diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,2578 +1,8551 @@
 {
-  "best_metric": 0.9901,
-  "best_model_checkpoint": "VIT_AI_image_detector/checkpoint-4218",
-  "epoch": 2.9994666666666667,
-  "global_step": 4218,
+  "best_metric": 0.9931,
+  "best_model_checkpoint": "VIT_AI_image_detector/checkpoint-12656",
+  "epoch": 9.998222222222223,
+  "global_step": 14060,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.01,
-      "learning_rate": 1.1848341232227488e-06,
-      "loss": 0.6878,
+      "learning_rate": 3.556187766714083e-07,
+      "loss": 0.7115,
       "step": 10
     },
     {
       "epoch": 0.01,
-      "learning_rate": 2.3696682464454976e-06,
-      "loss": 0.6907,
+      "learning_rate": 7.112375533428166e-07,
+      "loss": 0.7071,
       "step": 20
     },
     {
       "epoch": 0.02,
-      "learning_rate": 3.5545023696682464e-06,
-      "loss": 0.6836,
+      "learning_rate": 1.0668563300142247e-06,
+      "loss": 0.7087,
       "step": 30
     },
     {
       "epoch": 0.03,
-      "learning_rate": 4.739336492890995e-06,
-      "loss": 0.6766,
+      "learning_rate": 1.4224751066856332e-06,
+      "loss": 0.7038,
       "step": 40
     },
     {
       "epoch": 0.04,
-      "learning_rate": 5.924170616113745e-06,
-      "loss": 0.6704,
+      "learning_rate": 1.7780938833570414e-06,
+      "loss": 0.7043,
       "step": 50
     },
     {
       "epoch": 0.04,
-      "learning_rate": 7.109004739336493e-06,
-      "loss": 0.6556,
+      "learning_rate": 2.1337126600284495e-06,
+      "loss": 0.7001,
       "step": 60
     },
     {
       "epoch": 0.05,
-      "learning_rate": 8.293838862559241e-06,
-      "loss": 0.6506,
+      "learning_rate": 2.4893314366998577e-06,
+      "loss": 0.6982,
       "step": 70
     },
     {
       "epoch": 0.06,
-      "learning_rate": 9.47867298578199e-06,
-      "loss": 0.6276,
+      "learning_rate": 2.8449502133712664e-06,
+      "loss": 0.6987,
       "step": 80
     },
     {
       "epoch": 0.06,
-      "learning_rate": 1.066350710900474e-05,
-      "loss": 0.5957,
+      "learning_rate": 3.2005689900426746e-06,
+      "loss": 0.6925,
       "step": 90
     },
     {
       "epoch": 0.07,
-      "learning_rate": 1.184834123222749e-05,
-      "loss": 0.5524,
+      "learning_rate": 3.556187766714083e-06,
+      "loss": 0.6905,
       "step": 100
     },
     {
       "epoch": 0.08,
-      "learning_rate": 1.3033175355450238e-05,
-      "loss": 0.4691,
+      "learning_rate": 3.9118065433854916e-06,
+      "loss": 0.6865,
       "step": 110
     },
     {
       "epoch": 0.09,
-      "learning_rate": 1.4218009478672985e-05,
-      "loss": 0.4064,
+      "learning_rate": 4.267425320056899e-06,
+      "loss": 0.6767,
       "step": 120
     },
     {
       "epoch": 0.09,
-      "learning_rate": 1.5402843601895736e-05,
-      "loss": 0.3614,
+      "learning_rate": 4.623044096728307e-06,
+      "loss": 0.6693,
       "step": 130
     },
     {
       "epoch": 0.1,
-      "learning_rate": 1.6587677725118483e-05,
-      "loss": 0.3137,
+      "learning_rate": 4.9786628733997155e-06,
+      "loss": 0.67,
       "step": 140
     },
     {
       "epoch": 0.11,
-      "learning_rate": 1.7772511848341233e-05,
-      "loss": 0.3232,
+      "learning_rate": 5.334281650071124e-06,
+      "loss": 0.659,
       "step": 150
     },
     {
       "epoch": 0.11,
-      "learning_rate": 1.895734597156398e-05,
-      "loss": 0.3303,
+      "learning_rate": 5.689900426742533e-06,
+      "loss": 0.6514,
       "step": 160
     },
     {
       "epoch": 0.12,
-      "learning_rate": 2.014218009478673e-05,
-      "loss": 0.2869,
+      "learning_rate": 6.045519203413941e-06,
+      "loss": 0.638,
       "step": 170
     },
     {
       "epoch": 0.13,
-      "learning_rate": 2.132701421800948e-05,
-      "loss": 0.2903,
+      "learning_rate": 6.401137980085349e-06,
+      "loss": 0.6258,
       "step": 180
     },
     {
       "epoch": 0.14,
-      "learning_rate": 2.251184834123223e-05,
-      "loss": 0.2726,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 0.615,
       "step": 190
     },
     {
       "epoch": 0.14,
-      "learning_rate": 2.369668246445498e-05,
-      "loss": 0.2885,
+      "learning_rate": 7.112375533428166e-06,
+      "loss": 0.595,
       "step": 200
     },
     {
       "epoch": 0.15,
-      "learning_rate": 2.4881516587677726e-05,
-      "loss": 0.2507,
+      "learning_rate": 7.467994310099574e-06,
+      "loss": 0.572,
       "step": 210
     },
     {
       "epoch": 0.16,
-      "learning_rate": 2.6066350710900477e-05,
-      "loss": 0.2451,
+      "learning_rate": 7.823613086770983e-06,
+      "loss": 0.5214,
       "step": 220
     },
     {
       "epoch": 0.16,
-      "learning_rate": 2.7251184834123224e-05,
-      "loss": 0.239,
+      "learning_rate": 8.179231863442389e-06,
+      "loss": 0.4782,
       "step": 230
     },
     {
       "epoch": 0.17,
-      "learning_rate": 2.843601895734597e-05,
-      "loss": 0.2621,
+      "learning_rate": 8.534850640113798e-06,
+      "loss": 0.4596,
       "step": 240
     },
     {
       "epoch": 0.18,
-      "learning_rate": 2.962085308056872e-05,
-      "loss": 0.2509,
+      "learning_rate": 8.890469416785207e-06,
+      "loss": 0.3853,
       "step": 250
     },
     {
       "epoch": 0.18,
-      "learning_rate": 3.080568720379147e-05,
-      "loss": 0.2168,
+      "learning_rate": 9.246088193456614e-06,
+      "loss": 0.348,
       "step": 260
     },
     {
       "epoch": 0.19,
-      "learning_rate": 3.1990521327014215e-05,
-      "loss": 0.2547,
+      "learning_rate": 9.601706970128024e-06,
+      "loss": 0.3326,
       "step": 270
     },
     {
       "epoch": 0.2,
-      "learning_rate": 3.3175355450236966e-05,
-      "loss": 0.1996,
+      "learning_rate": 9.957325746799431e-06,
+      "loss": 0.3217,
       "step": 280
     },
     {
       "epoch": 0.21,
-      "learning_rate": 3.4360189573459716e-05,
-      "loss": 0.2337,
+      "learning_rate": 1.031294452347084e-05,
+      "loss": 0.2727,
       "step": 290
     },
     {
       "epoch": 0.21,
-      "learning_rate": 3.554502369668247e-05,
-      "loss": 0.2903,
+      "learning_rate": 1.0668563300142247e-05,
+      "loss": 0.2778,
       "step": 300
     },
     {
       "epoch": 0.22,
-      "learning_rate": 3.672985781990522e-05,
-      "loss": 0.2271,
+      "learning_rate": 1.1024182076813657e-05,
+      "loss": 0.2972,
       "step": 310
     },
     {
       "epoch": 0.23,
-      "learning_rate": 3.791469194312796e-05,
-      "loss": 0.2384,
+      "learning_rate": 1.1379800853485066e-05,
+      "loss": 0.2962,
       "step": 320
     },
     {
       "epoch": 0.23,
-      "learning_rate": 3.909952606635071e-05,
-      "loss": 0.2323,
+      "learning_rate": 1.1735419630156473e-05,
+      "loss": 0.3132,
       "step": 330
     },
     {
       "epoch": 0.24,
-      "learning_rate": 4.028436018957346e-05,
-      "loss": 0.208,
+      "learning_rate": 1.2091038406827882e-05,
+      "loss": 0.2449,
       "step": 340
     },
     {
       "epoch": 0.25,
-      "learning_rate": 4.146919431279621e-05,
-      "loss": 0.2671,
+      "learning_rate": 1.244665718349929e-05,
+      "loss": 0.2392,
       "step": 350
     },
     {
       "epoch": 0.26,
-      "learning_rate": 4.265402843601896e-05,
-      "loss": 0.2282,
+      "learning_rate": 1.2802275960170699e-05,
+      "loss": 0.2364,
       "step": 360
     },
     {
       "epoch": 0.26,
-      "learning_rate": 4.383886255924171e-05,
-      "loss": 0.1993,
+      "learning_rate": 1.3157894736842106e-05,
+      "loss": 0.2754,
       "step": 370
     },
     {
       "epoch": 0.27,
-      "learning_rate": 4.502369668246446e-05,
-      "loss": 0.225,
+      "learning_rate": 1.3513513513513515e-05,
+      "loss": 0.2719,
       "step": 380
     },
     {
       "epoch": 0.28,
-      "learning_rate": 4.620853080568721e-05,
-      "loss": 0.2417,
+      "learning_rate": 1.3869132290184922e-05,
+      "loss": 0.2573,
       "step": 390
     },
     {
       "epoch": 0.28,
-      "learning_rate": 4.739336492890996e-05,
-      "loss": 0.1922,
+      "learning_rate": 1.4224751066856332e-05,
+      "loss": 0.221,
       "step": 400
     },
     {
       "epoch": 0.29,
-      "learning_rate": 4.857819905213271e-05,
-      "loss": 0.208,
+      "learning_rate": 1.4580369843527739e-05,
+      "loss": 0.2255,
       "step": 410
     },
     {
       "epoch": 0.3,
-      "learning_rate": 4.976303317535545e-05,
-      "loss": 0.2107,
+      "learning_rate": 1.4935988620199148e-05,
+      "loss": 0.2563,
       "step": 420
     },
     {
       "epoch": 0.31,
-      "learning_rate": 4.989462592202318e-05,
-      "loss": 0.22,
+      "learning_rate": 1.5291607396870555e-05,
+      "loss": 0.2138,
       "step": 430
     },
     {
       "epoch": 0.31,
-      "learning_rate": 4.976290832455216e-05,
-      "loss": 0.2069,
+      "learning_rate": 1.5647226173541966e-05,
+      "loss": 0.2396,
       "step": 440
     },
     {
       "epoch": 0.32,
-      "learning_rate": 4.963119072708114e-05,
-      "loss": 0.1807,
+      "learning_rate": 1.6002844950213374e-05,
+      "loss": 0.2277,
       "step": 450
     },
     {
       "epoch": 0.33,
-      "learning_rate": 4.949947312961012e-05,
-      "loss": 0.1716,
+      "learning_rate": 1.6358463726884778e-05,
+      "loss": 0.2767,
       "step": 460
     },
     {
       "epoch": 0.33,
-      "learning_rate": 4.93677555321391e-05,
-      "loss": 0.2296,
+      "learning_rate": 1.671408250355619e-05,
+      "loss": 0.2262,
       "step": 470
     },
     {
       "epoch": 0.34,
-      "learning_rate": 4.923603793466807e-05,
-      "loss": 0.2679,
+      "learning_rate": 1.7069701280227596e-05,
+      "loss": 0.2236,
       "step": 480
     },
     {
       "epoch": 0.35,
-      "learning_rate": 4.910432033719705e-05,
-      "loss": 0.2065,
+      "learning_rate": 1.7425320056899007e-05,
+      "loss": 0.1985,
       "step": 490
     },
     {
       "epoch": 0.36,
-      "learning_rate": 4.8972602739726034e-05,
-      "loss": 0.1917,
+      "learning_rate": 1.7780938833570414e-05,
+      "loss": 0.2236,
       "step": 500
     },
     {
       "epoch": 0.36,
-      "learning_rate": 4.8840885142255003e-05,
-      "loss": 0.2167,
+      "learning_rate": 1.813655761024182e-05,
+      "loss": 0.2137,
       "step": 510
     },
     {
       "epoch": 0.37,
-      "learning_rate": 4.8709167544783986e-05,
-      "loss": 0.1904,
+      "learning_rate": 1.849217638691323e-05,
+      "loss": 0.2136,
       "step": 520
     },
     {
       "epoch": 0.38,
-      "learning_rate": 4.857744994731296e-05,
-      "loss": 0.1902,
+      "learning_rate": 1.884779516358464e-05,
+      "loss": 0.2386,
       "step": 530
     },
     {
       "epoch": 0.38,
-      "learning_rate": 4.8445732349841945e-05,
-      "loss": 0.2011,
+      "learning_rate": 1.9203413940256047e-05,
+      "loss": 0.2386,
       "step": 540
     },
     {
       "epoch": 0.39,
-      "learning_rate": 4.831401475237092e-05,
-      "loss": 0.2673,
+      "learning_rate": 1.9559032716927454e-05,
+      "loss": 0.1955,
       "step": 550
     },
     {
       "epoch": 0.4,
-      "learning_rate": 4.8182297154899896e-05,
-      "loss": 0.232,
+      "learning_rate": 1.9914651493598862e-05,
+      "loss": 0.2746,
       "step": 560
     },
     {
       "epoch": 0.41,
-      "learning_rate": 4.805057955742888e-05,
-      "loss": 0.2093,
+      "learning_rate": 2.0270270270270273e-05,
+      "loss": 0.254,
       "step": 570
     },
     {
       "epoch": 0.41,
-      "learning_rate": 4.791886195995785e-05,
-      "loss": 0.182,
+      "learning_rate": 2.062588904694168e-05,
+      "loss": 0.187,
       "step": 580
     },
     {
       "epoch": 0.42,
-      "learning_rate": 4.778714436248683e-05,
-      "loss": 0.194,
+      "learning_rate": 2.0981507823613087e-05,
+      "loss": 0.2168,
       "step": 590
     },
     {
       "epoch": 0.43,
-      "learning_rate": 4.7655426765015806e-05,
-      "loss": 0.2216,
+      "learning_rate": 2.1337126600284495e-05,
+      "loss": 0.1693,
       "step": 600
     },
     {
       "epoch": 0.43,
-      "learning_rate": 4.752370916754479e-05,
-      "loss": 0.1712,
+      "learning_rate": 2.1692745376955902e-05,
+      "loss": 0.1984,
       "step": 610
     },
     {
       "epoch": 0.44,
-      "learning_rate": 4.7391991570073765e-05,
-      "loss": 0.207,
+      "learning_rate": 2.2048364153627313e-05,
+      "loss": 0.2064,
       "step": 620
     },
     {
       "epoch": 0.45,
-      "learning_rate": 4.726027397260274e-05,
-      "loss": 0.1605,
+      "learning_rate": 2.240398293029872e-05,
+      "loss": 0.2282,
       "step": 630
     },
     {
       "epoch": 0.46,
-      "learning_rate": 4.712855637513172e-05,
-      "loss": 0.1532,
+      "learning_rate": 2.275960170697013e-05,
+      "loss": 0.2301,
       "step": 640
     },
     {
       "epoch": 0.46,
-      "learning_rate": 4.699683877766069e-05,
-      "loss": 0.2005,
+      "learning_rate": 2.3115220483641535e-05,
+      "loss": 0.1791,
       "step": 650
     },
     {
       "epoch": 0.47,
-      "learning_rate": 4.6865121180189675e-05,
-      "loss": 0.1555,
+      "learning_rate": 2.3470839260312946e-05,
+      "loss": 0.2016,
       "step": 660
     },
     {
       "epoch": 0.48,
-      "learning_rate": 4.673340358271866e-05,
-      "loss": 0.1744,
+      "learning_rate": 2.3826458036984353e-05,
+      "loss": 0.1897,
       "step": 670
     },
     {
       "epoch": 0.48,
-      "learning_rate": 4.6601685985247633e-05,
-      "loss": 0.1753,
+      "learning_rate": 2.4182076813655764e-05,
+      "loss": 0.2337,
       "step": 680
     },
     {
       "epoch": 0.49,
-      "learning_rate": 4.646996838777661e-05,
-      "loss": 0.1799,
+      "learning_rate": 2.453769559032717e-05,
+      "loss": 0.2104,
       "step": 690
     },
     {
       "epoch": 0.5,
-      "learning_rate": 4.6338250790305585e-05,
-      "loss": 0.1929,
+      "learning_rate": 2.489331436699858e-05,
+      "loss": 0.1817,
       "step": 700
     },
     {
       "epoch": 0.5,
-      "learning_rate": 4.620653319283457e-05,
-      "loss": 0.246,
+      "learning_rate": 2.524893314366999e-05,
+      "loss": 0.1538,
       "step": 710
     },
     {
       "epoch": 0.51,
-      "learning_rate": 4.6074815595363544e-05,
-      "loss": 0.1788,
+      "learning_rate": 2.5604551920341397e-05,
+      "loss": 0.2035,
       "step": 720
     },
     {
       "epoch": 0.52,
-      "learning_rate": 4.594309799789252e-05,
-      "loss": 0.1918,
+      "learning_rate": 2.5960170697012805e-05,
+      "loss": 0.1988,
       "step": 730
     },
     {
       "epoch": 0.53,
-      "learning_rate": 4.58113804004215e-05,
-      "loss": 0.1817,
+      "learning_rate": 2.6315789473684212e-05,
+      "loss": 0.2104,
       "step": 740
     },
     {
       "epoch": 0.53,
-      "learning_rate": 4.567966280295047e-05,
-      "loss": 0.1969,
+      "learning_rate": 2.6671408250355616e-05,
+      "loss": 0.2216,
       "step": 750
     },
     {
       "epoch": 0.54,
-      "learning_rate": 4.5547945205479454e-05,
-      "loss": 0.1572,
+      "learning_rate": 2.702702702702703e-05,
+      "loss": 0.2038,
       "step": 760
     },
     {
       "epoch": 0.55,
-      "learning_rate": 4.541622760800843e-05,
-      "loss": 0.1323,
+      "learning_rate": 2.7382645803698438e-05,
+      "loss": 0.1784,
       "step": 770
     },
     {
       "epoch": 0.55,
-      "learning_rate": 4.528451001053741e-05,
-      "loss": 0.1533,
+      "learning_rate": 2.7738264580369845e-05,
+      "loss": 0.1656,
       "step": 780
     },
     {
       "epoch": 0.56,
-      "learning_rate": 4.515279241306639e-05,
-      "loss": 0.1593,
+      "learning_rate": 2.8093883357041252e-05,
+      "loss": 0.2006,
       "step": 790
     },
     {
       "epoch": 0.57,
-      "learning_rate": 4.5021074815595364e-05,
-      "loss": 0.2544,
+      "learning_rate": 2.8449502133712663e-05,
+      "loss": 0.2099,
       "step": 800
     },
     {
       "epoch": 0.58,
-      "learning_rate": 4.488935721812435e-05,
-      "loss": 0.2024,
+      "learning_rate": 2.880512091038407e-05,
+      "loss": 0.1788,
       "step": 810
     },
     {
       "epoch": 0.58,
-      "learning_rate": 4.4757639620653316e-05,
-      "loss": 0.1798,
+      "learning_rate": 2.9160739687055478e-05,
+      "loss": 0.1604,
       "step": 820
     },
     {
       "epoch": 0.59,
-      "learning_rate": 4.46259220231823e-05,
-      "loss": 0.1406,
+      "learning_rate": 2.9516358463726885e-05,
+      "loss": 0.2234,
       "step": 830
     },
     {
       "epoch": 0.6,
-      "learning_rate": 4.449420442571128e-05,
-      "loss": 0.1335,
+      "learning_rate": 2.9871977240398296e-05,
+      "loss": 0.1596,
       "step": 840
     },
     {
       "epoch": 0.6,
-      "learning_rate": 4.436248682824026e-05,
-      "loss": 0.1829,
+      "learning_rate": 3.0227596017069704e-05,
+      "loss": 0.1874,
       "step": 850
     },
     {
       "epoch": 0.61,
-      "learning_rate": 4.423076923076923e-05,
-      "loss": 0.1353,
+      "learning_rate": 3.058321479374111e-05,
+      "loss": 0.183,
       "step": 860
     },
     {
       "epoch": 0.62,
-      "learning_rate": 4.409905163329821e-05,
-      "loss": 0.1598,
+      "learning_rate": 3.093883357041252e-05,
+      "loss": 0.183,
       "step": 870
     },
     {
       "epoch": 0.63,
-      "learning_rate": 4.396733403582719e-05,
-      "loss": 0.1518,
+      "learning_rate": 3.129445234708393e-05,
+      "loss": 0.1798,
       "step": 880
     },
     {
       "epoch": 0.63,
-      "learning_rate": 4.383561643835617e-05,
-      "loss": 0.1554,
+      "learning_rate": 3.165007112375534e-05,
+      "loss": 0.183,
       "step": 890
     },
     {
       "epoch": 0.64,
-      "learning_rate": 4.370389884088514e-05,
-      "loss": 0.1387,
+      "learning_rate": 3.200568990042675e-05,
+      "loss": 0.171,
       "step": 900
     },
     {
       "epoch": 0.65,
-      "learning_rate": 4.3572181243414125e-05,
-      "loss": 0.1769,
+      "learning_rate": 3.236130867709815e-05,
+      "loss": 0.2169,
       "step": 910
     },
     {
       "epoch": 0.65,
-      "learning_rate": 4.34404636459431e-05,
-      "loss": 0.1697,
+      "learning_rate": 3.2716927453769555e-05,
+      "loss": 0.1667,
       "step": 920
     },
     {
       "epoch": 0.66,
-      "learning_rate": 4.330874604847208e-05,
-      "loss": 0.1711,
+      "learning_rate": 3.307254623044097e-05,
+      "loss": 0.177,
       "step": 930
     },
     {
       "epoch": 0.67,
-      "learning_rate": 4.317702845100105e-05,
-      "loss": 0.1842,
+      "learning_rate": 3.342816500711238e-05,
+      "loss": 0.2029,
       "step": 940
     },
     {
       "epoch": 0.68,
-      "learning_rate": 4.3045310853530036e-05,
-      "loss": 0.1413,
+      "learning_rate": 3.3783783783783784e-05,
+      "loss": 0.1601,
       "step": 950
     },
     {
       "epoch": 0.68,
-      "learning_rate": 4.291359325605901e-05,
-      "loss": 0.1799,
+      "learning_rate": 3.413940256045519e-05,
+      "loss": 0.1539,
       "step": 960
     },
     {
       "epoch": 0.69,
-      "learning_rate": 4.278187565858799e-05,
-      "loss": 0.1479,
+      "learning_rate": 3.4495021337126606e-05,
+      "loss": 0.1943,
       "step": 970
     },
     {
       "epoch": 0.7,
-      "learning_rate": 4.265015806111697e-05,
-      "loss": 0.1873,
+      "learning_rate": 3.485064011379801e-05,
+      "loss": 0.2285,
       "step": 980
     },
     {
       "epoch": 0.7,
-      "learning_rate": 4.2518440463645946e-05,
-      "loss": 0.1741,
+      "learning_rate": 3.520625889046942e-05,
+      "loss": 0.1538,
       "step": 990
     },
     {
       "epoch": 0.71,
-      "learning_rate": 4.238672286617492e-05,
-      "loss": 0.1262,
+      "learning_rate": 3.556187766714083e-05,
+      "loss": 0.2079,
       "step": 1000
     },
     {
       "epoch": 0.72,
-      "learning_rate": 4.2255005268703904e-05,
-      "loss": 0.1856,
+      "learning_rate": 3.5917496443812235e-05,
+      "loss": 0.2049,
       "step": 1010
     },
     {
       "epoch": 0.73,
-      "learning_rate": 4.212328767123288e-05,
-      "loss": 0.1391,
+      "learning_rate": 3.627311522048364e-05,
+      "loss": 0.1354,
       "step": 1020
     },
     {
       "epoch": 0.73,
-      "learning_rate": 4.1991570073761856e-05,
-      "loss": 0.1748,
+      "learning_rate": 3.662873399715505e-05,
+      "loss": 0.1739,
       "step": 1030
     },
     {
       "epoch": 0.74,
-      "learning_rate": 4.185985247629083e-05,
-      "loss": 0.1404,
+      "learning_rate": 3.698435277382646e-05,
+      "loss": 0.1598,
       "step": 1040
     },
     {
       "epoch": 0.75,
-      "learning_rate": 4.1728134878819814e-05,
-      "loss": 0.1151,
+      "learning_rate": 3.7339971550497865e-05,
+      "loss": 0.1764,
       "step": 1050
     },
     {
       "epoch": 0.75,
-      "learning_rate": 4.159641728134879e-05,
-      "loss": 0.192,
+      "learning_rate": 3.769559032716928e-05,
+      "loss": 0.1805,
       "step": 1060
     },
     {
       "epoch": 0.76,
-      "learning_rate": 4.1464699683877766e-05,
-      "loss": 0.1731,
+      "learning_rate": 3.805120910384069e-05,
+      "loss": 0.1988,
       "step": 1070
     },
     {
       "epoch": 0.77,
-      "learning_rate": 4.133298208640675e-05,
-      "loss": 0.1551,
+      "learning_rate": 3.8406827880512094e-05,
+      "loss": 0.1605,
       "step": 1080
     },
     {
       "epoch": 0.78,
-      "learning_rate": 4.1201264488935725e-05,
-      "loss": 0.1545,
+      "learning_rate": 3.87624466571835e-05,
+      "loss": 0.1984,
       "step": 1090
     },
     {
       "epoch": 0.78,
-      "learning_rate": 4.10695468914647e-05,
-      "loss": 0.1238,
+      "learning_rate": 3.911806543385491e-05,
+      "loss": 0.1951,
       "step": 1100
     },
     {
       "epoch": 0.79,
-      "learning_rate": 4.0937829293993676e-05,
-      "loss": 0.1036,
+      "learning_rate": 3.9473684210526316e-05,
+      "loss": 0.1981,
       "step": 1110
     },
     {
       "epoch": 0.8,
-      "learning_rate": 4.080611169652266e-05,
-      "loss": 0.173,
+      "learning_rate": 3.9829302987197724e-05,
+      "loss": 0.1877,
       "step": 1120
     },
     {
       "epoch": 0.8,
-      "learning_rate": 4.0674394099051635e-05,
-      "loss": 0.123,
+      "learning_rate": 4.018492176386913e-05,
+      "loss": 0.1534,
       "step": 1130
     },
     {
       "epoch": 0.81,
-      "learning_rate": 4.054267650158061e-05,
-      "loss": 0.1379,
+      "learning_rate": 4.0540540540540545e-05,
+      "loss": 0.1715,
       "step": 1140
     },
     {
       "epoch": 0.82,
-      "learning_rate": 4.041095890410959e-05,
-      "loss": 0.1605,
+      "learning_rate": 4.089615931721195e-05,
+      "loss": 0.1782,
       "step": 1150
     },
     {
       "epoch": 0.82,
-      "learning_rate": 4.027924130663857e-05,
-      "loss": 0.1423,
+      "learning_rate": 4.125177809388336e-05,
+      "loss": 0.1625,
       "step": 1160
     },
     {
       "epoch": 0.83,
-      "learning_rate": 4.0147523709167545e-05,
-      "loss": 0.1369,
+      "learning_rate": 4.160739687055477e-05,
+      "loss": 0.1598,
       "step": 1170
     },
     {
       "epoch": 0.84,
-      "learning_rate": 4.001580611169653e-05,
-      "loss": 0.129,
+      "learning_rate": 4.1963015647226175e-05,
+      "loss": 0.1522,
       "step": 1180
     },
     {
       "epoch": 0.85,
-      "learning_rate": 3.98840885142255e-05,
-      "loss": 0.1163,
+      "learning_rate": 4.231863442389758e-05,
+      "loss": 0.1893,
       "step": 1190
     },
     {
       "epoch": 0.85,
-      "learning_rate": 3.975237091675448e-05,
-      "loss": 0.1269,
+      "learning_rate": 4.267425320056899e-05,
+      "loss": 0.1808,
       "step": 1200
     },
     {
       "epoch": 0.86,
-      "learning_rate": 3.9620653319283455e-05,
-      "loss": 0.1513,
+      "learning_rate": 4.30298719772404e-05,
+      "loss": 0.1721,
       "step": 1210
     },
     {
       "epoch": 0.87,
-      "learning_rate": 3.948893572181244e-05,
-      "loss": 0.1355,
+      "learning_rate": 4.3385490753911804e-05,
+      "loss": 0.2053,
       "step": 1220
     },
     {
       "epoch": 0.87,
-      "learning_rate": 3.9357218124341414e-05,
-      "loss": 0.1308,
+      "learning_rate": 4.374110953058322e-05,
+      "loss": 0.1406,
       "step": 1230
     },
     {
       "epoch": 0.88,
-      "learning_rate": 3.922550052687039e-05,
-      "loss": 0.1474,
+      "learning_rate": 4.4096728307254626e-05,
+      "loss": 0.161,
       "step": 1240
     },
     {
       "epoch": 0.89,
-      "learning_rate": 3.909378292939937e-05,
-      "loss": 0.1599,
+      "learning_rate": 4.4452347083926033e-05,
+      "loss": 0.161,
       "step": 1250
     },
     {
       "epoch": 0.9,
-      "learning_rate": 3.896206533192835e-05,
-      "loss": 0.1383,
+      "learning_rate": 4.480796586059744e-05,
+      "loss": 0.1412,
       "step": 1260
     },
     {
       "epoch": 0.9,
-      "learning_rate": 3.8830347734457324e-05,
-      "loss": 0.1245,
+      "learning_rate": 4.5163584637268855e-05,
+      "loss": 0.1472,
       "step": 1270
     },
     {
       "epoch": 0.91,
-      "learning_rate": 3.86986301369863e-05,
-      "loss": 0.1458,
+      "learning_rate": 4.551920341394026e-05,
+      "loss": 0.2432,
       "step": 1280
     },
     {
       "epoch": 0.92,
-      "learning_rate": 3.856691253951528e-05,
-      "loss": 0.1361,
+      "learning_rate": 4.587482219061166e-05,
+      "loss": 0.2071,
       "step": 1290
     },
     {
       "epoch": 0.92,
-      "learning_rate": 3.843519494204426e-05,
-      "loss": 0.1498,
+      "learning_rate": 4.623044096728307e-05,
+      "loss": 0.1838,
       "step": 1300
     },
     {
       "epoch": 0.93,
-      "learning_rate": 3.8303477344573234e-05,
-      "loss": 0.1605,
+      "learning_rate": 4.6586059743954485e-05,
+      "loss": 0.1968,
       "step": 1310
     },
     {
       "epoch": 0.94,
-      "learning_rate": 3.8171759747102217e-05,
-      "loss": 0.1485,
+      "learning_rate": 4.694167852062589e-05,
+      "loss": 0.1429,
       "step": 1320
     },
     {
       "epoch": 0.95,
-      "learning_rate": 3.804004214963119e-05,
-      "loss": 0.1245,
+      "learning_rate": 4.72972972972973e-05,
+      "loss": 0.1746,
       "step": 1330
     },
     {
       "epoch": 0.95,
-      "learning_rate": 3.790832455216017e-05,
-      "loss": 0.149,
+      "learning_rate": 4.765291607396871e-05,
+      "loss": 0.1518,
       "step": 1340
     },
     {
       "epoch": 0.96,
-      "learning_rate": 3.777660695468915e-05,
-      "loss": 0.1138,
+      "learning_rate": 4.8008534850640114e-05,
+      "loss": 0.1725,
       "step": 1350
     },
     {
       "epoch": 0.97,
-      "learning_rate": 3.764488935721813e-05,
-      "loss": 0.1203,
+      "learning_rate": 4.836415362731153e-05,
+      "loss": 0.1604,
       "step": 1360
     },
     {
       "epoch": 0.97,
-      "learning_rate": 3.751317175974711e-05,
-      "loss": 0.1144,
+      "learning_rate": 4.8719772403982936e-05,
+      "loss": 0.1572,
       "step": 1370
     },
     {
       "epoch": 0.98,
-      "learning_rate": 3.738145416227608e-05,
-      "loss": 0.1451,
+      "learning_rate": 4.907539118065434e-05,
+      "loss": 0.2229,
       "step": 1380
     },
     {
       "epoch": 0.99,
-      "learning_rate": 3.724973656480506e-05,
-      "loss": 0.1256,
+      "learning_rate": 4.9431009957325744e-05,
+      "loss": 0.144,
       "step": 1390
     },
     {
       "epoch": 1.0,
-      "learning_rate": 3.711801896733404e-05,
-      "loss": 0.1624,
+      "learning_rate": 4.978662873399716e-05,
+      "loss": 0.1943,
       "step": 1400
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.9799,
-      "eval_loss": 0.05523204430937767,
-      "eval_runtime": 133.9464,
-      "eval_samples_per_second": 74.657,
-      "eval_steps_per_second": 4.666,
+      "eval_accuracy": 0.9757,
+      "eval_loss": 0.06815742701292038,
+      "eval_runtime": 46.9879,
+      "eval_samples_per_second": 212.821,
+      "eval_steps_per_second": 13.301,
       "step": 1406
     },
     {
       "epoch": 1.0,
-      "learning_rate": 3.698630136986301e-05,
-      "loss": 0.1724,
+      "learning_rate": 4.998419472103683e-05,
+      "loss": 0.1712,
       "step": 1410
     },
     {
       "epoch": 1.01,
-      "learning_rate": 3.6854583772391995e-05,
-      "loss": 0.1486,
+      "learning_rate": 4.9944681523628895e-05,
+      "loss": 0.1729,
       "step": 1420
     },
     {
       "epoch": 1.02,
-      "learning_rate": 3.672286617492097e-05,
-      "loss": 0.1642,
+      "learning_rate": 4.990516832622096e-05,
+      "loss": 0.1283,
       "step": 1430
     },
     {
       "epoch": 1.02,
-      "learning_rate": 3.6591148577449954e-05,
-      "loss": 0.1472,
+      "learning_rate": 4.9865655128813026e-05,
+      "loss": 0.1521,
       "step": 1440
     },
     {
       "epoch": 1.03,
-      "learning_rate": 3.645943097997892e-05,
-      "loss": 0.118,
+      "learning_rate": 4.982614193140509e-05,
+      "loss": 0.1357,
       "step": 1450
     },
     {
       "epoch": 1.04,
-      "learning_rate": 3.6327713382507905e-05,
-      "loss": 0.1535,
+      "learning_rate": 4.978662873399716e-05,
+      "loss": 0.2002,
       "step": 1460
     },
     {
       "epoch": 1.05,
-      "learning_rate": 3.619599578503688e-05,
-      "loss": 0.1396,
+      "learning_rate": 4.9747115536589224e-05,
+      "loss": 0.1805,
       "step": 1470
     },
     {
       "epoch": 1.05,
-      "learning_rate": 3.606427818756586e-05,
-      "loss": 0.1082,
+      "learning_rate": 4.970760233918128e-05,
+      "loss": 0.1749,
       "step": 1480
     },
     {
       "epoch": 1.06,
-      "learning_rate": 3.593256059009484e-05,
-      "loss": 0.0991,
+      "learning_rate": 4.9668089141773355e-05,
+      "loss": 0.1852,
       "step": 1490
     },
     {
       "epoch": 1.07,
-      "learning_rate": 3.5800842992623816e-05,
-      "loss": 0.1091,
+      "learning_rate": 4.962857594436542e-05,
+      "loss": 0.1257,
       "step": 1500
     },
     {
       "epoch": 1.07,
-      "learning_rate": 3.566912539515279e-05,
-      "loss": 0.1155,
+      "learning_rate": 4.958906274695749e-05,
+      "loss": 0.1639,
       "step": 1510
     },
     {
       "epoch": 1.08,
-      "learning_rate": 3.5537407797681774e-05,
-      "loss": 0.1383,
+      "learning_rate": 4.954954954954955e-05,
+      "loss": 0.1342,
       "step": 1520
     },
     {
       "epoch": 1.09,
-      "learning_rate": 3.540569020021075e-05,
-      "loss": 0.0968,
+      "learning_rate": 4.951003635214162e-05,
+      "loss": 0.1566,
       "step": 1530
     },
     {
       "epoch": 1.1,
-      "learning_rate": 3.527397260273973e-05,
-      "loss": 0.1082,
+      "learning_rate": 4.9470523154733685e-05,
+      "loss": 0.1851,
       "step": 1540
     },
     {
       "epoch": 1.1,
-      "learning_rate": 3.51422550052687e-05,
-      "loss": 0.1382,
+      "learning_rate": 4.9431009957325744e-05,
+      "loss": 0.1835,
       "step": 1550
     },
     {
       "epoch": 1.11,
-      "learning_rate": 3.5010537407797684e-05,
-      "loss": 0.1374,
+      "learning_rate": 4.9391496759917816e-05,
+      "loss": 0.1461,
       "step": 1560
     },
     {
       "epoch": 1.12,
-      "learning_rate": 3.487881981032666e-05,
-      "loss": 0.1468,
+      "learning_rate": 4.935198356250988e-05,
+      "loss": 0.1731,
       "step": 1570
     },
     {
       "epoch": 1.12,
-      "learning_rate": 3.4747102212855636e-05,
-      "loss": 0.1012,
+      "learning_rate": 4.931247036510195e-05,
+      "loss": 0.1354,
       "step": 1580
     },
     {
       "epoch": 1.13,
-      "learning_rate": 3.461538461538462e-05,
-      "loss": 0.1253,
+      "learning_rate": 4.9272957167694014e-05,
+      "loss": 0.1748,
       "step": 1590
     },
     {
       "epoch": 1.14,
-      "learning_rate": 3.4483667017913594e-05,
-      "loss": 0.1533,
+      "learning_rate": 4.923344397028608e-05,
+      "loss": 0.1737,
       "step": 1600
     },
     {
       "epoch": 1.14,
-      "learning_rate": 3.435194942044258e-05,
-      "loss": 0.1365,
+      "learning_rate": 4.9193930772878146e-05,
+      "loss": 0.1697,
       "step": 1610
     },
     {
       "epoch": 1.15,
-      "learning_rate": 3.4220231822971546e-05,
-      "loss": 0.1451,
+      "learning_rate": 4.9154417575470205e-05,
+      "loss": 0.1296,
       "step": 1620
     },
     {
       "epoch": 1.16,
-      "learning_rate": 3.408851422550053e-05,
-      "loss": 0.1192,
+      "learning_rate": 4.911490437806228e-05,
+      "loss": 0.1612,
       "step": 1630
     },
     {
       "epoch": 1.17,
-      "learning_rate": 3.3956796628029505e-05,
-      "loss": 0.1206,
+      "learning_rate": 4.907539118065434e-05,
+      "loss": 0.1369,
       "step": 1640
     },
     {
       "epoch": 1.17,
-      "learning_rate": 3.382507903055848e-05,
-      "loss": 0.1222,
+      "learning_rate": 4.903587798324641e-05,
+      "loss": 0.16,
       "step": 1650
     },
     {
       "epoch": 1.18,
-      "learning_rate": 3.369336143308746e-05,
-      "loss": 0.1068,
+      "learning_rate": 4.8996364785838475e-05,
+      "loss": 0.1488,
       "step": 1660
     },
     {
       "epoch": 1.19,
-      "learning_rate": 3.356164383561644e-05,
-      "loss": 0.1541,
+      "learning_rate": 4.8956851588430534e-05,
+      "loss": 0.1858,
       "step": 1670
     },
     {
       "epoch": 1.19,
-      "learning_rate": 3.342992623814542e-05,
-      "loss": 0.1448,
+      "learning_rate": 4.8917338391022607e-05,
+      "loss": 0.1724,
       "step": 1680
     },
     {
       "epoch": 1.2,
-      "learning_rate": 3.32982086406744e-05,
-      "loss": 0.1205,
+      "learning_rate": 4.8877825193614666e-05,
+      "loss": 0.1517,
       "step": 1690
     },
     {
       "epoch": 1.21,
-      "learning_rate": 3.316649104320337e-05,
-      "loss": 0.1337,
+      "learning_rate": 4.883831199620674e-05,
+      "loss": 0.1406,
       "step": 1700
     },
     {
       "epoch": 1.22,
-      "learning_rate": 3.3034773445732356e-05,
-      "loss": 0.1059,
+      "learning_rate": 4.87987987987988e-05,
+      "loss": 0.1573,
       "step": 1710
     },
     {
       "epoch": 1.22,
-      "learning_rate": 3.2903055848261325e-05,
-      "loss": 0.1046,
+      "learning_rate": 4.875928560139087e-05,
+      "loss": 0.1929,
       "step": 1720
     },
     {
       "epoch": 1.23,
-      "learning_rate": 3.277133825079031e-05,
-      "loss": 0.1184,
+      "learning_rate": 4.8719772403982936e-05,
+      "loss": 0.1649,
       "step": 1730
     },
     {
       "epoch": 1.24,
-      "learning_rate": 3.2639620653319283e-05,
-      "loss": 0.1183,
+      "learning_rate": 4.8680259206574995e-05,
+      "loss": 0.1615,
       "step": 1740
     },
     {
       "epoch": 1.24,
-      "learning_rate": 3.2507903055848266e-05,
-      "loss": 0.0955,
+      "learning_rate": 4.864074600916707e-05,
+      "loss": 0.1738,
       "step": 1750
     },
     {
       "epoch": 1.25,
-      "learning_rate": 3.237618545837724e-05,
-      "loss": 0.1096,
+      "learning_rate": 4.8601232811759127e-05,
+      "loss": 0.1062,
       "step": 1760
     },
     {
       "epoch": 1.26,
-      "learning_rate": 3.224446786090622e-05,
-      "loss": 0.0932,
+      "learning_rate": 4.85617196143512e-05,
+      "loss": 0.1233,
       "step": 1770
     },
     {
       "epoch": 1.27,
-      "learning_rate": 3.21127502634352e-05,
-      "loss": 0.1468,
+      "learning_rate": 4.852220641694326e-05,
+      "loss": 0.1855,
       "step": 1780
     },
     {
       "epoch": 1.27,
-      "learning_rate": 3.198103266596417e-05,
-      "loss": 0.1202,
+      "learning_rate": 4.8482693219535324e-05,
+      "loss": 0.1745,
       "step": 1790
     },
     {
       "epoch": 1.28,
-      "learning_rate": 3.184931506849315e-05,
-      "loss": 0.1458,
+      "learning_rate": 4.84431800221274e-05,
+      "loss": 0.1482,
       "step": 1800
     },
     {
       "epoch": 1.29,
-      "learning_rate": 3.171759747102213e-05,
-      "loss": 0.1065,
+      "learning_rate": 4.8403666824719456e-05,
+      "loss": 0.1579,
       "step": 1810
     },
     {
       "epoch": 1.29,
-      "learning_rate": 3.1585879873551104e-05,
-      "loss": 0.1114,
+      "learning_rate": 4.836415362731153e-05,
+      "loss": 0.1364,
       "step": 1820
     },
     {
       "epoch": 1.3,
-      "learning_rate": 3.1454162276080086e-05,
-      "loss": 0.1282,
+      "learning_rate": 4.832464042990359e-05,
+      "loss": 0.1315,
       "step": 1830
     },
     {
       "epoch": 1.31,
-      "learning_rate": 3.132244467860906e-05,
-      "loss": 0.1271,
+      "learning_rate": 4.828512723249566e-05,
+      "loss": 0.2167,
       "step": 1840
     },
     {
       "epoch": 1.32,
-      "learning_rate": 3.1190727081138045e-05,
-      "loss": 0.111,
+      "learning_rate": 4.824561403508772e-05,
+      "loss": 0.11,
       "step": 1850
     },
     {
       "epoch": 1.32,
-      "learning_rate": 3.105900948366702e-05,
-      "loss": 0.1371,
+      "learning_rate": 4.8206100837679785e-05,
+      "loss": 0.1454,
       "step": 1860
     },
     {
       "epoch": 1.33,
-      "learning_rate": 3.0927291886195997e-05,
-      "loss": 0.1111,
+      "learning_rate": 4.816658764027185e-05,
+      "loss": 0.1258,
       "step": 1870
     },
     {
       "epoch": 1.34,
-      "learning_rate": 3.079557428872498e-05,
-      "loss": 0.1484,
+      "learning_rate": 4.812707444286392e-05,
+      "loss": 0.1144,
       "step": 1880
     },
     {
       "epoch": 1.34,
-      "learning_rate": 3.066385669125395e-05,
-      "loss": 0.1045,
+      "learning_rate": 4.808756124545599e-05,
+      "loss": 0.1276,
       "step": 1890
     },
     {
       "epoch": 1.35,
-      "learning_rate": 3.053213909378293e-05,
-      "loss": 0.1128,
+      "learning_rate": 4.804804804804805e-05,
+      "loss": 0.1447,
       "step": 1900
     },
     {
       "epoch": 1.36,
-      "learning_rate": 3.040042149631191e-05,
-      "loss": 0.1391,
+      "learning_rate": 4.8008534850640114e-05,
+      "loss": 0.1221,
       "step": 1910
     },
     {
       "epoch": 1.37,
-      "learning_rate": 3.026870389884089e-05,
-      "loss": 0.1183,
+      "learning_rate": 4.796902165323218e-05,
+      "loss": 0.1232,
       "step": 1920
     },
     {
       "epoch": 1.37,
-      "learning_rate": 3.0136986301369862e-05,
-      "loss": 0.0963,
+      "learning_rate": 4.7929508455824246e-05,
+      "loss": 0.1267,
       "step": 1930
     },
     {
       "epoch": 1.38,
-      "learning_rate": 3.000526870389884e-05,
-      "loss": 0.0996,
+      "learning_rate": 4.788999525841631e-05,
+      "loss": 0.1107,
       "step": 1940
     },
     {
       "epoch": 1.39,
-      "learning_rate": 2.987355110642782e-05,
-      "loss": 0.106,
+      "learning_rate": 4.785048206100838e-05,
+      "loss": 0.1353,
       "step": 1950
     },
     {
       "epoch": 1.39,
-      "learning_rate": 2.9741833508956796e-05,
-      "loss": 0.138,
+      "learning_rate": 4.781096886360045e-05,
+      "loss": 0.1466,
       "step": 1960
     },
     {
       "epoch": 1.4,
-      "learning_rate": 2.9610115911485775e-05,
-      "loss": 0.1299,
+      "learning_rate": 4.777145566619251e-05,
+      "loss": 0.141,
       "step": 1970
     },
     {
       "epoch": 1.41,
-      "learning_rate": 2.9478398314014755e-05,
-      "loss": 0.1163,
+      "learning_rate": 4.7731942468784575e-05,
+      "loss": 0.1489,
       "step": 1980
     },
     {
       "epoch": 1.42,
-      "learning_rate": 2.9346680716543734e-05,
-      "loss": 0.0867,
+      "learning_rate": 4.769242927137664e-05,
+      "loss": 0.1332,
       "step": 1990
     },
     {
       "epoch": 1.42,
-      "learning_rate": 2.921496311907271e-05,
-      "loss": 0.1192,
+      "learning_rate": 4.765291607396871e-05,
+      "loss": 0.1713,
       "step": 2000
     },
     {
       "epoch": 1.43,
-      "learning_rate": 2.908324552160169e-05,
-      "loss": 0.0969,
+      "learning_rate": 4.761340287656077e-05,
+      "loss": 0.1306,
       "step": 2010
     },
     {
       "epoch": 1.44,
-      "learning_rate": 2.8951527924130668e-05,
-      "loss": 0.1388,
+      "learning_rate": 4.757388967915284e-05,
+      "loss": 0.1145,
       "step": 2020
     },
     {
       "epoch": 1.44,
-      "learning_rate": 2.881981032665964e-05,
-      "loss": 0.1036,
+      "learning_rate": 4.753437648174491e-05,
+      "loss": 0.1578,
       "step": 2030
     },
     {
       "epoch": 1.45,
-      "learning_rate": 2.868809272918862e-05,
-      "loss": 0.09,
+      "learning_rate": 4.749486328433697e-05,
+      "loss": 0.1339,
       "step": 2040
     },
     {
       "epoch": 1.46,
-      "learning_rate": 2.85563751317176e-05,
-      "loss": 0.1042,
+      "learning_rate": 4.7455350086929036e-05,
+      "loss": 0.1509,
       "step": 2050
     },
     {
       "epoch": 1.46,
-      "learning_rate": 2.842465753424658e-05,
-      "loss": 0.1187,
+      "learning_rate": 4.74158368895211e-05,
+      "loss": 0.1377,
       "step": 2060
     },
     {
       "epoch": 1.47,
-      "learning_rate": 2.8292939936775554e-05,
-      "loss": 0.1013,
+      "learning_rate": 4.737632369211317e-05,
+      "loss": 0.1663,
       "step": 2070
     },
     {
       "epoch": 1.48,
-      "learning_rate": 2.8161222339304533e-05,
-      "loss": 0.1275,
+      "learning_rate": 4.7336810494705234e-05,
+      "loss": 0.1554,
       "step": 2080
     },
     {
       "epoch": 1.49,
-      "learning_rate": 2.8029504741833513e-05,
-      "loss": 0.0892,
+      "learning_rate": 4.72972972972973e-05,
+      "loss": 0.1503,
       "step": 2090
     },
     {
       "epoch": 1.49,
-      "learning_rate": 2.7897787144362485e-05,
-      "loss": 0.1066,
+      "learning_rate": 4.7257784099889365e-05,
+      "loss": 0.1727,
       "step": 2100
     },
     {
       "epoch": 1.5,
-      "learning_rate": 2.7766069546891464e-05,
-      "loss": 0.096,
+      "learning_rate": 4.721827090248143e-05,
+      "loss": 0.1174,
       "step": 2110
     },
     {
       "epoch": 1.51,
-      "learning_rate": 2.7634351949420444e-05,
-      "loss": 0.1263,
+      "learning_rate": 4.71787577050735e-05,
+      "loss": 0.125,
       "step": 2120
     },
     {
       "epoch": 1.51,
-      "learning_rate": 2.7502634351949423e-05,
-      "loss": 0.118,
+      "learning_rate": 4.713924450766556e-05,
+      "loss": 0.1155,
       "step": 2130
     },
     {
       "epoch": 1.52,
-      "learning_rate": 2.73709167544784e-05,
-      "loss": 0.1079,
+      "learning_rate": 4.709973131025763e-05,
+      "loss": 0.1423,
       "step": 2140
     },
     {
       "epoch": 1.53,
-      "learning_rate": 2.7239199157007378e-05,
-      "loss": 0.1137,
+      "learning_rate": 4.7060218112849694e-05,
+      "loss": 0.1535,
       "step": 2150
     },
     {
       "epoch": 1.54,
-      "learning_rate": 2.7107481559536357e-05,
-      "loss": 0.1092,
+      "learning_rate": 4.702070491544176e-05,
+      "loss": 0.1161,
       "step": 2160
     },
     {
       "epoch": 1.54,
-      "learning_rate": 2.6975763962065333e-05,
-      "loss": 0.1195,
+      "learning_rate": 4.6981191718033826e-05,
+      "loss": 0.1184,
       "step": 2170
     },
     {
       "epoch": 1.55,
-      "learning_rate": 2.6844046364594312e-05,
-      "loss": 0.1131,
+      "learning_rate": 4.694167852062589e-05,
+      "loss": 0.1196,
       "step": 2180
     },
     {
       "epoch": 1.56,
-      "learning_rate": 2.671232876712329e-05,
-      "loss": 0.1363,
+      "learning_rate": 4.690216532321796e-05,
+      "loss": 0.1115,
       "step": 2190
     },
     {
       "epoch": 1.56,
-      "learning_rate": 2.6580611169652264e-05,
-      "loss": 0.13,
+      "learning_rate": 4.6862652125810024e-05,
+      "loss": 0.2096,
       "step": 2200
     },
     {
       "epoch": 1.57,
-      "learning_rate": 2.6448893572181243e-05,
-      "loss": 0.1152,
+      "learning_rate": 4.682313892840209e-05,
+      "loss": 0.1405,
       "step": 2210
     },
     {
       "epoch": 1.58,
-      "learning_rate": 2.6317175974710222e-05,
-      "loss": 0.1155,
+      "learning_rate": 4.678362573099415e-05,
+      "loss": 0.1463,
       "step": 2220
     },
     {
       "epoch": 1.59,
-      "learning_rate": 2.61854583772392e-05,
-      "loss": 0.1159,
+      "learning_rate": 4.674411253358622e-05,
+      "loss": 0.1438,
       "step": 2230
     },
     {
       "epoch": 1.59,
-      "learning_rate": 2.6053740779768177e-05,
-      "loss": 0.1146,
+      "learning_rate": 4.670459933617829e-05,
+      "loss": 0.122,
       "step": 2240
     },
     {
       "epoch": 1.6,
-      "learning_rate": 2.5922023182297157e-05,
-      "loss": 0.1176,
+      "learning_rate": 4.666508613877035e-05,
+      "loss": 0.1434,
       "step": 2250
     },
     {
       "epoch": 1.61,
-      "learning_rate": 2.5790305584826136e-05,
-      "loss": 0.1327,
+      "learning_rate": 4.662557294136242e-05,
+      "loss": 0.1469,
       "step": 2260
     },
     {
       "epoch": 1.61,
-      "learning_rate": 2.565858798735511e-05,
-      "loss": 0.0938,
+      "learning_rate": 4.6586059743954485e-05,
+      "loss": 0.1753,
       "step": 2270
     },
     {
       "epoch": 1.62,
-      "learning_rate": 2.5526870389884088e-05,
-      "loss": 0.1234,
+      "learning_rate": 4.654654654654655e-05,
+      "loss": 0.1207,
       "step": 2280
     },
     {
       "epoch": 1.63,
-      "learning_rate": 2.5395152792413067e-05,
-      "loss": 0.1329,
+      "learning_rate": 4.650703334913861e-05,
+      "loss": 0.1152,
       "step": 2290
     },
     {
       "epoch": 1.64,
-      "learning_rate": 2.5263435194942046e-05,
-      "loss": 0.1093,
+      "learning_rate": 4.646752015173068e-05,
+      "loss": 0.1753,
       "step": 2300
     },
     {
       "epoch": 1.64,
-      "learning_rate": 2.5131717597471022e-05,
-      "loss": 0.1051,
+      "learning_rate": 4.642800695432275e-05,
+      "loss": 0.1196,
       "step": 2310
     },
     {
       "epoch": 1.65,
-      "learning_rate": 2.5e-05,
-      "loss": 0.1022,
+      "learning_rate": 4.6388493756914814e-05,
+      "loss": 0.1254,
       "step": 2320
     },
     {
       "epoch": 1.66,
-      "learning_rate": 2.4868282402528977e-05,
-      "loss": 0.1071,
+      "learning_rate": 4.634898055950688e-05,
+      "loss": 0.1075,
       "step": 2330
     },
     {
       "epoch": 1.66,
-      "learning_rate": 2.4736564805057956e-05,
-      "loss": 0.1097,
+      "learning_rate": 4.630946736209894e-05,
+      "loss": 0.1065,
       "step": 2340
     },
     {
       "epoch": 1.67,
-      "learning_rate": 2.4604847207586936e-05,
-      "loss": 0.0796,
+      "learning_rate": 4.626995416469101e-05,
+      "loss": 0.1408,
       "step": 2350
     },
     {
       "epoch": 1.68,
-      "learning_rate": 2.4473129610115915e-05,
-      "loss": 0.0795,
+      "learning_rate": 4.623044096728307e-05,
+      "loss": 0.1087,
       "step": 2360
     },
     {
       "epoch": 1.69,
-      "learning_rate": 2.434141201264489e-05,
-      "loss": 0.0999,
+      "learning_rate": 4.619092776987514e-05,
+      "loss": 0.1663,
       "step": 2370
     },
     {
       "epoch": 1.69,
-      "learning_rate": 2.420969441517387e-05,
-      "loss": 0.1133,
+      "learning_rate": 4.61514145724672e-05,
+      "loss": 0.0979,
       "step": 2380
     },
     {
       "epoch": 1.7,
-      "learning_rate": 2.4077976817702846e-05,
-      "loss": 0.0935,
+      "learning_rate": 4.6111901375059275e-05,
+      "loss": 0.1626,
       "step": 2390
     },
     {
       "epoch": 1.71,
-      "learning_rate": 2.394625922023182e-05,
-      "loss": 0.0992,
+      "learning_rate": 4.607238817765134e-05,
+      "loss": 0.1509,
       "step": 2400
     },
     {
       "epoch": 1.71,
-      "learning_rate": 2.3814541622760804e-05,
-      "loss": 0.1119,
+      "learning_rate": 4.60328749802434e-05,
+      "loss": 0.1337,
       "step": 2410
     },
     {
       "epoch": 1.72,
-      "learning_rate": 2.368282402528978e-05,
-      "loss": 0.1198,
+      "learning_rate": 4.599336178283547e-05,
+      "loss": 0.1147,
       "step": 2420
     },
     {
       "epoch": 1.73,
-      "learning_rate": 2.355110642781876e-05,
-      "loss": 0.0977,
+      "learning_rate": 4.595384858542753e-05,
+      "loss": 0.118,
       "step": 2430
     },
     {
       "epoch": 1.74,
-      "learning_rate": 2.3419388830347735e-05,
-      "loss": 0.1244,
+      "learning_rate": 4.5914335388019604e-05,
+      "loss": 0.1415,
       "step": 2440
     },
     {
       "epoch": 1.74,
-      "learning_rate": 2.328767123287671e-05,
-      "loss": 0.1181,
+      "learning_rate": 4.587482219061166e-05,
+      "loss": 0.1387,
       "step": 2450
     },
     {
       "epoch": 1.75,
-      "learning_rate": 2.315595363540569e-05,
-      "loss": 0.1042,
+      "learning_rate": 4.5835308993203736e-05,
+      "loss": 0.1285,
       "step": 2460
     },
     {
       "epoch": 1.76,
-      "learning_rate": 2.302423603793467e-05,
-      "loss": 0.1323,
+      "learning_rate": 4.57957957957958e-05,
+      "loss": 0.1746,
       "step": 2470
     },
     {
       "epoch": 1.76,
-      "learning_rate": 2.289251844046365e-05,
-      "loss": 0.1283,
+      "learning_rate": 4.575628259838786e-05,
+      "loss": 0.1029,
       "step": 2480
     },
     {
       "epoch": 1.77,
-      "learning_rate": 2.2760800842992624e-05,
-      "loss": 0.1031,
+      "learning_rate": 4.571676940097993e-05,
+      "loss": 0.1199,
       "step": 2490
     },
     {
       "epoch": 1.78,
-      "learning_rate": 2.2629083245521604e-05,
-      "loss": 0.0992,
+      "learning_rate": 4.567725620357199e-05,
+      "loss": 0.1041,
       "step": 2500
     },
     {
       "epoch": 1.78,
-      "learning_rate": 2.249736564805058e-05,
-      "loss": 0.083,
+      "learning_rate": 4.5637743006164065e-05,
+      "loss": 0.1304,
       "step": 2510
     },
     {
       "epoch": 1.79,
-      "learning_rate": 2.236564805057956e-05,
-      "loss": 0.1079,
+      "learning_rate": 4.5598229808756124e-05,
+      "loss": 0.121,
       "step": 2520
     },
     {
       "epoch": 1.8,
-      "learning_rate": 2.2233930453108538e-05,
-      "loss": 0.0877,
+      "learning_rate": 4.555871661134819e-05,
+      "loss": 0.1369,
       "step": 2530
     },
     {
       "epoch": 1.81,
-      "learning_rate": 2.2102212855637514e-05,
-      "loss": 0.1187,
+      "learning_rate": 4.551920341394026e-05,
+      "loss": 0.1516,
       "step": 2540
     },
     {
       "epoch": 1.81,
-      "learning_rate": 2.1970495258166493e-05,
-      "loss": 0.0835,
+      "learning_rate": 4.547969021653232e-05,
+      "loss": 0.1706,
       "step": 2550
     },
     {
       "epoch": 1.82,
-      "learning_rate": 2.183877766069547e-05,
-      "loss": 0.0926,
+      "learning_rate": 4.5440177019124394e-05,
+      "loss": 0.1014,
       "step": 2560
     },
     {
       "epoch": 1.83,
-      "learning_rate": 2.1707060063224448e-05,
-      "loss": 0.1383,
+      "learning_rate": 4.540066382171645e-05,
+      "loss": 0.1303,
       "step": 2570
     },
     {
       "epoch": 1.83,
-      "learning_rate": 2.1575342465753427e-05,
-      "loss": 0.1143,
+      "learning_rate": 4.5361150624308526e-05,
+      "loss": 0.0914,
       "step": 2580
     },
     {
       "epoch": 1.84,
-      "learning_rate": 2.1443624868282403e-05,
-      "loss": 0.1197,
+      "learning_rate": 4.5321637426900585e-05,
+      "loss": 0.1159,
       "step": 2590
     },
     {
       "epoch": 1.85,
-      "learning_rate": 2.1311907270811383e-05,
-      "loss": 0.0983,
+      "learning_rate": 4.528212422949265e-05,
+      "loss": 0.1496,
       "step": 2600
     },
     {
       "epoch": 1.86,
-      "learning_rate": 2.118018967334036e-05,
-      "loss": 0.1105,
+      "learning_rate": 4.5242611032084717e-05,
+      "loss": 0.1396,
       "step": 2610
     },
     {
       "epoch": 1.86,
-      "learning_rate": 2.1048472075869338e-05,
-      "loss": 0.0993,
+      "learning_rate": 4.520309783467678e-05,
+      "loss": 0.1284,
       "step": 2620
     },
     {
       "epoch": 1.87,
-      "learning_rate": 2.0916754478398313e-05,
-      "loss": 0.1156,
+      "learning_rate": 4.5163584637268855e-05,
+      "loss": 0.144,
       "step": 2630
     },
     {
       "epoch": 1.88,
-      "learning_rate": 2.0785036880927293e-05,
-      "loss": 0.1111,
+      "learning_rate": 4.5124071439860914e-05,
+      "loss": 0.1135,
       "step": 2640
     },
     {
       "epoch": 1.88,
-      "learning_rate": 2.0653319283456272e-05,
-      "loss": 0.0834,
+      "learning_rate": 4.508455824245298e-05,
+      "loss": 0.1112,
       "step": 2650
     },
     {
       "epoch": 1.89,
-      "learning_rate": 2.0521601685985248e-05,
-      "loss": 0.0993,
+      "learning_rate": 4.5045045045045046e-05,
+      "loss": 0.158,
       "step": 2660
     },
     {
       "epoch": 1.9,
-      "learning_rate": 2.0389884088514227e-05,
-      "loss": 0.0803,
+      "learning_rate": 4.500553184763711e-05,
+      "loss": 0.1285,
       "step": 2670
     },
     {
       "epoch": 1.91,
-      "learning_rate": 2.0258166491043203e-05,
-      "loss": 0.1206,
+      "learning_rate": 4.496601865022918e-05,
+      "loss": 0.1628,
       "step": 2680
     },
     {
       "epoch": 1.91,
-      "learning_rate": 2.0126448893572182e-05,
-      "loss": 0.09,
+      "learning_rate": 4.492650545282124e-05,
+      "loss": 0.1085,
       "step": 2690
     },
     {
       "epoch": 1.92,
-      "learning_rate": 1.999473129610116e-05,
-      "loss": 0.0954,
+      "learning_rate": 4.4886992255413316e-05,
+      "loss": 0.1082,
       "step": 2700
     },
     {
       "epoch": 1.93,
-      "learning_rate": 1.9863013698630137e-05,
-      "loss": 0.1062,
+      "learning_rate": 4.4847479058005375e-05,
+      "loss": 0.1529,
       "step": 2710
     },
     {
       "epoch": 1.93,
-      "learning_rate": 1.9731296101159116e-05,
-      "loss": 0.1117,
+      "learning_rate": 4.480796586059744e-05,
+      "loss": 0.1051,
       "step": 2720
     },
     {
       "epoch": 1.94,
-      "learning_rate": 1.9599578503688092e-05,
-      "loss": 0.1496,
+      "learning_rate": 4.476845266318951e-05,
+      "loss": 0.1181,
       "step": 2730
     },
     {
       "epoch": 1.95,
-      "learning_rate": 1.946786090621707e-05,
-      "loss": 0.0816,
+      "learning_rate": 4.472893946578157e-05,
+      "loss": 0.1345,
       "step": 2740
     },
     {
       "epoch": 1.96,
-      "learning_rate": 1.933614330874605e-05,
-      "loss": 0.1257,
+      "learning_rate": 4.468942626837364e-05,
+      "loss": 0.1143,
       "step": 2750
     },
     {
       "epoch": 1.96,
-      "learning_rate": 1.9204425711275027e-05,
-      "loss": 0.1006,
+      "learning_rate": 4.4649913070965704e-05,
+      "loss": 0.1242,
       "step": 2760
     },
     {
       "epoch": 1.97,
-      "learning_rate": 1.9072708113804006e-05,
-      "loss": 0.0912,
+      "learning_rate": 4.461039987355777e-05,
+      "loss": 0.078,
       "step": 2770
     },
     {
       "epoch": 1.98,
-      "learning_rate": 1.894099051633298e-05,
-      "loss": 0.0774,
+      "learning_rate": 4.4570886676149836e-05,
+      "loss": 0.1616,
       "step": 2780
     },
     {
       "epoch": 1.98,
-      "learning_rate": 1.880927291886196e-05,
-      "loss": 0.0785,
+      "learning_rate": 4.45313734787419e-05,
+      "loss": 0.1086,
       "step": 2790
     },
     {
       "epoch": 1.99,
-      "learning_rate": 1.8677555321390937e-05,
-      "loss": 0.063,
+      "learning_rate": 4.449186028133397e-05,
+      "loss": 0.1138,
       "step": 2800
     },
     {
       "epoch": 2.0,
-      "learning_rate": 1.8545837723919916e-05,
-      "loss": 0.1376,
+      "learning_rate": 4.4452347083926033e-05,
+      "loss": 0.1288,
       "step": 2810
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.9895,
-      "eval_loss": 0.03449298441410065,
-      "eval_runtime": 133.6612,
-      "eval_samples_per_second": 74.816,
-      "eval_steps_per_second": 4.676,
+      "eval_accuracy": 0.9852,
+      "eval_loss": 0.042292360216379166,
+      "eval_runtime": 47.0561,
+      "eval_samples_per_second": 212.512,
+      "eval_steps_per_second": 13.282,
       "step": 2812
     },
     {
       "epoch": 2.01,
-      "learning_rate": 1.8414120126448895e-05,
-      "loss": 0.0691,
+      "learning_rate": 4.44128338865181e-05,
+      "loss": 0.1454,
       "step": 2820
     },
     {
       "epoch": 2.01,
-      "learning_rate": 1.828240252897787e-05,
-      "loss": 0.0683,
+      "learning_rate": 4.4373320689110165e-05,
+      "loss": 0.1065,
       "step": 2830
     },
     {
       "epoch": 2.02,
-      "learning_rate": 1.815068493150685e-05,
-      "loss": 0.0809,
+      "learning_rate": 4.433380749170223e-05,
+      "loss": 0.097,
       "step": 2840
     },
     {
       "epoch": 2.03,
-      "learning_rate": 1.8018967334035826e-05,
-      "loss": 0.0951,
+      "learning_rate": 4.42942942942943e-05,
+      "loss": 0.1325,
       "step": 2850
     },
     {
       "epoch": 2.03,
-      "learning_rate": 1.7887249736564805e-05,
-      "loss": 0.0883,
+      "learning_rate": 4.425478109688636e-05,
+      "loss": 0.1473,
       "step": 2860
     },
     {
       "epoch": 2.04,
-      "learning_rate": 1.7755532139093785e-05,
-      "loss": 0.0633,
+      "learning_rate": 4.421526789947843e-05,
+      "loss": 0.1265,
       "step": 2870
     },
     {
       "epoch": 2.05,
-      "learning_rate": 1.7623814541622764e-05,
-      "loss": 0.0828,
+      "learning_rate": 4.4175754702070494e-05,
+      "loss": 0.0941,
       "step": 2880
     },
     {
       "epoch": 2.06,
-      "learning_rate": 1.749209694415174e-05,
-      "loss": 0.0937,
+      "learning_rate": 4.413624150466256e-05,
+      "loss": 0.0755,
       "step": 2890
     },
     {
       "epoch": 2.06,
-      "learning_rate": 1.7360379346680716e-05,
-      "loss": 0.1109,
+      "learning_rate": 4.4096728307254626e-05,
+      "loss": 0.0979,
       "step": 2900
     },
     {
       "epoch": 2.07,
-      "learning_rate": 1.7228661749209695e-05,
-      "loss": 0.081,
+      "learning_rate": 4.405721510984669e-05,
+      "loss": 0.1223,
       "step": 2910
     },
     {
       "epoch": 2.08,
-      "learning_rate": 1.7096944151738674e-05,
-      "loss": 0.0974,
+      "learning_rate": 4.401770191243876e-05,
+      "loss": 0.1535,
       "step": 2920
     },
     {
       "epoch": 2.08,
-      "learning_rate": 1.6965226554267653e-05,
-      "loss": 0.0826,
+      "learning_rate": 4.3978188715030824e-05,
+      "loss": 0.1229,
       "step": 2930
     },
     {
       "epoch": 2.09,
-      "learning_rate": 1.683350895679663e-05,
-      "loss": 0.0593,
+      "learning_rate": 4.393867551762289e-05,
+      "loss": 0.117,
       "step": 2940
     },
     {
       "epoch": 2.1,
-      "learning_rate": 1.6701791359325605e-05,
-      "loss": 0.073,
+      "learning_rate": 4.3899162320214955e-05,
+      "loss": 0.093,
       "step": 2950
     },
     {
       "epoch": 2.1,
-      "learning_rate": 1.6570073761854584e-05,
-      "loss": 0.1086,
+      "learning_rate": 4.3859649122807014e-05,
+      "loss": 0.1124,
       "step": 2960
     },
     {
       "epoch": 2.11,
-      "learning_rate": 1.643835616438356e-05,
-      "loss": 0.0972,
+      "learning_rate": 4.382013592539909e-05,
+      "loss": 0.0885,
       "step": 2970
     },
     {
       "epoch": 2.12,
-      "learning_rate": 1.630663856691254e-05,
-      "loss": 0.0953,
+      "learning_rate": 4.378062272799115e-05,
+      "loss": 0.1006,
       "step": 2980
     },
     {
       "epoch": 2.13,
-      "learning_rate": 1.617492096944152e-05,
-      "loss": 0.0819,
+      "learning_rate": 4.374110953058322e-05,
+      "loss": 0.0977,
       "step": 2990
     },
     {
       "epoch": 2.13,
-      "learning_rate": 1.6043203371970498e-05,
-      "loss": 0.0631,
+      "learning_rate": 4.3701596333175284e-05,
+      "loss": 0.1124,
       "step": 3000
     },
     {
       "epoch": 2.14,
-      "learning_rate": 1.5911485774499474e-05,
-      "loss": 0.1025,
+      "learning_rate": 4.366208313576735e-05,
+      "loss": 0.111,
       "step": 3010
     },
     {
       "epoch": 2.15,
-      "learning_rate": 1.577976817702845e-05,
-      "loss": 0.0983,
+      "learning_rate": 4.3622569938359416e-05,
+      "loss": 0.1109,
       "step": 3020
     },
     {
       "epoch": 2.15,
-      "learning_rate": 1.564805057955743e-05,
-      "loss": 0.1099,
+      "learning_rate": 4.3583056740951475e-05,
+      "loss": 0.0898,
       "step": 3030
     },
     {
       "epoch": 2.16,
-      "learning_rate": 1.5516332982086408e-05,
-      "loss": 0.0981,
+      "learning_rate": 4.354354354354355e-05,
+      "loss": 0.1237,
       "step": 3040
     },
     {
       "epoch": 2.17,
-      "learning_rate": 1.5384615384615387e-05,
-      "loss": 0.0666,
+      "learning_rate": 4.350403034613561e-05,
+      "loss": 0.1249,
       "step": 3050
     },
     {
       "epoch": 2.18,
-      "learning_rate": 1.5252897787144363e-05,
-      "loss": 0.0664,
+      "learning_rate": 4.346451714872768e-05,
+      "loss": 0.1043,
       "step": 3060
     },
     {
       "epoch": 2.18,
-      "learning_rate": 1.5121180189673342e-05,
-      "loss": 0.0918,
+      "learning_rate": 4.3425003951319745e-05,
+      "loss": 0.1029,
       "step": 3070
     },
     {
       "epoch": 2.19,
-      "learning_rate": 1.498946259220232e-05,
-      "loss": 0.0968,
+      "learning_rate": 4.3385490753911804e-05,
+      "loss": 0.1117,
       "step": 3080
     },
     {
       "epoch": 2.2,
-      "learning_rate": 1.4857744994731296e-05,
-      "loss": 0.0992,
+      "learning_rate": 4.334597755650388e-05,
+      "loss": 0.1324,
       "step": 3090
     },
     {
       "epoch": 2.2,
-      "learning_rate": 1.4726027397260275e-05,
-      "loss": 0.0951,
+      "learning_rate": 4.3306464359095936e-05,
+      "loss": 0.1076,
       "step": 3100
     },
     {
       "epoch": 2.21,
-      "learning_rate": 1.4594309799789252e-05,
-      "loss": 0.0966,
+      "learning_rate": 4.326695116168801e-05,
+      "loss": 0.1281,
       "step": 3110
     },
     {
       "epoch": 2.22,
-      "learning_rate": 1.4462592202318232e-05,
-      "loss": 0.0805,
+      "learning_rate": 4.322743796428007e-05,
+      "loss": 0.1039,
       "step": 3120
     },
     {
       "epoch": 2.23,
-      "learning_rate": 1.4330874604847208e-05,
-      "loss": 0.0969,
+      "learning_rate": 4.318792476687214e-05,
+      "loss": 0.1492,
       "step": 3130
     },
     {
       "epoch": 2.23,
-      "learning_rate": 1.4199157007376185e-05,
-      "loss": 0.0749,
+      "learning_rate": 4.3148411569464206e-05,
+      "loss": 0.1167,
       "step": 3140
     },
     {
       "epoch": 2.24,
-      "learning_rate": 1.4067439409905164e-05,
-      "loss": 0.0712,
+      "learning_rate": 4.3108898372056265e-05,
+      "loss": 0.1031,
       "step": 3150
     },
     {
       "epoch": 2.25,
-      "learning_rate": 1.3935721812434142e-05,
-      "loss": 0.068,
+      "learning_rate": 4.306938517464834e-05,
+      "loss": 0.0887,
       "step": 3160
     },
     {
       "epoch": 2.25,
-      "learning_rate": 1.3804004214963121e-05,
-      "loss": 0.1018,
+      "learning_rate": 4.30298719772404e-05,
+      "loss": 0.1235,
       "step": 3170
     },
     {
       "epoch": 2.26,
-      "learning_rate": 1.3672286617492097e-05,
-      "loss": 0.0862,
+      "learning_rate": 4.299035877983247e-05,
+      "loss": 0.1069,
       "step": 3180
     },
     {
       "epoch": 2.27,
-      "learning_rate": 1.3540569020021076e-05,
-      "loss": 0.0855,
+      "learning_rate": 4.295084558242453e-05,
+      "loss": 0.1044,
       "step": 3190
     },
     {
       "epoch": 2.28,
-      "learning_rate": 1.3408851422550054e-05,
-      "loss": 0.0979,
+      "learning_rate": 4.2911332385016595e-05,
+      "loss": 0.1262,
       "step": 3200
     },
     {
       "epoch": 2.28,
-      "learning_rate": 1.327713382507903e-05,
-      "loss": 0.0842,
+      "learning_rate": 4.287181918760867e-05,
+      "loss": 0.1243,
       "step": 3210
     },
     {
       "epoch": 2.29,
-      "learning_rate": 1.3145416227608009e-05,
-      "loss": 0.0732,
+      "learning_rate": 4.2832305990200726e-05,
+      "loss": 0.1174,
       "step": 3220
     },
     {
       "epoch": 2.3,
-      "learning_rate": 1.3013698630136986e-05,
-      "loss": 0.0867,
+      "learning_rate": 4.27927927927928e-05,
+      "loss": 0.1238,
       "step": 3230
     },
     {
       "epoch": 2.3,
-      "learning_rate": 1.2881981032665966e-05,
-      "loss": 0.0735,
+      "learning_rate": 4.275327959538486e-05,
+      "loss": 0.11,
       "step": 3240
     },
     {
       "epoch": 2.31,
-      "learning_rate": 1.2750263435194943e-05,
-      "loss": 0.0785,
+      "learning_rate": 4.271376639797693e-05,
+      "loss": 0.1224,
       "step": 3250
     },
     {
       "epoch": 2.32,
-      "learning_rate": 1.2618545837723922e-05,
-      "loss": 0.0638,
+      "learning_rate": 4.267425320056899e-05,
+      "loss": 0.1092,
       "step": 3260
     },
     {
       "epoch": 2.33,
-      "learning_rate": 1.2486828240252898e-05,
-      "loss": 0.0971,
+      "learning_rate": 4.2634740003161055e-05,
+      "loss": 0.0728,
       "step": 3270
     },
     {
       "epoch": 2.33,
-      "learning_rate": 1.2355110642781877e-05,
-      "loss": 0.093,
+      "learning_rate": 4.259522680575312e-05,
+      "loss": 0.103,
       "step": 3280
     },
     {
       "epoch": 2.34,
-      "learning_rate": 1.2223393045310855e-05,
-      "loss": 0.0736,
+      "learning_rate": 4.255571360834519e-05,
+      "loss": 0.0897,
       "step": 3290
     },
     {
       "epoch": 2.35,
-      "learning_rate": 1.209167544783983e-05,
-      "loss": 0.0776,
+      "learning_rate": 4.251620041093726e-05,
+      "loss": 0.132,
       "step": 3300
     },
     {
       "epoch": 2.35,
-      "learning_rate": 1.195995785036881e-05,
-      "loss": 0.0852,
+      "learning_rate": 4.247668721352932e-05,
+      "loss": 0.0878,
       "step": 3310
     },
     {
       "epoch": 2.36,
-      "learning_rate": 1.1828240252897788e-05,
-      "loss": 0.1035,
+      "learning_rate": 4.243717401612139e-05,
+      "loss": 0.1151,
       "step": 3320
     },
     {
       "epoch": 2.37,
-      "learning_rate": 1.1696522655426765e-05,
-      "loss": 0.0769,
+      "learning_rate": 4.239766081871345e-05,
+      "loss": 0.1183,
       "step": 3330
     },
     {
       "epoch": 2.38,
-      "learning_rate": 1.1564805057955744e-05,
-      "loss": 0.0953,
+      "learning_rate": 4.2358147621305516e-05,
+      "loss": 0.1016,
       "step": 3340
     },
     {
       "epoch": 2.38,
-      "learning_rate": 1.1433087460484722e-05,
-      "loss": 0.0646,
+      "learning_rate": 4.231863442389758e-05,
+      "loss": 0.1043,
       "step": 3350
     },
     {
       "epoch": 2.39,
-      "learning_rate": 1.1301369863013698e-05,
-      "loss": 0.0785,
+      "learning_rate": 4.227912122648965e-05,
+      "loss": 0.1006,
       "step": 3360
     },
     {
       "epoch": 2.4,
-      "learning_rate": 1.1169652265542677e-05,
-      "loss": 0.0738,
+      "learning_rate": 4.223960802908172e-05,
+      "loss": 0.1159,
       "step": 3370
     },
     {
       "epoch": 2.4,
-      "learning_rate": 1.1037934668071655e-05,
-      "loss": 0.0638,
+      "learning_rate": 4.220009483167378e-05,
+      "loss": 0.1107,
       "step": 3380
     },
     {
       "epoch": 2.41,
-      "learning_rate": 1.0906217070600632e-05,
-      "loss": 0.0948,
+      "learning_rate": 4.2160581634265846e-05,
+      "loss": 0.0833,
       "step": 3390
     },
     {
       "epoch": 2.42,
-      "learning_rate": 1.0774499473129611e-05,
-      "loss": 0.1331,
+      "learning_rate": 4.212106843685791e-05,
+      "loss": 0.0788,
       "step": 3400
     },
     {
       "epoch": 2.42,
-      "learning_rate": 1.0642781875658589e-05,
-      "loss": 0.1084,
+      "learning_rate": 4.208155523944998e-05,
+      "loss": 0.0945,
       "step": 3410
     },
     {
       "epoch": 2.43,
-      "learning_rate": 1.0511064278187566e-05,
-      "loss": 0.0919,
+      "learning_rate": 4.204204204204204e-05,
+      "loss": 0.108,
       "step": 3420
     },
     {
       "epoch": 2.44,
-      "learning_rate": 1.0379346680716544e-05,
-      "loss": 0.0666,
+      "learning_rate": 4.200252884463411e-05,
+      "loss": 0.0832,
       "step": 3430
     },
     {
       "epoch": 2.45,
-      "learning_rate": 1.0247629083245521e-05,
-      "loss": 0.0911,
+      "learning_rate": 4.1963015647226175e-05,
+      "loss": 0.1148,
       "step": 3440
     },
     {
       "epoch": 2.45,
-      "learning_rate": 1.01159114857745e-05,
-      "loss": 0.0706,
+      "learning_rate": 4.192350244981824e-05,
+      "loss": 0.1272,
       "step": 3450
     },
     {
       "epoch": 2.46,
-      "learning_rate": 9.984193888303478e-06,
-      "loss": 0.1086,
+      "learning_rate": 4.1883989252410307e-05,
+      "loss": 0.1329,
       "step": 3460
     },
     {
       "epoch": 2.47,
-      "learning_rate": 9.852476290832456e-06,
-      "loss": 0.0729,
+      "learning_rate": 4.184447605500237e-05,
+      "loss": 0.0954,
       "step": 3470
     },
     {
       "epoch": 2.47,
-      "learning_rate": 9.720758693361433e-06,
-      "loss": 0.1097,
+      "learning_rate": 4.180496285759444e-05,
+      "loss": 0.0966,
       "step": 3480
     },
     {
       "epoch": 2.48,
-      "learning_rate": 9.589041095890411e-06,
-      "loss": 0.0737,
+      "learning_rate": 4.1765449660186504e-05,
+      "loss": 0.1018,
       "step": 3490
     },
     {
       "epoch": 2.49,
-      "learning_rate": 9.457323498419388e-06,
-      "loss": 0.0839,
+      "learning_rate": 4.172593646277857e-05,
+      "loss": 0.1111,
       "step": 3500
     },
     {
       "epoch": 2.5,
-      "learning_rate": 9.325605900948368e-06,
-      "loss": 0.1214,
+      "learning_rate": 4.1686423265370636e-05,
+      "loss": 0.0898,
       "step": 3510
     },
     {
       "epoch": 2.5,
-      "learning_rate": 9.193888303477345e-06,
-      "loss": 0.1008,
+      "learning_rate": 4.16469100679627e-05,
+      "loss": 0.1106,
       "step": 3520
     },
     {
       "epoch": 2.51,
-      "learning_rate": 9.062170706006323e-06,
-      "loss": 0.0774,
+      "learning_rate": 4.160739687055477e-05,
+      "loss": 0.1366,
       "step": 3530
     },
     {
       "epoch": 2.52,
-      "learning_rate": 8.930453108535302e-06,
-      "loss": 0.0631,
+      "learning_rate": 4.156788367314683e-05,
+      "loss": 0.0975,
       "step": 3540
     },
     {
       "epoch": 2.52,
-      "learning_rate": 8.798735511064278e-06,
-      "loss": 0.0705,
+      "learning_rate": 4.15283704757389e-05,
+      "loss": 0.0902,
       "step": 3550
     },
     {
       "epoch": 2.53,
-      "learning_rate": 8.667017913593255e-06,
-      "loss": 0.1071,
+      "learning_rate": 4.1488857278330965e-05,
+      "loss": 0.1274,
       "step": 3560
     },
     {
       "epoch": 2.54,
-      "learning_rate": 8.535300316122235e-06,
-      "loss": 0.0645,
+      "learning_rate": 4.144934408092303e-05,
+      "loss": 0.1242,
       "step": 3570
     },
     {
       "epoch": 2.55,
-      "learning_rate": 8.403582718651212e-06,
-      "loss": 0.0713,
+      "learning_rate": 4.14098308835151e-05,
+      "loss": 0.1585,
       "step": 3580
     },
     {
       "epoch": 2.55,
-      "learning_rate": 8.27186512118019e-06,
-      "loss": 0.0961,
+      "learning_rate": 4.137031768610716e-05,
+      "loss": 0.1124,
       "step": 3590
     },
     {
       "epoch": 2.56,
-      "learning_rate": 8.140147523709169e-06,
-      "loss": 0.0805,
+      "learning_rate": 4.133080448869923e-05,
+      "loss": 0.1349,
       "step": 3600
     },
     {
       "epoch": 2.57,
-      "learning_rate": 8.008429926238145e-06,
-      "loss": 0.0885,
+      "learning_rate": 4.1291291291291294e-05,
+      "loss": 0.1257,
       "step": 3610
     },
     {
       "epoch": 2.57,
-      "learning_rate": 7.876712328767124e-06,
-      "loss": 0.0623,
+      "learning_rate": 4.125177809388336e-05,
+      "loss": 0.1136,
       "step": 3620
     },
     {
       "epoch": 2.58,
-      "learning_rate": 7.744994731296102e-06,
-      "loss": 0.1018,
+      "learning_rate": 4.121226489647542e-05,
+      "loss": 0.125,
       "step": 3630
     },
     {
       "epoch": 2.59,
-      "learning_rate": 7.613277133825079e-06,
-      "loss": 0.0997,
+      "learning_rate": 4.117275169906749e-05,
+      "loss": 0.0887,
       "step": 3640
     },
     {
       "epoch": 2.6,
-      "learning_rate": 7.4815595363540575e-06,
-      "loss": 0.0845,
+      "learning_rate": 4.113323850165956e-05,
+      "loss": 0.0922,
       "step": 3650
     },
     {
       "epoch": 2.6,
-      "learning_rate": 7.349841938883036e-06,
-      "loss": 0.0419,
+      "learning_rate": 4.1093725304251623e-05,
+      "loss": 0.1087,
       "step": 3660
     },
     {
       "epoch": 2.61,
-      "learning_rate": 7.218124341412013e-06,
-      "loss": 0.0906,
+      "learning_rate": 4.105421210684369e-05,
+      "loss": 0.1114,
       "step": 3670
     },
     {
       "epoch": 2.62,
-      "learning_rate": 7.08640674394099e-06,
-      "loss": 0.0784,
+      "learning_rate": 4.1014698909435755e-05,
+      "loss": 0.1162,
       "step": 3680
     },
     {
       "epoch": 2.62,
-      "learning_rate": 6.9546891464699685e-06,
-      "loss": 0.0744,
+      "learning_rate": 4.097518571202782e-05,
+      "loss": 0.1538,
       "step": 3690
     },
     {
       "epoch": 2.63,
-      "learning_rate": 6.822971548998947e-06,
-      "loss": 0.0621,
+      "learning_rate": 4.093567251461988e-05,
+      "loss": 0.1118,
       "step": 3700
     },
     {
       "epoch": 2.64,
-      "learning_rate": 6.6912539515279244e-06,
-      "loss": 0.0877,
+      "learning_rate": 4.089615931721195e-05,
+      "loss": 0.1053,
       "step": 3710
     },
     {
       "epoch": 2.65,
-      "learning_rate": 6.559536354056903e-06,
-      "loss": 0.0639,
+      "learning_rate": 4.085664611980402e-05,
+      "loss": 0.1349,
       "step": 3720
     },
     {
       "epoch": 2.65,
-      "learning_rate": 6.42781875658588e-06,
-      "loss": 0.0977,
+      "learning_rate": 4.0817132922396084e-05,
+      "loss": 0.1274,
       "step": 3730
     },
     {
       "epoch": 2.66,
-      "learning_rate": 6.296101159114858e-06,
-      "loss": 0.0833,
+      "learning_rate": 4.077761972498815e-05,
+      "loss": 0.1044,
       "step": 3740
     },
     {
       "epoch": 2.67,
-      "learning_rate": 6.1643835616438354e-06,
-      "loss": 0.0595,
+      "learning_rate": 4.0738106527580216e-05,
+      "loss": 0.0914,
       "step": 3750
     },
     {
       "epoch": 2.67,
-      "learning_rate": 6.032665964172814e-06,
-      "loss": 0.0697,
+      "learning_rate": 4.069859333017228e-05,
+      "loss": 0.086,
       "step": 3760
     },
     {
       "epoch": 2.68,
-      "learning_rate": 5.900948366701791e-06,
-      "loss": 0.0869,
+      "learning_rate": 4.065908013276434e-05,
+      "loss": 0.1263,
       "step": 3770
     },
     {
       "epoch": 2.69,
-      "learning_rate": 5.76923076923077e-06,
-      "loss": 0.0671,
+      "learning_rate": 4.0619566935356414e-05,
+      "loss": 0.1042,
       "step": 3780
     },
     {
       "epoch": 2.7,
-      "learning_rate": 5.637513171759747e-06,
-      "loss": 0.0699,
+      "learning_rate": 4.058005373794847e-05,
+      "loss": 0.0893,
       "step": 3790
     },
     {
       "epoch": 2.7,
-      "learning_rate": 5.505795574288726e-06,
-      "loss": 0.0711,
+      "learning_rate": 4.0540540540540545e-05,
+      "loss": 0.0991,
       "step": 3800
     },
     {
       "epoch": 2.71,
-      "learning_rate": 5.374077976817703e-06,
-      "loss": 0.0665,
+      "learning_rate": 4.050102734313261e-05,
+      "loss": 0.1228,
       "step": 3810
     },
     {
       "epoch": 2.72,
-      "learning_rate": 5.242360379346681e-06,
-      "loss": 0.0904,
+      "learning_rate": 4.046151414572467e-05,
+      "loss": 0.0995,
       "step": 3820
     },
     {
       "epoch": 2.72,
-      "learning_rate": 5.110642781875659e-06,
-      "loss": 0.0876,
+      "learning_rate": 4.042200094831674e-05,
+      "loss": 0.0887,
       "step": 3830
     },
     {
       "epoch": 2.73,
-      "learning_rate": 4.978925184404637e-06,
-      "loss": 0.0891,
+      "learning_rate": 4.03824877509088e-05,
+      "loss": 0.0963,
       "step": 3840
     },
     {
       "epoch": 2.74,
-      "learning_rate": 4.847207586933614e-06,
-      "loss": 0.0819,
+      "learning_rate": 4.0342974553500874e-05,
+      "loss": 0.1189,
       "step": 3850
     },
     {
       "epoch": 2.74,
-      "learning_rate": 4.715489989462593e-06,
-      "loss": 0.0673,
+      "learning_rate": 4.0303461356092934e-05,
+      "loss": 0.1028,
       "step": 3860
     },
     {
       "epoch": 2.75,
-      "learning_rate": 4.58377239199157e-06,
-      "loss": 0.0519,
+      "learning_rate": 4.0263948158685006e-05,
+      "loss": 0.1343,
       "step": 3870
     },
     {
       "epoch": 2.76,
-      "learning_rate": 4.452054794520548e-06,
-      "loss": 0.087,
+      "learning_rate": 4.022443496127707e-05,
+      "loss": 0.134,
       "step": 3880
     },
     {
       "epoch": 2.77,
-      "learning_rate": 4.320337197049526e-06,
-      "loss": 0.0651,
+      "learning_rate": 4.018492176386913e-05,
+      "loss": 0.0978,
       "step": 3890
     },
     {
       "epoch": 2.77,
-      "learning_rate": 4.188619599578504e-06,
-      "loss": 0.0633,
+      "learning_rate": 4.0145408566461204e-05,
+      "loss": 0.1245,
       "step": 3900
     },
     {
       "epoch": 2.78,
-      "learning_rate": 4.056902002107481e-06,
-      "loss": 0.0846,
+      "learning_rate": 4.010589536905326e-05,
+      "loss": 0.1202,
       "step": 3910
     },
     {
       "epoch": 2.79,
-      "learning_rate": 3.92518440463646e-06,
-      "loss": 0.0721,
+      "learning_rate": 4.0066382171645335e-05,
+      "loss": 0.0695,
       "step": 3920
     },
     {
       "epoch": 2.79,
-      "learning_rate": 3.7934668071654376e-06,
-      "loss": 0.1049,
+      "learning_rate": 4.0026868974237394e-05,
+      "loss": 0.0789,
       "step": 3930
     },
     {
       "epoch": 2.8,
-      "learning_rate": 3.661749209694415e-06,
-      "loss": 0.1,
+      "learning_rate": 3.998735577682946e-05,
+      "loss": 0.0869,
       "step": 3940
     },
     {
       "epoch": 2.81,
-      "learning_rate": 3.530031612223393e-06,
-      "loss": 0.086,
+      "learning_rate": 3.9947842579421526e-05,
+      "loss": 0.1131,
       "step": 3950
     },
     {
       "epoch": 2.82,
-      "learning_rate": 3.398314014752371e-06,
-      "loss": 0.0802,
+      "learning_rate": 3.990832938201359e-05,
+      "loss": 0.0855,
       "step": 3960
     },
     {
       "epoch": 2.82,
-      "learning_rate": 3.2665964172813486e-06,
-      "loss": 0.0789,
+      "learning_rate": 3.9868816184605665e-05,
+      "loss": 0.1407,
       "step": 3970
     },
     {
       "epoch": 2.83,
-      "learning_rate": 3.1348788198103265e-06,
-      "loss": 0.0612,
+      "learning_rate": 3.9829302987197724e-05,
+      "loss": 0.1111,
       "step": 3980
     },
     {
       "epoch": 2.84,
-      "learning_rate": 3.003161222339305e-06,
-      "loss": 0.0697,
+      "learning_rate": 3.9789789789789796e-05,
+      "loss": 0.1307,
       "step": 3990
     },
     {
       "epoch": 2.84,
-      "learning_rate": 2.8714436248682825e-06,
-      "loss": 0.0862,
+      "learning_rate": 3.9750276592381855e-05,
+      "loss": 0.0966,
       "step": 4000
     },
     {
       "epoch": 2.85,
-      "learning_rate": 2.7397260273972604e-06,
-      "loss": 0.0541,
+      "learning_rate": 3.971076339497392e-05,
+      "loss": 0.0976,
       "step": 4010
     },
     {
       "epoch": 2.86,
-      "learning_rate": 2.6080084299262384e-06,
-      "loss": 0.0559,
+      "learning_rate": 3.967125019756599e-05,
+      "loss": 0.0978,
       "step": 4020
     },
     {
       "epoch": 2.87,
-      "learning_rate": 2.476290832455216e-06,
-      "loss": 0.0611,
+      "learning_rate": 3.963173700015805e-05,
+      "loss": 0.1109,
       "step": 4030
     },
     {
       "epoch": 2.87,
-      "learning_rate": 2.3445732349841943e-06,
-      "loss": 0.0657,
+      "learning_rate": 3.9592223802750125e-05,
+      "loss": 0.0837,
       "step": 4040
     },
     {
       "epoch": 2.88,
-      "learning_rate": 2.212855637513172e-06,
-      "loss": 0.049,
+      "learning_rate": 3.9552710605342185e-05,
+      "loss": 0.0762,
       "step": 4050
     },
     {
       "epoch": 2.89,
-      "learning_rate": 2.08113804004215e-06,
-      "loss": 0.0733,
+      "learning_rate": 3.951319740793425e-05,
+      "loss": 0.0911,
       "step": 4060
     },
     {
       "epoch": 2.89,
-      "learning_rate": 1.949420442571128e-06,
-      "loss": 0.0608,
+      "learning_rate": 3.9473684210526316e-05,
+      "loss": 0.0966,
       "step": 4070
     },
     {
       "epoch": 2.9,
-      "learning_rate": 1.8177028451001056e-06,
-      "loss": 0.0718,
+      "learning_rate": 3.943417101311838e-05,
+      "loss": 0.1033,
       "step": 4080
     },
     {
       "epoch": 2.91,
-      "learning_rate": 1.6859852476290833e-06,
-      "loss": 0.0728,
+      "learning_rate": 3.939465781571045e-05,
+      "loss": 0.094,
       "step": 4090
     },
     {
       "epoch": 2.92,
-      "learning_rate": 1.554267650158061e-06,
-      "loss": 0.0627,
+      "learning_rate": 3.9355144618302514e-05,
+      "loss": 0.081,
       "step": 4100
     },
     {
       "epoch": 2.92,
-      "learning_rate": 1.422550052687039e-06,
-      "loss": 0.0953,
+      "learning_rate": 3.9315631420894586e-05,
+      "loss": 0.1094,
       "step": 4110
     },
     {
       "epoch": 2.93,
-      "learning_rate": 1.290832455216017e-06,
-      "loss": 0.0794,
+      "learning_rate": 3.9276118223486645e-05,
+      "loss": 0.0941,
       "step": 4120
     },
     {
       "epoch": 2.94,
-      "learning_rate": 1.1591148577449948e-06,
-      "loss": 0.0844,
+      "learning_rate": 3.923660502607871e-05,
+      "loss": 0.1206,
       "step": 4130
     },
     {
       "epoch": 2.94,
-      "learning_rate": 1.0273972602739725e-06,
-      "loss": 0.0601,
+      "learning_rate": 3.919709182867078e-05,
+      "loss": 0.1134,
       "step": 4140
     },
     {
       "epoch": 2.95,
-      "learning_rate": 8.956796628029505e-07,
-      "loss": 0.0775,
+      "learning_rate": 3.915757863126284e-05,
+      "loss": 0.1231,
       "step": 4150
     },
     {
       "epoch": 2.96,
-      "learning_rate": 7.639620653319284e-07,
-      "loss": 0.0666,
+      "learning_rate": 3.911806543385491e-05,
+      "loss": 0.106,
       "step": 4160
     },
     {
       "epoch": 2.97,
-      "learning_rate": 6.322444678609063e-07,
-      "loss": 0.0676,
+      "learning_rate": 3.9078552236446975e-05,
+      "loss": 0.1186,
       "step": 4170
     },
     {
       "epoch": 2.97,
-      "learning_rate": 5.00526870389884e-07,
-      "loss": 0.1339,
+      "learning_rate": 3.903903903903904e-05,
+      "loss": 0.0887,
       "step": 4180
     },
     {
       "epoch": 2.98,
-      "learning_rate": 3.6880927291886196e-07,
-      "loss": 0.0614,
+      "learning_rate": 3.8999525841631106e-05,
+      "loss": 0.0892,
       "step": 4190
     },
     {
       "epoch": 2.99,
-      "learning_rate": 2.3709167544783985e-07,
-      "loss": 0.092,
+      "learning_rate": 3.896001264422317e-05,
+      "loss": 0.101,
       "step": 4200
     },
     {
       "epoch": 2.99,
-      "learning_rate": 1.0537407797681771e-07,
-      "loss": 0.086,
+      "learning_rate": 3.892049944681524e-05,
+      "loss": 0.0952,
       "step": 4210
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9901,
-      "eval_loss": 0.03231855481863022,
-      "eval_runtime": 133.5993,
-      "eval_samples_per_second": 74.851,
-      "eval_steps_per_second": 4.678,
+      "eval_accuracy": 0.9866,
+      "eval_loss": 0.03929213061928749,
+      "eval_runtime": 47.224,
+      "eval_samples_per_second": 211.757,
+      "eval_steps_per_second": 13.235,
       "step": 4218
     },
     {
       "epoch": 3.0,
-      "step": 4218,
-      "total_flos": 2.091911758043834e+19,
-      "train_loss": 0.13970869080807824,
-      "train_runtime": 10165.3024,
-      "train_samples_per_second": 26.561,
-      "train_steps_per_second": 0.415
+      "learning_rate": 3.8880986249407304e-05,
+      "loss": 0.1122,
+      "step": 4220
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.884147305199937e-05,
+      "loss": 0.0891,
+      "step": 4230
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 3.8801959854591436e-05,
+      "loss": 0.0916,
+      "step": 4240
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 3.87624466571835e-05,
+      "loss": 0.0918,
+      "step": 4250
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 3.872293345977557e-05,
+      "loss": 0.1045,
+      "step": 4260
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 3.868342026236763e-05,
+      "loss": 0.1045,
+      "step": 4270
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 3.86439070649597e-05,
+      "loss": 0.1048,
+      "step": 4280
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 3.8604393867551765e-05,
+      "loss": 0.0775,
+      "step": 4290
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 3.856488067014383e-05,
+      "loss": 0.0826,
+      "step": 4300
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 3.8525367472735897e-05,
+      "loss": 0.0963,
+      "step": 4310
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 3.848585427532796e-05,
+      "loss": 0.0746,
+      "step": 4320
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 3.844634107792003e-05,
+      "loss": 0.0766,
+      "step": 4330
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 3.8406827880512094e-05,
+      "loss": 0.0796,
+      "step": 4340
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 3.836731468310416e-05,
+      "loss": 0.0986,
+      "step": 4350
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 3.8327801485696226e-05,
+      "loss": 0.0758,
+      "step": 4360
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 3.8288288288288285e-05,
+      "loss": 0.0918,
+      "step": 4370
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 3.824877509088036e-05,
+      "loss": 0.0934,
+      "step": 4380
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 3.820926189347242e-05,
+      "loss": 0.1022,
+      "step": 4390
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 3.816974869606449e-05,
+      "loss": 0.1022,
+      "step": 4400
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 3.8130235498656555e-05,
+      "loss": 0.1082,
+      "step": 4410
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 3.809072230124862e-05,
+      "loss": 0.0822,
+      "step": 4420
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 3.805120910384069e-05,
+      "loss": 0.1129,
+      "step": 4430
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 3.8011695906432746e-05,
+      "loss": 0.1,
+      "step": 4440
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 3.797218270902482e-05,
+      "loss": 0.0807,
+      "step": 4450
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 3.793266951161688e-05,
+      "loss": 0.1072,
+      "step": 4460
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 3.789315631420895e-05,
+      "loss": 0.0561,
+      "step": 4470
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 3.7853643116801016e-05,
+      "loss": 0.1078,
+      "step": 4480
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 3.7814129919393075e-05,
+      "loss": 0.0851,
+      "step": 4490
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 3.777461672198515e-05,
+      "loss": 0.1027,
+      "step": 4500
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 3.773510352457721e-05,
+      "loss": 0.0823,
+      "step": 4510
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 3.769559032716928e-05,
+      "loss": 0.0732,
+      "step": 4520
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 3.765607712976134e-05,
+      "loss": 0.0994,
+      "step": 4530
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 3.761656393235341e-05,
+      "loss": 0.1032,
+      "step": 4540
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 3.757705073494548e-05,
+      "loss": 0.0924,
+      "step": 4550
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 3.7537537537537536e-05,
+      "loss": 0.0803,
+      "step": 4560
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 3.749802434012961e-05,
+      "loss": 0.0757,
+      "step": 4570
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 3.745851114272167e-05,
+      "loss": 0.0747,
+      "step": 4580
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 3.741899794531374e-05,
+      "loss": 0.0798,
+      "step": 4590
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 3.73794847479058e-05,
+      "loss": 0.0887,
+      "step": 4600
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 3.7339971550497865e-05,
+      "loss": 0.0975,
+      "step": 4610
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 3.730045835308993e-05,
+      "loss": 0.0873,
+      "step": 4620
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 3.7260945155682e-05,
+      "loss": 0.1188,
+      "step": 4630
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 3.722143195827407e-05,
+      "loss": 0.0884,
+      "step": 4640
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 3.718191876086613e-05,
+      "loss": 0.0999,
+      "step": 4650
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 3.71424055634582e-05,
+      "loss": 0.1047,
+      "step": 4660
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 3.710289236605026e-05,
+      "loss": 0.081,
+      "step": 4670
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 3.7063379168642326e-05,
+      "loss": 0.1047,
+      "step": 4680
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 3.702386597123439e-05,
+      "loss": 0.1071,
+      "step": 4690
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 3.698435277382646e-05,
+      "loss": 0.0926,
+      "step": 4700
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 3.694483957641853e-05,
+      "loss": 0.1022,
+      "step": 4710
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 3.690532637901059e-05,
+      "loss": 0.0997,
+      "step": 4720
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 3.686581318160266e-05,
+      "loss": 0.0984,
+      "step": 4730
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 3.682629998419472e-05,
+      "loss": 0.0726,
+      "step": 4740
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 3.678678678678679e-05,
+      "loss": 0.0868,
+      "step": 4750
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 3.674727358937885e-05,
+      "loss": 0.079,
+      "step": 4760
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 3.670776039197092e-05,
+      "loss": 0.0658,
+      "step": 4770
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 3.666824719456299e-05,
+      "loss": 0.1381,
+      "step": 4780
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 3.662873399715505e-05,
+      "loss": 0.0849,
+      "step": 4790
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 3.6589220799747116e-05,
+      "loss": 0.0859,
+      "step": 4800
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 3.654970760233918e-05,
+      "loss": 0.1095,
+      "step": 4810
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 3.651019440493125e-05,
+      "loss": 0.0631,
+      "step": 4820
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 3.6470681207523314e-05,
+      "loss": 0.0819,
+      "step": 4830
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 3.643116801011538e-05,
+      "loss": 0.0856,
+      "step": 4840
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 3.6391654812707445e-05,
+      "loss": 0.0906,
+      "step": 4850
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 3.635214161529951e-05,
+      "loss": 0.1255,
+      "step": 4860
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 3.631262841789158e-05,
+      "loss": 0.0781,
+      "step": 4870
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 3.627311522048364e-05,
+      "loss": 0.0984,
+      "step": 4880
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 3.623360202307571e-05,
+      "loss": 0.0811,
+      "step": 4890
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 3.6194088825667775e-05,
+      "loss": 0.0748,
+      "step": 4900
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 3.615457562825984e-05,
+      "loss": 0.0807,
+      "step": 4910
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 3.6115062430851906e-05,
+      "loss": 0.1111,
+      "step": 4920
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 3.607554923344397e-05,
+      "loss": 0.1307,
+      "step": 4930
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 3.603603603603604e-05,
+      "loss": 0.1032,
+      "step": 4940
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 3.5996522838628104e-05,
+      "loss": 0.0957,
+      "step": 4950
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 3.595700964122017e-05,
+      "loss": 0.0788,
+      "step": 4960
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 3.5917496443812235e-05,
+      "loss": 0.098,
+      "step": 4970
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 3.58779832464043e-05,
+      "loss": 0.0912,
+      "step": 4980
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 3.583847004899637e-05,
+      "loss": 0.0651,
+      "step": 4990
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 3.579895685158843e-05,
+      "loss": 0.0789,
+      "step": 5000
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 3.57594436541805e-05,
+      "loss": 0.0838,
+      "step": 5010
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 3.5719930456772565e-05,
+      "loss": 0.0902,
+      "step": 5020
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 3.568041725936463e-05,
+      "loss": 0.0834,
+      "step": 5030
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 3.564090406195669e-05,
+      "loss": 0.081,
+      "step": 5040
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 3.560139086454876e-05,
+      "loss": 0.0911,
+      "step": 5050
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 3.556187766714083e-05,
+      "loss": 0.0834,
+      "step": 5060
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 3.5522364469732894e-05,
+      "loss": 0.0749,
+      "step": 5070
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 3.548285127232496e-05,
+      "loss": 0.0767,
+      "step": 5080
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 3.5443338074917026e-05,
+      "loss": 0.0996,
+      "step": 5090
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 3.540382487750909e-05,
+      "loss": 0.0914,
+      "step": 5100
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 3.536431168010115e-05,
+      "loss": 0.0836,
+      "step": 5110
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 3.532479848269322e-05,
+      "loss": 0.1035,
+      "step": 5120
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 3.528528528528528e-05,
+      "loss": 0.0721,
+      "step": 5130
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 3.5245772087877355e-05,
+      "loss": 0.1058,
+      "step": 5140
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 3.520625889046942e-05,
+      "loss": 0.1103,
+      "step": 5150
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 3.5166745693061487e-05,
+      "loss": 0.0879,
+      "step": 5160
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 3.512723249565355e-05,
+      "loss": 0.1053,
+      "step": 5170
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 3.508771929824561e-05,
+      "loss": 0.0802,
+      "step": 5180
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 3.5048206100837684e-05,
+      "loss": 0.0789,
+      "step": 5190
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 3.500869290342974e-05,
+      "loss": 0.0882,
+      "step": 5200
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 3.4969179706021816e-05,
+      "loss": 0.1045,
+      "step": 5210
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 3.492966650861388e-05,
+      "loss": 0.0927,
+      "step": 5220
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 3.489015331120594e-05,
+      "loss": 0.0997,
+      "step": 5230
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 3.485064011379801e-05,
+      "loss": 0.1027,
+      "step": 5240
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 3.481112691639007e-05,
+      "loss": 0.092,
+      "step": 5250
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 3.4771613718982145e-05,
+      "loss": 0.1051,
+      "step": 5260
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 3.4732100521574204e-05,
+      "loss": 0.0802,
+      "step": 5270
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 3.469258732416628e-05,
+      "loss": 0.1104,
+      "step": 5280
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 3.465307412675834e-05,
+      "loss": 0.0839,
+      "step": 5290
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 3.46135609293504e-05,
+      "loss": 0.0744,
+      "step": 5300
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 3.4574047731942474e-05,
+      "loss": 0.0745,
+      "step": 5310
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 3.453453453453453e-05,
+      "loss": 0.0929,
+      "step": 5320
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 3.4495021337126606e-05,
+      "loss": 0.0858,
+      "step": 5330
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 3.4455508139718665e-05,
+      "loss": 0.0799,
+      "step": 5340
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 3.441599494231073e-05,
+      "loss": 0.087,
+      "step": 5350
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 3.43764817449028e-05,
+      "loss": 0.0659,
+      "step": 5360
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 3.433696854749486e-05,
+      "loss": 0.0949,
+      "step": 5370
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 3.4297455350086935e-05,
+      "loss": 0.0959,
+      "step": 5380
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 3.4257942152678994e-05,
+      "loss": 0.0623,
+      "step": 5390
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 3.421842895527107e-05,
+      "loss": 0.1071,
+      "step": 5400
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 3.4178915757863126e-05,
+      "loss": 0.1068,
+      "step": 5410
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 3.413940256045519e-05,
+      "loss": 0.0902,
+      "step": 5420
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 3.409988936304726e-05,
+      "loss": 0.091,
+      "step": 5430
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 3.4060376165639323e-05,
+      "loss": 0.1093,
+      "step": 5440
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 3.4020862968231396e-05,
+      "loss": 0.1063,
+      "step": 5450
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 3.3981349770823455e-05,
+      "loss": 0.0608,
+      "step": 5460
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 3.394183657341552e-05,
+      "loss": 0.1039,
+      "step": 5470
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 3.390232337600759e-05,
+      "loss": 0.0801,
+      "step": 5480
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 3.386281017859965e-05,
+      "loss": 0.0907,
+      "step": 5490
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 3.382329698119172e-05,
+      "loss": 0.1046,
+      "step": 5500
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 3.3783783783783784e-05,
+      "loss": 0.0935,
+      "step": 5510
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 3.374427058637585e-05,
+      "loss": 0.107,
+      "step": 5520
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 3.3704757388967916e-05,
+      "loss": 0.1194,
+      "step": 5530
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 3.366524419155998e-05,
+      "loss": 0.0943,
+      "step": 5540
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 3.362573099415205e-05,
+      "loss": 0.1067,
+      "step": 5550
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 3.3586217796744114e-05,
+      "loss": 0.0751,
+      "step": 5560
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 3.354670459933618e-05,
+      "loss": 0.0813,
+      "step": 5570
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 3.3507191401928245e-05,
+      "loss": 0.0844,
+      "step": 5580
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 3.346767820452031e-05,
+      "loss": 0.0691,
+      "step": 5590
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 3.342816500711238e-05,
+      "loss": 0.0996,
+      "step": 5600
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 3.338865180970444e-05,
+      "loss": 0.1017,
+      "step": 5610
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.334913861229651e-05,
+      "loss": 0.0743,
+      "step": 5620
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9866,
+      "eval_loss": 0.04100377485156059,
+      "eval_runtime": 46.9117,
+      "eval_samples_per_second": 213.166,
+      "eval_steps_per_second": 13.323,
+      "step": 5625
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.3309625414888574e-05,
+      "loss": 0.0728,
+      "step": 5630
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 3.327011221748064e-05,
+      "loss": 0.0633,
+      "step": 5640
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 3.3230599020072706e-05,
+      "loss": 0.1061,
+      "step": 5650
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 3.319108582266477e-05,
+      "loss": 0.077,
+      "step": 5660
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.315157262525684e-05,
+      "loss": 0.0922,
+      "step": 5670
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 3.3112059427848904e-05,
+      "loss": 0.103,
+      "step": 5680
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 3.307254623044097e-05,
+      "loss": 0.0772,
+      "step": 5690
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 3.3033033033033035e-05,
+      "loss": 0.1065,
+      "step": 5700
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.29935198356251e-05,
+      "loss": 0.077,
+      "step": 5710
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 3.295400663821717e-05,
+      "loss": 0.0862,
+      "step": 5720
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 3.291449344080923e-05,
+      "loss": 0.088,
+      "step": 5730
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.28749802434013e-05,
+      "loss": 0.0672,
+      "step": 5740
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.2835467045993365e-05,
+      "loss": 0.0677,
+      "step": 5750
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.279595384858543e-05,
+      "loss": 0.1157,
+      "step": 5760
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.2756440651177496e-05,
+      "loss": 0.0869,
+      "step": 5770
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 3.2716927453769555e-05,
+      "loss": 0.0622,
+      "step": 5780
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 3.267741425636163e-05,
+      "loss": 0.0911,
+      "step": 5790
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 3.2637901058953694e-05,
+      "loss": 0.099,
+      "step": 5800
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 3.259838786154576e-05,
+      "loss": 0.1056,
+      "step": 5810
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 3.2558874664137825e-05,
+      "loss": 0.0808,
+      "step": 5820
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 3.251936146672989e-05,
+      "loss": 0.0563,
+      "step": 5830
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 3.247984826932196e-05,
+      "loss": 0.0829,
+      "step": 5840
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 3.2440335071914016e-05,
+      "loss": 0.0967,
+      "step": 5850
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 3.240082187450609e-05,
+      "loss": 0.0896,
+      "step": 5860
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 3.236130867709815e-05,
+      "loss": 0.0942,
+      "step": 5870
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 3.232179547969022e-05,
+      "loss": 0.1141,
+      "step": 5880
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 3.2282282282282286e-05,
+      "loss": 0.0837,
+      "step": 5890
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 3.2242769084874345e-05,
+      "loss": 0.1076,
+      "step": 5900
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 3.220325588746642e-05,
+      "loss": 0.0751,
+      "step": 5910
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 3.216374269005848e-05,
+      "loss": 0.0594,
+      "step": 5920
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 3.212422949265055e-05,
+      "loss": 0.0634,
+      "step": 5930
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 3.208471629524261e-05,
+      "loss": 0.0795,
+      "step": 5940
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 3.204520309783468e-05,
+      "loss": 0.0732,
+      "step": 5950
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 3.200568990042675e-05,
+      "loss": 0.0787,
+      "step": 5960
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 3.1966176703018806e-05,
+      "loss": 0.1017,
+      "step": 5970
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 3.192666350561088e-05,
+      "loss": 0.0746,
+      "step": 5980
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 3.188715030820294e-05,
+      "loss": 0.0703,
+      "step": 5990
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 3.184763711079501e-05,
+      "loss": 0.0426,
+      "step": 6000
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 3.180812391338707e-05,
+      "loss": 0.0722,
+      "step": 6010
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 3.176861071597914e-05,
+      "loss": 0.0627,
+      "step": 6020
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 3.17290975185712e-05,
+      "loss": 0.0825,
+      "step": 6030
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 3.168958432116327e-05,
+      "loss": 0.085,
+      "step": 6040
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 3.165007112375534e-05,
+      "loss": 0.1109,
+      "step": 6050
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 3.16105579263474e-05,
+      "loss": 0.0854,
+      "step": 6060
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 3.157104472893947e-05,
+      "loss": 0.0747,
+      "step": 6070
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 3.153153153153153e-05,
+      "loss": 0.0662,
+      "step": 6080
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 3.1492018334123597e-05,
+      "loss": 0.0766,
+      "step": 6090
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 3.145250513671566e-05,
+      "loss": 0.078,
+      "step": 6100
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 3.141299193930773e-05,
+      "loss": 0.0936,
+      "step": 6110
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 3.13734787418998e-05,
+      "loss": 0.0909,
+      "step": 6120
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 3.133396554449186e-05,
+      "loss": 0.0806,
+      "step": 6130
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 3.129445234708393e-05,
+      "loss": 0.0767,
+      "step": 6140
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 3.125493914967599e-05,
+      "loss": 0.0869,
+      "step": 6150
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 3.121542595226806e-05,
+      "loss": 0.0644,
+      "step": 6160
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 3.117591275486012e-05,
+      "loss": 0.0581,
+      "step": 6170
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 3.113639955745219e-05,
+      "loss": 0.0675,
+      "step": 6180
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 3.109688636004426e-05,
+      "loss": 0.093,
+      "step": 6190
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 3.105737316263632e-05,
+      "loss": 0.0881,
+      "step": 6200
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 3.101785996522839e-05,
+      "loss": 0.0658,
+      "step": 6210
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 3.097834676782045e-05,
+      "loss": 0.1012,
+      "step": 6220
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 3.093883357041252e-05,
+      "loss": 0.0931,
+      "step": 6230
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.0899320373004584e-05,
+      "loss": 0.0877,
+      "step": 6240
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.085980717559665e-05,
+      "loss": 0.0826,
+      "step": 6250
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.0820293978188716e-05,
+      "loss": 0.0729,
+      "step": 6260
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 3.078078078078078e-05,
+      "loss": 0.0964,
+      "step": 6270
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 3.074126758337285e-05,
+      "loss": 0.0896,
+      "step": 6280
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 3.0701754385964913e-05,
+      "loss": 0.0796,
+      "step": 6290
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 3.066224118855698e-05,
+      "loss": 0.0608,
+      "step": 6300
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 3.0622727991149045e-05,
+      "loss": 0.0915,
+      "step": 6310
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 3.058321479374111e-05,
+      "loss": 0.0835,
+      "step": 6320
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 3.054370159633318e-05,
+      "loss": 0.0692,
+      "step": 6330
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 3.0504188398925243e-05,
+      "loss": 0.0883,
+      "step": 6340
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 3.0464675201517305e-05,
+      "loss": 0.0718,
+      "step": 6350
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 3.0425162004109374e-05,
+      "loss": 0.078,
+      "step": 6360
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 3.038564880670144e-05,
+      "loss": 0.085,
+      "step": 6370
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 3.034613560929351e-05,
+      "loss": 0.0651,
+      "step": 6380
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 3.0306622411885572e-05,
+      "loss": 0.074,
+      "step": 6390
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 3.0267109214477634e-05,
+      "loss": 0.0633,
+      "step": 6400
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 3.0227596017069704e-05,
+      "loss": 0.0712,
+      "step": 6410
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 3.0188082819661766e-05,
+      "loss": 0.0814,
+      "step": 6420
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 3.0148569622253835e-05,
+      "loss": 0.0792,
+      "step": 6430
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 3.01090564248459e-05,
+      "loss": 0.1052,
+      "step": 6440
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 3.0069543227437967e-05,
+      "loss": 0.0687,
+      "step": 6450
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 3.0030030030030033e-05,
+      "loss": 0.0605,
+      "step": 6460
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.9990516832622095e-05,
+      "loss": 0.0661,
+      "step": 6470
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.9951003635214164e-05,
+      "loss": 0.0547,
+      "step": 6480
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.9911490437806227e-05,
+      "loss": 0.1164,
+      "step": 6490
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.9871977240398296e-05,
+      "loss": 0.084,
+      "step": 6500
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 2.983246404299036e-05,
+      "loss": 0.0812,
+      "step": 6510
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.9792950845582424e-05,
+      "loss": 0.087,
+      "step": 6520
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.9753437648174494e-05,
+      "loss": 0.0839,
+      "step": 6530
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.9713924450766556e-05,
+      "loss": 0.073,
+      "step": 6540
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.9674411253358625e-05,
+      "loss": 0.0701,
+      "step": 6550
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.9634898055950688e-05,
+      "loss": 0.0868,
+      "step": 6560
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.9595384858542757e-05,
+      "loss": 0.079,
+      "step": 6570
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.955587166113482e-05,
+      "loss": 0.0714,
+      "step": 6580
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.9516358463726885e-05,
+      "loss": 0.0761,
+      "step": 6590
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.9476845266318955e-05,
+      "loss": 0.0999,
+      "step": 6600
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 2.9437332068911017e-05,
+      "loss": 0.0585,
+      "step": 6610
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.9397818871503086e-05,
+      "loss": 0.0842,
+      "step": 6620
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.935830567409515e-05,
+      "loss": 0.0792,
+      "step": 6630
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 2.931879247668721e-05,
+      "loss": 0.0827,
+      "step": 6640
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 2.927927927927928e-05,
+      "loss": 0.0692,
+      "step": 6650
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 2.9239766081871346e-05,
+      "loss": 0.0725,
+      "step": 6660
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 2.9200252884463412e-05,
+      "loss": 0.0735,
+      "step": 6670
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 2.9160739687055478e-05,
+      "loss": 0.0565,
+      "step": 6680
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.9121226489647547e-05,
+      "loss": 0.0552,
+      "step": 6690
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.908171329223961e-05,
+      "loss": 0.0698,
+      "step": 6700
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 2.9042200094831672e-05,
+      "loss": 0.0622,
+      "step": 6710
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 2.900268689742374e-05,
+      "loss": 0.0717,
+      "step": 6720
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 2.8963173700015804e-05,
+      "loss": 0.0744,
+      "step": 6730
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 2.8923660502607873e-05,
+      "loss": 0.1094,
+      "step": 6740
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 2.888414730519994e-05,
+      "loss": 0.0621,
+      "step": 6750
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 2.8844634107792e-05,
+      "loss": 0.1005,
+      "step": 6760
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 2.880512091038407e-05,
+      "loss": 0.0954,
+      "step": 6770
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 2.8765607712976133e-05,
+      "loss": 0.1011,
+      "step": 6780
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 2.8726094515568202e-05,
+      "loss": 0.0689,
+      "step": 6790
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 2.8686581318160265e-05,
+      "loss": 0.073,
+      "step": 6800
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 2.8647068120752334e-05,
+      "loss": 0.0695,
+      "step": 6810
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 2.86075549233444e-05,
+      "loss": 0.0895,
+      "step": 6820
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 2.8568041725936462e-05,
+      "loss": 0.1054,
+      "step": 6830
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 2.852852852852853e-05,
+      "loss": 0.0608,
+      "step": 6840
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 2.8489015331120594e-05,
+      "loss": 0.096,
+      "step": 6850
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 2.8449502133712663e-05,
+      "loss": 0.0765,
+      "step": 6860
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 2.8409988936304726e-05,
+      "loss": 0.0955,
+      "step": 6870
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 2.837047573889679e-05,
+      "loss": 0.0799,
+      "step": 6880
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 2.833096254148886e-05,
+      "loss": 0.061,
+      "step": 6890
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 2.8291449344080923e-05,
+      "loss": 0.0992,
+      "step": 6900
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 2.8251936146672992e-05,
+      "loss": 0.0562,
+      "step": 6910
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 2.8212422949265055e-05,
+      "loss": 0.0716,
+      "step": 6920
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 2.8172909751857124e-05,
+      "loss": 0.089,
+      "step": 6930
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 2.8133396554449187e-05,
+      "loss": 0.0784,
+      "step": 6940
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 2.8093883357041252e-05,
+      "loss": 0.096,
+      "step": 6950
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 2.8054370159633318e-05,
+      "loss": 0.0734,
+      "step": 6960
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.8014856962225384e-05,
+      "loss": 0.0727,
+      "step": 6970
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.7975343764817453e-05,
+      "loss": 0.0526,
+      "step": 6980
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 2.7935830567409516e-05,
+      "loss": 0.0671,
+      "step": 6990
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 2.7896317370001585e-05,
+      "loss": 0.0621,
+      "step": 7000
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 2.7856804172593647e-05,
+      "loss": 0.1225,
+      "step": 7010
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 2.781729097518571e-05,
+      "loss": 0.0957,
+      "step": 7020
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.0587,
+      "step": 7030
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.9889,
+      "eval_loss": 0.033243872225284576,
+      "eval_runtime": 47.0954,
+      "eval_samples_per_second": 212.335,
+      "eval_steps_per_second": 13.271,
+      "step": 7031
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 2.7738264580369845e-05,
+      "loss": 0.0917,
+      "step": 7040
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 2.7698751382961914e-05,
+      "loss": 0.0604,
+      "step": 7050
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 2.7659238185553977e-05,
+      "loss": 0.0547,
+      "step": 7060
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 2.761972498814604e-05,
+      "loss": 0.0855,
+      "step": 7070
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 2.758021179073811e-05,
+      "loss": 0.0837,
+      "step": 7080
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 2.754069859333017e-05,
+      "loss": 0.079,
+      "step": 7090
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 2.750118539592224e-05,
+      "loss": 0.067,
+      "step": 7100
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 2.7461672198514306e-05,
+      "loss": 0.1076,
+      "step": 7110
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 2.7422159001106372e-05,
+      "loss": 0.0531,
+      "step": 7120
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 2.7382645803698438e-05,
+      "loss": 0.0918,
+      "step": 7130
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 2.73431326062905e-05,
+      "loss": 0.0625,
+      "step": 7140
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 2.730361940888257e-05,
+      "loss": 0.0695,
+      "step": 7150
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 2.7264106211474632e-05,
+      "loss": 0.0614,
+      "step": 7160
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 2.72245930140667e-05,
+      "loss": 0.0643,
+      "step": 7170
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 2.7185079816658763e-05,
+      "loss": 0.0404,
+      "step": 7180
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 2.714556661925083e-05,
+      "loss": 0.0743,
+      "step": 7190
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 2.71060534218429e-05,
+      "loss": 0.043,
+      "step": 7200
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 2.706654022443496e-05,
+      "loss": 0.0638,
+      "step": 7210
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 2.702702702702703e-05,
+      "loss": 0.0623,
+      "step": 7220
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 2.6987513829619093e-05,
+      "loss": 0.0735,
+      "step": 7230
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 2.6948000632211162e-05,
+      "loss": 0.0615,
+      "step": 7240
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 2.6908487434803224e-05,
+      "loss": 0.0706,
+      "step": 7250
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 2.686897423739529e-05,
+      "loss": 0.1066,
+      "step": 7260
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 2.682946103998736e-05,
+      "loss": 0.0724,
+      "step": 7270
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 2.6789947842579422e-05,
+      "loss": 0.0693,
+      "step": 7280
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 2.675043464517149e-05,
+      "loss": 0.0642,
+      "step": 7290
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 2.6710921447763554e-05,
+      "loss": 0.0848,
+      "step": 7300
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 2.6671408250355616e-05,
+      "loss": 0.071,
+      "step": 7310
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 2.6631895052947685e-05,
+      "loss": 0.0804,
+      "step": 7320
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 2.659238185553975e-05,
+      "loss": 0.0772,
+      "step": 7330
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 2.655286865813182e-05,
+      "loss": 0.0702,
+      "step": 7340
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 2.6513355460723883e-05,
+      "loss": 0.0862,
+      "step": 7350
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 2.6473842263315952e-05,
+      "loss": 0.0953,
+      "step": 7360
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 2.6434329065908014e-05,
+      "loss": 0.0666,
+      "step": 7370
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 2.6394815868500077e-05,
+      "loss": 0.0852,
+      "step": 7380
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 2.6355302671092146e-05,
+      "loss": 0.0769,
+      "step": 7390
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 2.6315789473684212e-05,
+      "loss": 0.0812,
+      "step": 7400
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 2.6276276276276278e-05,
+      "loss": 0.0551,
+      "step": 7410
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 2.6236763078868344e-05,
+      "loss": 0.0709,
+      "step": 7420
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 2.6197249881460413e-05,
+      "loss": 0.0623,
+      "step": 7430
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 2.6157736684052475e-05,
+      "loss": 0.047,
+      "step": 7440
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 2.6118223486644538e-05,
+      "loss": 0.0708,
+      "step": 7450
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 2.6078710289236607e-05,
+      "loss": 0.0586,
+      "step": 7460
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 2.603919709182867e-05,
+      "loss": 0.0763,
+      "step": 7470
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 2.599968389442074e-05,
+      "loss": 0.0942,
+      "step": 7480
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 2.5960170697012805e-05,
+      "loss": 0.059,
+      "step": 7490
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 2.5920657499604867e-05,
+      "loss": 0.0774,
+      "step": 7500
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 2.5881144302196936e-05,
+      "loss": 0.069,
+      "step": 7510
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 2.5841631104789e-05,
+      "loss": 0.0573,
+      "step": 7520
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 2.5802117907381068e-05,
+      "loss": 0.0571,
+      "step": 7530
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 2.576260470997313e-05,
+      "loss": 0.0776,
+      "step": 7540
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 2.57230915125652e-05,
+      "loss": 0.066,
+      "step": 7550
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 2.5683578315157265e-05,
+      "loss": 0.0584,
+      "step": 7560
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 2.5644065117749328e-05,
+      "loss": 0.0561,
+      "step": 7570
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 2.5604551920341397e-05,
+      "loss": 0.062,
+      "step": 7580
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 2.556503872293346e-05,
+      "loss": 0.0703,
+      "step": 7590
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 2.552552552552553e-05,
+      "loss": 0.0589,
+      "step": 7600
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 2.548601232811759e-05,
+      "loss": 0.0509,
+      "step": 7610
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 2.5446499130709657e-05,
+      "loss": 0.1007,
+      "step": 7620
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 2.5406985933301723e-05,
+      "loss": 0.0642,
+      "step": 7630
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 2.536747273589379e-05,
+      "loss": 0.0704,
+      "step": 7640
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 2.5327959538485858e-05,
+      "loss": 0.0493,
+      "step": 7650
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 2.528844634107792e-05,
+      "loss": 0.0727,
+      "step": 7660
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 2.524893314366999e-05,
+      "loss": 0.0633,
+      "step": 7670
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 2.5209419946262052e-05,
+      "loss": 0.0876,
+      "step": 7680
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 2.5169906748854115e-05,
+      "loss": 0.0804,
+      "step": 7690
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 2.5130393551446184e-05,
+      "loss": 0.0917,
+      "step": 7700
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 2.509088035403825e-05,
+      "loss": 0.0649,
+      "step": 7710
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 2.505136715663032e-05,
+      "loss": 0.0862,
+      "step": 7720
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 2.501185395922238e-05,
+      "loss": 0.0854,
+      "step": 7730
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 2.4972340761814447e-05,
+      "loss": 0.0446,
+      "step": 7740
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 2.4932827564406513e-05,
+      "loss": 0.0911,
+      "step": 7750
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 2.489331436699858e-05,
+      "loss": 0.0714,
+      "step": 7760
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 2.485380116959064e-05,
+      "loss": 0.1171,
+      "step": 7770
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 2.481428797218271e-05,
+      "loss": 0.0676,
+      "step": 7780
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 2.4774774774774777e-05,
+      "loss": 0.0819,
+      "step": 7790
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 2.4735261577366842e-05,
+      "loss": 0.0796,
+      "step": 7800
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 2.4695748379958908e-05,
+      "loss": 0.0526,
+      "step": 7810
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 2.4656235182550974e-05,
+      "loss": 0.0596,
+      "step": 7820
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 2.461672198514304e-05,
+      "loss": 0.0883,
+      "step": 7830
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 2.4577208787735102e-05,
+      "loss": 0.0824,
+      "step": 7840
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 2.453769559032717e-05,
+      "loss": 0.0825,
+      "step": 7850
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 2.4498182392919237e-05,
+      "loss": 0.0539,
+      "step": 7860
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 2.4458669195511303e-05,
+      "loss": 0.0676,
+      "step": 7870
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 2.441915599810337e-05,
+      "loss": 0.0799,
+      "step": 7880
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 2.4379642800695435e-05,
+      "loss": 0.0831,
+      "step": 7890
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 2.4340129603287497e-05,
+      "loss": 0.0737,
+      "step": 7900
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 2.4300616405879563e-05,
+      "loss": 0.102,
+      "step": 7910
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 2.426110320847163e-05,
+      "loss": 0.0361,
+      "step": 7920
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 2.42215900110637e-05,
+      "loss": 0.076,
+      "step": 7930
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 2.4182076813655764e-05,
+      "loss": 0.0768,
+      "step": 7940
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 2.414256361624783e-05,
+      "loss": 0.0743,
+      "step": 7950
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 2.4103050418839892e-05,
+      "loss": 0.0632,
+      "step": 7960
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 2.406353722143196e-05,
+      "loss": 0.0636,
+      "step": 7970
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 2.4024024024024024e-05,
+      "loss": 0.0889,
+      "step": 7980
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 2.398451082661609e-05,
+      "loss": 0.0446,
+      "step": 7990
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 2.3944997629208156e-05,
+      "loss": 0.0691,
+      "step": 8000
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 2.3905484431800225e-05,
+      "loss": 0.0783,
+      "step": 8010
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 2.3865971234392288e-05,
+      "loss": 0.0875,
+      "step": 8020
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 2.3826458036984353e-05,
+      "loss": 0.0657,
+      "step": 8030
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 2.378694483957642e-05,
+      "loss": 0.0625,
+      "step": 8040
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 2.3747431642168485e-05,
+      "loss": 0.0639,
+      "step": 8050
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 2.370791844476055e-05,
+      "loss": 0.0549,
+      "step": 8060
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 2.3668405247352617e-05,
+      "loss": 0.0379,
+      "step": 8070
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 2.3628892049944683e-05,
+      "loss": 0.072,
+      "step": 8080
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 2.358937885253675e-05,
+      "loss": 0.0521,
+      "step": 8090
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 2.3549865655128814e-05,
+      "loss": 0.0606,
+      "step": 8100
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 2.351035245772088e-05,
+      "loss": 0.0577,
+      "step": 8110
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 2.3470839260312946e-05,
+      "loss": 0.0692,
+      "step": 8120
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 2.3431326062905012e-05,
+      "loss": 0.0833,
+      "step": 8130
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 2.3391812865497074e-05,
+      "loss": 0.0646,
+      "step": 8140
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 2.3352299668089144e-05,
+      "loss": 0.0464,
+      "step": 8150
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 2.331278647068121e-05,
+      "loss": 0.0695,
+      "step": 8160
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 2.3273273273273275e-05,
+      "loss": 0.0976,
+      "step": 8170
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 2.323376007586534e-05,
+      "loss": 0.0662,
+      "step": 8180
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 2.3194246878457407e-05,
+      "loss": 0.0573,
+      "step": 8190
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 2.315473368104947e-05,
+      "loss": 0.0828,
+      "step": 8200
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 2.3115220483641535e-05,
+      "loss": 0.0655,
+      "step": 8210
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 2.30757072862336e-05,
+      "loss": 0.0464,
+      "step": 8220
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 2.303619408882567e-05,
+      "loss": 0.0753,
+      "step": 8230
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 2.2996680891417736e-05,
+      "loss": 0.0613,
+      "step": 8240
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 2.2957167694009802e-05,
+      "loss": 0.0579,
+      "step": 8250
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 2.2917654496601868e-05,
+      "loss": 0.0657,
+      "step": 8260
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 2.287814129919393e-05,
+      "loss": 0.0758,
+      "step": 8270
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 2.2838628101785996e-05,
+      "loss": 0.0393,
+      "step": 8280
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 2.2799114904378062e-05,
+      "loss": 0.0675,
+      "step": 8290
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 2.275960170697013e-05,
+      "loss": 0.0668,
+      "step": 8300
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 2.2720088509562197e-05,
+      "loss": 0.0583,
+      "step": 8310
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 2.2680575312154263e-05,
+      "loss": 0.0639,
+      "step": 8320
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 2.2641062114746325e-05,
+      "loss": 0.0747,
+      "step": 8330
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 2.260154891733839e-05,
+      "loss": 0.0576,
+      "step": 8340
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 2.2562035719930457e-05,
+      "loss": 0.0409,
+      "step": 8350
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 2.2522522522522523e-05,
+      "loss": 0.1121,
+      "step": 8360
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 2.248300932511459e-05,
+      "loss": 0.083,
+      "step": 8370
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 2.2443496127706658e-05,
+      "loss": 0.0847,
+      "step": 8380
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 2.240398293029872e-05,
+      "loss": 0.0877,
+      "step": 8390
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 2.2364469732890786e-05,
+      "loss": 0.0688,
+      "step": 8400
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 2.2324956535482852e-05,
+      "loss": 0.0648,
+      "step": 8410
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 2.2285443338074918e-05,
+      "loss": 0.0779,
+      "step": 8420
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 2.2245930140666984e-05,
+      "loss": 0.0493,
+      "step": 8430
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.9919,
+      "eval_loss": 0.025265106931328773,
+      "eval_runtime": 46.8181,
+      "eval_samples_per_second": 213.593,
+      "eval_steps_per_second": 13.35,
+      "step": 8437
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 2.220641694325905e-05,
+      "loss": 0.0709,
+      "step": 8440
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 2.2166903745851115e-05,
+      "loss": 0.0567,
+      "step": 8450
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 2.212739054844318e-05,
+      "loss": 0.0635,
+      "step": 8460
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 2.2087877351035247e-05,
+      "loss": 0.057,
+      "step": 8470
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 2.2048364153627313e-05,
+      "loss": 0.0707,
+      "step": 8480
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 2.200885095621938e-05,
+      "loss": 0.0609,
+      "step": 8490
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 2.1969337758811445e-05,
+      "loss": 0.0488,
+      "step": 8500
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 2.1929824561403507e-05,
+      "loss": 0.0635,
+      "step": 8510
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 2.1890311363995576e-05,
+      "loss": 0.0435,
+      "step": 8520
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 2.1850798166587642e-05,
+      "loss": 0.0634,
+      "step": 8530
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 2.1811284969179708e-05,
+      "loss": 0.0782,
+      "step": 8540
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 2.1771771771771774e-05,
+      "loss": 0.0559,
+      "step": 8550
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 2.173225857436384e-05,
+      "loss": 0.0539,
+      "step": 8560
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 2.1692745376955902e-05,
+      "loss": 0.0423,
+      "step": 8570
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 2.1653232179547968e-05,
+      "loss": 0.0547,
+      "step": 8580
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 2.1613718982140034e-05,
+      "loss": 0.0739,
+      "step": 8590
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 2.1574205784732103e-05,
+      "loss": 0.0896,
+      "step": 8600
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 2.153469258732417e-05,
+      "loss": 0.0824,
+      "step": 8610
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 2.1495179389916235e-05,
+      "loss": 0.0456,
+      "step": 8620
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 2.1455666192508297e-05,
+      "loss": 0.085,
+      "step": 8630
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 2.1416152995100363e-05,
+      "loss": 0.0883,
+      "step": 8640
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 2.137663979769243e-05,
+      "loss": 0.0715,
+      "step": 8650
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 2.1337126600284495e-05,
+      "loss": 0.0646,
+      "step": 8660
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 2.129761340287656e-05,
+      "loss": 0.1001,
+      "step": 8670
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 2.125810020546863e-05,
+      "loss": 0.0465,
+      "step": 8680
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 2.1218587008060696e-05,
+      "loss": 0.062,
+      "step": 8690
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 2.1179073810652758e-05,
+      "loss": 0.0759,
+      "step": 8700
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 2.1139560613244824e-05,
+      "loss": 0.0511,
+      "step": 8710
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 2.110004741583689e-05,
+      "loss": 0.0574,
+      "step": 8720
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 2.1060534218428956e-05,
+      "loss": 0.0794,
+      "step": 8730
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 2.102102102102102e-05,
+      "loss": 0.0701,
+      "step": 8740
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 2.0981507823613087e-05,
+      "loss": 0.0593,
+      "step": 8750
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 2.0941994626205153e-05,
+      "loss": 0.0558,
+      "step": 8760
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 2.090248142879722e-05,
+      "loss": 0.0759,
+      "step": 8770
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 2.0862968231389285e-05,
+      "loss": 0.0598,
+      "step": 8780
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 2.082345503398135e-05,
+      "loss": 0.0838,
+      "step": 8790
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 2.0783941836573417e-05,
+      "loss": 0.0728,
+      "step": 8800
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 2.0744428639165482e-05,
+      "loss": 0.0728,
+      "step": 8810
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 2.070491544175755e-05,
+      "loss": 0.0663,
+      "step": 8820
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 2.0665402244349614e-05,
+      "loss": 0.0564,
+      "step": 8830
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 2.062588904694168e-05,
+      "loss": 0.0792,
+      "step": 8840
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 2.0586375849533746e-05,
+      "loss": 0.0437,
+      "step": 8850
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 2.0546862652125812e-05,
+      "loss": 0.0518,
+      "step": 8860
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 2.0507349454717878e-05,
+      "loss": 0.0564,
+      "step": 8870
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 2.046783625730994e-05,
+      "loss": 0.0702,
+      "step": 8880
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 2.042832305990201e-05,
+      "loss": 0.0575,
+      "step": 8890
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 2.0388809862494075e-05,
+      "loss": 0.0505,
+      "step": 8900
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 2.034929666508614e-05,
+      "loss": 0.0646,
+      "step": 8910
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 2.0309783467678207e-05,
+      "loss": 0.0769,
+      "step": 8920
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 2.0270270270270273e-05,
+      "loss": 0.0455,
+      "step": 8930
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 2.0230757072862335e-05,
+      "loss": 0.0648,
+      "step": 8940
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 2.01912438754544e-05,
+      "loss": 0.059,
+      "step": 8950
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 2.0151730678046467e-05,
+      "loss": 0.0724,
+      "step": 8960
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 2.0112217480638536e-05,
+      "loss": 0.0573,
+      "step": 8970
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 2.0072704283230602e-05,
+      "loss": 0.0871,
+      "step": 8980
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 2.0033191085822668e-05,
+      "loss": 0.0417,
+      "step": 8990
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 1.999367788841473e-05,
+      "loss": 0.0444,
+      "step": 9000
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 1.9954164691006796e-05,
+      "loss": 0.0425,
+      "step": 9010
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 1.9914651493598862e-05,
+      "loss": 0.0854,
+      "step": 9020
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 1.9875138296190928e-05,
+      "loss": 0.0751,
+      "step": 9030
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 1.9835625098782994e-05,
+      "loss": 0.063,
+      "step": 9040
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 1.9796111901375063e-05,
+      "loss": 0.0665,
+      "step": 9050
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 1.9756598703967125e-05,
+      "loss": 0.0374,
+      "step": 9060
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 1.971708550655919e-05,
+      "loss": 0.0536,
+      "step": 9070
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 1.9677572309151257e-05,
+      "loss": 0.0603,
+      "step": 9080
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 1.9638059111743323e-05,
+      "loss": 0.0584,
+      "step": 9090
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 1.959854591433539e-05,
+      "loss": 0.0636,
+      "step": 9100
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 1.9559032716927454e-05,
+      "loss": 0.0432,
+      "step": 9110
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 1.951951951951952e-05,
+      "loss": 0.0619,
+      "step": 9120
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 1.9480006322111586e-05,
+      "loss": 0.0614,
+      "step": 9130
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 1.9440493124703652e-05,
+      "loss": 0.0729,
+      "step": 9140
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 1.9400979927295718e-05,
+      "loss": 0.0665,
+      "step": 9150
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 1.9361466729887784e-05,
+      "loss": 0.0761,
+      "step": 9160
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 1.932195353247985e-05,
+      "loss": 0.0766,
+      "step": 9170
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 1.9282440335071915e-05,
+      "loss": 0.0425,
+      "step": 9180
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 1.924292713766398e-05,
+      "loss": 0.0439,
+      "step": 9190
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 1.9203413940256047e-05,
+      "loss": 0.0633,
+      "step": 9200
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 1.9163900742848113e-05,
+      "loss": 0.0679,
+      "step": 9210
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 1.912438754544018e-05,
+      "loss": 0.0906,
+      "step": 9220
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 1.9084874348032245e-05,
+      "loss": 0.0765,
+      "step": 9230
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 1.904536115062431e-05,
+      "loss": 0.0554,
+      "step": 9240
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 1.9005847953216373e-05,
+      "loss": 0.0811,
+      "step": 9250
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 1.896633475580844e-05,
+      "loss": 0.047,
+      "step": 9260
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 1.8926821558400508e-05,
+      "loss": 0.0809,
+      "step": 9270
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 1.8887308360992574e-05,
+      "loss": 0.0619,
+      "step": 9280
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 1.884779516358464e-05,
+      "loss": 0.0701,
+      "step": 9290
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 1.8808281966176705e-05,
+      "loss": 0.0905,
+      "step": 9300
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 1.8768768768768768e-05,
+      "loss": 0.0487,
+      "step": 9310
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 1.8729255571360834e-05,
+      "loss": 0.0544,
+      "step": 9320
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 1.86897423739529e-05,
+      "loss": 0.0708,
+      "step": 9330
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 1.8650229176544965e-05,
+      "loss": 0.0444,
+      "step": 9340
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 1.8610715979137035e-05,
+      "loss": 0.0799,
+      "step": 9350
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 1.85712027817291e-05,
+      "loss": 0.0631,
+      "step": 9360
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 1.8531689584321163e-05,
+      "loss": 0.0551,
+      "step": 9370
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 1.849217638691323e-05,
+      "loss": 0.0648,
+      "step": 9380
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 1.8452663189505295e-05,
+      "loss": 0.0673,
+      "step": 9390
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 1.841314999209736e-05,
+      "loss": 0.0635,
+      "step": 9400
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 1.8373636794689426e-05,
+      "loss": 0.0652,
+      "step": 9410
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 1.8334123597281496e-05,
+      "loss": 0.0532,
+      "step": 9420
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 1.8294610399873558e-05,
+      "loss": 0.0352,
+      "step": 9430
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 1.8255097202465624e-05,
+      "loss": 0.0708,
+      "step": 9440
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 1.821558400505769e-05,
+      "loss": 0.05,
+      "step": 9450
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 1.8176070807649756e-05,
+      "loss": 0.0476,
+      "step": 9460
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 1.813655761024182e-05,
+      "loss": 0.0681,
+      "step": 9470
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 1.8097044412833887e-05,
+      "loss": 0.0635,
+      "step": 9480
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 1.8057531215425953e-05,
+      "loss": 0.0644,
+      "step": 9490
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 1.801801801801802e-05,
+      "loss": 0.0715,
+      "step": 9500
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 1.7978504820610085e-05,
+      "loss": 0.0493,
+      "step": 9510
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 1.793899162320215e-05,
+      "loss": 0.0726,
+      "step": 9520
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 1.7899478425794217e-05,
+      "loss": 0.0618,
+      "step": 9530
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 1.7859965228386282e-05,
+      "loss": 0.0592,
+      "step": 9540
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 1.7820452030978345e-05,
+      "loss": 0.0618,
+      "step": 9550
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 1.7780938833570414e-05,
+      "loss": 0.0462,
+      "step": 9560
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 1.774142563616248e-05,
+      "loss": 0.0533,
+      "step": 9570
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 1.7701912438754546e-05,
+      "loss": 0.07,
+      "step": 9580
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 1.766239924134661e-05,
+      "loss": 0.0549,
+      "step": 9590
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 1.7622886043938677e-05,
+      "loss": 0.0337,
+      "step": 9600
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 1.7583372846530743e-05,
+      "loss": 0.0576,
+      "step": 9610
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 1.7543859649122806e-05,
+      "loss": 0.0674,
+      "step": 9620
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 1.750434645171487e-05,
+      "loss": 0.0346,
+      "step": 9630
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 1.746483325430694e-05,
+      "loss": 0.0648,
+      "step": 9640
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 1.7425320056899007e-05,
+      "loss": 0.058,
+      "step": 9650
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 1.7385806859491072e-05,
+      "loss": 0.0557,
+      "step": 9660
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 1.734629366208314e-05,
+      "loss": 0.0285,
+      "step": 9670
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 1.73067804646752e-05,
+      "loss": 0.0493,
+      "step": 9680
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 1.7267267267267267e-05,
+      "loss": 0.0611,
+      "step": 9690
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 1.7227754069859332e-05,
+      "loss": 0.0844,
+      "step": 9700
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 1.71882408724514e-05,
+      "loss": 0.0495,
+      "step": 9710
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 1.7148727675043468e-05,
+      "loss": 0.0554,
+      "step": 9720
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 1.7109214477635533e-05,
+      "loss": 0.0509,
+      "step": 9730
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 1.7069701280227596e-05,
+      "loss": 0.0629,
+      "step": 9740
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 1.7030188082819662e-05,
+      "loss": 0.0487,
+      "step": 9750
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 1.6990674885411728e-05,
+      "loss": 0.0705,
+      "step": 9760
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 1.6951161688003793e-05,
+      "loss": 0.06,
+      "step": 9770
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 1.691164849059586e-05,
+      "loss": 0.0616,
+      "step": 9780
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 1.6872135293187925e-05,
+      "loss": 0.0499,
+      "step": 9790
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 1.683262209577999e-05,
+      "loss": 0.0739,
+      "step": 9800
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 1.6793108898372057e-05,
+      "loss": 0.0346,
+      "step": 9810
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 1.6753595700964123e-05,
+      "loss": 0.0602,
+      "step": 9820
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 1.671408250355619e-05,
+      "loss": 0.0745,
+      "step": 9830
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 1.6674569306148254e-05,
+      "loss": 0.06,
+      "step": 9840
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.9922,
+      "eval_loss": 0.027898238971829414,
+      "eval_runtime": 46.8519,
+      "eval_samples_per_second": 213.439,
+      "eval_steps_per_second": 13.34,
+      "step": 9843
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 1.663505610874032e-05,
+      "loss": 0.0531,
+      "step": 9850
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 1.6595542911332386e-05,
+      "loss": 0.0427,
+      "step": 9860
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 1.6556029713924452e-05,
+      "loss": 0.0741,
+      "step": 9870
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 1.6516516516516518e-05,
+      "loss": 0.0555,
+      "step": 9880
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 1.6477003319108584e-05,
+      "loss": 0.0658,
+      "step": 9890
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 1.643749012170065e-05,
+      "loss": 0.0705,
+      "step": 9900
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 1.6397976924292715e-05,
+      "loss": 0.0614,
+      "step": 9910
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 1.6358463726884778e-05,
+      "loss": 0.047,
+      "step": 9920
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 1.6318950529476847e-05,
+      "loss": 0.0721,
+      "step": 9930
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 1.6279437332068913e-05,
+      "loss": 0.0717,
+      "step": 9940
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 1.623992413466098e-05,
+      "loss": 0.0732,
+      "step": 9950
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 1.6200410937253044e-05,
+      "loss": 0.06,
+      "step": 9960
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 1.616089773984511e-05,
+      "loss": 0.0674,
+      "step": 9970
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 1.6121384542437173e-05,
+      "loss": 0.0826,
+      "step": 9980
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 1.608187134502924e-05,
+      "loss": 0.0693,
+      "step": 9990
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 1.6042358147621304e-05,
+      "loss": 0.0629,
+      "step": 10000
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 1.6002844950213374e-05,
+      "loss": 0.0538,
+      "step": 10010
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 1.596333175280544e-05,
+      "loss": 0.0707,
+      "step": 10020
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 1.5923818555397505e-05,
+      "loss": 0.0721,
+      "step": 10030
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 1.588430535798957e-05,
+      "loss": 0.0523,
+      "step": 10040
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 1.5844792160581634e-05,
+      "loss": 0.0517,
+      "step": 10050
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 1.58052789631737e-05,
+      "loss": 0.0491,
+      "step": 10060
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 1.5765765765765765e-05,
+      "loss": 0.0889,
+      "step": 10070
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 1.572625256835783e-05,
+      "loss": 0.0763,
+      "step": 10080
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 1.56867393709499e-05,
+      "loss": 0.0687,
+      "step": 10090
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 1.5647226173541966e-05,
+      "loss": 0.0411,
+      "step": 10100
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 1.560771297613403e-05,
+      "loss": 0.0443,
+      "step": 10110
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 1.5568199778726095e-05,
+      "loss": 0.0398,
+      "step": 10120
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 1.552868658131816e-05,
+      "loss": 0.0445,
+      "step": 10130
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 1.5489173383910226e-05,
+      "loss": 0.052,
+      "step": 10140
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 1.5449660186502292e-05,
+      "loss": 0.0646,
+      "step": 10150
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 1.5410146989094358e-05,
+      "loss": 0.0466,
+      "step": 10160
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 1.5370633791686424e-05,
+      "loss": 0.062,
+      "step": 10170
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 1.533112059427849e-05,
+      "loss": 0.0632,
+      "step": 10180
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 1.5291607396870555e-05,
+      "loss": 0.0697,
+      "step": 10190
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 1.5252094199462621e-05,
+      "loss": 0.0344,
+      "step": 10200
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 1.5212581002054687e-05,
+      "loss": 0.0469,
+      "step": 10210
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 1.5173067804646755e-05,
+      "loss": 0.0461,
+      "step": 10220
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 1.5133554607238817e-05,
+      "loss": 0.0525,
+      "step": 10230
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 1.5094041409830883e-05,
+      "loss": 0.0397,
+      "step": 10240
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 1.505452821242295e-05,
+      "loss": 0.0843,
+      "step": 10250
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 1.5015015015015016e-05,
+      "loss": 0.0672,
+      "step": 10260
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 1.4975501817607082e-05,
+      "loss": 0.0723,
+      "step": 10270
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 1.4935988620199148e-05,
+      "loss": 0.0634,
+      "step": 10280
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 1.4896475422791212e-05,
+      "loss": 0.0557,
+      "step": 10290
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 1.4856962225383278e-05,
+      "loss": 0.072,
+      "step": 10300
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 1.4817449027975344e-05,
+      "loss": 0.0578,
+      "step": 10310
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 1.477793583056741e-05,
+      "loss": 0.0488,
+      "step": 10320
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 1.4738422633159477e-05,
+      "loss": 0.0673,
+      "step": 10330
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 1.4698909435751543e-05,
+      "loss": 0.046,
+      "step": 10340
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 1.4659396238343606e-05,
+      "loss": 0.053,
+      "step": 10350
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 1.4619883040935673e-05,
+      "loss": 0.0628,
+      "step": 10360
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 1.4580369843527739e-05,
+      "loss": 0.0613,
+      "step": 10370
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 1.4540856646119805e-05,
+      "loss": 0.0452,
+      "step": 10380
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 1.450134344871187e-05,
+      "loss": 0.0381,
+      "step": 10390
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 1.4461830251303937e-05,
+      "loss": 0.0832,
+      "step": 10400
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 1.4422317053896e-05,
+      "loss": 0.0623,
+      "step": 10410
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 1.4382803856488066e-05,
+      "loss": 0.0475,
+      "step": 10420
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 1.4343290659080132e-05,
+      "loss": 0.0506,
+      "step": 10430
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 1.43037774616722e-05,
+      "loss": 0.0398,
+      "step": 10440
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 1.4264264264264266e-05,
+      "loss": 0.0695,
+      "step": 10450
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 1.4224751066856332e-05,
+      "loss": 0.0584,
+      "step": 10460
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 1.4185237869448396e-05,
+      "loss": 0.061,
+      "step": 10470
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 1.4145724672040462e-05,
+      "loss": 0.0519,
+      "step": 10480
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 1.4106211474632527e-05,
+      "loss": 0.0528,
+      "step": 10490
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 1.4066698277224593e-05,
+      "loss": 0.0947,
+      "step": 10500
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 1.4027185079816659e-05,
+      "loss": 0.0604,
+      "step": 10510
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 1.3987671882408727e-05,
+      "loss": 0.0526,
+      "step": 10520
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 1.3948158685000792e-05,
+      "loss": 0.0572,
+      "step": 10530
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 1.3908645487592855e-05,
+      "loss": 0.0744,
+      "step": 10540
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 1.3869132290184922e-05,
+      "loss": 0.0749,
+      "step": 10550
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 1.3829619092776988e-05,
+      "loss": 0.0487,
+      "step": 10560
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 1.3790105895369054e-05,
+      "loss": 0.0383,
+      "step": 10570
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 1.375059269796112e-05,
+      "loss": 0.0688,
+      "step": 10580
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 1.3711079500553186e-05,
+      "loss": 0.0795,
+      "step": 10590
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 1.367156630314525e-05,
+      "loss": 0.0517,
+      "step": 10600
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 1.3632053105737316e-05,
+      "loss": 0.0694,
+      "step": 10610
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 1.3592539908329382e-05,
+      "loss": 0.0669,
+      "step": 10620
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 1.355302671092145e-05,
+      "loss": 0.0448,
+      "step": 10630
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 1.3513513513513515e-05,
+      "loss": 0.0636,
+      "step": 10640
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 1.3474000316105581e-05,
+      "loss": 0.0583,
+      "step": 10650
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 1.3434487118697645e-05,
+      "loss": 0.0733,
+      "step": 10660
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 1.3394973921289711e-05,
+      "loss": 0.0666,
+      "step": 10670
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 1.3355460723881777e-05,
+      "loss": 0.0646,
+      "step": 10680
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 1.3315947526473843e-05,
+      "loss": 0.0594,
+      "step": 10690
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 1.327643432906591e-05,
+      "loss": 0.0641,
+      "step": 10700
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 1.3236921131657976e-05,
+      "loss": 0.0416,
+      "step": 10710
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 1.3197407934250038e-05,
+      "loss": 0.0404,
+      "step": 10720
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 1.3157894736842106e-05,
+      "loss": 0.0723,
+      "step": 10730
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 1.3118381539434172e-05,
+      "loss": 0.0461,
+      "step": 10740
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 1.3078868342026238e-05,
+      "loss": 0.0684,
+      "step": 10750
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 1.3039355144618304e-05,
+      "loss": 0.0509,
+      "step": 10760
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 1.299984194721037e-05,
+      "loss": 0.0511,
+      "step": 10770
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 1.2960328749802434e-05,
+      "loss": 0.051,
+      "step": 10780
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 1.29208155523945e-05,
+      "loss": 0.0537,
+      "step": 10790
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 1.2881302354986565e-05,
+      "loss": 0.048,
+      "step": 10800
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 1.2841789157578633e-05,
+      "loss": 0.043,
+      "step": 10810
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 1.2802275960170699e-05,
+      "loss": 0.0776,
+      "step": 10820
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 1.2762762762762764e-05,
+      "loss": 0.0683,
+      "step": 10830
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 1.2723249565354829e-05,
+      "loss": 0.0487,
+      "step": 10840
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 1.2683736367946894e-05,
+      "loss": 0.0442,
+      "step": 10850
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 1.264422317053896e-05,
+      "loss": 0.0417,
+      "step": 10860
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 1.2604709973131026e-05,
+      "loss": 0.0851,
+      "step": 10870
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 1.2565196775723092e-05,
+      "loss": 0.0382,
+      "step": 10880
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 1.252568357831516e-05,
+      "loss": 0.057,
+      "step": 10890
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 1.2486170380907224e-05,
+      "loss": 0.0765,
+      "step": 10900
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 1.244665718349929e-05,
+      "loss": 0.0647,
+      "step": 10910
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 1.2407143986091355e-05,
+      "loss": 0.0513,
+      "step": 10920
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 1.2367630788683421e-05,
+      "loss": 0.0454,
+      "step": 10930
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 1.2328117591275487e-05,
+      "loss": 0.0691,
+      "step": 10940
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 1.2288604393867551e-05,
+      "loss": 0.0552,
+      "step": 10950
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 1.2249091196459619e-05,
+      "loss": 0.0662,
+      "step": 10960
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 1.2209577999051685e-05,
+      "loss": 0.0516,
+      "step": 10970
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 1.2170064801643749e-05,
+      "loss": 0.0669,
+      "step": 10980
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 1.2130551604235815e-05,
+      "loss": 0.0421,
+      "step": 10990
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 1.2091038406827882e-05,
+      "loss": 0.0584,
+      "step": 11000
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 1.2051525209419946e-05,
+      "loss": 0.0641,
+      "step": 11010
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 1.2012012012012012e-05,
+      "loss": 0.0386,
+      "step": 11020
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 1.1972498814604078e-05,
+      "loss": 0.0683,
+      "step": 11030
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 1.1932985617196144e-05,
+      "loss": 0.051,
+      "step": 11040
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 1.189347241978821e-05,
+      "loss": 0.0564,
+      "step": 11050
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 1.1853959222380275e-05,
+      "loss": 0.0627,
+      "step": 11060
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 1.1814446024972341e-05,
+      "loss": 0.0551,
+      "step": 11070
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 1.1774932827564407e-05,
+      "loss": 0.0608,
+      "step": 11080
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 1.1735419630156473e-05,
+      "loss": 0.0401,
+      "step": 11090
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 1.1695906432748537e-05,
+      "loss": 0.0596,
+      "step": 11100
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 1.1656393235340605e-05,
+      "loss": 0.0514,
+      "step": 11110
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 1.161688003793267e-05,
+      "loss": 0.0572,
+      "step": 11120
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 1.1577366840524735e-05,
+      "loss": 0.0511,
+      "step": 11130
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 1.15378536431168e-05,
+      "loss": 0.066,
+      "step": 11140
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 1.1498340445708868e-05,
+      "loss": 0.0607,
+      "step": 11150
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 1.1458827248300934e-05,
+      "loss": 0.0754,
+      "step": 11160
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 1.1419314050892998e-05,
+      "loss": 0.0446,
+      "step": 11170
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 1.1379800853485066e-05,
+      "loss": 0.0462,
+      "step": 11180
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 1.1340287656077131e-05,
+      "loss": 0.0585,
+      "step": 11190
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 1.1300774458669196e-05,
+      "loss": 0.0707,
+      "step": 11200
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 1.1261261261261261e-05,
+      "loss": 0.0708,
+      "step": 11210
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 1.1221748063853329e-05,
+      "loss": 0.0652,
+      "step": 11220
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 1.1182234866445393e-05,
+      "loss": 0.0657,
+      "step": 11230
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 1.1142721669037459e-05,
+      "loss": 0.0709,
+      "step": 11240
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 1.1103208471629525e-05,
+      "loss": 0.0738,
+      "step": 11250
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.9907,
+      "eval_loss": 0.03255428001284599,
+      "eval_runtime": 46.804,
+      "eval_samples_per_second": 213.657,
+      "eval_steps_per_second": 13.354,
+      "step": 11250
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 1.106369527422159e-05,
+      "loss": 0.0715,
+      "step": 11260
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 1.1024182076813657e-05,
+      "loss": 0.051,
+      "step": 11270
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 1.0984668879405722e-05,
+      "loss": 0.0518,
+      "step": 11280
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 1.0945155681997788e-05,
+      "loss": 0.0199,
+      "step": 11290
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 1.0905642484589854e-05,
+      "loss": 0.0467,
+      "step": 11300
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 1.086612928718192e-05,
+      "loss": 0.0283,
+      "step": 11310
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 1.0826616089773984e-05,
+      "loss": 0.0627,
+      "step": 11320
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 1.0787102892366052e-05,
+      "loss": 0.0257,
+      "step": 11330
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 1.0747589694958117e-05,
+      "loss": 0.0285,
+      "step": 11340
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 1.0708076497550182e-05,
+      "loss": 0.0614,
+      "step": 11350
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 1.0668563300142247e-05,
+      "loss": 0.0527,
+      "step": 11360
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 1.0629050102734315e-05,
+      "loss": 0.0461,
+      "step": 11370
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 1.0589536905326379e-05,
+      "loss": 0.0491,
+      "step": 11380
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 1.0550023707918445e-05,
+      "loss": 0.0643,
+      "step": 11390
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 1.051051051051051e-05,
+      "loss": 0.0441,
+      "step": 11400
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 1.0470997313102577e-05,
+      "loss": 0.0408,
+      "step": 11410
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 1.0431484115694642e-05,
+      "loss": 0.0537,
+      "step": 11420
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 1.0391970918286708e-05,
+      "loss": 0.064,
+      "step": 11430
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 1.0352457720878774e-05,
+      "loss": 0.0396,
+      "step": 11440
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 1.031294452347084e-05,
+      "loss": 0.0445,
+      "step": 11450
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 1.0273431326062906e-05,
+      "loss": 0.0667,
+      "step": 11460
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 1.023391812865497e-05,
+      "loss": 0.067,
+      "step": 11470
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 1.0194404931247038e-05,
+      "loss": 0.0543,
+      "step": 11480
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 1.0154891733839103e-05,
+      "loss": 0.0441,
+      "step": 11490
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 1.0115378536431168e-05,
+      "loss": 0.0396,
+      "step": 11500
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 1.0075865339023233e-05,
+      "loss": 0.0686,
+      "step": 11510
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 1.0036352141615301e-05,
+      "loss": 0.0614,
+      "step": 11520
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 9.996838944207365e-06,
+      "loss": 0.0485,
+      "step": 11530
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 9.957325746799431e-06,
+      "loss": 0.0453,
+      "step": 11540
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 9.917812549391497e-06,
+      "loss": 0.077,
+      "step": 11550
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 9.878299351983563e-06,
+      "loss": 0.0373,
+      "step": 11560
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 9.838786154575628e-06,
+      "loss": 0.0733,
+      "step": 11570
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 9.799272957167694e-06,
+      "loss": 0.0391,
+      "step": 11580
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 9.75975975975976e-06,
+      "loss": 0.0335,
+      "step": 11590
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 9.720246562351826e-06,
+      "loss": 0.0369,
+      "step": 11600
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 9.680733364943892e-06,
+      "loss": 0.0607,
+      "step": 11610
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 9.641220167535958e-06,
+      "loss": 0.0397,
+      "step": 11620
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 9.601706970128024e-06,
+      "loss": 0.0409,
+      "step": 11630
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 9.56219377272009e-06,
+      "loss": 0.048,
+      "step": 11640
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 9.522680575312155e-06,
+      "loss": 0.0488,
+      "step": 11650
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 9.48316737790422e-06,
+      "loss": 0.063,
+      "step": 11660
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 9.443654180496287e-06,
+      "loss": 0.0417,
+      "step": 11670
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 9.404140983088353e-06,
+      "loss": 0.0263,
+      "step": 11680
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 9.364627785680417e-06,
+      "loss": 0.0335,
+      "step": 11690
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 9.325114588272483e-06,
+      "loss": 0.039,
+      "step": 11700
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 9.28560139086455e-06,
+      "loss": 0.0727,
+      "step": 11710
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 9.246088193456614e-06,
+      "loss": 0.0468,
+      "step": 11720
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 9.20657499604868e-06,
+      "loss": 0.0774,
+      "step": 11730
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 9.167061798640748e-06,
+      "loss": 0.0424,
+      "step": 11740
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 9.127548601232812e-06,
+      "loss": 0.0465,
+      "step": 11750
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 9.088035403824878e-06,
+      "loss": 0.0438,
+      "step": 11760
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 9.048522206416944e-06,
+      "loss": 0.0611,
+      "step": 11770
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 9.00900900900901e-06,
+      "loss": 0.0581,
+      "step": 11780
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 8.969495811601075e-06,
+      "loss": 0.0414,
+      "step": 11790
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 8.929982614193141e-06,
+      "loss": 0.0577,
+      "step": 11800
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 8.890469416785207e-06,
+      "loss": 0.0656,
+      "step": 11810
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 8.850956219377273e-06,
+      "loss": 0.0373,
+      "step": 11820
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 8.811443021969339e-06,
+      "loss": 0.0406,
+      "step": 11830
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 8.771929824561403e-06,
+      "loss": 0.051,
+      "step": 11840
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 8.73241662715347e-06,
+      "loss": 0.0647,
+      "step": 11850
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 8.692903429745536e-06,
+      "loss": 0.0293,
+      "step": 11860
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 8.6533902323376e-06,
+      "loss": 0.0738,
+      "step": 11870
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 8.613877034929666e-06,
+      "loss": 0.0569,
+      "step": 11880
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 8.574363837521734e-06,
+      "loss": 0.0478,
+      "step": 11890
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 8.534850640113798e-06,
+      "loss": 0.043,
+      "step": 11900
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 8.495337442705864e-06,
+      "loss": 0.0529,
+      "step": 11910
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 8.45582424529793e-06,
+      "loss": 0.0409,
+      "step": 11920
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 8.416311047889995e-06,
+      "loss": 0.038,
+      "step": 11930
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 8.376797850482061e-06,
+      "loss": 0.0508,
+      "step": 11940
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 8.337284653074127e-06,
+      "loss": 0.0554,
+      "step": 11950
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 8.297771455666193e-06,
+      "loss": 0.0602,
+      "step": 11960
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 8.258258258258259e-06,
+      "loss": 0.0382,
+      "step": 11970
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 8.218745060850325e-06,
+      "loss": 0.0473,
+      "step": 11980
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 8.179231863442389e-06,
+      "loss": 0.0724,
+      "step": 11990
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 8.139718666034456e-06,
+      "loss": 0.0527,
+      "step": 12000
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 8.100205468626522e-06,
+      "loss": 0.0457,
+      "step": 12010
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 8.060692271218586e-06,
+      "loss": 0.0631,
+      "step": 12020
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 8.021179073810652e-06,
+      "loss": 0.0381,
+      "step": 12030
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 7.98166587640272e-06,
+      "loss": 0.0779,
+      "step": 12040
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 7.942152678994786e-06,
+      "loss": 0.0408,
+      "step": 12050
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 7.90263948158685e-06,
+      "loss": 0.0433,
+      "step": 12060
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 7.863126284178916e-06,
+      "loss": 0.0717,
+      "step": 12070
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 7.823613086770983e-06,
+      "loss": 0.0606,
+      "step": 12080
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 7.784099889363047e-06,
+      "loss": 0.0512,
+      "step": 12090
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 7.744586691955113e-06,
+      "loss": 0.0527,
+      "step": 12100
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 7.705073494547179e-06,
+      "loss": 0.057,
+      "step": 12110
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 7.665560297139245e-06,
+      "loss": 0.0668,
+      "step": 12120
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 7.626047099731311e-06,
+      "loss": 0.0453,
+      "step": 12130
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 7.586533902323377e-06,
+      "loss": 0.052,
+      "step": 12140
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 7.5470207049154415e-06,
+      "loss": 0.0524,
+      "step": 12150
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 7.507507507507508e-06,
+      "loss": 0.0504,
+      "step": 12160
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 7.467994310099574e-06,
+      "loss": 0.0634,
+      "step": 12170
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 7.428481112691639e-06,
+      "loss": 0.0407,
+      "step": 12180
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 7.388967915283705e-06,
+      "loss": 0.057,
+      "step": 12190
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 7.3494547178757716e-06,
+      "loss": 0.0459,
+      "step": 12200
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 7.3099415204678366e-06,
+      "loss": 0.0447,
+      "step": 12210
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 7.270428323059902e-06,
+      "loss": 0.0483,
+      "step": 12220
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 7.230915125651968e-06,
+      "loss": 0.0626,
+      "step": 12230
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 7.191401928244033e-06,
+      "loss": 0.0274,
+      "step": 12240
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 7.1518887308361e-06,
+      "loss": 0.0515,
+      "step": 12250
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 7.112375533428166e-06,
+      "loss": 0.0374,
+      "step": 12260
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 7.072862336020231e-06,
+      "loss": 0.0572,
+      "step": 12270
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 7.033349138612297e-06,
+      "loss": 0.0712,
+      "step": 12280
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 6.993835941204363e-06,
+      "loss": 0.0516,
+      "step": 12290
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 6.9543227437964275e-06,
+      "loss": 0.0384,
+      "step": 12300
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 6.914809546388494e-06,
+      "loss": 0.0334,
+      "step": 12310
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 6.87529634898056e-06,
+      "loss": 0.0639,
+      "step": 12320
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 6.835783151572625e-06,
+      "loss": 0.0595,
+      "step": 12330
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 6.796269954164691e-06,
+      "loss": 0.0546,
+      "step": 12340
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 0.0501,
+      "step": 12350
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 6.7172435593488225e-06,
+      "loss": 0.0774,
+      "step": 12360
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 6.677730361940888e-06,
+      "loss": 0.0603,
+      "step": 12370
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 6.638217164532955e-06,
+      "loss": 0.0615,
+      "step": 12380
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 6.598703967125019e-06,
+      "loss": 0.0507,
+      "step": 12390
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 6.559190769717086e-06,
+      "loss": 0.056,
+      "step": 12400
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 6.519677572309152e-06,
+      "loss": 0.0573,
+      "step": 12410
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 6.480164374901217e-06,
+      "loss": 0.0435,
+      "step": 12420
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 6.440651177493283e-06,
+      "loss": 0.0643,
+      "step": 12430
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 6.401137980085349e-06,
+      "loss": 0.0403,
+      "step": 12440
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 6.361624782677414e-06,
+      "loss": 0.037,
+      "step": 12450
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 6.32211158526948e-06,
+      "loss": 0.0282,
+      "step": 12460
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 6.282598387861546e-06,
+      "loss": 0.0445,
+      "step": 12470
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 6.243085190453612e-06,
+      "loss": 0.0559,
+      "step": 12480
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 6.203571993045678e-06,
+      "loss": 0.0594,
+      "step": 12490
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 6.1640587956377435e-06,
+      "loss": 0.0491,
+      "step": 12500
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 6.124545598229809e-06,
+      "loss": 0.0578,
+      "step": 12510
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 6.085032400821874e-06,
+      "loss": 0.039,
+      "step": 12520
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 6.045519203413941e-06,
+      "loss": 0.0562,
+      "step": 12530
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 6.006006006006006e-06,
+      "loss": 0.0346,
+      "step": 12540
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 5.966492808598072e-06,
+      "loss": 0.047,
+      "step": 12550
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 5.926979611190138e-06,
+      "loss": 0.0487,
+      "step": 12560
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 5.8874664137822036e-06,
+      "loss": 0.0598,
+      "step": 12570
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 5.8479532163742686e-06,
+      "loss": 0.0219,
+      "step": 12580
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 5.808440018966335e-06,
+      "loss": 0.0768,
+      "step": 12590
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 5.7689268215584e-06,
+      "loss": 0.0258,
+      "step": 12600
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 5.729413624150467e-06,
+      "loss": 0.0708,
+      "step": 12610
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 5.689900426742533e-06,
+      "loss": 0.0536,
+      "step": 12620
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 5.650387229334598e-06,
+      "loss": 0.0318,
+      "step": 12630
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 5.6108740319266645e-06,
+      "loss": 0.0414,
+      "step": 12640
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 5.5713608345187295e-06,
+      "loss": 0.065,
+      "step": 12650
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.9931,
+      "eval_loss": 0.02777969278395176,
+      "eval_runtime": 46.9096,
+      "eval_samples_per_second": 213.176,
+      "eval_steps_per_second": 13.323,
+      "step": 12656
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 5.531847637110795e-06,
+      "loss": 0.0585,
+      "step": 12660
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 5.492334439702861e-06,
+      "loss": 0.0373,
+      "step": 12670
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 5.452821242294927e-06,
+      "loss": 0.045,
+      "step": 12680
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 5.413308044886992e-06,
+      "loss": 0.0316,
+      "step": 12690
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 5.373794847479059e-06,
+      "loss": 0.0273,
+      "step": 12700
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 5.334281650071124e-06,
+      "loss": 0.0388,
+      "step": 12710
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 5.2947684526631895e-06,
+      "loss": 0.0353,
+      "step": 12720
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 5.255255255255255e-06,
+      "loss": 0.055,
+      "step": 12730
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 5.215742057847321e-06,
+      "loss": 0.0595,
+      "step": 12740
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 5.176228860439387e-06,
+      "loss": 0.0548,
+      "step": 12750
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 5.136715663031453e-06,
+      "loss": 0.0523,
+      "step": 12760
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 5.097202465623519e-06,
+      "loss": 0.0536,
+      "step": 12770
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 5.057689268215584e-06,
+      "loss": 0.0421,
+      "step": 12780
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 5.0181760708076505e-06,
+      "loss": 0.0306,
+      "step": 12790
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 4.9786628733997155e-06,
+      "loss": 0.0535,
+      "step": 12800
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 4.939149675991781e-06,
+      "loss": 0.0355,
+      "step": 12810
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 4.899636478583847e-06,
+      "loss": 0.0537,
+      "step": 12820
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 4.860123281175913e-06,
+      "loss": 0.0475,
+      "step": 12830
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 4.820610083767979e-06,
+      "loss": 0.051,
+      "step": 12840
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 4.781096886360045e-06,
+      "loss": 0.0521,
+      "step": 12850
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 4.74158368895211e-06,
+      "loss": 0.0274,
+      "step": 12860
+    },
+    {
+      "epoch": 9.15,
+      "learning_rate": 4.702070491544176e-06,
+      "loss": 0.0516,
+      "step": 12870
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 4.662557294136241e-06,
+      "loss": 0.0275,
+      "step": 12880
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 4.623044096728307e-06,
+      "loss": 0.0562,
+      "step": 12890
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 4.583530899320374e-06,
+      "loss": 0.04,
+      "step": 12900
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 4.544017701912439e-06,
+      "loss": 0.0313,
+      "step": 12910
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 4.504504504504505e-06,
+      "loss": 0.0261,
+      "step": 12920
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 4.464991307096571e-06,
+      "loss": 0.0235,
+      "step": 12930
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 4.4254781096886364e-06,
+      "loss": 0.0479,
+      "step": 12940
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 4.3859649122807014e-06,
+      "loss": 0.0404,
+      "step": 12950
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 4.346451714872768e-06,
+      "loss": 0.0583,
+      "step": 12960
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 4.306938517464833e-06,
+      "loss": 0.0594,
+      "step": 12970
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 4.267425320056899e-06,
+      "loss": 0.0318,
+      "step": 12980
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 4.227912122648965e-06,
+      "loss": 0.0533,
+      "step": 12990
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 4.188398925241031e-06,
+      "loss": 0.0497,
+      "step": 13000
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 4.1488857278330965e-06,
+      "loss": 0.037,
+      "step": 13010
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 4.109372530425162e-06,
+      "loss": 0.0369,
+      "step": 13020
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 4.069859333017228e-06,
+      "loss": 0.0585,
+      "step": 13030
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 4.030346135609293e-06,
+      "loss": 0.0397,
+      "step": 13040
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 3.99083293820136e-06,
+      "loss": 0.0319,
+      "step": 13050
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 3.951319740793425e-06,
+      "loss": 0.0656,
+      "step": 13060
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 3.9118065433854916e-06,
+      "loss": 0.0635,
+      "step": 13070
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 3.8722933459775566e-06,
+      "loss": 0.0458,
+      "step": 13080
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 3.832780148569622e-06,
+      "loss": 0.0597,
+      "step": 13090
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 3.7932669511616887e-06,
+      "loss": 0.0584,
+      "step": 13100
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 3.753753753753754e-06,
+      "loss": 0.0445,
+      "step": 13110
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 3.7142405563458195e-06,
+      "loss": 0.0331,
+      "step": 13120
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 3.6747273589378858e-06,
+      "loss": 0.0439,
+      "step": 13130
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 3.635214161529951e-06,
+      "loss": 0.0414,
+      "step": 13140
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 3.5957009641220166e-06,
+      "loss": 0.069,
+      "step": 13150
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 3.556187766714083e-06,
+      "loss": 0.0434,
+      "step": 13160
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 3.5166745693061483e-06,
+      "loss": 0.0492,
+      "step": 13170
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 3.4771613718982137e-06,
+      "loss": 0.0437,
+      "step": 13180
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 3.43764817449028e-06,
+      "loss": 0.0601,
+      "step": 13190
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 3.3981349770823454e-06,
+      "loss": 0.0405,
+      "step": 13200
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 3.3586217796744113e-06,
+      "loss": 0.037,
+      "step": 13210
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 3.3191085822664775e-06,
+      "loss": 0.0515,
+      "step": 13220
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 3.279595384858543e-06,
+      "loss": 0.0642,
+      "step": 13230
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 3.2400821874506084e-06,
+      "loss": 0.0317,
+      "step": 13240
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 3.2005689900426746e-06,
+      "loss": 0.043,
+      "step": 13250
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 3.16105579263474e-06,
+      "loss": 0.064,
+      "step": 13260
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 3.121542595226806e-06,
+      "loss": 0.0375,
+      "step": 13270
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 3.0820293978188718e-06,
+      "loss": 0.0424,
+      "step": 13280
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 3.042516200410937e-06,
+      "loss": 0.0637,
+      "step": 13290
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 3.003003003003003e-06,
+      "loss": 0.0675,
+      "step": 13300
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 2.963489805595069e-06,
+      "loss": 0.0533,
+      "step": 13310
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 2.9239766081871343e-06,
+      "loss": 0.0419,
+      "step": 13320
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 2.8844634107792e-06,
+      "loss": 0.0513,
+      "step": 13330
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 2.8449502133712664e-06,
+      "loss": 0.0591,
+      "step": 13340
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 2.8054370159633322e-06,
+      "loss": 0.0617,
+      "step": 13350
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 2.7659238185553977e-06,
+      "loss": 0.0238,
+      "step": 13360
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 2.7264106211474635e-06,
+      "loss": 0.0516,
+      "step": 13370
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 2.6868974237395294e-06,
+      "loss": 0.066,
+      "step": 13380
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 2.6473842263315948e-06,
+      "loss": 0.0527,
+      "step": 13390
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 2.6078710289236606e-06,
+      "loss": 0.0346,
+      "step": 13400
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 2.5683578315157265e-06,
+      "loss": 0.0412,
+      "step": 13410
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 2.528844634107792e-06,
+      "loss": 0.043,
+      "step": 13420
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 2.4893314366998577e-06,
+      "loss": 0.0676,
+      "step": 13430
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 2.4498182392919236e-06,
+      "loss": 0.0318,
+      "step": 13440
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 2.4103050418839894e-06,
+      "loss": 0.0422,
+      "step": 13450
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 2.370791844476055e-06,
+      "loss": 0.033,
+      "step": 13460
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 2.3312786470681207e-06,
+      "loss": 0.0319,
+      "step": 13470
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 2.291765449660187e-06,
+      "loss": 0.0652,
+      "step": 13480
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 2.2522522522522524e-06,
+      "loss": 0.0298,
+      "step": 13490
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 2.2127390548443182e-06,
+      "loss": 0.0277,
+      "step": 13500
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 2.173225857436384e-06,
+      "loss": 0.0445,
+      "step": 13510
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 2.1337126600284495e-06,
+      "loss": 0.0472,
+      "step": 13520
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 2.0941994626205153e-06,
+      "loss": 0.0409,
+      "step": 13530
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 2.054686265212581e-06,
+      "loss": 0.0517,
+      "step": 13540
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 2.0151730678046466e-06,
+      "loss": 0.0329,
+      "step": 13550
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 1.9756598703967124e-06,
+      "loss": 0.0255,
+      "step": 13560
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 1.9361466729887783e-06,
+      "loss": 0.0367,
+      "step": 13570
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 1.8966334755808443e-06,
+      "loss": 0.0432,
+      "step": 13580
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 1.8571202781729098e-06,
+      "loss": 0.02,
+      "step": 13590
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 1.8176070807649756e-06,
+      "loss": 0.0538,
+      "step": 13600
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 1.7780938833570414e-06,
+      "loss": 0.0402,
+      "step": 13610
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 1.7385806859491069e-06,
+      "loss": 0.0452,
+      "step": 13620
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 1.6990674885411727e-06,
+      "loss": 0.0288,
+      "step": 13630
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 1.6595542911332388e-06,
+      "loss": 0.062,
+      "step": 13640
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 1.6200410937253042e-06,
+      "loss": 0.051,
+      "step": 13650
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 1.58052789631737e-06,
+      "loss": 0.0172,
+      "step": 13660
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 1.5410146989094359e-06,
+      "loss": 0.0482,
+      "step": 13670
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 1.5015015015015015e-06,
+      "loss": 0.0499,
+      "step": 13680
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 1.4619883040935671e-06,
+      "loss": 0.0369,
+      "step": 13690
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 1.4224751066856332e-06,
+      "loss": 0.0491,
+      "step": 13700
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 1.3829619092776988e-06,
+      "loss": 0.0257,
+      "step": 13710
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 1.3434487118697647e-06,
+      "loss": 0.0306,
+      "step": 13720
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 1.3039355144618303e-06,
+      "loss": 0.0633,
+      "step": 13730
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 1.264422317053896e-06,
+      "loss": 0.0587,
+      "step": 13740
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 1.2249091196459618e-06,
+      "loss": 0.0492,
+      "step": 13750
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 1.1853959222380274e-06,
+      "loss": 0.0379,
+      "step": 13760
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 1.1458827248300935e-06,
+      "loss": 0.0437,
+      "step": 13770
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 1.1063695274221591e-06,
+      "loss": 0.0558,
+      "step": 13780
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 1.0668563300142247e-06,
+      "loss": 0.048,
+      "step": 13790
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 1.0273431326062906e-06,
+      "loss": 0.0338,
+      "step": 13800
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 9.878299351983562e-07,
+      "loss": 0.0376,
+      "step": 13810
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 9.483167377904222e-07,
+      "loss": 0.0452,
+      "step": 13820
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 9.088035403824878e-07,
+      "loss": 0.0792,
+      "step": 13830
+    },
+    {
+      "epoch": 9.84,
+      "learning_rate": 8.692903429745534e-07,
+      "loss": 0.0402,
+      "step": 13840
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 8.297771455666194e-07,
+      "loss": 0.0442,
+      "step": 13850
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 7.90263948158685e-07,
+      "loss": 0.0418,
+      "step": 13860
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 7.507507507507508e-07,
+      "loss": 0.042,
+      "step": 13870
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 7.112375533428166e-07,
+      "loss": 0.0699,
+      "step": 13880
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 6.717243559348823e-07,
+      "loss": 0.029,
+      "step": 13890
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 6.32211158526948e-07,
+      "loss": 0.0411,
+      "step": 13900
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 5.926979611190137e-07,
+      "loss": 0.061,
+      "step": 13910
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 5.531847637110796e-07,
+      "loss": 0.0332,
+      "step": 13920
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 5.136715663031453e-07,
+      "loss": 0.035,
+      "step": 13930
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 4.741583688952111e-07,
+      "loss": 0.0566,
+      "step": 13940
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 4.346451714872767e-07,
+      "loss": 0.0373,
+      "step": 13950
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 3.951319740793425e-07,
+      "loss": 0.0403,
+      "step": 13960
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 3.556187766714083e-07,
+      "loss": 0.0628,
+      "step": 13970
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 3.16105579263474e-07,
+      "loss": 0.0534,
+      "step": 13980
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 2.765923818555398e-07,
+      "loss": 0.0508,
+      "step": 13990
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 2.3707918444760554e-07,
+      "loss": 0.0316,
+      "step": 14000
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 1.9756598703967125e-07,
+      "loss": 0.0687,
+      "step": 14010
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 1.58052789631737e-07,
+      "loss": 0.0399,
+      "step": 14020
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 1.1853959222380277e-07,
+      "loss": 0.0329,
+      "step": 14030
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 7.90263948158685e-08,
+      "loss": 0.0349,
+      "step": 14040
+    },
+    {
+      "epoch": 9.99,
+      "learning_rate": 3.951319740793425e-08,
+      "loss": 0.0491,
+      "step": 14050
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 0.0,
+      "loss": 0.045,
+      "step": 14060
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.9928,
+      "eval_loss": 0.027864722535014153,
+      "eval_runtime": 46.9525,
+      "eval_samples_per_second": 212.981,
+      "eval_steps_per_second": 13.311,
+      "step": 14060
+    },
+    {
+      "epoch": 10.0,
+      "step": 14060,
+      "total_flos": 6.9730391934794465e+19,
+      "train_loss": 0.0988586475932242,
+      "train_runtime": 11106.5671,
+      "train_samples_per_second": 81.033,
+      "train_steps_per_second": 1.266
     }
   ],
-  "max_steps": 4218,
-  "num_train_epochs": 3,
-  "total_flos": 2.091911758043834e+19,
+  "max_steps": 14060,
+  "num_train_epochs": 10,
+  "total_flos": 6.9730391934794465e+19,
   "trial_name": null,
   "trial_params": null
 }