16940
+    },
+    {
+      "epoch": 0.9731498669932827,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.460061959586063e-05,
+      "loss": 6.8968,
+      "step": 16950
+    },
+    {
+      "epoch": 0.973723996708323,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.455432506053562e-05,
+      "loss": 6.8586,
+      "step": 16960
+    },
+    {
+      "epoch": 0.9742981264233632,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.450802658790821e-05,
+      "loss": 6.8524,
+      "step": 16970
+    },
+    {
+      "epoch": 0.9748722561384036,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.4461724218004386e-05,
+      "loss": 6.9294,
+      "step": 16980
+    },
+    {
+      "epoch": 0.9754463858534438,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.441541799085341e-05,
+      "loss": 6.8962,
+      "step": 16990
+    },
+    {
+      "epoch": 0.9760205155684841,
+      "grad_norm": 1.625,
+      "learning_rate": 5.436910794648794e-05,
+      "loss": 6.8631,
+      "step": 17000
+    },
+    {
+      "epoch": 0.9765946452835244,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.432279412494386e-05,
+      "loss": 6.8757,
+      "step": 17010
+    },
+    {
+      "epoch": 0.9771687749985647,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.4276476566260426e-05,
+      "loss": 6.9315,
+      "step": 17020
+    },
+    {
+      "epoch": 0.9777429047136049,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.423015531048003e-05,
+      "loss": 6.9108,
+      "step": 17030
+    },
+    {
+      "epoch": 0.9783170344286453,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.418383039764833e-05,
+      "loss": 6.851,
+      "step": 17040
+    },
+    {
+      "epoch": 0.9788911641436855,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.413750186781406e-05,
+      "loss": 6.8727,
+      "step": 17050
+    },
+    {
+      "epoch": 0.9794652938587258,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.409116976102916e-05,
+      "loss": 6.8615,
+      "step": 17060
+    },
+    {
+      "epoch": 0.980039423573766,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.4044834117348643e-05,
+      "loss": 6.8752,
+      "step": 17070
+    },
+    {
+      "epoch": 0.9806135532888064,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.3998494976830574e-05,
+      "loss": 6.8833,
+      "step": 17080
+    },
+    {
+      "epoch": 0.9811876830038466,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.395215237953601e-05,
+      "loss": 6.8956,
+      "step": 17090
+    },
+    {
+      "epoch": 0.981761812718887,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.390580636552904e-05,
+      "loss": 6.8642,
+      "step": 17100
+    },
+    {
+      "epoch": 0.9823359424339272,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.385945697487672e-05,
+      "loss": 6.9134,
+      "step": 17110
+    },
+    {
+      "epoch": 0.9829100721489675,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.3813104247648973e-05,
+      "loss": 6.8912,
+      "step": 17120
+    },
+    {
+      "epoch": 0.9834842018640078,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.376674822391861e-05,
+      "loss": 6.8854,
+      "step": 17130
+    },
+    {
+      "epoch": 0.9840583315790481,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.372038894376135e-05,
+      "loss": 6.9169,
+      "step": 17140
+    },
+    {
+      "epoch": 0.9846324612940883,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.367402644725566e-05,
+      "loss": 6.8841,
+      "step": 17150
+    },
+    {
+      "epoch": 0.9852065910091287,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.3627660774482846e-05,
+      "loss": 6.9321,
+      "step": 17160
+    },
+    {
+      "epoch": 0.985780720724169,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.3581291965526924e-05,
+      "loss": 6.9105,
+      "step": 17170
+    },
+    {
+      "epoch": 0.9863548504392092,
+      "grad_norm": 1.5,
+      "learning_rate": 5.353492006047461e-05,
+      "loss": 6.9344,
+      "step": 17180
+    },
+    {
+      "epoch": 0.9869289801542496,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.348854509941533e-05,
+      "loss": 6.8562,
+      "step": 17190
+    },
+    {
+      "epoch": 0.9875031098692898,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.3442167122441145e-05,
+      "loss": 6.9075,
+      "step": 17200
+    },
+    {
+      "epoch": 0.9880772395843301,
+      "grad_norm": 1.75,
+      "learning_rate": 5.33957861696467e-05,
+      "loss": 6.894,
+      "step": 17210
+    },
+    {
+      "epoch": 0.9886513692993704,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.3349402281129246e-05,
+      "loss": 6.8863,
+      "step": 17220
+    },
+    {
+      "epoch": 0.9892254990144107,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.330301549698853e-05,
+      "loss": 6.8475,
+      "step": 17230
+    },
+    {
+      "epoch": 0.9897996287294509,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.325662585732683e-05,
+      "loss": 6.8824,
+      "step": 17240
+    },
+    {
+      "epoch": 0.9903737584444913,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.321023340224893e-05,
+      "loss": 6.8266,
+      "step": 17250
+    },
+    {
+      "epoch": 0.9909478881595315,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.316383817186196e-05,
+      "loss": 6.8925,
+      "step": 17260
+    },
+    {
+      "epoch": 0.9915220178745718,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.3117440206275504e-05,
+      "loss": 6.916,
+      "step": 17270
+    },
+    {
+      "epoch": 0.9920961475896121,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.307103954560153e-05,
+      "loss": 6.9125,
+      "step": 17280
+    },
+    {
+      "epoch": 0.9926702773046524,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.302463622995429e-05,
+      "loss": 6.8616,
+      "step": 17290
+    },
+    {
+      "epoch": 0.9932444070196926,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.297823029945036e-05,
+      "loss": 6.9259,
+      "step": 17300
+    },
+    {
+      "epoch": 0.993818536734733,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.293182179420855e-05,
+      "loss": 6.8977,
+      "step": 17310
+    },
+    {
+      "epoch": 0.9943926664497732,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.288541075434992e-05,
+      "loss": 6.9003,
+      "step": 17320
+    },
+    {
+      "epoch": 0.9949667961648135,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.283899721999772e-05,
+      "loss": 6.8966,
+      "step": 17330
+    },
+    {
+      "epoch": 0.9955409258798538,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.279258123127735e-05,
+      "loss": 6.9307,
+      "step": 17340
+    },
+    {
+      "epoch": 0.9961150555948941,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.2746162828316334e-05,
+      "loss": 6.9119,
+      "step": 17350
+    },
+    {
+      "epoch": 0.9966891853099343,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.269974205124426e-05,
+      "loss": 6.8631,
+      "step": 17360
+    },
+    {
+      "epoch": 0.9972633150249747,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.265331894019283e-05,
+      "loss": 6.8766,
+      "step": 17370
+    },
+    {
+      "epoch": 0.9978374447400149,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.260689353529571e-05,
+      "loss": 6.8461,
+      "step": 17380
+    },
+    {
+      "epoch": 0.9984115744550552,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.256046587668855e-05,
+      "loss": 6.8429,
+      "step": 17390
+    },
+    {
+      "epoch": 0.9989857041700955,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.251403600450895e-05,
+      "loss": 6.8993,
+      "step": 17400
+    },
+    {
+      "epoch": 0.9995598338851358,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.246760395889646e-05,
+      "loss": 6.8517,
+      "step": 17410
+    },
+    {
+      "epoch": 1.0001148259430082,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.2421169779992486e-05,
+      "loss": 6.8916,
+      "step": 17420
+    },
+    {
+      "epoch": 1.0006889556580483,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.237473350794026e-05,
+      "loss": 6.8312,
+      "step": 17430
+    },
+    {
+      "epoch": 1.0012630853730886,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.232829518288487e-05,
+      "loss": 6.8256,
+      "step": 17440
+    },
+    {
+      "epoch": 1.001837215088129,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.22818548449731e-05,
+      "loss": 6.8157,
+      "step": 17450
+    },
+    {
+      "epoch": 1.0024113448031693,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.223541253435356e-05,
+      "loss": 6.8387,
+      "step": 17460
+    },
+    {
+      "epoch": 1.0029854745182094,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.2188968291176524e-05,
+      "loss": 6.7973,
+      "step": 17470
+    },
+    {
+      "epoch": 1.0035596042332497,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.214252215559393e-05,
+      "loss": 6.8648,
+      "step": 17480
+    },
+    {
+      "epoch": 1.00413373394829,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.209607416775937e-05,
+      "loss": 6.8186,
+      "step": 17490
+    },
+    {
+      "epoch": 1.0047078636633304,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.204962436782802e-05,
+      "loss": 6.8089,
+      "step": 17500
+    },
+    {
+      "epoch": 1.0052819933783705,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.200317279595666e-05,
+      "loss": 6.7777,
+      "step": 17510
+    },
+    {
+      "epoch": 1.0058561230934109,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.1956719492303554e-05,
+      "loss": 6.7948,
+      "step": 17520
+    },
+    {
+      "epoch": 1.0064302528084512,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.191026449702848e-05,
+      "loss": 6.8556,
+      "step": 17530
+    },
+    {
+      "epoch": 1.0070043825234916,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.186380785029269e-05,
+      "loss": 6.8099,
+      "step": 17540
+    },
+    {
+      "epoch": 1.0075785122385317,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.181734959225886e-05,
+      "loss": 6.8332,
+      "step": 17550
+    },
+    {
+      "epoch": 1.008152641953572,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.177088976309106e-05,
+      "loss": 6.8506,
+      "step": 17560
+    },
+    {
+      "epoch": 1.0087267716686124,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.1724428402954694e-05,
+      "loss": 6.805,
+      "step": 17570
+    },
+    {
+      "epoch": 1.0093009013836527,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.1677965552016515e-05,
+      "loss": 6.8179,
+      "step": 17580
+    },
+    {
+      "epoch": 1.0098750310986928,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.163150125044458e-05,
+      "loss": 6.8217,
+      "step": 17590
+    },
+    {
+      "epoch": 1.0104491608137331,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.1585035538408155e-05,
+      "loss": 6.7778,
+      "step": 17600
+    },
+    {
+      "epoch": 1.0110232905287735,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.153856845607776e-05,
+      "loss": 6.8724,
+      "step": 17610
+    },
+    {
+      "epoch": 1.0115974202438138,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.149210004362508e-05,
+      "loss": 6.8117,
+      "step": 17620
+    },
+    {
+      "epoch": 1.012171549958854,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.1445630341222984e-05,
+      "loss": 6.7884,
+      "step": 17630
+    },
+    {
+      "epoch": 1.0127456796738943,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.1399159389045406e-05,
+      "loss": 6.8375,
+      "step": 17640
+    },
+    {
+      "epoch": 1.0133198093889346,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.1352687227267395e-05,
+      "loss": 6.8501,
+      "step": 17650
+    },
+    {
+      "epoch": 1.013893939103975,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.1306213896065024e-05,
+      "loss": 6.8114,
+      "step": 17660
+    },
+    {
+      "epoch": 1.014468068819015,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.12597394356154e-05,
+      "loss": 6.8388,
+      "step": 17670
+    },
+    {
+      "epoch": 1.0150421985340554,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.121326388609661e-05,
+      "loss": 6.8205,
+      "step": 17680
+    },
+    {
+      "epoch": 1.0156163282490958,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.116678728768764e-05,
+      "loss": 6.8352,
+      "step": 17690
+    },
+    {
+      "epoch": 1.016190457964136,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.112030968056843e-05,
+      "loss": 6.8312,
+      "step": 17700
+    },
+    {
+      "epoch": 1.0167645876791762,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.107383110491978e-05,
+      "loss": 6.8543,
+      "step": 17710
+    },
+    {
+      "epoch": 1.0173387173942166,
+      "grad_norm": 1.625,
+      "learning_rate": 5.102735160092329e-05,
+      "loss": 6.7808,
+      "step": 17720
+    },
+    {
+      "epoch": 1.017912847109257,
+      "grad_norm": 2.125,
+      "learning_rate": 5.0980871208761427e-05,
+      "loss": 6.815,
+      "step": 17730
+    },
+    {
+      "epoch": 1.0184869768242972,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.0934389968617366e-05,
+      "loss": 6.8204,
+      "step": 17740
+    },
+    {
+      "epoch": 1.0190611065393373,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.088790792067506e-05,
+      "loss": 6.8204,
+      "step": 17750
+    },
+    {
+      "epoch": 1.0196352362543777,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.084142510511911e-05,
+      "loss": 6.7875,
+      "step": 17760
+    },
+    {
+      "epoch": 1.020209365969418,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.079494156213485e-05,
+      "loss": 6.8449,
+      "step": 17770
+    },
+    {
+      "epoch": 1.0207834956844584,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.0748457331908186e-05,
+      "loss": 6.7516,
+      "step": 17780
+    },
+    {
+      "epoch": 1.0213576253994985,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.070197245462564e-05,
+      "loss": 6.7899,
+      "step": 17790
+    },
+    {
+      "epoch": 1.0219317551145388,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.065548697047429e-05,
+      "loss": 6.7944,
+      "step": 17800
+    },
+    {
+      "epoch": 1.0225058848295792,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.060900091964174e-05,
+      "loss": 6.7772,
+      "step": 17810
+    },
+    {
+      "epoch": 1.0230800145446195,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.056251434231607e-05,
+      "loss": 6.8124,
+      "step": 17820
+    },
+    {
+      "epoch": 1.0236541442596598,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.051602727868585e-05,
+      "loss": 6.7964,
+      "step": 17830
+    },
+    {
+      "epoch": 1.0242282739747,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.046953976894002e-05,
+      "loss": 6.8475,
+      "step": 17840
+    },
+    {
+      "epoch": 1.0248024036897403,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.0423051853267965e-05,
+      "loss": 6.8388,
+      "step": 17850
+    },
+    {
+      "epoch": 1.0253765334047806,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.037656357185938e-05,
+      "loss": 6.824,
+      "step": 17860
+    },
+    {
+      "epoch": 1.025950663119821,
+      "grad_norm": 1.5,
+      "learning_rate": 5.0330074964904273e-05,
+      "loss": 6.7823,
+      "step": 17870
+    },
+    {
+      "epoch": 1.026524792834861,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.028358607259297e-05,
+      "loss": 6.8141,
+      "step": 17880
+    },
+    {
+      "epoch": 1.0270989225499014,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.023709693511598e-05,
+      "loss": 6.8081,
+      "step": 17890
+    },
+    {
+      "epoch": 1.0276730522649418,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.019060759266411e-05,
+      "loss": 6.7904,
+      "step": 17900
+    },
+    {
+      "epoch": 1.028247181979982,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.014411808542827e-05,
+      "loss": 6.8179,
+      "step": 17910
+    },
+    {
+      "epoch": 1.0288213116950222,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.009762845359954e-05,
+      "loss": 6.7808,
+      "step": 17920
+    },
+    {
+      "epoch": 1.0293954414100626,
+      "grad_norm": 1.4765625,
+      "learning_rate": 5.005113873736914e-05,
+      "loss": 6.784,
+      "step": 17930
+    },
+    {
+      "epoch": 1.029969571125103,
+      "grad_norm": 1.7890625,
+      "learning_rate": 5.000464897692829e-05,
+      "loss": 6.7981,
+      "step": 17940
+    },
+    {
+      "epoch": 1.0305437008401432,
+      "grad_norm": 1.5,
+      "learning_rate": 4.9958159212468335e-05,
+      "loss": 6.8296,
+      "step": 17950
+    },
+    {
+      "epoch": 1.0311178305551834,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.991166948418054e-05,
+      "loss": 6.836,
+      "step": 17960
+    },
+    {
+      "epoch": 1.0316919602702237,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.9865179832256174e-05,
+      "loss": 6.8217,
+      "step": 17970
+    },
+    {
+      "epoch": 1.032266089985264,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.9818690296886475e-05,
+      "loss": 6.86,
+      "step": 17980
+    },
+    {
+      "epoch": 1.0328402197003044,
+      "grad_norm": 1.7421875,
+      "learning_rate": 4.977220091826253e-05,
+      "loss": 6.8151,
+      "step": 17990
+    },
+    {
+      "epoch": 1.0334143494153445,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.972571173657531e-05,
+      "loss": 6.7877,
+      "step": 18000
+    },
+    {
+      "epoch": 1.0339884791303848,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.96792227920156e-05,
+      "loss": 6.7953,
+      "step": 18010
+    },
+    {
+      "epoch": 1.0345626088454252,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.963273412477403e-05,
+      "loss": 6.8301,
+      "step": 18020
+    },
+    {
+      "epoch": 1.0351367385604655,
+      "grad_norm": 1.640625,
+      "learning_rate": 4.9586245775040926e-05,
+      "loss": 6.8222,
+      "step": 18030
+    },
+    {
+      "epoch": 1.0357108682755056,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.9539757783006376e-05,
+      "loss": 6.805,
+      "step": 18040
+    },
+    {
+      "epoch": 1.036284997990546,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.949327018886013e-05,
+      "loss": 6.778,
+      "step": 18050
+    },
+    {
+      "epoch": 1.0368591277055863,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.944678303279166e-05,
+      "loss": 6.7922,
+      "step": 18060
+    },
+    {
+      "epoch": 1.0374332574206266,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.9400296354989974e-05,
+      "loss": 6.8341,
+      "step": 18070
+    },
+    {
+      "epoch": 1.0380073871356668,
+      "grad_norm": 1.6484375,
+      "learning_rate": 4.935381019564374e-05,
+      "loss": 6.853,
+      "step": 18080
+    },
+    {
+      "epoch": 1.038581516850707,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.930732459494113e-05,
+      "loss": 6.8091,
+      "step": 18090
+    },
+    {
+      "epoch": 1.0391556465657474,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.926083959306984e-05,
+      "loss": 6.8279,
+      "step": 18100
+    },
+    {
+      "epoch": 1.0397297762807878,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.9214355230217054e-05,
+      "loss": 6.844,
+      "step": 18110
+    },
+    {
+      "epoch": 1.040303905995828,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.916787154656942e-05,
+      "loss": 6.8196,
+      "step": 18120
+    },
+    {
+      "epoch": 1.0408780357108682,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.912138858231297e-05,
+      "loss": 6.8181,
+      "step": 18130
+    },
+    {
+      "epoch": 1.0414521654259086,
+      "grad_norm": 1.609375,
+      "learning_rate": 4.907490637763314e-05,
+      "loss": 6.7709,
+      "step": 18140
+    },
+    {
+      "epoch": 1.042026295140949,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.90284249727147e-05,
+      "loss": 6.7914,
+      "step": 18150
+    },
+    {
+      "epoch": 1.042600424855989,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.8981944407741704e-05,
+      "loss": 6.7891,
+      "step": 18160
+    },
+    {
+      "epoch": 1.0431745545710294,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.8935464722897525e-05,
+      "loss": 6.8074,
+      "step": 18170
+    },
+    {
+      "epoch": 1.0437486842860697,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.888898595836475e-05,
+      "loss": 6.8193,
+      "step": 18180
+    },
+    {
+      "epoch": 1.04432281400111,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.884250815432516e-05,
+      "loss": 6.8144,
+      "step": 18190
+    },
+    {
+      "epoch": 1.0448969437161502,
+      "grad_norm": 1.625,
+      "learning_rate": 4.879603135095972e-05,
+      "loss": 6.8019,
+      "step": 18200
+    },
+    {
+      "epoch": 1.0454710734311905,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.874955558844854e-05,
+      "loss": 6.8075,
+      "step": 18210
+    },
+    {
+      "epoch": 1.0460452031462308,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.870308090697082e-05,
+      "loss": 6.799,
+      "step": 18220
+    },
+    {
+      "epoch": 1.0466193328612712,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.8656607346704805e-05,
+      "loss": 6.7688,
+      "step": 18230
+    },
+    {
+      "epoch": 1.0471934625763115,
+      "grad_norm": 1.65625,
+      "learning_rate": 4.861013494782779e-05,
+      "loss": 6.8544,
+      "step": 18240
+    },
+    {
+      "epoch": 1.0477675922913516,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.8563663750516094e-05,
+      "loss": 6.7944,
+      "step": 18250
+    },
+    {
+      "epoch": 1.048341722006392,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.8517193794944945e-05,
+      "loss": 6.8,
+      "step": 18260
+    },
+    {
+      "epoch": 1.0489158517214323,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.847072512128852e-05,
+      "loss": 6.8211,
+      "step": 18270
+    },
+    {
+      "epoch": 1.0494899814364727,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.8424257769719885e-05,
+      "loss": 6.8798,
+      "step": 18280
+    },
+    {
+      "epoch": 1.0500641111515128,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.8377791780411e-05,
+      "loss": 6.7913,
+      "step": 18290
+    },
+    {
+      "epoch": 1.050638240866553,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.8331327193532585e-05,
+      "loss": 6.8235,
+      "step": 18300
+    },
+    {
+      "epoch": 1.0512123705815934,
+      "grad_norm": 1.6640625,
+      "learning_rate": 4.828486404925418e-05,
+      "loss": 6.822,
+      "step": 18310
+    },
+    {
+      "epoch": 1.0517865002966338,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.823840238774408e-05,
+      "loss": 6.7757,
+      "step": 18320
+    },
+    {
+      "epoch": 1.052360630011674,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.8191942249169266e-05,
+      "loss": 6.8006,
+      "step": 18330
+    },
+    {
+      "epoch": 1.0529347597267142,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.8145483673695476e-05,
+      "loss": 6.7836,
+      "step": 18340
+    },
+    {
+      "epoch": 1.0535088894417546,
+      "grad_norm": 1.6328125,
+      "learning_rate": 4.8099026701487024e-05,
+      "loss": 6.8307,
+      "step": 18350
+    },
+    {
+      "epoch": 1.054083019156795,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.8052571372706866e-05,
+      "loss": 6.7965,
+      "step": 18360
+    },
+    {
+      "epoch": 1.054657148871835,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.8006117727516525e-05,
+      "loss": 6.8181,
+      "step": 18370
+    },
+    {
+      "epoch": 1.0552312785868754,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.79596658060761e-05,
+      "loss": 6.851,
+      "step": 18380
+    },
+    {
+      "epoch": 1.0558054083019157,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.791321564854417e-05,
+      "loss": 6.8201,
+      "step": 18390
+    },
+    {
+      "epoch": 1.056379538016956,
+      "grad_norm": 1.609375,
+      "learning_rate": 4.786676729507779e-05,
+      "loss": 6.8427,
+      "step": 18400
+    },
+    {
+      "epoch": 1.0569536677319962,
+      "grad_norm": 1.640625,
+      "learning_rate": 4.7820320785832454e-05,
+      "loss": 6.7823,
+      "step": 18410
+    },
+    {
+      "epoch": 1.0575277974470365,
+      "grad_norm": 1.640625,
+      "learning_rate": 4.777387616096209e-05,
+      "loss": 6.7862,
+      "step": 18420
+    },
+    {
+      "epoch": 1.0581019271620768,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.772743346061898e-05,
+      "loss": 6.845,
+      "step": 18430
+    },
+    {
+      "epoch": 1.0586760568771172,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.768099272495373e-05,
+      "loss": 6.8106,
+      "step": 18440
+    },
+    {
+      "epoch": 1.0592501865921573,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.763455399411524e-05,
+      "loss": 6.8431,
+      "step": 18450
+    },
+    {
+      "epoch": 1.0598243163071976,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.7588117308250725e-05,
+      "loss": 6.8614,
+      "step": 18460
+    },
+    {
+      "epoch": 1.060398446022238,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.7541682707505584e-05,
+      "loss": 6.8251,
+      "step": 18470
+    },
+    {
+      "epoch": 1.0609725757372783,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.7495250232023435e-05,
+      "loss": 6.863,
+      "step": 18480
+    },
+    {
+      "epoch": 1.0615467054523184,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.744881992194605e-05,
+      "loss": 6.8241,
+      "step": 18490
+    },
+    {
+      "epoch": 1.0621208351673588,
+      "grad_norm": 1.5,
+      "learning_rate": 4.7402391817413324e-05,
+      "loss": 6.8399,
+      "step": 18500
+    },
+    {
+      "epoch": 1.0626949648823991,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.735596595856327e-05,
+      "loss": 6.8467,
+      "step": 18510
+    },
+    {
+      "epoch": 1.0632690945974395,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.730954238553194e-05,
+      "loss": 6.8012,
+      "step": 18520
+    },
+    {
+      "epoch": 1.0638432243124796,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.7263121138453406e-05,
+      "loss": 6.7794,
+      "step": 18530
+    },
+    {
+      "epoch": 1.06441735402752,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.721670225745974e-05,
+      "loss": 6.7886,
+      "step": 18540
+    },
+    {
+      "epoch": 1.0649914837425603,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.717028578268097e-05,
+      "loss": 6.8604,
+      "step": 18550
+    },
+    {
+      "epoch": 1.0655656134576006,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.712387175424504e-05,
+      "loss": 6.7841,
+      "step": 18560
+    },
+    {
+      "epoch": 1.0661397431726407,
+      "grad_norm": 1.609375,
+      "learning_rate": 4.7077460212277775e-05,
+      "loss": 6.8609,
+      "step": 18570
+    },
+    {
+      "epoch": 1.066713872887681,
+      "grad_norm": 1.6328125,
+      "learning_rate": 4.703105119690283e-05,
+      "loss": 6.8164,
+      "step": 18580
+    },
+    {
+      "epoch": 1.0672880026027214,
+      "grad_norm": 1.65625,
+      "learning_rate": 4.6984644748241744e-05,
+      "loss": 6.817,
+      "step": 18590
+    },
+    {
+      "epoch": 1.0678621323177617,
+      "grad_norm": 1.640625,
+      "learning_rate": 4.693824090641375e-05,
+      "loss": 6.8193,
+      "step": 18600
+    },
+    {
+      "epoch": 1.0684362620328018,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.689183971153591e-05,
+      "loss": 6.7969,
+      "step": 18610
+    },
+    {
+      "epoch": 1.0690103917478422,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.6845441203722926e-05,
+      "loss": 6.8536,
+      "step": 18620
+    },
+    {
+      "epoch": 1.0695845214628825,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.6799045423087225e-05,
+      "loss": 6.8249,
+      "step": 18630
+    },
+    {
+      "epoch": 1.0701586511779229,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.675265240973883e-05,
+      "loss": 6.8164,
+      "step": 18640
+    },
+    {
+      "epoch": 1.0707327808929632,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.670626220378545e-05,
+      "loss": 6.8229,
+      "step": 18650
+    },
+    {
+      "epoch": 1.0713069106080033,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.665987484533229e-05,
+      "loss": 6.8113,
+      "step": 18660
+    },
+    {
+      "epoch": 1.0718810403230437,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.661349037448211e-05,
+      "loss": 6.8164,
+      "step": 18670
+    },
+    {
+      "epoch": 1.072455170038084,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.6567108831335196e-05,
+      "loss": 6.8221,
+      "step": 18680
+    },
+    {
+      "epoch": 1.073029299753124,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.65207302559893e-05,
+      "loss": 6.816,
+      "step": 18690
+    },
+    {
+      "epoch": 1.0736034294681644,
+      "grad_norm": 1.6875,
+      "learning_rate": 4.6474354688539596e-05,
+      "loss": 6.8139,
+      "step": 18700
+    },
+    {
+      "epoch": 1.0741775591832048,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.642798216907866e-05,
+      "loss": 6.8234,
+      "step": 18710
+    },
+    {
+      "epoch": 1.0747516888982451,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.638161273769643e-05,
+      "loss": 6.8128,
+      "step": 18720
+    },
+    {
+      "epoch": 1.0753258186132855,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.633524643448017e-05,
+      "loss": 6.7798,
+      "step": 18730
+    },
+    {
+      "epoch": 1.0758999483283256,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.6288883299514486e-05,
+      "loss": 6.8403,
+      "step": 18740
+    },
+    {
+      "epoch": 1.076474078043366,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.624252337288117e-05,
+      "loss": 6.8104,
+      "step": 18750
+    },
+    {
+      "epoch": 1.0770482077584063,
+      "grad_norm": 1.6484375,
+      "learning_rate": 4.619616669465929e-05,
+      "loss": 6.8176,
+      "step": 18760
+    },
+    {
+      "epoch": 1.0776223374734466,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.614981330492509e-05,
+      "loss": 6.8101,
+      "step": 18770
+    },
+    {
+      "epoch": 1.0781964671884867,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.6103463243751995e-05,
+      "loss": 6.7995,
+      "step": 18780
+    },
+    {
+      "epoch": 1.078770596903527,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.6057116551210506e-05,
+      "loss": 6.8148,
+      "step": 18790
+    },
+    {
+      "epoch": 1.0793447266185674,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.601077326736825e-05,
+      "loss": 6.8365,
+      "step": 18800
+    },
+    {
+      "epoch": 1.0799188563336077,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.5964433432289885e-05,
+      "loss": 6.8396,
+      "step": 18810
+    },
+    {
+      "epoch": 1.0804929860486479,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.5918097086037116e-05,
+      "loss": 6.7852,
+      "step": 18820
+    },
+    {
+      "epoch": 1.0810671157636882,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.5871764268668603e-05,
+      "loss": 6.7954,
+      "step": 18830
+    },
+    {
+      "epoch": 1.0816412454787285,
+      "grad_norm": 1.6640625,
+      "learning_rate": 4.582543502023996e-05,
+      "loss": 6.8054,
+      "step": 18840
+    },
+    {
+      "epoch": 1.0822153751937689,
+      "grad_norm": 1.6875,
+      "learning_rate": 4.577910938080372e-05,
+      "loss": 6.8172,
+      "step": 18850
+    },
+    {
+      "epoch": 1.082789504908809,
+      "grad_norm": 1.609375,
+      "learning_rate": 4.573278739040932e-05,
+      "loss": 6.826,
+      "step": 18860
+    },
+    {
+      "epoch": 1.0833636346238493,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.568646908910299e-05,
+      "loss": 6.8217,
+      "step": 18870
+    },
+    {
+      "epoch": 1.0839377643388897,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.564015451692782e-05,
+      "loss": 6.7828,
+      "step": 18880
+    },
+    {
+      "epoch": 1.08451189405393,
+      "grad_norm": 1.625,
+      "learning_rate": 4.5593843713923645e-05,
+      "loss": 6.7962,
+      "step": 18890
+    },
+    {
+      "epoch": 1.0850860237689701,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.5547536720127045e-05,
+      "loss": 6.8068,
+      "step": 18900
+    },
+    {
+      "epoch": 1.0856601534840105,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.5501233575571337e-05,
+      "loss": 6.831,
+      "step": 18910
+    },
+    {
+      "epoch": 1.0862342831990508,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.545493432028648e-05,
+      "loss": 6.8396,
+      "step": 18920
+    },
+    {
+      "epoch": 1.0868084129140911,
+      "grad_norm": 1.6953125,
+      "learning_rate": 4.5408638994299066e-05,
+      "loss": 6.8095,
+      "step": 18930
+    },
+    {
+      "epoch": 1.0873825426291313,
+      "grad_norm": 1.6796875,
+      "learning_rate": 4.536234763763231e-05,
+      "loss": 6.7861,
+      "step": 18940
+    },
+    {
+      "epoch": 1.0879566723441716,
+      "grad_norm": 1.640625,
+      "learning_rate": 4.5316060290305996e-05,
+      "loss": 6.7677,
+      "step": 18950
+    },
+    {
+      "epoch": 1.088530802059212,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.526977699233643e-05,
+      "loss": 6.7941,
+      "step": 18960
+    },
+    {
+      "epoch": 1.0891049317742523,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.522349778373641e-05,
+      "loss": 6.7881,
+      "step": 18970
+    },
+    {
+      "epoch": 1.0896790614892924,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.517722270451521e-05,
+      "loss": 6.8236,
+      "step": 18980
+    },
+    {
+      "epoch": 1.0902531912043327,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.513095179467855e-05,
+      "loss": 6.8218,
+      "step": 18990
+    },
+    {
+      "epoch": 1.090827320919373,
+      "grad_norm": 1.640625,
+      "learning_rate": 4.5084685094228524e-05,
+      "loss": 6.8067,
+      "step": 19000
+    },
+    {
+      "epoch": 1.0914014506344134,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.503842264316359e-05,
+      "loss": 6.8452,
+      "step": 19010
+    },
+    {
+      "epoch": 1.0919755803494535,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.499216448147852e-05,
+      "loss": 6.8308,
+      "step": 19020
+    },
+    {
+      "epoch": 1.0925497100644939,
+      "grad_norm": 1.6640625,
+      "learning_rate": 4.494591064916441e-05,
+      "loss": 6.8074,
+      "step": 19030
+    },
+    {
+      "epoch": 1.0931238397795342,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.489966118620859e-05,
+      "loss": 6.7477,
+      "step": 19040
+    },
+    {
+      "epoch": 1.0936979694945745,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.485341613259462e-05,
+      "loss": 6.8247,
+      "step": 19050
+    },
+    {
+      "epoch": 1.0942720992096147,
+      "grad_norm": 1.6640625,
+      "learning_rate": 4.4807175528302234e-05,
+      "loss": 6.808,
+      "step": 19060
+    },
+    {
+      "epoch": 1.094846228924655,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.4760939413307355e-05,
+      "loss": 6.8517,
+      "step": 19070
+    },
+    {
+      "epoch": 1.0954203586396953,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.4714707827581994e-05,
+      "loss": 6.8317,
+      "step": 19080
+    },
+    {
+      "epoch": 1.0959944883547357,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.466848081109424e-05,
+      "loss": 6.833,
+      "step": 19090
+    },
+    {
+      "epoch": 1.0965686180697758,
+      "grad_norm": 1.625,
+      "learning_rate": 4.4622258403808226e-05,
+      "loss": 6.8285,
+      "step": 19100
+    },
+    {
+      "epoch": 1.0971427477848161,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.4576040645684174e-05,
+      "loss": 6.8295,
+      "step": 19110
+    },
+    {
+      "epoch": 1.0977168774998565,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.452982757667821e-05,
+      "loss": 6.7981,
+      "step": 19120
+    },
+    {
+      "epoch": 1.0982910072148968,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.44836192367424e-05,
+      "loss": 6.8319,
+      "step": 19130
+    },
+    {
+      "epoch": 1.0988651369299371,
+      "grad_norm": 1.640625,
+      "learning_rate": 4.44374156658248e-05,
+      "loss": 6.8486,
+      "step": 19140
+    },
+    {
+      "epoch": 1.0994392666449773,
+      "grad_norm": 1.6328125,
+      "learning_rate": 4.439121690386926e-05,
+      "loss": 6.7847,
+      "step": 19150
+    },
+    {
+      "epoch": 1.1000133963600176,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.434502299081551e-05,
+      "loss": 6.7814,
+      "step": 19160
+    },
+    {
+      "epoch": 1.100587526075058,
+      "grad_norm": 1.6328125,
+      "learning_rate": 4.429883396659908e-05,
+      "loss": 6.7948,
+      "step": 19170
+    },
+    {
+      "epoch": 1.101161655790098,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.42526498711513e-05,
+      "loss": 6.7628,
+      "step": 19180
+    },
+    {
+      "epoch": 1.1017357855051384,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.42064707443992e-05,
+      "loss": 6.8418,
+      "step": 19190
+    },
+    {
+      "epoch": 1.1023099152201787,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.416029662626553e-05,
+      "loss": 6.7977,
+      "step": 19200
+    },
+    {
+      "epoch": 1.102884044935219,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.41141275566687e-05,
+      "loss": 6.7637,
+      "step": 19210
+    },
+    {
+      "epoch": 1.1034581746502594,
+      "grad_norm": 1.609375,
+      "learning_rate": 4.40679635755228e-05,
+      "loss": 6.8116,
+      "step": 19220
+    },
+    {
+      "epoch": 1.1040323043652995,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.4021804722737466e-05,
+      "loss": 6.8296,
+      "step": 19230
+    },
+    {
+      "epoch": 1.1046064340803399,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.3975651038217916e-05,
+      "loss": 6.7715,
+      "step": 19240
+    },
+    {
+      "epoch": 1.1051805637953802,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.39295025618649e-05,
+      "loss": 6.821,
+      "step": 19250
+    },
+    {
+      "epoch": 1.1057546935104205,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.3883359333574695e-05,
+      "loss": 6.8011,
+      "step": 19260
+    },
+    {
+      "epoch": 1.1063288232254607,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.3837221393239015e-05,
+      "loss": 6.8172,
+      "step": 19270
+    },
+    {
+      "epoch": 1.106902952940501,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.3791088780744984e-05,
+      "loss": 6.7852,
+      "step": 19280
+    },
+    {
+      "epoch": 1.1074770826555413,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.374496153597514e-05,
+      "loss": 6.8689,
+      "step": 19290
+    },
+    {
+      "epoch": 1.1080512123705817,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.3698839698807415e-05,
+      "loss": 6.7963,
+      "step": 19300
+    },
+    {
+      "epoch": 1.1086253420856218,
+      "grad_norm": 1.609375,
+      "learning_rate": 4.3652723309115e-05,
+      "loss": 6.7908,
+      "step": 19310
+    },
+    {
+      "epoch": 1.1091994718006621,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.360661240676642e-05,
+      "loss": 6.8105,
+      "step": 19320
+    },
+    {
+      "epoch": 1.1097736015157025,
+      "grad_norm": 1.609375,
+      "learning_rate": 4.3560507031625435e-05,
+      "loss": 6.7767,
+      "step": 19330
+    },
+    {
+      "epoch": 1.1103477312307428,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.351440722355104e-05,
+      "loss": 6.7945,
+      "step": 19340
+    },
+    {
+      "epoch": 1.110921860945783,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.346831302239743e-05,
+      "loss": 6.8243,
+      "step": 19350
+    },
+    {
+      "epoch": 1.1114959906608233,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.342222446801392e-05,
+      "loss": 6.8118,
+      "step": 19360
+    },
+    {
+      "epoch": 1.1120701203758636,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.3376141600244957e-05,
+      "loss": 6.822,
+      "step": 19370
+    },
+    {
+      "epoch": 1.112644250090904,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.3330064458930076e-05,
+      "loss": 6.8116,
+      "step": 19380
+    },
+    {
+      "epoch": 1.113218379805944,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.328399308390387e-05,
+      "loss": 6.8233,
+      "step": 19390
+    },
+    {
+      "epoch": 1.1137925095209844,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.323792751499593e-05,
+      "loss": 6.8053,
+      "step": 19400
+    },
+    {
+      "epoch": 1.1143666392360247,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.3191867792030834e-05,
+      "loss": 6.8136,
+      "step": 19410
+    },
+    {
+      "epoch": 1.114940768951065,
+      "grad_norm": 1.7109375,
+      "learning_rate": 4.314581395482809e-05,
+      "loss": 6.7782,
+      "step": 19420
+    },
+    {
+      "epoch": 1.1155148986661052,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.309976604320217e-05,
+      "loss": 6.7782,
+      "step": 19430
+    },
+    {
+      "epoch": 1.1160890283811455,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.305372409696236e-05,
+      "loss": 6.8324,
+      "step": 19440
+    },
+    {
+      "epoch": 1.1166631580961859,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.300768815591282e-05,
+      "loss": 6.8208,
+      "step": 19450
+    },
+    {
+      "epoch": 1.1172372878112262,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.296165825985251e-05,
+      "loss": 6.8429,
+      "step": 19460
+    },
+    {
+      "epoch": 1.1178114175262663,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.2915634448575184e-05,
+      "loss": 6.8359,
+      "step": 19470
+    },
+    {
+      "epoch": 1.1183855472413067,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.2869616761869304e-05,
+      "loss": 6.818,
+      "step": 19480
+    },
+    {
+      "epoch": 1.118959676956347,
+      "grad_norm": 1.53125,
+      "learning_rate": 4.282360523951806e-05,
+      "loss": 6.7769,
+      "step": 19490
+    },
+    {
+      "epoch": 1.1195338066713874,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.2777599921299304e-05,
+      "loss": 6.7761,
+      "step": 19500
+    },
+    {
+      "epoch": 1.1201079363864275,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.273160084698552e-05,
+      "loss": 6.7693,
+      "step": 19510
+    },
+    {
+      "epoch": 1.1206820661014678,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.268560805634382e-05,
+      "loss": 6.7573,
+      "step": 19520
+    },
+    {
+      "epoch": 1.1212561958165081,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.2639621589135845e-05,
+      "loss": 6.8521,
+      "step": 19530
+    },
+    {
+      "epoch": 1.1218303255315485,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.259364148511779e-05,
+      "loss": 6.787,
+      "step": 19540
+    },
+    {
+      "epoch": 1.1224044552465888,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.254766778404034e-05,
+      "loss": 6.8294,
+      "step": 19550
+    },
+    {
+      "epoch": 1.122978584961629,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.250170052564868e-05,
+      "loss": 6.8143,
+      "step": 19560
+    },
+    {
+      "epoch": 1.1235527146766693,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.2455739749682374e-05,
+      "loss": 6.8112,
+      "step": 19570
+    },
+    {
+      "epoch": 1.1241268443917096,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.24097854958754e-05,
+      "loss": 6.8071,
+      "step": 19580
+    },
+    {
+      "epoch": 1.1247009741067497,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.2363837803956115e-05,
+      "loss": 6.8058,
+      "step": 19590
+    },
+    {
+      "epoch": 1.12527510382179,
+      "grad_norm": 1.71875,
+      "learning_rate": 4.2317896713647185e-05,
+      "loss": 6.821,
+      "step": 19600
+    },
+    {
+      "epoch": 1.1258492335368304,
+      "grad_norm": 1.609375,
+      "learning_rate": 4.2271962264665575e-05,
+      "loss": 6.8413,
+      "step": 19610
+    },
+    {
+      "epoch": 1.1264233632518708,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.222603449672249e-05,
+      "loss": 6.7789,
+      "step": 19620
+    },
+    {
+      "epoch": 1.126997492966911,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.218011344952341e-05,
+      "loss": 6.7947,
+      "step": 19630
+    },
+    {
+      "epoch": 1.1275716226819512,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.2134199162767956e-05,
+      "loss": 6.8209,
+      "step": 19640
+    },
+    {
+      "epoch": 1.1281457523969916,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.208829167614991e-05,
+      "loss": 6.7714,
+      "step": 19650
+    },
+    {
+      "epoch": 1.128719882112032,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.20423910293572e-05,
+      "loss": 6.8094,
+      "step": 19660
+    },
+    {
+      "epoch": 1.129294011827072,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.199649726207181e-05,
+      "loss": 6.8415,
+      "step": 19670
+    },
+    {
+      "epoch": 1.1298681415421123,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.1950610413969814e-05,
+      "loss": 6.7762,
+      "step": 19680
+    },
+    {
+      "epoch": 1.1304422712571527,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.190473052472125e-05,
+      "loss": 6.8122,
+      "step": 19690
+    },
+    {
+      "epoch": 1.131016400972193,
+      "grad_norm": 1.6875,
+      "learning_rate": 4.1858857633990204e-05,
+      "loss": 6.7658,
+      "step": 19700
+    },
+    {
+      "epoch": 1.1315905306872334,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.181299178143467e-05,
+      "loss": 6.7999,
+      "step": 19710
+    },
+    {
+      "epoch": 1.1321646604022735,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.1767133006706555e-05,
+      "loss": 6.8033,
+      "step": 19720
+    },
+    {
+      "epoch": 1.1327387901173138,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.172128134945167e-05,
+      "loss": 6.8088,
+      "step": 19730
+    },
+    {
+      "epoch": 1.1333129198323542,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.167543684930966e-05,
+      "loss": 6.7741,
+      "step": 19740
+    },
+    {
+      "epoch": 1.1338870495473945,
+      "grad_norm": 1.625,
+      "learning_rate": 4.162959954591399e-05,
+      "loss": 6.8355,
+      "step": 19750
+    },
+    {
+      "epoch": 1.1344611792624346,
+      "grad_norm": 1.65625,
+      "learning_rate": 4.1583769478891885e-05,
+      "loss": 6.7529,
+      "step": 19760
+    },
+    {
+      "epoch": 1.135035308977475,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.153794668786435e-05,
+      "loss": 6.8089,
+      "step": 19770
+    },
+    {
+      "epoch": 1.1356094386925153,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.149213121244604e-05,
+      "loss": 6.7903,
+      "step": 19780
+    },
+    {
+      "epoch": 1.1361835684075556,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.144632309224536e-05,
+      "loss": 6.8052,
+      "step": 19790
+    },
+    {
+      "epoch": 1.1367576981225957,
+      "grad_norm": 1.625,
+      "learning_rate": 4.1400522366864306e-05,
+      "loss": 6.8578,
+      "step": 19800
+    },
+    {
+      "epoch": 1.137331827837636,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.135472907589849e-05,
+      "loss": 6.8058,
+      "step": 19810
+    },
+    {
+      "epoch": 1.1379059575526764,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.130894325893708e-05,
+      "loss": 6.8482,
+      "step": 19820
+    },
+    {
+      "epoch": 1.1384800872677168,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.126316495556284e-05,
+      "loss": 6.7568,
+      "step": 19830
+    },
+    {
+      "epoch": 1.1390542169827569,
+      "grad_norm": 1.6796875,
+      "learning_rate": 4.121739420535199e-05,
+      "loss": 6.8415,
+      "step": 19840
+    },
+    {
+      "epoch": 1.1396283466977972,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.117163104787422e-05,
+      "loss": 6.803,
+      "step": 19850
+    },
+    {
+      "epoch": 1.1402024764128376,
+      "grad_norm": 1.609375,
+      "learning_rate": 4.112587552269267e-05,
+      "loss": 6.7807,
+      "step": 19860
+    },
+    {
+      "epoch": 1.140776606127878,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.108012766936389e-05,
+      "loss": 6.8434,
+      "step": 19870
+    },
+    {
+      "epoch": 1.141350735842918,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.103438752743778e-05,
+      "loss": 6.8533,
+      "step": 19880
+    },
+    {
+      "epoch": 1.1419248655579584,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.0988655136457583e-05,
+      "loss": 6.8059,
+      "step": 19890
+    },
+    {
+      "epoch": 1.1424989952729987,
+      "grad_norm": 1.625,
+      "learning_rate": 4.094293053595983e-05,
+      "loss": 6.7928,
+      "step": 19900
+    },
+    {
+      "epoch": 1.143073124988039,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.089721376547433e-05,
+      "loss": 6.8311,
+      "step": 19910
+    },
+    {
+      "epoch": 1.1436472547030792,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.085150486452412e-05,
+      "loss": 6.8427,
+      "step": 19920
+    },
+    {
+      "epoch": 1.1442213844181195,
+      "grad_norm": 1.6953125,
+      "learning_rate": 4.0805803872625434e-05,
+      "loss": 6.7816,
+      "step": 19930
+    },
+    {
+      "epoch": 1.1447955141331598,
+      "grad_norm": 1.6796875,
+      "learning_rate": 4.076011082928766e-05,
+      "loss": 6.8232,
+      "step": 19940
+    },
+    {
+      "epoch": 1.1453696438482002,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.071442577401331e-05,
+      "loss": 6.7945,
+      "step": 19950
+    },
+    {
+      "epoch": 1.1459437735632405,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.0668748746298026e-05,
+      "loss": 6.813,
+      "step": 19960
+    },
+    {
+      "epoch": 1.1465179032782806,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.062307978563047e-05,
+      "loss": 6.8593,
+      "step": 19970
+    },
+    {
+      "epoch": 1.147092032993321,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.057741893149234e-05,
+      "loss": 6.81,
+      "step": 19980
+    },
+    {
+      "epoch": 1.1476661627083613,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.053176622335834e-05,
+      "loss": 6.7886,
+      "step": 19990
+    },
+    {
+      "epoch": 1.1482402924234014,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.048612170069612e-05,
+      "loss": 6.8227,
+      "step": 20000
+    },
+    {
+      "epoch": 1.1488144221384418,
+      "grad_norm": 1.6328125,
+      "learning_rate": 4.0440485402966254e-05,
+      "loss": 6.8347,
+      "step": 20010
+    },
+    {
+      "epoch": 1.149388551853482,
+      "grad_norm": 1.6484375,
+      "learning_rate": 4.039485736962221e-05,
+      "loss": 6.7834,
+      "step": 20020
+    },
+    {
+      "epoch": 1.1499626815685224,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.034923764011029e-05,
+      "loss": 6.8269,
+      "step": 20030
+    },
+    {
+      "epoch": 1.1505368112835628,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.0303626253869655e-05,
+      "loss": 6.8141,
+      "step": 20040
+    },
+    {
+      "epoch": 1.151110940998603,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.0258023250332235e-05,
+      "loss": 6.7619,
+      "step": 20050
+    },
+    {
+      "epoch": 1.1516850707136432,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.02124286689227e-05,
+      "loss": 6.8245,
+      "step": 20060
+    },
+    {
+      "epoch": 1.1522592004286836,
+      "grad_norm": 1.6953125,
+      "learning_rate": 4.016684254905845e-05,
+      "loss": 6.795,
+      "step": 20070
+    },
+    {
+      "epoch": 1.1528333301437237,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.012126493014957e-05,
+      "loss": 6.7862,
+      "step": 20080
+    },
+    {
+      "epoch": 1.153407459858764,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.007569585159881e-05,
+      "loss": 6.8271,
+      "step": 20090
+    },
+    {
+      "epoch": 1.1539815895738044,
+      "grad_norm": 1.6171875,
+      "learning_rate": 4.0030135352801505e-05,
+      "loss": 6.7944,
+      "step": 20100
+    },
+    {
+      "epoch": 1.1545557192888447,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.99845834731456e-05,
+      "loss": 6.8388,
+      "step": 20110
+    },
+    {
+      "epoch": 1.155129849003885,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.993904025201157e-05,
+      "loss": 6.7978,
+      "step": 20120
+    },
+    {
+      "epoch": 1.1557039787189252,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.9893505728772423e-05,
+      "loss": 6.7911,
+      "step": 20130
+    },
+    {
+      "epoch": 1.1562781084339655,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.984797994279363e-05,
+      "loss": 6.8024,
+      "step": 20140
+    },
+    {
+      "epoch": 1.1568522381490058,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.9802462933433106e-05,
+      "loss": 6.7964,
+      "step": 20150
+    },
+    {
+      "epoch": 1.1574263678640462,
+      "grad_norm": 1.65625,
+      "learning_rate": 3.975695474004123e-05,
+      "loss": 6.7825,
+      "step": 20160
+    },
+    {
+      "epoch": 1.1580004975790863,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.9711455401960675e-05,
+      "loss": 6.7821,
+      "step": 20170
+    },
+    {
+      "epoch": 1.1585746272941266,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.9665964958526516e-05,
+      "loss": 6.8397,
+      "step": 20180
+    },
+    {
+      "epoch": 1.159148757009167,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.962048344906612e-05,
+      "loss": 6.7868,
+      "step": 20190
+    },
+    {
+      "epoch": 1.1597228867242073,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.957501091289916e-05,
+      "loss": 6.821,
+      "step": 20200
+    },
+    {
+      "epoch": 1.1602970164392474,
+      "grad_norm": 1.6484375,
+      "learning_rate": 3.95295473893375e-05,
+      "loss": 6.8652,
+      "step": 20210
+    },
+    {
+      "epoch": 1.1608711461542878,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.9484092917685214e-05,
+      "loss": 6.8374,
+      "step": 20220
+    },
+    {
+      "epoch": 1.161445275869328,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.943864753723863e-05,
+      "loss": 6.7714,
+      "step": 20230
+    },
+    {
+      "epoch": 1.1620194055843684,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.939321128728613e-05,
+      "loss": 6.8082,
+      "step": 20240
+    },
+    {
+      "epoch": 1.1625935352994086,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.934778420710824e-05,
+      "loss": 6.8283,
+      "step": 20250
+    },
+    {
+      "epoch": 1.163167665014449,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.9302366335977535e-05,
+      "loss": 6.8022,
+      "step": 20260
+    },
+    {
+      "epoch": 1.1637417947294892,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.925695771315867e-05,
+      "loss": 6.7978,
+      "step": 20270
+    },
+    {
+      "epoch": 1.1643159244445296,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.921155837790828e-05,
+      "loss": 6.8335,
+      "step": 20280
+    },
+    {
+      "epoch": 1.1648900541595697,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.916616836947495e-05,
+      "loss": 6.7864,
+      "step": 20290
+    },
+    {
+      "epoch": 1.16546418387461,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.9120787727099226e-05,
+      "loss": 6.7743,
+      "step": 20300
+    },
+    {
+      "epoch": 1.1660383135896504,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.9075416490013573e-05,
+      "loss": 6.7906,
+      "step": 20310
+    },
+    {
+      "epoch": 1.1666124433046907,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.90300546974423e-05,
+      "loss": 6.7976,
+      "step": 20320
+    },
+    {
+      "epoch": 1.1671865730197308,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.8984702388601544e-05,
+      "loss": 6.8192,
+      "step": 20330
+    },
+    {
+      "epoch": 1.1677607027347712,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.893935960269927e-05,
+      "loss": 6.8099,
+      "step": 20340
+    },
+    {
+      "epoch": 1.1683348324498115,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.889402637893518e-05,
+      "loss": 6.8522,
+      "step": 20350
+    },
+    {
+      "epoch": 1.1689089621648519,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.8848702756500736e-05,
+      "loss": 6.862,
+      "step": 20360
+    },
+    {
+      "epoch": 1.169483091879892,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.88033887745791e-05,
+      "loss": 6.8081,
+      "step": 20370
+    },
+    {
+      "epoch": 1.1700572215949323,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.8758084472345064e-05,
+      "loss": 6.8136,
+      "step": 20380
+    },
+    {
+      "epoch": 1.1706313513099726,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.871278988896508e-05,
+      "loss": 6.8325,
+      "step": 20390
+    },
+    {
+      "epoch": 1.171205481025013,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.8667505063597215e-05,
+      "loss": 6.8118,
+      "step": 20400
+    },
+    {
+      "epoch": 1.171779610740053,
+      "grad_norm": 1.6640625,
+      "learning_rate": 3.862223003539107e-05,
+      "loss": 6.7854,
+      "step": 20410
+    },
+    {
+      "epoch": 1.1723537404550934,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.857696484348777e-05,
+      "loss": 6.7856,
+      "step": 20420
+    },
+    {
+      "epoch": 1.1729278701701338,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.853170952701996e-05,
+      "loss": 6.8181,
+      "step": 20430
+    },
+    {
+      "epoch": 1.1735019998851741,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.848646412511175e-05,
+      "loss": 6.7811,
+      "step": 20440
+    },
+    {
+      "epoch": 1.1740761296002145,
+      "grad_norm": 1.625,
+      "learning_rate": 3.844122867687867e-05,
+      "loss": 6.8026,
+      "step": 20450
+    },
+    {
+      "epoch": 1.1746502593152546,
+      "grad_norm": 1.640625,
+      "learning_rate": 3.839600322142762e-05,
+      "loss": 6.8377,
+      "step": 20460
+    },
+    {
+      "epoch": 1.175224389030295,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.835078779785689e-05,
+      "loss": 6.7971,
+      "step": 20470
+    },
+    {
+      "epoch": 1.1757985187453353,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.830558244525611e-05,
+      "loss": 6.7962,
+      "step": 20480
+    },
+    {
+      "epoch": 1.1763726484603754,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.826038720270616e-05,
+      "loss": 6.8334,
+      "step": 20490
+    },
+    {
+      "epoch": 1.1769467781754157,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.821520210927922e-05,
+      "loss": 6.8429,
+      "step": 20500
+    },
+    {
+      "epoch": 1.177520907890456,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.817002720403868e-05,
+      "loss": 6.8043,
+      "step": 20510
+    },
+    {
+      "epoch": 1.1780950376054964,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.812486252603909e-05,
+      "loss": 6.8029,
+      "step": 20520
+    },
+    {
+      "epoch": 1.1786691673205367,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.807970811432625e-05,
+      "loss": 6.8094,
+      "step": 20530
+    },
+    {
+      "epoch": 1.1792432970355768,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.803456400793698e-05,
+      "loss": 6.7932,
+      "step": 20540
+    },
+    {
+      "epoch": 1.1798174267506172,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.798943024589924e-05,
+      "loss": 6.8151,
+      "step": 20550
+    },
+    {
+      "epoch": 1.1803915564656575,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.794430686723205e-05,
+      "loss": 6.8063,
+      "step": 20560
+    },
+    {
+      "epoch": 1.1809656861806976,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.789919391094546e-05,
+      "loss": 6.8127,
+      "step": 20570
+    },
+    {
+      "epoch": 1.181539815895738,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.7854091416040475e-05,
+      "loss": 6.8294,
+      "step": 20580
+    },
+    {
+      "epoch": 1.1821139456107783,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.780899942150908e-05,
+      "loss": 6.8127,
+      "step": 20590
+    },
+    {
+      "epoch": 1.1826880753258187,
+      "grad_norm": 1.6484375,
+      "learning_rate": 3.776391796633418e-05,
+      "loss": 6.8079,
+      "step": 20600
+    },
+    {
+      "epoch": 1.183262205040859,
+      "grad_norm": 1.640625,
+      "learning_rate": 3.7718847089489584e-05,
+      "loss": 6.8004,
+      "step": 20610
+    },
+    {
+      "epoch": 1.1838363347558991,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.7673786829939924e-05,
+      "loss": 6.7828,
+      "step": 20620
+    },
+    {
+      "epoch": 1.1844104644709395,
+      "grad_norm": 1.6640625,
+      "learning_rate": 3.762873722664067e-05,
+      "loss": 6.8034,
+      "step": 20630
+    },
+    {
+      "epoch": 1.1849845941859798,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.758369831853806e-05,
+      "loss": 6.8227,
+      "step": 20640
+    },
+    {
+      "epoch": 1.1855587239010201,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.753867014456914e-05,
+      "loss": 6.8206,
+      "step": 20650
+    },
+    {
+      "epoch": 1.1861328536160602,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.749365274366163e-05,
+      "loss": 6.7616,
+      "step": 20660
+    },
+    {
+      "epoch": 1.1867069833311006,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.744864615473391e-05,
+      "loss": 6.86,
+      "step": 20670
+    },
+    {
+      "epoch": 1.187281113046141,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.74036504166951e-05,
+      "loss": 6.7983,
+      "step": 20680
+    },
+    {
+      "epoch": 1.1878552427611813,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.7358665568444864e-05,
+      "loss": 6.8181,
+      "step": 20690
+    },
+    {
+      "epoch": 1.1884293724762214,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.731369164887347e-05,
+      "loss": 6.7961,
+      "step": 20700
+    },
+    {
+      "epoch": 1.1890035021912617,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.726872869686176e-05,
+      "loss": 6.8288,
+      "step": 20710
+    },
+    {
+      "epoch": 1.189577631906302,
+      "grad_norm": 1.609375,
+      "learning_rate": 3.722377675128108e-05,
+      "loss": 6.8376,
+      "step": 20720
+    },
+    {
+      "epoch": 1.1901517616213424,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.717883585099324e-05,
+      "loss": 6.7921,
+      "step": 20730
+    },
+    {
+      "epoch": 1.1907258913363825,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.713390603485053e-05,
+      "loss": 6.8358,
+      "step": 20740
+    },
+    {
+      "epoch": 1.1913000210514229,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.708898734169563e-05,
+      "loss": 6.8181,
+      "step": 20750
+    },
+    {
+      "epoch": 1.1918741507664632,
+      "grad_norm": 1.6484375,
+      "learning_rate": 3.704407981036167e-05,
+      "loss": 6.781,
+      "step": 20760
+    },
+    {
+      "epoch": 1.1924482804815035,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.699918347967204e-05,
+      "loss": 6.818,
+      "step": 20770
+    },
+    {
+      "epoch": 1.1930224101965436,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.6954298388440494e-05,
+      "loss": 6.781,
+      "step": 20780
+    },
+    {
+      "epoch": 1.193596539911584,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.690942457547106e-05,
+      "loss": 6.8087,
+      "step": 20790
+    },
+    {
+      "epoch": 1.1941706696266243,
+      "grad_norm": 1.6640625,
+      "learning_rate": 3.686456207955805e-05,
+      "loss": 6.8211,
+      "step": 20800
+    },
+    {
+      "epoch": 1.1947447993416647,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.681971093948594e-05,
+      "loss": 6.8242,
+      "step": 20810
+    },
+    {
+      "epoch": 1.1953189290567048,
+      "grad_norm": 1.625,
+      "learning_rate": 3.677487119402941e-05,
+      "loss": 6.7892,
+      "step": 20820
+    },
+    {
+      "epoch": 1.1958930587717451,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.673004288195328e-05,
+      "loss": 6.8204,
+      "step": 20830
+    },
+    {
+      "epoch": 1.1964671884867855,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.668522604201252e-05,
+      "loss": 6.822,
+      "step": 20840
+    },
+    {
+      "epoch": 1.1970413182018258,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.664042071295214e-05,
+      "loss": 6.8165,
+      "step": 20850
+    },
+    {
+      "epoch": 1.1976154479168661,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.659562693350723e-05,
+      "loss": 6.8037,
+      "step": 20860
+    },
+    {
+      "epoch": 1.1981895776319063,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.655084474240286e-05,
+      "loss": 6.8077,
+      "step": 20870
+    },
+    {
+      "epoch": 1.1987637073469466,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.650607417835412e-05,
+      "loss": 6.839,
+      "step": 20880
+    },
+    {
+      "epoch": 1.199337837061987,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.646131528006604e-05,
+      "loss": 6.8457,
+      "step": 20890
+    },
+    {
+      "epoch": 1.199911966777027,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.641656808623353e-05,
+      "loss": 6.8387,
+      "step": 20900
+    },
+    {
+      "epoch": 1.2004860964920674,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.637183263554143e-05,
+      "loss": 6.8158,
+      "step": 20910
+    },
+    {
+      "epoch": 1.2010602262071077,
+      "grad_norm": 1.640625,
+      "learning_rate": 3.632710896666437e-05,
+      "loss": 6.797,
+      "step": 20920
+    },
+    {
+      "epoch": 1.201634355922148,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.6282397118266876e-05,
+      "loss": 6.8049,
+      "step": 20930
+    },
+    {
+      "epoch": 1.2022084856371884,
+      "grad_norm": 1.625,
+      "learning_rate": 3.623769712900319e-05,
+      "loss": 6.7645,
+      "step": 20940
+    },
+    {
+      "epoch": 1.2027826153522285,
+      "grad_norm": 1.625,
+      "learning_rate": 3.6193009037517314e-05,
+      "loss": 6.7906,
+      "step": 20950
+    },
+    {
+      "epoch": 1.2033567450672689,
+      "grad_norm": 1.5,
+      "learning_rate": 3.614833288244295e-05,
+      "loss": 6.8026,
+      "step": 20960
+    },
+    {
+      "epoch": 1.2039308747823092,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.6103668702403546e-05,
+      "loss": 6.8169,
+      "step": 20970
+    },
+    {
+      "epoch": 1.2045050044973493,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.6059016536012124e-05,
+      "loss": 6.8261,
+      "step": 20980
+    },
+    {
+      "epoch": 1.2050791342123897,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.601437642187135e-05,
+      "loss": 6.8157,
+      "step": 20990
+    },
+    {
+      "epoch": 1.20565326392743,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.5969748398573474e-05,
+      "loss": 6.7973,
+      "step": 21000
+    },
+    {
+      "epoch": 1.2062273936424703,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.5925132504700286e-05,
+      "loss": 6.7928,
+      "step": 21010
+    },
+    {
+      "epoch": 1.2068015233575107,
+      "grad_norm": 1.5,
+      "learning_rate": 3.58805287788231e-05,
+      "loss": 6.8132,
+      "step": 21020
+    },
+    {
+      "epoch": 1.2073756530725508,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.583593725950268e-05,
+      "loss": 6.8118,
+      "step": 21030
+    },
+    {
+      "epoch": 1.2079497827875911,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.5791357985289277e-05,
+      "loss": 6.821,
+      "step": 21040
+    },
+    {
+      "epoch": 1.2085239125026315,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.5746790994722534e-05,
+      "loss": 6.8024,
+      "step": 21050
+    },
+    {
+      "epoch": 1.2090980422176718,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.570223632633148e-05,
+      "loss": 6.8069,
+      "step": 21060
+    },
+    {
+      "epoch": 1.209672171932712,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.56576940186345e-05,
+      "loss": 6.7979,
+      "step": 21070
+    },
+    {
+      "epoch": 1.2102463016477523,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.5613164110139275e-05,
+      "loss": 6.7888,
+      "step": 21080
+    },
+    {
+      "epoch": 1.2108204313627926,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.556864663934275e-05,
+      "loss": 6.8396,
+      "step": 21090
+    },
+    {
+      "epoch": 1.211394561077833,
+      "grad_norm": 1.609375,
+      "learning_rate": 3.552414164473118e-05,
+      "loss": 6.8034,
+      "step": 21100
+    },
+    {
+      "epoch": 1.211968690792873,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.547964916477998e-05,
+      "loss": 6.8231,
+      "step": 21110
+    },
+    {
+      "epoch": 1.2125428205079134,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.543516923795377e-05,
+      "loss": 6.8226,
+      "step": 21120
+    },
+    {
+      "epoch": 1.2131169502229537,
+      "grad_norm": 1.6875,
+      "learning_rate": 3.539070190270629e-05,
+      "loss": 6.8056,
+      "step": 21130
+    },
+    {
+      "epoch": 1.213691079937994,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.534624719748043e-05,
+      "loss": 6.7595,
+      "step": 21140
+    },
+    {
+      "epoch": 1.2142652096530342,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.530180516070815e-05,
+      "loss": 6.7858,
+      "step": 21150
+    },
+    {
+      "epoch": 1.2148393393680745,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.525737583081044e-05,
+      "loss": 6.8228,
+      "step": 21160
+    },
+    {
+      "epoch": 1.2154134690831149,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.521295924619731e-05,
+      "loss": 6.7693,
+      "step": 21170
+    },
+    {
+      "epoch": 1.2159875987981552,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.516855544526779e-05,
+      "loss": 6.7931,
+      "step": 21180
+    },
+    {
+      "epoch": 1.2165617285131953,
+      "grad_norm": 1.7109375,
+      "learning_rate": 3.51241644664098e-05,
+      "loss": 6.7883,
+      "step": 21190
+    },
+    {
+      "epoch": 1.2171358582282357,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.50797863480002e-05,
+      "loss": 6.8149,
+      "step": 21200
+    },
+    {
+      "epoch": 1.217709987943276,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.503542112840476e-05,
+      "loss": 6.8273,
+      "step": 21210
+    },
+    {
+      "epoch": 1.2182841176583163,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.4991068845978056e-05,
+      "loss": 6.7913,
+      "step": 21220
+    },
+    {
+      "epoch": 1.2188582473733565,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.494672953906349e-05,
+      "loss": 6.7956,
+      "step": 21230
+    },
+    {
+      "epoch": 1.2194323770883968,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.490240324599328e-05,
+      "loss": 6.8015,
+      "step": 21240
+    },
+    {
+      "epoch": 1.2200065068034371,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.485809000508834e-05,
+      "loss": 6.8228,
+      "step": 21250
+    },
+    {
+      "epoch": 1.2205806365184775,
+      "grad_norm": 1.609375,
+      "learning_rate": 3.481378985465833e-05,
+      "loss": 6.8111,
+      "step": 21260
+    },
+    {
+      "epoch": 1.2211547662335176,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.476950283300159e-05,
+      "loss": 6.779,
+      "step": 21270
+    },
+    {
+      "epoch": 1.221728895948558,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.472522897840512e-05,
+      "loss": 6.808,
+      "step": 21280
+    },
+    {
+      "epoch": 1.2223030256635983,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.468096832914452e-05,
+      "loss": 6.801,
+      "step": 21290
+    },
+    {
+      "epoch": 1.2228771553786386,
+      "grad_norm": 1.5,
+      "learning_rate": 3.463672092348399e-05,
+      "loss": 6.8156,
+      "step": 21300
+    },
+    {
+      "epoch": 1.2234512850936787,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.4592486799676256e-05,
+      "loss": 6.8023,
+      "step": 21310
+    },
+    {
+      "epoch": 1.224025414808719,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.454826599596256e-05,
+      "loss": 6.7724,
+      "step": 21320
+    },
+    {
+      "epoch": 1.2245995445237594,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.450405855057268e-05,
+      "loss": 6.8353,
+      "step": 21330
+    },
+    {
+      "epoch": 1.2251736742387997,
+      "grad_norm": 1.640625,
+      "learning_rate": 3.44598645017248e-05,
+      "loss": 6.8009,
+      "step": 21340
+    },
+    {
+      "epoch": 1.22574780395384,
+      "grad_norm": 1.6484375,
+      "learning_rate": 3.441568388762553e-05,
+      "loss": 6.8197,
+      "step": 21350
+    },
+    {
+      "epoch": 1.2263219336688802,
+      "grad_norm": 1.640625,
+      "learning_rate": 3.4371516746469847e-05,
+      "loss": 6.7526,
+      "step": 21360
+    },
+    {
+      "epoch": 1.2268960633839205,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.4327363116441136e-05,
+      "loss": 6.8039,
+      "step": 21370
+    },
+    {
+      "epoch": 1.2274701930989609,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.4283223035711045e-05,
+      "loss": 6.842,
+      "step": 21380
+    },
+    {
+      "epoch": 1.228044322814001,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.423909654243954e-05,
+      "loss": 6.8283,
+      "step": 21390
+    },
+    {
+      "epoch": 1.2286184525290413,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.4194983674774805e-05,
+      "loss": 6.8216,
+      "step": 21400
+    },
+    {
+      "epoch": 1.2291925822440817,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.415088447085332e-05,
+      "loss": 6.8379,
+      "step": 21410
+    },
+    {
+      "epoch": 1.229766711959122,
+      "grad_norm": 1.609375,
+      "learning_rate": 3.410679896879966e-05,
+      "loss": 6.7836,
+      "step": 21420
+    },
+    {
+      "epoch": 1.2303408416741624,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.4062727206726606e-05,
+      "loss": 6.817,
+      "step": 21430
+    },
+    {
+      "epoch": 1.2309149713892025,
+      "grad_norm": 1.625,
+      "learning_rate": 3.4018669222735054e-05,
+      "loss": 6.7849,
+      "step": 21440
+    },
+    {
+      "epoch": 1.2314891011042428,
+      "grad_norm": 1.5234375,
+      "learning_rate": 3.3974625054914e-05,
+      "loss": 6.8206,
+      "step": 21450
+    },
+    {
+      "epoch": 1.2320632308192832,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.393059474134047e-05,
+      "loss": 6.7817,
+      "step": 21460
+    },
+    {
+      "epoch": 1.2326373605343233,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.388657832007951e-05,
+      "loss": 6.8132,
+      "step": 21470
+    },
+    {
+      "epoch": 1.2332114902493636,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.384257582918418e-05,
+      "loss": 6.8044,
+      "step": 21480
+    },
+    {
+      "epoch": 1.233785619964404,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.379858730669551e-05,
+      "loss": 6.8424,
+      "step": 21490
+    },
+    {
+      "epoch": 1.2343597496794443,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.375461279064239e-05,
+      "loss": 6.8208,
+      "step": 21500
+    },
+    {
+      "epoch": 1.2349338793944846,
+      "grad_norm": 1.5,
+      "learning_rate": 3.371065231904168e-05,
+      "loss": 6.8039,
+      "step": 21510
+    },
+    {
+      "epoch": 1.2355080091095247,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.366670592989803e-05,
+      "loss": 6.7942,
+      "step": 21520
+    },
+    {
+      "epoch": 1.236082138824565,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.362277366120397e-05,
+      "loss": 6.7946,
+      "step": 21530
+    },
+    {
+      "epoch": 1.2366562685396054,
+      "grad_norm": 1.625,
+      "learning_rate": 3.357885555093978e-05,
+      "loss": 6.8253,
+      "step": 21540
+    },
+    {
+      "epoch": 1.2372303982546458,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.353495163707353e-05,
+      "loss": 6.8448,
+      "step": 21550
+    },
+    {
+      "epoch": 1.2378045279696859,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.349106195756101e-05,
+      "loss": 6.8034,
+      "step": 21560
+    },
+    {
+      "epoch": 1.2383786576847262,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.344718655034568e-05,
+      "loss": 6.7819,
+      "step": 21570
+    },
+    {
+      "epoch": 1.2389527873997666,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.340332545335869e-05,
+      "loss": 6.8123,
+      "step": 21580
+    },
+    {
+      "epoch": 1.239526917114807,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.335947870451882e-05,
+      "loss": 6.8522,
+      "step": 21590
+    },
+    {
+      "epoch": 1.240101046829847,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.331564634173243e-05,
+      "loss": 6.8463,
+      "step": 21600
+    },
+    {
+      "epoch": 1.2406751765448873,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.327182840289343e-05,
+      "loss": 6.8039,
+      "step": 21610
+    },
+    {
+      "epoch": 1.2412493062599277,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.32280249258833e-05,
+      "loss": 6.8103,
+      "step": 21620
+    },
+    {
+      "epoch": 1.241823435974968,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.3184235948570983e-05,
+      "loss": 6.8569,
+      "step": 21630
+    },
+    {
+      "epoch": 1.2423975656900081,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.3140461508812914e-05,
+      "loss": 6.7996,
+      "step": 21640
+    },
+    {
+      "epoch": 1.2429716954050485,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.309670164445292e-05,
+      "loss": 6.8066,
+      "step": 21650
+    },
+    {
+      "epoch": 1.2435458251200888,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.3052956393322287e-05,
+      "loss": 6.7836,
+      "step": 21660
+    },
+    {
+      "epoch": 1.2441199548351292,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.30092257932396e-05,
+      "loss": 6.8482,
+      "step": 21670
+    },
+    {
+      "epoch": 1.2446940845501693,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.296550988201083e-05,
+      "loss": 6.7908,
+      "step": 21680
+    },
+    {
+      "epoch": 1.2452682142652096,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.292180869742924e-05,
+      "loss": 6.774,
+      "step": 21690
+    },
+    {
+      "epoch": 1.24584234398025,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.2878122277275313e-05,
+      "loss": 6.803,
+      "step": 21700
+    },
+    {
+      "epoch": 1.2464164736952903,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.283445065931685e-05,
+      "loss": 6.7836,
+      "step": 21710
+    },
+    {
+      "epoch": 1.2469906034103304,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.279079388130877e-05,
+      "loss": 6.8021,
+      "step": 21720
+    },
+    {
+      "epoch": 1.2475647331253708,
+      "grad_norm": 1.6484375,
+      "learning_rate": 3.274715198099324e-05,
+      "loss": 6.7848,
+      "step": 21730
+    },
+    {
+      "epoch": 1.248138862840411,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.270352499609952e-05,
+      "loss": 6.8375,
+      "step": 21740
+    },
+    {
+      "epoch": 1.2487129925554514,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.265991296434399e-05,
+      "loss": 6.7839,
+      "step": 21750
+    },
+    {
+      "epoch": 1.2492871222704918,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.261631592343008e-05,
+      "loss": 6.8102,
+      "step": 21760
+    },
+    {
+      "epoch": 1.2498612519855319,
+      "grad_norm": 1.625,
+      "learning_rate": 3.257273391104829e-05,
+      "loss": 6.8041,
+      "step": 21770
+    },
+    {
+      "epoch": 1.2504353817005722,
+      "grad_norm": 1.6953125,
+      "learning_rate": 3.2529166964876115e-05,
+      "loss": 6.8141,
+      "step": 21780
+    },
+    {
+      "epoch": 1.2510095114156126,
+      "grad_norm": 1.5,
+      "learning_rate": 3.248561512257802e-05,
+      "loss": 6.8004,
+      "step": 21790
+    },
+    {
+      "epoch": 1.2515836411306527,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.244207842180542e-05,
+      "loss": 6.7802,
+      "step": 21800
+    },
+    {
+      "epoch": 1.252157770845693,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.2398556900196636e-05,
+      "loss": 6.8308,
+      "step": 21810
+    },
+    {
+      "epoch": 1.2527319005607334,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.235505059537688e-05,
+      "loss": 6.7989,
+      "step": 21820
+    },
+    {
+      "epoch": 1.2533060302757737,
+      "grad_norm": 1.5,
+      "learning_rate": 3.2311559544958174e-05,
+      "loss": 6.7923,
+      "step": 21830
+    },
+    {
+      "epoch": 1.253880159990814,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.226808378653938e-05,
+      "loss": 6.8128,
+      "step": 21840
+    },
+    {
+      "epoch": 1.2544542897058542,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.222462335770615e-05,
+      "loss": 6.8132,
+      "step": 21850
+    },
+    {
+      "epoch": 1.2550284194208945,
+      "grad_norm": 1.5859375,
+      "learning_rate": 3.218117829603087e-05,
+      "loss": 6.8116,
+      "step": 21860
+    },
+    {
+      "epoch": 1.2556025491359348,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.213774863907262e-05,
+      "loss": 6.7959,
+      "step": 21870
+    },
+    {
+      "epoch": 1.256176678850975,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.2094334424377176e-05,
+      "loss": 6.8238,
+      "step": 21880
+    },
+    {
+      "epoch": 1.2567508085660153,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.205093568947699e-05,
+      "loss": 6.7817,
+      "step": 21890
+    },
+    {
+      "epoch": 1.2573249382810556,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.200755247189111e-05,
+      "loss": 6.8276,
+      "step": 21900
+    },
+    {
+      "epoch": 1.257899067996096,
+      "grad_norm": 1.5078125,
+      "learning_rate": 3.196418480912515e-05,
+      "loss": 6.7674,
+      "step": 21910
+    },
+    {
+      "epoch": 1.2584731977111363,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.192083273867131e-05,
+      "loss": 6.8002,
+      "step": 21920
+    },
+    {
+      "epoch": 1.2590473274261764,
+      "grad_norm": 1.671875,
+      "learning_rate": 3.187749629800829e-05,
+      "loss": 6.7855,
+      "step": 21930
+    },
+    {
+      "epoch": 1.2596214571412168,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.183417552460129e-05,
+      "loss": 6.8001,
+      "step": 21940
+    },
+    {
+      "epoch": 1.260195586856257,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.179087045590196e-05,
+      "loss": 6.7425,
+      "step": 21950
+    },
+    {
+      "epoch": 1.2607697165712972,
+      "grad_norm": 1.546875,
+      "learning_rate": 3.174758112934836e-05,
+      "loss": 6.7388,
+      "step": 21960
+    },
+    {
+      "epoch": 1.2613438462863376,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.170430758236495e-05,
+      "loss": 6.816,
+      "step": 21970
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 34834,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 999,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.04593957210827e+19,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}